Article body

We may start from the fact that the intellectual enterprise – or ‘effort’ as Henri Bergson would say (see Kristian Bankov 2000) – of ‘musical semiotics’ exists. It is no longer in the state of a Saussurian prophecy such that, “Puisqu’elle n’existe pas encore, on ne peut dire ce qu’elle sera; mais elle a droit à l’existence, sa place est déterminée d’avance”. (Since it does not yet exist [Saussure is speaking here of semiology], no one can say what it would be; but it has a right to existence, a place staked out in advance; Saussure 1916/1995 : 33). In its actual form musical semiotics exists either as a branch of general semiotics, as the application of some general semiotic theory to music, or as a subdiscipline of musicology, the science of music, which studies both European and non-European, ‘classical’ and popular forms of music. Since its inception in the 1960s and 1970s, musical semiotics has now become institutionalized in all three of these avenues. World congresses of Semiotics, held by the IASS/AIS, have shown so much through the many sessions and roundtable discussions organized on the topic. The same holds true for world congresses of musicology, where even the most conservative-minded scholars have had to admit the emergence of this new field of musical studies. However, if musical semiotics exists, what, we may ask, are its aims?

To be sure, it studies music, yet which music? Also, where is its object : ‘music’? Some say, it does not exist in a particular place as a fixed object at all. If you read, for instance, À la recherche du temps perdu by Marcel Proust and particularly the famous portrayal of a concert at the Parisian salon of Madame Verdurin at the turn of the century, you find a genial and profound conception of the essence of music itself. It is simply omnipresent at all levels of communication in this context, as the Parisian high society has gathered to listen to a posthumous work by the fictive composer Vinteuil. Here, music exists, of course, in the score which identifies the work and whose establishment was the result of a great labor by the daughter of Vinteuil (and her friend) from her father’s sketches. Yet, it also exists in the aesthetics of Vinteuil : what did he want his musical work to convey, express, signify – to use the three categories of musical meaning as discussed by Roger Scruton. (Scruton 1997 : 118-141)? To what other musical works does it refer to as possible ‘interpretants’, to use the Peircean terminology? Who is performing the music, i.e., rendering the notation and abstract structure into a sonorous, sensory and sensual reality? By which kind of corporeal activity do they effectuate it? (We know of Proust’s wonderful theory of performance whereby, when a musician starts to play on a stage, they step into another body than their normal physical body; it is their intentional body [Proust 2003; Tarasti 2015 : 222]). How does the music sound? How is it understood by the listeners (or, rather, misunderstood, since Madame Verdurin’s salon is full of Parisian snobs who only pretend to understand music, but in reality do not care for it)? Which social factors were present in the context of the concert and caused it? M. de Charlus wanted his favorite violinist, Morel, to be decorated and so he served as the initiator of the event (Proust 2003). Here, as we can see, the musical reality that is portrayed combines a host of extremely diverse phenomena, yet without any one of them in this ‘chain of communication’ being more privileged than the other. As John Cage once put it : “there is no correct understanding of music anywhere” (1981 : 90).

If music is so extremely manifold (and even ephemeral) in its existence, how could it reasonably be studied? All musical semiotics start from the premise that music carries meanings, that it possesses sense and that it constitutes a meaningful activity. So it can accept Eduard Hanslick’s formalist theory of music as tönend bewegte Formen (moving sound forms), but only as what constitutes the minimal or primal aspect of musical signification. This was clearly stated by Boris Gasparow in his early work on musical semantics (1976). And if, for a semiotician, music concerns meaning, then we are tempted to investigate it as a form of communication. The Swedish musicologist Ingmar Bengtsson proposed in the 1970s his ‘chain of musical communication’ as the starting point for any musicological work (1977 : 17). But he did not consider the problem of what it is that is being communicated in the chain, i.e., the musical signs themselves and their meanings.

When we scrutinize the vast spectrum of semiotic approaches as they appear in the works of classic authors, or else in textbooks or encyclopedias of semiotics, it is noteworthy how few of the field’s founding fathers have said anything about music. Greimas? Nothing. Juri Lotman? Nothing. Saussure? Nothing. Peirce? Here there are a few things : he was musical, played clarinet and translated the libretti of Wagner’s operas into English, and he also mentioned that the performance of a piece of concerted music was a sign, while discussing his concept of the “emotional interpretant” (CP 5.475). He, elsewhere, illustrated his category of Firstness with the example of a “prolonged musical note” (CP 1.303). Next, there is Umberto Eco who, in his Struttura assente (1968) claimed that a C note on a musical score denotes the C of a musical instrument, but does not possess any connotations (yet, C and C major have a lot of connotations; think of the C major fugue in the finale of Mozart’s Jupiter Symphony, or the apotheosis of C in the neoclassical music of Stravinsky and Prokofiev). Later, Eco wrote about medieval music and its numeric symbolism. He was also very musical himself, capable of directing improvised choirs as I have had occasion to see and hear at his castle in Monte Cerignone.

If we turn to France, the situation is slightly different. We may ignore Julia Kristeva (though her sister, Ivanka Stoianova developed a form musical semiotics, see Stoianova 1978, 1996) and Michel Foucault, neither of whom spoke about music. Instead, think of Lévi-Strauss – his work is filled with musical references and it is well known that he once wished to become a composer but was unable to do with music what he later achieved with myths! His Mythologiques has many sections whose titles allude to music from “Well-Tempered Astronomy” and “Fugue of Five Senses” to “Pastoral Symphony” and “Sonata of Good Manners”, and ending with an analysis of Maurice Ravel’s Bolero (Lévi-Strauss 1964-1971). The role of music was also decisive in his classical study of the myth of Oedipus. From this source grew the whole Nattiez/Ruwet school of musical semiotics and the method of ‘paradigmatic analysis’ which for a long time – until the 1980s – seemed to offer the final word on the topic (Nattiez 1990). And then there was Roland Barthes, whose writings were saturated with musical sensitivity, and who published many inspiring essays on music, like his essay on Robert Schumann’s Kreisleriana (1975) or “The Grain of the Voice” (1977) – moreover Barthes was good pianist himself with nice elegant hands, as I remember.

Now, what of applying general semiotic theories to music? I would dare to say that if one were to move directly from semiotic theory to musical practice, one would most likely fail. For the theory must first be reinterpreted in musical terms, i.e., there must be an intermediary level of analysis, one taking into account all the achievements of traditional musical research on musical meaning and communication. Otherwise, the danger of musical semiotics moving immediately from abstract semiotic concepts to musical reality is that it might seek to forcefully adapt music to concepts which are alien to it. This has been the problem of much musical semiotics developed from linguistic theories that started from the premise that music is a ‘language’, as tempting as this hypothesis may first seem to be. In music and language, of course, we are dealing with linear chains, or syntagma, of signs, put one after the other according to a syntax, so that ‘phrases’, whether linguistic or musical, become ‘grammatical and ‘well-formed’, with both being operative when ‘uttered’ through an audio channel. Yet, too much is omitted when the non-verbal nature of music is neglected. In what follows I shall present two such alternatives from my personal research, namely 1) musical semiotics based on the Paris School, 2) musical semiotics based on my existential semiotic theory.

There exists, however, another avenue for musical semiotics, one that starts from the music itself, namely those musical practices and behaviors (to use Alan P. Merriam’s terminology 1964) that have appeared in various cultures and in musical history. I do not completely subscribe to this approach since the amount of data to be examined is endless, moving in all sorts of directions at once, and since without a small, even slight, ‘pre-understanding’ of the issue or without some theoretical guidance to begin with, it would simply be impossible to study this data reasonably or systematically. Some theoretical grounding is therefore necessary. Yet our approach must remain close to musical reality so that, in following through on the idea of musical signification through the ages, we do not altogether alienate ourselves from the musical life itself. Moreover, in this fashion, our theories will be more easily understood by anyone inclined toward the intellectual pondering of music and possessing enough musical experience to have a kind of musical competence at that level (let the latter be divided into two groups as Gino Stefani proposes, i.e., high culture and popular competence, competenza colta/popolare [1982]).

Musical Semiotics via Musical Practice

Here I gather only some of the most obvious cases from the history of the Western musical ‘canon’ of art music. I should perhaps say ‘classical music’, which is a huge category of musical phenomena in our contemporary global marketing and consumption society, stemming from what was once defined as the ‘classical style’ in Vienna. Guido Adler (1911) saw it as a musical art in which form and content coincided and appearing in three historical periods : in the polyphonic structures of linear vocal art of Palestrina in the 16th century, in the baroque composers such as J.S. Bach and G.Fr. Handel (with fugue as the central technique), and in the so-called Viennese-classical style of Haydn, Mozart and Beethoven. This ‘hardcore’ of the Western musical canon, which ethnomusicologists want so willingly to ‘decanonize’, became the ‘classical music tradition’ in the academic world after World War II. It was studied in detail by the Austrian musicologist Heinrich Schenker using a method of analysis stemming from what he called Ursatz (fundamental structure – the most abstract form of the music consisting of the two part contrapuntal progression or motion which forms the background of all tonal music), Urlinie (the upper voice of the Ursatz consisting of a linear progression) and Bassbrechung (the background bass structure that spans complete movements in a piece) (one of the best introductions in the English language on the bass line was written by the young British scholar Tom Pankhurst [2008]). Schenker’s Ursatz actually came from Goethe and his organic philosophy, as it appeared, amongst other things, in his theory of plants where he spoke of the mysterious Urpflanze.

This theoretical dimension is important. In German musicology, such scholars as H.G Eggebrecht (1996) claimed that only this kind of music (Western art music), has, after all, been ‘victorious’ in the history of music, and that it was ‘theoriefähig’, or capable of a theoretical formulation. Carl Dahlhaus noted that in music history, the most abstract theories have had the most practical consequences in the end. One does not need to be a Hegelian speculative philosopher to admit that this holds true for the whole history of humanity. (Simply consider what power theologico-philosophical worldviews have in current world politics, for instance).

Moreover, this classical music tradition has now become commercialized by electronic technology : the whole world, regardless of culture or country, can now listen to it. The category was extended from the cultures of 16th – 18th century to the whole Western art music tradition, as we witness it by visiting music stores or reading magazines such as Classical Music. In the end, this turned out to be a kind of music that all cultures have wanted to play and listen. Its success has been enormous from Venezuela to China, and beyond. One should perhaps not see it radically as a phenomenon of global colonization, since if there are true innovations in human history, it would make sense that all nations, peoples, and cultures should have access to it. Therefore I do not see Western art music only as a fateful symptom of Western hegemony.

But then what are the musical meanings and significations of this vast and widespread musical practice such that it has become appealing to music lovers everywhere? If we consider Western art music in the terms set by UNESCO, then we would have to argue that it corresponds to more than a material cultural heritage – actual sounds fixed in an aural form – for it also constitutes an immaterial one in that it carries sense and values of an aesthetic and, even, moral mature.


In the so-called Classical Age, i.e., the 17th century, all scholars, philosophers etc., were required to master music and its theory. Therefore, it is not surprising to have treatises on music by Descartes or Leibniz. They often consider the meaning of music intervals, like in the Compendium musicae by Descartes, wherein he writes about musical sound that its goal is to please and introduce various passions in us (1987 : 54). (An obviously appealing idea for any semiotician interested in the theory of passions as developed by the Paris school of semiotics! – see Greimas & Fontanille 1993). Next, he investigates the means or signs whereby this happens :

En ce qui concerne la variété des passions que la musique peut exciter par la variété de la mesure, je dis qu’en général une mesure lente excite en nous également des passions lentes, comme le sont la langueur, la tristesse, la crainte , l’orgueuil, etc. et que la mesure rapide fait naître aussi des passions rapides, commes la joie etc. (Descartes op. cit. : 62). (With regards to the variety of passions music can excite owing to the variety of tempos, I say that a slow tempo generally excites in us slow passions such as languidness, sadness, fright, pride, etc., and that a quick tempo gives birth to rapid passions, such as joy, etc. [My translation.)

Thereafter, Descartes studies intervals and their denotations : “De la quarte : cette consonance est la plus malheureuse de toutes, et n’est jamais utilisée dans les airs sinon par accident et avec le secours des autres”. (ibid. : 82). (Of the fourth : this consonance is the saddest of all, and is never used in melodies unless by accident and with the help of others).

One of the greatest erudites of the time was the Jesuit Athanasius Kircher, whose Musurgia universalis (1650) was the main source of all musical treatises published in Europe. It contains all the knowledge the baroque period had hitherto reached and introduced us to what became one of the most central theories of musical meaning, namely rhetoric. For instance, Kircher proposed three fundamental rhetorical figures in the chapter De figuris musicis, namely anabasis, catabasis and circulation or Kyklosis. He explained these as follows :

Anabasis sive Ascensio est periodus harmonica, qua exaltationem, ascensionem v el res alteras et eminentes exprimimus , ut illud Moralis “scendens Christus in altum” (Anabasis or ascension is a harmonic period which expresses rise to other and eminent things, as Moralis said, “Christ rising to the height”).

Catabasis sive descensus periodus harmonica est, qua oppositos priori affectus pronunciamus servitutis, humilitatis, depressionis affectus , atque infimis rebus exprimendis ut illud ...Massentii “Descenderunt in inferum viventes” (Katabasis or descent is a harmonic period which contrarily expresses the effect of service, humility, depression, and going down to spheres of which Massentius said, “They descended alive to inferno”).

Kyklosis sive circulatio est periodus harmonica qua voces quasi in circulumu agit videntur, servitque verbis actionem circularem exprimentibus [....] (Kyklosis or circulation is a harmonic period which seems to move voices in a circle expressing circular activities). (Quoted from Unger 2000 : 94)

So, if the text to be composed possessed any of these movements, rising, sinking or moving around, then they had to appear in their musical equivalent as well. Unger, in his basic introduction to musical rhetoric (Unger 1941/2000 : 94), has given diagrams of various figures in which we see the quantity of such musical expressions : (See Figure 1 on annex pp. 50-51).

This was a whole thesaurus of common musical expressions mastered by musicians and listeners alike. They provided musical practice, which had just emancipated into purely instrumental form and was moving towards what was later known (especially in German musicology) as ‘absolute music’, with figures of musical signification. The preludes and fugues of J.S. Bach grew out of this context. Of course, no one has ever been able to make explicit, say through linguistic statements, the relationship of preludes and fugues in each of the keys relative to The Well-Tempered Clavier, yet, the music itself, which may sound like a pure mathematical structure, is nonetheless highly meaningful.

An example of such meaning is dialogue. The latter is often embedded in the music, creating specific meaningful situations : for instance, it can happen as a simultaneous relation between different parts and levels, like in the prelude in C minor of Well-Tempered Cl I, or as the dialogue and almost polemic fighting between theme and countersubject in the Fugue in F sharp major, where the latter ‘wins’ in the final resolution. It can also appear in the form of a cross in the C sharp minor fugue Well-Tempered I, which iconically evokes Christ and the crucifixion. It can appear as a vocal gesture, like the imitation of the ribattuta singing techniques (a form of ornement) in the opening trill of the Prelude in G minor. It can be passus duriusculus, the diabolic element and katabasis in the theme of the Fugue in F minor. It can be found in the orchestration of a fugal texture like in the same fugue in bars 34-36 and 40-43, in which dense texture (like that of the trombone or the trumpets) manifest an ‘appearance’, such as that of an angel announcing the end of times. Sometimes the meaning emerges as a breaking of the rules, like in the F minor fugue by Handel in his suite, where suddenly the theme of the fugue arises into an unexpected operatic lyrical power, which should, by no means, happen in any ‘grammatically’ well-formed fugue!









Music Example 1 (a-g.)

-> See the list of figures

All this goes to show that when music became “absolute”, it also became filled with various kinds of meanings and signs. Yet sense was yielded to music from other sources as well, such as Jakobson’s referential function, for instance. Indeed, Gino Stefani, a leading Italian music semiotician, richly illustrated in his early work how baroque music involved the idea of a fest of baroque times, offering almost an early form of Gesamtkunstwerk (see Stefani 1974 : 18-23). Moreover, baroque was the period of emerging national differences in musical practices, albeit, not ‘nationalism’ in the proper sense. Great composers like Bach, at the end of the period, mastered Italian, French, and German styles. Musicians also understood the possibility of differences between written music and the actual sound of music : Francois Couperin uttered the classical formula : Nous écrivons différemment de ce que nous éxécutons (We write differently than what we play; Veilhan 1977 : i). Furthermore, and especially in France, instrumental music became programmatic, as the preludes for harpsichord by Couperin show. These small character pieces followed the aesthetics of the ‘précieuses’ of the Parisian salons of the early 17th century (see Madeleine Scudery’s Le Grand Cyrus 1647-1650) among others, and appeared under various psychological titles such as La langueur, La coquetterie, Les vieux galants, Les coucous bénévoles, La jalousie, Taciturne, La frénesie, Le Désespoir, L’attendrissant (Couperin : Pièces de Clavecin III 1970), L’enchanteresse, La laborieuse, Les idées heureuses, La diligente, La ténébreuse, La lugubre, La dangereuse (Couperin : Pièces de Clavecin I; 196); or Les barricades mystérieuses – which was supposed to be linked to paintings by Watteau, where court people are portrayed in masquerading –, and where the term ‘barricades’ signifies a mask and was expressed with corresponding musical signs (see Tarasti 2015).


The major theory of the Viennese classical period onwards is that of topics. Recently it has also been a major concern of music semioticians, particularly in the British and American contexts. The notion of topics was considered by Leonard C. Ratner, in his book, Classic Music. Expression, Form, and Style (Ratner 1980), who conceived them to be “[...] subjects of musical discourse”. He wrote :

From its contacts with worship, poetry, drama, entertainment, dance, ceremony, the military, the hunt, and the life of the lower classes, music in the early 18th century developed a thesaurus of characteristic figures, which formed a rich hierarchy for classical composers. Some [...] were associated with various feelings and affections, others had a picturesque flavor [...] Topics appear as fully worked out pieces i.e. types, or as figures and progressions within a piece i.e. styles.

Ratner op. cit. : 9

Therefore, a list of topics might include : dances from minuet to sarabande, marches, military and hunt music, the singing style, the brilliant style, the French overture, Musette, Pastoral, Turkish music, Storm and Stress, Sensibility, Empfindsamkeit (Sensitive Style), the Strict style, the Learned style, Fantasia, etc. However, the use and application of such topics is not mechanical at all. Moreover, they could be made to appear in a piece of music as contrasts both in consecution and concomitantly. Topics can be seen as already paving the way toward the idea of musical narrativity or storytelling, but they also lead into what can be called, in semiotic terms, ‘complex isotopies’, i.e., superimposed topics and combinations of all kinds of affections. For instance, in Mozart, we find the Turkish topic clearly in the opera Abduction from Seraglio. However, in the opening theme of the piano sonata in A minor, we find the percussive Turkish janitshar topic combined not with a comical or grotesque musical character but with the tragic :

Music Example 2 Mozart

-> See the list of figures

A similar situation can be found in Beethoven’s Les adieux : the opening hunting signal becomes the romantic farewell by its turn to the deceptive cadence. As for the topic of the hunting horns, they also appear in Richard Strauss’ Rosenkavalier as nostalgic signs of the past.

Music Example 3 R Strauss

-> See the list of figures

The topic of galloping horses, i.e., a musical figure that imitates riding, appears throughout symphonic and operatic music. In Beethoven’s Piano Sonata op . 7 in E flat major, the transition theme is a riding motif – it is put in conflict with the chorale topic (see bars 25-32, and 59-96). Wagner’s Valkyries also appear with a famous riding motif. Even Sibelius uses the riding topic as galloping horses in his early piano quintette from 1891, when he was studying in Berlin.

Music Example 4

-> See the list of figures

The same topic appears in Mendelssohn’s Midsummer Night prelude. Mendelssohn is also a good example of a composer who creates his own topics. Thus at the beginning of Midsummer Night he elaborates a fairy tale topic with full note chords. This practice was later imitated by Wagner with his Brynnhilde’s awakening motif, and by Rimsky-Korsakov in the opening storytelling of Scheherazade, as well as by Tchaikovsky in Romeo and Juliette. In any case, most semioticians of music have recently focused on topics as an efficient tool of musical meaning. This group includes scholars such as Raymond Monelle, or Robert S. Hatten (in his Beethoven and Schubert studies), as well as Michael Spitzer, Byron Almén, Nicholas McKay, Ben Curry and others. In the same spirit, Hatten has lodged rhetorical figures at the center of music analysis as the study of tropes, metaphors, and more particularly gestures. Here, ‘metaphor’ means that a musical phrase remains the same while its meaning changes. The ‘trope’, on the other hand, refers to the fact that the musical signifier changes though the meaning stays the same. Hatten’s own definition is as follows : “Troping in music may be defined as the bringing together of two otherwise incompatible style types in a single location to produce a unique expressive meaning from their collision or fusion” (Hatten 2004 : 68). The use of topics is therefore a musical practice that possesses a strong evocative power.







Music Example 5 a-e

-> See the list of figures

As mentioned earlier, for German musicology a sonata or symphony, written by Beethoven, was considered to be the purest manifestation of ‘absolute music’, mostly on the ground of historical evidence and in line with Hanslick’s musical formalism. Nevertheless, even here, at the core of the Western canon of art music, which was supposed to consist purely and merely of organic sound, we see that we encounter the presence of meaning. The Viennese professor Hartmus Krones has studied this in his essay, “Von der ‘Beethoven’schen Redekunst am Pianoforte’. Beethovens Klavierwerke in der Sicht Czernys, Kannes und Schindlers” (Krones 1983). Indeed, it was generally thought that musicians should imitate the great poets. Carl Czerny, the student of Beethoven, confirmed this by saying that one could explain many great instrumental compositions from this standpoint. While a poet utters his ideas by reciting and declamating them with accents, the same happens in music : it should consist of a ‘speaking’ articulation uttered or performed in the proper situation. Thus, many musicians believed that Beethoven had provided his piano sonatas with contents. This was presumed by his secretary Schindler as well as by Franz Liszt and Hector Berlioz. Once, Schindler told about the deep impact that Beethoven’s sonatas, D minor and F minor op 31 and 57, had made upon him, he took the courage to ask the Master himself if there was a key or ‘code’ to understanding these pieces. Beethoven is said to have answered : Lesen Sie nur Shakespeare’s Sturm, (read only Shakespeare’s The Tempest!). Krones believes that certain passages in these sonatas actually portray the dialogue between Prospero and Miranda, as plaintiff suspiration (a rhetorical figure!) and Kyklosis figures the chromatic surrounding of certain notes :

Music Example 6 - Beginning of The Tempest Sonata in D minor.

-> See the list of figures

This brings to mind the possibility of hearing music as a ‘theater in sound’, along with the omnipresent dialogical principle.

All of it is quite amazing since generally we consider the thematic (actorial) process in Beethoven as something that is purely structural. However, even then we are besieged by meanings flowing from many connections and references. This should not, however, be brought to the extreme, as was the case with Arnold Schering who, in his Beethoven und die Dichtung, claimed that every Beethoven sonata or symphony had a literary starting point in Goethe or Schiller (1936 ).


The sources of meaning in music start to multiply with Romanticism to such an extent that music becomes truly the central art form subordinating all the others of the entire period. Musical discourse lives here in symbiosis with the other arts like poetry, theater, painting, and literature (mythology). There emerges a special, mythical, style manifesting itself through various ‘semes’, or aesthetic units of myth (see Tarasti 1979), such as the ‘nature mythical’, the ‘hero-mythical’, the ‘fabulous’, the ‘sacred’, the ‘pastoral’, the ‘legendary’, the ‘balladic’, the ‘demonic’, and the ‘national-musical’, etc.

In this context, leitmotifs appear in Richard Wagner as proper musical signs of operatic discourse. However, the idea charting out the various leitmotifs based on their dictionary meanings was not at all Wagner’s own idea but that of his pupil and admirer Hans von Wolzogen. This was rooted, however, in a manner of listening to the music that consists in following its course in a narrative sense. Although, even at the time of Wagner and of opera, older musical formations maintained their influence, we can still question whether an opera by Wagner is better described as a symphony or a ‘theater’ piece. The Romantic period is one where narrativity was foregrounded in the musical discourse. While in the Classical Age, the key semiotic elements of music resided in its rule-based nature and in its grammar or ‘langue’ – elements which Monelle considered to be the proper field of the semiotic study of music, in the Romantic Age, expression becomes central to musical meaning. This means that the signified of the musical sign lies ‘behind’ or ‘underneath’ its aural form. In fact, as of recently, this principle was the starting point for many scholars in the so-called American new musicology movement. These scholars have adopted semiotics not in its Saussurian form, as a discipline based on the study of langue, but instead according to the ideas of Julia Kristeva in her semiotic theory of khora. Following Kristeva, they understand khora as the primal non-verbal kinetic field of music, its kinetic energy, Freudian desire, gesturality, for which the patriarchal symbolic order only later brings a compelling structure. The domain of khora is like the large part of an iceberg as it lies under the sea level, its visible part being merely the superficial top of it. The khora stands for our archaic ego with its modal elements. Richard Taruskin, for instance, interprets the motif of Polovetsian girls in Borodin’s opera as a typical khora, the effeminate East, the oriental, and this is also called ‘the semiotical’ following Kristeva (Taruskin 1997 : 80). It is important to notice that these ‘postmodern’ semiotic theories often adopt their concepts in their own manner, which is also the case with British cultural theory and its applications to music. Such is, for instance, the ‘semiotic moment in the everyday musical life’, as recently defined by Tia DeNora. In speaking of music and society, she begins with the ‘grand tradition’ of Adorno but seeks to distinguish her approach from traditional ‘semiotic studies’ of music (or, rather, what she believes them to be) (DeNora 2000). Instead, she launches the ideas of music as ‘semiotic power’ and a ‘technology of self’ – ideas which, by the way, are entirely compatible with what I call an ‘existential semiotic’ approach to music (more on this in the final two sections of this essay).

The blending together of the arts intensified with the Romantic movement in France. It lead, along within Wagnerism, to the theory of the correspondence of the arts by Baudelaire and to the idea of synesthesia as a particular technique used by different artists. In music, such notions are linked to the primal element of music, namely sound (or Klang as the Wagner conductor Christian Thielemann puts it; Thielemann 2012 : 138). Timbre and keys are considered equal to colors. This prompted attempts at charting their correspondences as signs of this type of synesthetic language by Rimsky-Korsakov and Scriabin. In the latter, the use of these techniques becomes systematic as in his Prometheus chord composed of the smallest units of intervals of fourths and their characters.

Finally, with the emergence of Modernism, artists adopted a new attitude and conceived of such devices as belonging to the technical aspect of artistic creation, as in Russian formalism. Different phenomena, like Kandinsky’s non-figurative painting or else the notion that ‘point’, ‘line’ and ‘plane’ constitute the smallest units of any artwork, music, architecture or dance, begin to emerge (Kandinsky 1926/1970). Here, we are already nearing structuralism.

The development of structuralism owes a great deal to Russian formalism, which included the ‘narratological’ studies of Vladimir Propp. And it was out of structuralism that, some years later, musical semiotics emerged. Structuralists conceived all forms of signification and communication under the heading of ‘language’. As a consequence, phenomena that could not be conceived as languages were rejected as objects of study. This position lead to debates between Lévi-Strauss and Berio and others on whether nonfigurative painting or atonal music were languages. The answer usually given was that they were not languages (at least not a ‘langue ’, i.e., a language system) since they didn’t fulfill the principles of double articulation as described by the French linguist by André Martinet.

Musical Semiotics Emerging via Theories of General Semiotics

As mentioned earlier, the semiotic field is vast and variegated. A major tendency in musical semiotics consists in keeping a close contact with semiotic epistemology and theory, with the goal of directly applying them to music. So if we think of major and dominant theories in semiotics, we find : the Peircean approach, the Paris School of Greimas and his followers, Roland Barthes’s theory of codes, psychoanalysis, as well as some cognitive approaches (though in this latter case the references to semiotic vocabulary may be less obvious). There are also scholars, we should add, who develop their own methodologies from what they have found in the arsenal of the history of semiotics, following Derrida or Deleuze for instance, while others still follow quite new and contemporary theories like existential semiotics, a field that has been greatly expanding in the last few years. Some researchers also see semiotic potential in already established musicological theories such as Asafievian intonation theory. Others apply only fragments of semiotics to their empirical studies. Jean-Jacques Nattiez, for example, has based his theory on the epistemological principles developed by Jean Molino, but otherwise has not strictly followed any of the classical semiotic theories. Questions of methodology that are important for him concern the segmentation of musical texts, which is a valuable topic since we almost never can discuss an entire composition at once but need to attend to it by cutting it up into smaller sections. Another problem he has raised is that of pertinence. Indeed, as much as one wishes to proceed in an ‘objective’ and ‘empirical’ manner in musical semiotics, it is simply impossible to find what constitutes the proper criteria of pertinence for analysis without first considering the entire output of a given composer, as Nattiez himself argues in his study of Debussy’s Syrinx (see Nattiez 1975).

Paris school semiotics being one possible avenue for musical semiotics, I have myself opted for this methodology in my 1994 book A Theory of Musical Semiotics. Others have also made use of it, most notably Marta Grabócz from the University of Strasbourg. In her doctoral dissertation at the Budapest Music Academy, she studied the morphology of Franz Liszt’s piano works (1986) focusing especially on their programmatic content. She also borrowed from Greimas the concepts of ‘seme’ and ‘isotopy’ which she then applied to music. In her 2009 book, Musique, narrativité, signification (prefaced by Charles Rosen), Grabócz used the gamut of Greimassian concepts, including the semiotic square and the elementary structure of signification, applying them to Mozart’s symphonies[2] : (See Figure 2 next page).

My own first studies using Paris school methodology examined the works of Chopin. The very first one was an analysis of Chopin’s Polonaise Phantasy (Tarasti 1994 : 138-154). It started out from a segmentation of the piece into what I called its ‘narrative programs’, following the classical idea of Propp’s ‘functions’ of a story. Yet, the narrative arch, or tension, behind this surface structure consisted of the semiotic square S1, S2, non-S1 and non-S2 (which in this case were S2) : opening chords : plunging; non-S1 = opening arpeggio : rising. What is so exciting in this particular piece is that the composer starts with negative articulations of these categories letting the listener wait until the end of the piece to finally hear the true rising, absent at the beginning. Indeed, at the end, the galloping motif of the opening arpeggio is heard as a kind of ‘affirmation’, while the negation of the plunging chords is absent. It was found in the rising thirds in octaves as an upbeat in bars 271-272. This led me to conclude that in using the semiotic square for music the essential thing is to discover the correct sonorous counterparts which make sense. When successful the profound logic of the piece can be made to appear. The next problem was to consider how this synchronic square is discursively temporalized and made syntagmatic as the music unfolds.

Fig. 2

Exemple 8 : Symphonie K. 504, IIe mouvement (Andante) (schéma global de l'articulation des affects : structure élémentaire de la signification ou carré sémiotique)

Grabócz op. cit. : 149

-> See the list of figures

My second analysis of Chopin, namely a study of his Ballade in G minor, led to the most formalized analysis I have executed so far. We should keep in mind that the ideal form of scientificity the structuralist semiotician strives for is an extreme formalization and reduction of the surface manifestation of a phenomena into its fundamental (or basic) structural elements. The method used must be sufficiently rigorous as to automatically produce correct interpretations. However, in order to bring Greimas’s concepts to music much preparatory theoretical work is needed in order to consider what ‘isotopies’, ‘actors’, ‘space’, ‘time’, etc., mean in musical terms. In this sense traditional musicological theories, from Asafiev to Ernst Kurth, H. Schenker and L.B. Meyer and Charles Seeger, are needed to ensure this disciplinary mediation.

Greimas himself had put all his new theoretical concepts into a framework which he the called ‘generative path’ (parcours génératif) – the idea of ‘generativity’ was fashionable in the late 1970s and early 1980s. However, for musical purposes, I needed to reinterpret Greimas’s model, being satisfied with ‘only’ four levels of generation : isotopies; spatiality, temporality and actoriality; modalities; phemes/semes/figures (Tarasti 1994 : 48-49). The whole approach started from isotopies as the deepest level of signification determining all subsequent levels. In the next phase the distinction of outer/inner (or exteroceptive/interoceptive) was applied to musical space (outer = registers, inner = tonal tensions), time (outer = metric, inner = temporality as ‘durée’) and actors (outer = theme actors, inner = thematicity in general). The most radical innovation in the whole outline was, however, the level of ‘modalities’ which Greimas had called the ‘third semiotic revolution’. Modality here did not mean any Church tone, rather it was to be understood in the same way as in linguistics where it denotes the manners whereby a speaker is animated by their speech or utterance thanks to their expectations, emotions, certainties/uncertainties, obligations, abilities, etc. Modalization is thus the process through which an initially neutral message becomes ‘human’ merely by the intonations of speech. In music theory Vincent d’Indy, in his Cours de composition musicale (1897), had already noted the expressive importance of modality (what he referred to as ‘accents’), distinguishing three manners of uttering a sentence : neutral (‘indifférente’), interrogative (‘interrogative’) and affirmative (‘affirmative’). Indeed, modalization proved to be the most important source for musical meanings and semantics. And it concerned musical performance particularly. Thus in the study of a music like Chopin’s Ballade in G minor one could end up with a ‘modal grammar’ in which the originally linguistic modalities, e.g., ‘want’, ‘know’, ‘can’, ‘must’, and ‘believe’, were evaluated along with the piece. They were digitalized with five values : ++ , +. 0 , - ,- -, whether they appeared ‘excessively’, ‘sufficiently’, ‘neutrally’, ‘in-excessively’, or ‘insufficiently’. The starting point were the basic modalities of ‘being’ (être) and ‘doing’ (faire), being = consonant, static, doing = dissonant, dynamic (Tarasti 1994 : 177-179). Therefore at the end the modal ‘grammar’ of the whole Ballade could be written as such :

Fig. 3

Modal Activity in Chopin, Ballade in G minor, Op. 23

-> See the list of figures

These results could then be further elaborated into a more narrative form if the modalities of ‘being’, ‘doing’, ‘not-being’, ‘not-doing’ were interpreted as conjunctions or disjunctions of ‘subject’ and ‘object’ of the ‘story’ : Being = S ^ O, i.e., subject is conjuncted (for instance in a story the hero marries the princess, or acquires a golden ring, etc.) S v O, i.e., subject is disjuncted (the hero does not retrieve the princess, etc.), S non-v O, i.e., subject is going to be conjuncted, and S non ^ O i.e. subject is going to be disjuncted. The advantage of this approach for music semiotics was obvious : musical narrativity could now be compared to other forms of narrativity and the musical realm, even in the case of ‘absolute music’!, could be studied as a bonafide semiotic phenomenon alongside literature, theater, opera, cinema etc.

Fig. 4

-> See the list of figures

In Tarasti 1994 I applied this methodology, always with some variations, to various composers chosen from different style periods i.e. from Beethoven’s Waldstein Sonata to Musorgsky, Fauré, Liszt, Sibelius, Debussy, Martinu, and even to minimalist works. All this with the purpose of showing that what was involved was indeed a ‘methodology’, i.e., an analytical ‘device’ or ‘tool’ which could be used or applied to diverse styles and periods in the history of music. Yet, regardless of the success in achieving this goal, the following criticism could be formulated : the logical symbols stemming from modal logic and used by Greimas are difficult to read and to understand by a musician without prior philosophical training. Therefore, further attempts at using this method should be such as to be readable by notations which are closer to standard musical notation. It ought to become something like the Schenker method, using its own symbols within the frame set by accepted or standard notation! It turns out that one younger scholar, Tom Pankhurst, has already made important strides in this direction, yet one should go further still. It should be noted that in the U.S., Byron Almén has used many similar ideas in his study A Theory of Musical Narrative (Almén 2008). His highly erudite work has managed to bring Greimassian-type narrative analysis closer to the writings of the new musicology movement all the while making new interpretations of musical works. He also developed the analysis of actors and actants and investigated many subspecies of musical narratives such as tragic ironic and comic ones. This work shows how the narratological tradition, which started out with Propp’s Morphology of the Folktale before influencing Lévi-Strauss and Greimas, has now joined the semiotic analysis of music.

Existential Theory

Some twenty years ago my own work started moving into a different direction when I launched a novel theoretical and epistemological approach I have called ‘existential semiotics’.[3] In 2012, in my book Semiotics of Classical Music. How Mozart, Brahms and Wagner Talk To Us, I have sought to apply this new approach of mine to the study of music. As we shall see, this has led to a radical change of paradigm from French structuralism, towards a more philosophical and continental approach in line with thinkers such as Kant, Hegel, Kierkegaard, Heidegger, Jaspers, Sartre, Wahl, and Marcel.

It goes without saying that with this change in theoretical perspective my analyses of musical works have also undergone a decisive shift. It has been noted (by Robert Wilson, for instance) that my approach to music now bears some resemblance to Adorno’s. Of course, this is a flattering comment not the least because in Adorno’s case, even in his most concrete and detailed musical studies, there looms in the background the Hegelian tradition of thought. One simply cannot understand his judgments, may they concern Beethoven Wagner or Mahler, Stravinsky, Schönberg, or popular music (or even, among his colossal errors, his take on Sibelius) without knowledge of his philosophical doctrine and his intentions. Moreover, there are moments where Adorno’s reasoning comes remarkably close to semiotics. However, one should always avoid the arrogance with which Adorno often wrote or his belief that the philosopher is a person of ‘higher’ value than the musician (after all he dared criticize Beethoven, claiming, for instance, that the andante’s side theme of the 5th Symphony, was in his mind, too bombastic).

Now, because existential semiotics reconsiders the role of the subject and of subjectivity it could be seen as postmodern in some respect. To my mind, however, it might be better labeled (if one seeks such a thing) as neo-semiotics rather than anything post-. Yet even the label ‘neo-’ seems problematic since the prefix seems ever more to call to mind scientific endeavors that only aim at utilitarian purposes or else at economic profit in the context of global markets (as in ‘neo-liberalism’). As far as I’m concerned, then, existential semiotics is simply a further extension of classical, or even universal, semiotics.

The domains which concern existential semiotics are twofold : firstly, the study of Dasein, the living world of subjects and objects, and, secondly, that of transcendence beyond Dasein. Clearly, the latter area is the most radical innovation existential semiotics brings to semiotics, for it might lead one to question whether such a domain at all falls within the purview of this discipline. As for the first domain of study, it is important to note that existential semiotics does not seek to cancel or negate earlier semiotic approaches. Quite the opposite in fact : one can well continue being a Greimassian (I still occasionally refer to Greimas in my work), a Lotmanian, or even a Peircean, or whatsoever else (!) within the limits of Dasein. The difference pertains to the second domain of investigation with its idea of transcendence which seeks to offer a fuller perspective on semiotic phenomena, “une perspective englobante” (as Greimas might have said).

Although existential semiotics is still in many ways incomplete in its theorization, a work in progress if you will, I have nonetheless made an attempt at applying it to music (Tarasti 2000). Two pieces were given special consideration : Mozart’s Phantasy in D minor and Schumann’s Phantasy in C major.[4] In the next section I will offer an analysis and interpretation of Schumann’s Phantasy based on my existential semiotic theories and philosophies.

From Modes of Being to the Z Model and its Temporalization[5]

We begin with Hegel’s Logik der Wissenschaft (Science of Logic), specifically his distinction between being-in-itself and being-for-itself. As we know, the first of these is something that stands by its pure inner qualities, “an sich” – of which, according to Kant, we can know nothing. However, when this entity an sich enters into a social context, it becomes determined by it, and becomes für sich. For instance, someone who is good with his hands (an sich) may become a tailor, a pianist, or a baker (für sich); one who is good at speaking may become a teacher, politician, or priest. And so on. To this Hegelian distinction we add the point of view of the subject in the proper sense, what is in French called the “Moi” (Me). It follows that we can derive two more categories : an-mir sein (being in-myself) and für-mich sein (being for-myself).

Fig. 5

Varieties of Being in the Semiotic Square.

-> See the list of figures

The remaining category is that of “Soi”, or the self as determined by society : being-for-oneself and being-in-oneself. This last means society with its abstract values and norms; the former designates exemplifications of those values and norms via institutions and other social practices. For instance, if the value of being-in-oneself is Beauty, it can manifest itself through the institution of a conservatory if it involves aural aesthetics; if the value is visual, then through an arts academy; if patriotism, its corresponding practice might be the army. These two aspects of Moi and Soi represent two sides of our inner subjectivity : the social has impact on and power over us only because it has been internalized in our minds. In this way, the Soi can impose itself on our Moi.

Fig. 6

-> See the list of figures

Ultimately, we have four cases in this taxonomic model of ‘being’ : Moi1 (S4) constitutes our primary body, kinetic energy, desire, khora, our chaotic and fleshly physical existence. Moi2 (S3) means that, via habit, education, and dialogue with others, our subjectivity reaches a more permanent stability and identity; in other words, it becomes a person. Moi3 (= Soi2) signifies a Moi that is so reduced as to become subordinated to the Soi, i.e., to social practices and institutions. M4 (or Soi1) embodies the abstractly normative aspect of our society, community and culture.

This four-place model resembles the so-called “semiotic square” developed by Greimas, but is in fact rather different. As with the semiotic square, there is in our scheme an inherent dynamic tension between Moi and Soi : Moi gradually becomes socialized, the materiality of our ego is more and more sublimated, until it is capable of serving and representing the most intellectual and spiritual aspects of our society. On the other hand, the Soi urges the subject to become more concrete, physical and material, forming the kind of personalities that social practices and institutions can recruit for their purposes.

Fig. 7

-> See the list of figures

Of course, our most physical acts of Moi1 also contain a tiny trace of the Soi, just as in the most abstract categories of our social existence there also exists a small hint of our physical ego. Hence our square has become what I call a ‘Z model,’ the letter Z portraying these two counterforces as they encounter each other, zig-zag-like, within these four modes of being. Here it is possible to see merely a model of “organic growth” from one state of being into another, but we may also use the model to conceive of what an individual might do in shifts from one state to another. Perhaps most critical is always the step from S2 to M2, i.e., from the predominantly social to the predominantly subjective and individual : the eternal conflict between Ich und Gesellschaft (self and society), as Adorno described it.

Mieczyslaw Tomaszewski (2010) also elaborates an interesting musical application of an individual’s development through various phases. He speaks of moments in an artist’s creative life as stages of adopting a heritage : initial fascination, conflict, significant encounter, threat of existence, loneliness. These are similar to moments of our Moi 2, which appear in stylistic transformations of early, mature, climactic, and late periods of creativity. We may therefore view the movements and transformations of Moi/Soi as kinds of Bildungsprozesse, or educational phases (Tomaszewski 2010 : 96-98). Our model can serve as a taxonomy of almost anything that happens in our minds and in our society.

Fig. 8

The Z –model of Moi and Soi.

-> See the list of figures

We cannot think of any musical utterance in which we cannot distinguish among these four cases, and often all of them simultaneously. In every musical text, discourse or expression, we have these four aspects : (1) Concrete physical material appearing as musical gestures and musical desires, more or less chaotic wavelike formations of the kinetic energy of music. (2) Musical “actors” or persons, which are the more stable anthropomorphic entities, like musical motifs and themes; these have clear-cut profiles and constitute units of musical narration. (3) In music we have social norms, manifesting themselves in (a) rhetorical figures, (b) topics, i.e., style features internalized from musical or nonmusical social fields, and (c) genres, the social frameworks necessary for musical communication. (4) Every musical unit has an aesthetic aspect of content, being the values or ideas which the music conveys or signifies. For example, take Schumann’s Aufschwung piano piece : M1 = its kinetic volatile energy; M2 = its basic triple-rhythm musical motif; S2 = it belongs to the genre of Charakterstücke; S1 = it portrays Romantic aesthetics, a movement towards transcendence.

We should go further in the application of our Z model to determine how it may apply to linear and temporal texts, such as music and musical performance. Here, instead of speaking about the four different cases or positions of the square, we should talk about four different levels, which unfold in time and can proceed simultaneously, each with its own characteristic entities and formations. The M1 level consists of musical gestures and their consequences; these flow in “waves” of organic growth, as more or less chaotic manifestations of our primary will. The M2 level is comprised of units called motifs and themes, which already constitute a kind of narrative chain; i.e., they are identifiable and memorable entities of musical discourse, if still conceived in their physical and corporeal aspects. M3 = S2 articulates the musical course according to traditional formal schemes and topics such as sonata form, rondo, fugue, chaconne, concerto, opera, waltz, march, and so on. These may follow a rhetorical logic, as well, in the form of musical ‘figures of speech’. Finally, Soi1 = M4 represents the aesthetic ideas behind all these musical processes. Instead of a static, logical scheme such as the semiotic square, our model may be conceived as a generative one having proper levels of meaning. As such, it evokes similar models in the history of semiotics : Chomsky’s “tree”, which Lerdahl and Jackendoff (1983), among others, have applied to music; Greimas’s parcours génératif and its application to music (Tarasti 1994); the model for reconstructing ancient Slavic texts, as sketched by Juri Lotman and others in their famous essay, “Theses on the semiotic study of cultures” (1975). This last has levels going from lowest to highest in the form of a pyramid : phonetics, metrics, syntax, semantics and symbols. Conversely, the unifying principle proceeds from top to bottom in the “semantic gesture” theorized by the Prague School as does the “dominant” in Roman Jakobson’s model. Such generative models are “organic” in nature, such that no gaps or leaps occur in the shifts from one level to the next.

Fig. 9

The Z-model and its levels as temporalized.

-> See the list of figures

Fig. 10

Lotman et alii : Reconstruction of Texts (above); The Z-model as a Generation (below).

-> See the list of figures

I doubt, however, the validity of such a “quasi-organic” course; for leaps and ruptures occur in the move from deep structure to surface. Not all sign processes proceed uninterrupted from a guiding principle to subsequent formations on other levels. For example, let us reorient our four-level Z structure with the lowest level as M1, the second M2 and so on to the top of the pyramid, S1. This arrangement brings out the conflicts and oppositions among the different levels. For instance, something may happen on level M1 which negates the impact of S1; or the shift from S2 to M2 may not proceed unobstructed, but instead engage in struggle. Concerning Schumann’s Phantasie, much debate has taken place about whether or not it is a sonata. Against the sonata-form argument stands the Im Legendenton section, the developmental aspects of which are ambiguous; Rosen even claims, paradoxically, that it both is and is not developmental at the same time. In any case, a conflict arises : what happens on the level of M does not adapt to or fit the schemes offered by S2, namely the expectations of the sonata genre. Such incongruities can take place among any levels of the model. In fact, the possibility of differences, gaps, and negations can even serve as the constitutive force of a living musical “organism”.

Especially for such a radically innovative work as Schumann’s Phantasie, this can prove to be a significant source of musical meaning and form. If on its S level we put various kinds of aesthetic ideas prevailing in Germany at that time, say, Friedrich Schlegel’s theory of fragments or his theory of Universal Poetry, or the idea of an arabesque, or the Goethean ideology of eternal striving forwards, then we can observe the extent to which such principles manifest on subsequent levels, until arriving at the most corporeal and gestural level of M1. To do so, one must first analyze the piece on each level separately, deriving suitable concepts and perhaps finding proper notation for each type of analytic unit.

Fig. 11

Clashes Between the Levels in the Z-Model

-> See the list of figures


I have presented here some of the most important approaches in musical semiotics. It is a discipline which seems to progress rapidly and is often in touch with major contemporary scientific paradigms such as cognitive studies on one hand, or British cultural theory, on the other. As a practice it has now gained a solid foothold in academic institutions and in the schools’s curricula. Take, for instance, one aspect of academic work that often remains hidden from public view or else in the background, namely the master’s and doctoral theses that our graduate students produce. As we know, few of them become books. Yet the number of theses that call on the methodologies of semiotics to study music is in constant growth (something I have witnessed myself year afer year). And there is perhaps no better indicator of the health of a discipline than graduate work, for it reveals the deeper impact of theories and their settling into a scholarly heritage.

What then is to be the future of musical semiotics? One should imagine that it will acquire an increase of ‘social capital’ thanks in good measure to the advances of the ‘neo-oral’ culture that is flourishing around us in a world of digitization and electronic communications. One may of course ask whether musical life needs anything like the intellectual discourse offered by musical semiotics. Considering here only classical music, it seems likely that the more it is consumed and practiced by non-European cultures, the greater will be the need to discuss its history, its aesthetics and its practices, and to develop a proper metalanguage to convey its basic ideas. Musical semiotics is ideally suited to meet this challenge. And let music be ‘sui generis’, in its own realm, a musician should nonetheless know about the world of ideas that surrounds it at all levels. Richard Wagner said once that a musician who only makes music without knowing the world of ideas is like a cow, he would be better off doing something else (Cosima’s Diaries, see 1976 : 181). Musical semiotics, as we have seen, can help us uncover the deepest of musical isotopies so that we can teach them to others in musical education. Finally, musical semiotics has a claim to establish a common or universal methodology and metalanguage in that it concerns all cultures and all styles of music, and is able to account for them with its own vocabulary.