Siglind Bruhn’s Personal Home Page











Some Thoughts Towards a Theory of Musical Ekphrasis

Siglind Bruhn 

1 Art about Art?


" y a de bizarre, et même d'inquiétant, dans le fait d'une inspiration de seconde main, cherchée dans les oeuvres d'autrui, et cherchée dans un art dont les buts et les moyens sont très différents de ceux qui charact‚risent l'art poétique. Est-ce vraiment légitime? Est-ce vraiment utile et fécond?" [Étienne Souriau, La poésie française et la peinture (London 1966), p. 6]

"...there is something odd, and even disturbing, in second-hand inspiration, sought in the works of someone else, and sought in an art form of which the aims and the means are very different from that which characterize poetry. Is this really legitimate? Is this truly useful and fruitful?"

There are various ways in which one art form can fruitfully relate to another. Coexistence is much more frequent--and apparently much less disturbing for an audience--than the declared attempt at a "transformation" or new representation in another sign system. Does this "second-hand inspiration," as Souriau called it, constitute a genuine creative act? To overemphasize what seems to be his question: is there a risk that the "representation of a representation" might suck the blood and life force from the first work, or to come out as a merely derivative, bloodless response? What do artists mean when they say that the new work can be cherished alone but fully understood and appreciated only in light of the earlier work on which it reflects?

When composing his piano cycle Gaspard de la nuit, Ravel not only chose for his three pieces the titles of three of Bertrand's poems, but actually reprinted each poem on the page facing the beginning of the musical piece that refers to it. While Ravel's music is no doubt beautiful and self-sufficient when appreciated without knowledge of the literary source (as is usually the case in today's concert practice), the listeners' insight into the depth of the musical message increases dramatically once the music is comprehended in light of the poem.

Let me briefly recall the central piece, Gibet. Bertrand, in asking us to witness the death of a hanged man, draws our attention to two facets of a transitional space. On the one hand, there is the very moment between life and death; the two framing verses clearly stake out this ground. The question that pervades all six stanzas of his poem asks after the origin and nature of a sound--a sound that, after having been suspected to come from the man himself or from the insects that surround his head, turns out to be the tolling of the death-knell. At the beginning, the lyrical "I" is wondering whether the sound may be the sigh of the hanged man; there may still be life. But the end speaks unequivocally of a carcass, a corpse. The entire poem can thus be read as an unfolding of that moment between almost-no-life and definite death. On the other hand, Bertrand elicits, in the four central stanzas, the interaction between the living and the not-quite-dead. Significantly, the creatures proposed as possible sources of the puzzling sound are not animals whom a man could look in the eye, but insects--representatives of transition. Cricket, fly, beetle, and spider all relate to the hanged man in ways that evolve from the innocuously disinterested to the downright morose.

Ravel captures many of the nuances expressed through Bertrand's poem in his piano piece. As in the poem, the tolling of the bell is the unifying feature. The tolling never pauses and never changes its pitch. Its rhythm, however, makes it clear that all is not in order here. Against this incessant sounding of the death-knell, Ravel proceeds to lay out his melodic material which, in four ever more emotionally loaded steps, moves further and further away from any meaningful relationship to the central scene and the dignity we expect in the context of a death-knell. In the image drawn by Bertrand, this musical development corresponds with the increasingly disrespectful way in which the creatures of transitional space relate to the hanged man. While Ravel's piano piece is undoubtedly beautiful when heard as absolute music without any connection to an extra-musical stimulus, the listener gains access to its full depth only when appreciating it as a transmedialization of Bertrand's poem.


 2 Musical Ekphrasis in Intermedial Space

As the brief description shows, Ravel's piano cycle on Bertrand's poems in no way constitutes vaguely impressionistic "program music." Instead, this is a case of a transformation of a message--in content and form, imagery and suggested symbolic signification--from one medium into another. For this phenomenon we seem to lack a specific term; I will make a case for calling it "the musical equivalent to ekphrasis." Not surprisingly given the lack of a distinctive term, no methodology seems to have been developed that would allow us to differentiate within what I will argue is a unified and highly sophisticated genre, or to define the genre within the larger fields in which it is situated.

These fields can be imagined as surrounding musical ekphrasis, linked to it at various points of interaction or by way of the questions asked in aesthetic theory about assumptions underlying all of them. (In the graphic overview I single out two of music's sister arts--painting and literature--to stand for what is of course a much richer texture of interactions, encompassing not only other forms of visual art but also dance and mime as well as many hybrid forms of artistic expression. By the same token, the aesthetic theories that question musical ekphrasis with regard to its concepts, touch on many more issues than the few that I have listed in the diagram.)


Among the possible pairings between two art forms that express themselves in different sign systems (verbal, pictorial, sonic, kinetic, etc.), the relationship between words and images is the one that is most widely explored. And in fact, the most securely established terminology is found in a field that has experienced a significant revival in recent years: ekphrasis or, more particularly, ekphrastic poetry: poems inspired by paintings or other works of visual art, including etchings and drawings, sculptures and architecture, photographs, films, etc. The field is amazingly broad and varied both historically and geographically. In his three-volume study Das Bildgedicht, the German scholar of ekphrasis, Gisbert Kranz, lists 5764 authors of ekphrastic poetry representing thirty-five languages and twenty-eight centuries (from Homer to our days)! Altogether fifty thousand poems on visual art are referenced in his 1500-page bibliography. James Heffernan in his seminal book of 1993, Museum of Words: The Poetics of Ekphrasis from Homer to Ashbery, defines ekphrasis as "the verbal representation of visual representation"; Claus Clüver, in his 1997 article "Ekphrasis Reconsidered: On Verbal Representations of Non-Verbal Texts," expands the definition fortuitously to "the verbal representation of a real or fictitious text composed in a non-verbal sign system."(FN1)

The musical equivalent of ekphrasis is a much more recent phenomenon. Moreover, the first examples of the budding new genre, written in the last years of the 19th century, were mostly not distinguished from the broader category of "program music."; Musical compositions with explicit reference--whether verbal in titles and accompanying notes or onomatopoeic--have existed for much of the history of Western music; yet, I claim, musical ekphrasis has not.


3 Musical Ekphrasis versus Program Music

This brings me to an important task in approaching the subject matter of this study, that of defining the criteria along which we can agree to distinguish between musical ekphrasis on the one hand and what is generally known as "program music" on the other. The two genres belong to the same general species: both denote purely instrumental music that has its raison d'être in a definite literary or pictorial scheme; both have variously been described as "illustrative" or "representative" music.While the term "program music" is considered by many to be simply the umbrella term for both kinds, I will argue that it is not only meaningful, but essential for a full understanding of music of the "ekphrastic" kind to attempt a distinction.

In literature, the equivalent is the distinction between ekphrasis proper and "word painting" or "Beschreibungsliteratur." One way of approaching the difference is to ask whose fictional reality is being represented. "Program music" narrates or paints, suggests or represents scenes or stories (and, by extension, events or characters) that may or may not exist out there but enter the music from the composer's mind. The range of application for the term "program music" is wide, spanning from the biographical (Strauss's Aus Italien) and the emotional expression associated with nature near or far (from Beethoven's "Erwachen heiterer Gefühle bei der Ankunft auf dem Lande" in the Pastoral Symphony to Holst's The Planets) through the depiction of an historical or literary character (Berlioz's King Lear, Liszt's Hamlet) all the way to a musical impression of a philosophically created "world" (Strauss's Also sprach Zarathustra).

The musical equivalent to ekphrasis, by contrast, narrates or paints a fictional reality created by an artist other than the composer of the music: a painter or a poet. Also, ekphrastic music usually relates not only to the content of the poetically or pictorially conveyed fictional reality, but also to the form and style of representation in which this content was cast in its primary medium.

The generous grouping and lack of distinction between program music and musical ekphrasis affected composers as well as listeners and scholars. Composers, particularly at the beginning of the 20th century when "program music" was gaining a bad reputation in comparison to "absolute" or "pure" music, often obfuscated their full intent in the hope to be taken seriously. Such concealment happened not only with programs of the more general kind (one is reminded of Mahler's withdrawing his poetic outlines for his symphonies), but also and particularly in the case of music based on extant works of art. Thus Schoenberg originally denied that his Pelleas und Melisande was more than only vaguely inspired by the topic of Maeterlinck's Symbolist drama, acknowledging only decades later how exact a "transformation"; he had actually tried to achieve here. The fact that listeners and scholars were discouraged from making a distinction between the two adjacent categories of music resulted in a considerable delay between the first occurrence of the phenomenon of musical ekphrasis and its proper recognition.

I am interested in finding an answer to the question what it may mean if composers claim to be inspired by a poem or painting, a drama or sculpture, and to have transformed the essence of that art work's features and message, including their personal reaction to it, into their own medium: the musical language. I expect to find as many responses to the challenge of interartistic transformation as there are works in the genre. Thus, while my investigations will be guided by the search for a methodological framework within which all such transpositions find their place, I admit that my fascination with the variety of approaches taken and solutions developed overrides my interest in the grid on which I may eventually lay them out.


 4 Attempt at an Analog Definition

When pursuing the above-mentioned question what exactly we mean when we talk about a transmedialization of a work of literature or art into music, I begin with the assumption that the creative process that applies in the step from a painting to its poetic rendering can usefully be compared to that which leads from a poem or painting to its rendering in music; in fact I maintain that they correspond to a degree that justifies adapting the terminology developed in the adjacent field.

In view of this wider application, I would thus like to offer a third definition of ekphrasis which further generalizes Claus Clüver's wording. Ekphrasis in this wider sense would then be defined as "a representation in one medium of a real or fictitious text composed in another medium." As I understand it, what must be present in every case of ekphrasis is a three-tiered structure of reality and its artistic transformation:



a scene or story--fictitious or real,


a representation of that scene or story in visual form (a painting or drawing, photograph, carving, or sculpture (or, for that matter, in film or dance;(FN2) in any mode that reaches us primarily through our visual perception), and


a rendering of that representation in poetic language.


The poetic rendering can and should do more than merely describe the visual image. Characteristically, it evokes interpretations or additional layers of meaning, changes the viewers' focus, or guides our eyes towards details and contexts we might otherwise overlook. Correspondingly, what must be present in every case of what I will refer to as "the musical equivalent to ekphrasis" is



a scene or story--fictitious or real,


its representation in a visual or a verbal text, and


a rendering of that representation in musical language.


5 Depiction and Reference

Expanding from here, I wish to argue that what and how music communicates about any extra-musical stimulus falls into two categories that can be seen as analogous with those pertinent in the context of painting and poetry, namely, depiction and reference. I will use depiction by musical means as encompassing not only instances of mimicry, but also, and more importantly, emotions and feelings. Correspondingly, reference by musical means, just like reference by verbal and pictorial means, will be understood as relying on cultural and historical conventions. In this context, Leonard Meyer speaks of connotations, which he defines as "those associations which are shared in common by a group of individuals within a culture." Thus, he continues, "[c]onnotations are the result of the associations made between some aspect of the musical organization and extramusical experience."(FN3)

In music, a representation of "the extant world" is not quite so obvious. Schopenhauer objects fervently to the notion of musical imitations of "phenomena of the world of perception,"(FN4) and Tovey concurs with so many musicologists of his time who maintain that programmatic elements in "serious" music are irrelevant to its value as music.(FN5) One believes that music has only very limited mimetic potential, while the other declares any musical representation as undesirable. This was not always the dominant view. Rousseau when writing his Dictionnaire de musique in the 18th century clearly did not think so. Under the heading "imitation" he included two entries, apparently conflating "mimesis" and "imitatio." The second entry deals with the expected, technical device of "the same aire, or one similar, in many parts," while the more prominent first entry explores the field of music imitating things extra-musical, clearly arguing that this art is no less capable of emulation than its sister arts.

Conventions established between the parties engaging in communication through representation need not, and in fact do not, end with verbal language. The musical language--our primary concern in this study--has developed a highly sophisticated catalogue of signifiers that are agreed, within our cultural tradition, to be understood as "pointing towards" non-musical objects. Among the most well-known are


the semantic interpretation of brief musical units its representation in a visual or a verbal text, and as "gestures" on the basis of their kinesthetic shape,(FN6)


the figures of musical rhetoric developed in the 15th and 16th centuries,


the retracing of a visual object (like the Cross) in the pitch outline, and


the letter-name representation of or allusion to persons--from Bach's famous pitch signature and those of Schumann, Shostakovich,(FN7) Schoenberg, Berg, Webern, etc. to the acrostic bows of reverence to a patron (Schumann's ABEGG) or a lover (Berg's HF).

These four basic categories actually constitute intrinsically different ways of music's "referring to" a non-musical object. Rhetorical figures, while modeled after (verbal) oratory, do not rely on a mediator to be understood by those familiar with them; they function almost like a semantic vocabulary. Gestures need Einfühlung on the part of the individual listener, who perceptively links a certain structure with a kinesthetic image to arrive at an affective connotation. Suggestive pitch contours are (usually clumsy) translations of visual silhouettes and represent an object only insofar as the listener attaches the (metaphoric) concepts of "high" and "low" to what is heard as faster or slower vibration;(FN8) and letter-name allusions rely on the prior translation of the musically received message into its notational equivalent and its basically arbitrary, though conventionally prescribed alphabetic signifiers in order to be decodable.

Yet even the latter two cases of mediated representation can turn into convention. The listeners' experience of a correlation between certain musical tropes and implied meanings develops from unexpected recognition--or the recognition of unexpectedness(FN9) --via repeated exposure to anticipation, thus establishing a set of conventions that may gradually come to bypass the original mediator, even develop into forms where the mediator is actually inaccessible. Similarly, the Germanic naming of pitches (with B and H as well as the suffix-inflected Fis for F# and Es for Eb) is self-evident neither for the Romance-language terms for pitches, which are based on do-re-mi and modified by idiosyncratic words for "sharp" and "flat," nor for the Anglo-Saxon scale lettered A-B-C-D-E-F-G. As a consequence, it is a matter of learned convention, and thus of the "joy of literacy," if lovers of Western music across language barriers recognize that A-Eb-C-B stands for Arnold SCHoenberg (on the basis of the Germanic spelling of the letters as A-S-C-H.)

Conversely, composers using musical tropes to represent non-musical objects and concepts employ a great variety of mimetic, descriptive, suggestive, allusive, and symbolic means. Single components (motifs or musical formulas) and their syntactic organization, vertical texture and horizontal structure, tonal organization and timbral coloring are entrusted with suggesting depiction. Quotations of pre-existing musical material may add allusive reference, and allow for modifications of context, medium, or tonal environment that successfully express defamiliarization or irony. Last but by no means least, countable units--from notes to beats, bars, or sections --invite play with numerical symbols both traditional and innovative. These latter cases move ever further into the realm described by the phrase "the joy of literacy": not only do such significations remain hidden to the uninitiated, we no longer expect them to be accessible even to insiders through the means of primary sensory perception, but only to skilled readers of the score.

Furthermore, music, as I hope to show convincingly, is capable of a kind of descriptive effect that Wendy Steiner, writing about the poetry of e.e.cummings and others, refers to as the "embodying of the still-movement paradox." Even more than language, music can do so without compromising its intrinsic logic. The reason for this greater flexibility is that music, while resembling verbal texts in that it develops in time, at the same time "paints." Like the media of visual art, it conveys to its audience the sensual experience of colors and textures, rather than referring to them as language does. Both its range of register and its compositional textures (polyphony above all) create a spatiality to which literary modes can only allude.


 6 Instrumental Music and Narrativity

But can non-texted music also narrate? In his work on Mahler, Anthony Newcomb (drawing on definitions by Paul Ricoeur) maintains that it can and does. Following a work of music entails, he believes, the same basic activity as following a story: the interpretation of a succession of events as a meaningful configuration. Carolyn Abbate urges us to differentiate the nineteenth-century claim that certain linear elements of music can be regarded in analogy to the events in a dramatic plot (music is perceived as generating expectations on the basis of culturally established paradigms; it moves through tension and release towards closure). She argues that music should be "seen not merely as "acting out" or "representing" events as if it were a sort of unscrolling and noisy tapestry that mimes actions not visually but sonically, but also as occasionally respeaking an object in a morally distancing act of narration."(FN10) However, she cautions, such "moments of diegesis" are far from normal or universal in non-texted instrumental music. Since Abbate is here referring to music that does not, by its title, claim to be a representation of an extra-musical reality, the allowance for "narrative acts of music" is extremely encouraging. Non-texted music may not be able to differentiate the details of a plotline because it cannot establish the non-musical specifications of the characters and props in the fictional world. But as the aesthetician Kendall Walton in his work on the representational qualities of music confirms, mere titles often suffice to provide this essential factual skeleton and make music patently representational--and even narrative.

Having thus argued that music, like art and literature, is capable of depicting and referring to things, including things in a world outside its own sonic realm, and that what is represented in a pictorial, literary, or musical medium may be image or story, design or narrative, I now turn to the more specific question how music may represent something that does not belong to the primary reality "in the world out there" or "in the soul in here," but has previously been represented in a work of visual art or literature.

This leads me to a question that, I hope, addresses the delineation and definition of what the musical equivalent to ekphrasis may be from a different perspective: the question whether what we find in different cases can be described as poems or paintings... and music? poems or paintings in music? or poems or paintings into music? In this brief exploration I wish to address the question how a poetic or pictorial source text relates to--and possibly makes its way into--a musical composition.

In order to develop pertinent categories that may help me, generally, to deal with the musical material in a systematic way and, specifically, to know what to exclude and why, I turn (as I do so often these days) to the already established methodologies in (literary) ekphrasis.  


7 Variations of Artistic Interaction

In this context, the thoughts of the Scandinavian interarts researcher Hans Lund, as far as we can glean them from the only work of his that has been translated into English, Text as Picture: Studies in the Literary Transformation of Pictures (Swedish 1982, English 1992) prove exceedingly helpful. In his chapter "The Picture in the Poem: A Theoretical Discussion," Lund offers a very useful scheme of defining what stance the author of the secondary representation (here, a poet; in our case, a composer) may be adopting towards the work of art (here, a painting; in our case, a painting/poem/drama) that constitutes the primary representation of the scene or story. Lund establishes three main categories for the relation of text to picture: combination, integration, and transformation. (In my discussion of the equivalents in music's relationship to the sister arts, I will further differentiate within two of them.) Here are Lund's definitions one by one, and my own adaptations for the field of music.


(a) Literature or Painting ... and Music

By combination I mean a coexistence, at best a cooperation between words and pictures. It is, then, a question of a bi-medial communication, where the media are intended to add to and comment on each other. The old emblematic writing belongs to this category. Here, too, are found certain works by authors traditionally called "Doppelbegabungen" by German critics, i.e. authors who combine and to a certain degree master the literary as well as the pictorial medium. Examples are William Blake, Dante Gabriel Rossetti and Günter Grass. Works which are the results of a creative cooperation between a writer and a pictorial artist [...] are also found here. Illustrations made afterwards to match literary texts are not primarily a concern for literary scholars but for art historians.

What Lund is sketching here amounts, it seems to me, to two somewhat different genres in the case of musical composition: setting and collaboration; both answer the questions put in the heading of this chapter with "Poems or paintings and music." Collaborations involving music as one of the key components include works like Parade (by Cocteau + Satie + Massine + Picasso) and L'histoire du soldat (Stravinsky + Ramuz), to name only two outstanding cases here. Collaborations with music as one component differ essentially from transformations of a painting or poem into music, whereby a structured entity with all its constituent parts and many layers of message is recreated on another plane. They are excluded from this study for two reasons. First, it is usually unclear which sign system, if indeed any of those involved, should be considered primary, and which constitutes the transposition. Second, one may assume that the myriad aspects of communication, which would otherwise be expressed within a single artistic text, are conceived as being shared among the collaborating arts here. We are, then, not dealing with the transformation of form and content from one artistic representation into another, but instead with a sort of "synthetic effect" whereby the various arts contribute the constituent parts of a single artistic gestalt and message. In such a joint venture, individual components complement one another but could often not stand on their own.

Music knows few cases that correspond directly to the phenomena of "emblematic writing"; or the dual art work of "Doppelbegabungen." A composer like Arnold Schoenberg, who was also a gifted artist, nevertheless did not, to my knowledge, create any work in which expressions of his dual talent combine in such a way as to engender a single overarching artistic message. The closest analog in recent music is probably Erik Satie. Many of his piano scores (see, e.g., Sports et Divertissements, published as facsimile) tread a fine line between musical score and artwork. The brief pieces are prefaced with drawings by Charles Martin and may have been intended, or so Satie scholars believe, to be looked at as much as performed. From the time when emblematic writing itself blossomed, one composition at least seems to function as a musical analog. In the early 17th century, Michael Maier created a work under the title Atalanta fugiens which consists of fifty musical settings in an imitative style accompanied by emblems and epigrams. (Also known as "Michael Maier's alchemical emblem book," the work is specifically intended to be appreciated "per oculis et intellectui").(FN11)

On the other hand, the field of music encompasses compositions that are manifestations of a combination of talents that is much rarer than the dual aptitude for poetry and painting, composition and painting, or music and poetry writing: synaesthesia. In correspondence with some painters who claim to be putting on canvas the hues communicated to them in musical sounds, composers endowed with this gift of seeing colors when hearing pitches or chords may purport to be creating a composition consisting of sound and color. In the case of composers who, like Olivier Messiaen, expected his audience to see with their inner eye the hues expressed in his chords, the visual component is, for most of us, beyond our perceptive abilities and thus beyond verification; these works thus do not literally involve two media. The composer's assertion refers to a very private reality which is not easily shared with an audience and the details of which have to be taken at face value. In compositions like Alexander Scriabin's Prometheus by contrast, notated for clavier à lumières in addition to the instruments of musical performance, the audience does enjoy a bi-medial performance. Moreover, analysis reveals that the correlations of sounds and colors are part of a complex system of spiritual symbolism.

Settings of one text in another medium, while often intriguing in themselves, also constitute a hybrid form in comparison to the phenomenon I am studying here. Whenever a poetic text is set as vocal music, or a dramatic text as opera (or, for that matter, a musical composition as ballet), the original medium is inflected rather than transformed. Granted, in vocal music, intonation--one of the many features of vocal language--is modified; secondary features dependent upon or related to intonation, like speech tempo, word spacing, etc., may be more or less effected, and structure may occasionally be expanded by repetitions. All other aspects of the original text, however--vocabulary and syntax, metaphors and allusions, the mode of expression and the objects spoken of--will characteristically remain completely untouched. The instrumental accompaniment may be anything from servant to partner (and, in recent times, even competitor) to the vocal part, but it is not typically entrusted with creating a self-contained musical transformation of as many aspects of the poetic model as possible. Rather, we often speak of it as "supporting" the vocal line or "painting a backdrop" for it. Such accompaniment acts as a musical illustration of and to the poetic text. The case is somewhat more complex when a choreographer chooses a piece of music to which to compose a ballet. One would want to distinguish in what cases the music is used primarily as an aesthetically satisfying vehicle for the choreography, and in what cases it actually inspires a conceptual interpretation.

Ideally, in order to make such a distinction with authority, one would need to create an artificial situation in which one could focus on choreographies in a silent performance--as, for instance, on video recordings without tone. The question would then be whether such a purely kinetic work could be intuited as a transformation of (essential aspects of) the musical compositions in any of the myriad ways in which ekphrastic poems--often read without the model being present--relate to the works of visual art to which they owe their being. This brings me back to Lund and his second definition.


 (b) Literature or Painting ... in Music

The second sector of my field of research I call integration. Here a pictorial element is a part of the visual shape of a literary work. Whereas pictorial elements in a combination have relatively independent functions, a pictorial element in an integration cannot be removed without destroying the verbal structure. Integration means that verbal and visual elements constitute an overall unity which is not reducible to the sum of the constituting elements. In this sector we find stanzas in the shape of a goblet or hour-glass and the like in the pattern poems of baroque poetry, as well as Apollinaire's Calligrammes and the concrete poetry of Modernism.

The integration of verbal and visual expressions into musical compositions includes many examples that need little reflection: neither verbal performance instructions nor the visual element of the musical notation itself would normally prompt us to think that we are dealing with a relationship between two art forms, although both instances meet the condition: both will not be encountered independently of the musical contents. Musical notation would not be in existence without the medium it aims to perpetuate, and compositions would not have survived --or at least not in a condition as close to their original design--without the help of some means of record-keeping. Similarly, performance indications detached from the music to be performed make no sense, while music conceived with expressive nuances that cannot be specified unequivocally outside the verbal medium loses a valuable dimension when deprived of these directions.

While these examples of integration hardly concern us in the context I have set out to examine here, there are several other cases that would require answering our initial question with "the visual or the verbal in music." Music knows the equivalent to "stanzas in the shape of a goblet."(FN12) Conversely, Kurt Schwitters's famous Ursonate and many works of Hugo Ball have shown us that "poems in the form of musical sound patterns" are equally possible. Then there are cases in which visual elements that originate outside music appear integrated into a piece of music. One example occurs in scores whose visual presentation follows shapes the outlines of which suggest depicted objects.(FN13) In other cases, a constituent part of the musical language is based on a linguistic component which would not necessarily appear independently in a poem or drama; themes shaped on the basis of letter-name allusions (B-A-C-H etc.) fall into this category. Finally, as if in combination of the implicit graphic aspect and the implicit letter names, a musical score may contain elements that are graphically both musical and verbal text. The most striking example that comes to my mind is the title page of a composition for male chorus written in the ghetto Terezín by one of its inmates, the composer Pavel Haas. Besides the title itself, Al Sifod, and the usual information regarding composer, poet--Jakov Simoni--and scoring, Haas decorates the title page with musical notes that, while they are carefully placed on their staves, are actually adapted to look like Hebrew letters. The power that be in the camp would hardly have recognized this, but the ones for whom the message was intended did: it reads "Kizkeret lejon hasana harison vemuacharon begalut Terezín"--In remembrance of the first and at the same time the last anniversary of the Terezín exile.)(FN14)

One step further, musical scores may be accompanied by verbal and visual texts in the form of epigrams and illustrations. Since epigrams are frequently quotations from extant literary works, they could, of course, stand alone and do not concern us here. Illustrations in musical manuscripts, however, form a category of their own. Before Satie's sketches in his own pieces at the beginning of our century, they were known primarily from manuscripts of late medieval and Renaissance music. An illustrative example is the famous Chansonnier Cordiforme, the "heart-shaped chansonnier." More fanciful than useful for music making, it is a kind of troubadour song written into a preciously illuminated heart (topped with four instead of two semicircles). Similarly, the visual, verbal, and musical components appear almost inseparably integrated, and the artistic ingeniously blended with the practical, in the manuscript pages of fifteenth-century canons.

In a four-part untexted canon by Bartolom‚ Ramos de Pareja (c1440-1491), the single staff containing the musical sequence is bent into a circular shape and set, in golden ink, against a background of deep sky blue. Wind spirits blowing from the four sides of the page into the notes indicate the entry of the four voices, while the calligraphy fitted into the circle betrays the composer as a music theorist, who informs singers about the modes they will detect in the four-part harmony resulting from the proper execution of this canon.

Opera as a genre typically relies on integrating a verbal text into the composition in such a way that both elements, lyrics and music, when represented separately, seem to be lacking an essential complement. However, the degree to which the component parts of opera--the libretto on the one hand and the "pure" music on the other&--are capable of also functioning independently is often greater than in the cases Lund mentions. While the hour-glass shape of a poem is really nothing but an empty line drawing (and usually a fuzzy one, for that matter) once the words are taken out, the same cannot be said for librettos. Many of them may be of the rather more unimaginative kind when taken as dramatic works; but, as testified in the now established term "Literaturoper," there are a number of literary works that originated as dramas and continue to stand as such, before and after they are used by a composer. And as, for instance, Hindemith's symphonies Mathis der Maler and Die Harmonie der Welt prove, even the music can sometimes function as a fully valid artistic testimony when taken on its own. Yet these cases are exceptions rather than the rule, and the "music alone" or "drama alone" typically differs from the corresponding component that forms a constituent part of the opera.

This brings me to Lund's third definition.


(c) Literature or Painting ... into Music

In the third category--which I call transformation--no pictorial element is combined with or integrated into the verbal text. The text refers to an element or a combination of elements in pictures not present before the reader's eyes. The information to the reader about the picture is given exclusively by the verbal language.

 This, then, is the case of a poem or painting being transformed into music--the focus of this study. Where transformations appear in poetry or prose on painting, they are referred to as ekphrasis. In music, such ekphrasis can take as its object a work of literature (as is the case in Ravel's piano piece Gibet, briefly described above) or a work of visual art. Compositions that deal respectively with these two sides of musical ekphrasis include, on the one hand, symphonic compositions on Symbolist drama (Charles Martin Loeffler's and Bohuslav Martinu's compositions on Maurice Maeterlinck's marionette play La mort de Tintagiles, and Schoenberg's work on Maeterlinck's Pelleas und Melisande) as well as musical works about poems short and long (Schoenberg's sextet Verklärte Nacht and Elliott Carter's transmedialization, in Concerto for Orchestra, of Saint-John-Perse's epic poem Vents, and to a lesser degree, in A Symphony for Three Orchestras, on Hart Crane's epos The Bridge.(FN15) On the other hand, I explore music on paintings: the musical "triptychs" twentieth-century composers form from works of quattrocento artists (Ottorino Respighi's Trittico botticelliano and Bohuslav Martinu's Les Fresques de Piero della Francesca), two transmedializations of a Romantic painting (Serge Rachmaninov and Max Reger on Böcklin's Isle of the Dead), three musical ekphrases of the same early modern work by Paul Klee (the English composer Peter Maxwell Davies's, the American Gunther Schuller's, and the German Giselher Klebe's compositions under the title The Twittering Machine), and two contemporary Danish composers' reactions to drawings by M.C. Escher (Per Norgard's Ant Fugue from Prelude and Ant Fugue [with Crab Canon]: Hommage a M.C. Escher and Hans Abraham's Three Worlds).

When transformation of a work of visual art is brought onto the theatrical stage and wedded with the miming aspect of that genre, we speak of enactments. Here I know of at least three compositions based on serial paintings that can be shown to contain distinct elements of enactment. This is particularly intriguing given the fact that neither composition is strictly theatrical in its focus. Music knows a few cases where something corresponding to the typical dual transformation--from the visual to the verbal to the mimed--occurs outside the opera. However, since the operatic environment is the characteristic one for this sub-genre, I would like to introduce musical enactment using as an example the three scenes from act VI ("Sechstes Bild") of Hindemith's opera Mathis der Maler.

In the first of these scenes, Hindemith's painter Mathis attempts to soothe the distraught young girl Regina with a narration of what he claims to see in a picture portraying three angels. His verbal depiction leads us to one of the second-tier panels of the Isenheim Altarpiece, the masterpiece of the operatic protagonist's historical model, Grünewald. At this juncture, Hindemith the librettist puts into the mouth of his character Mathis a most intriguing tripartite description and interpretation of the panel that the historical "Master Mathis" painted ten or more years prior to the year into which this fictional conversation is placed. In the way in which Mathis tells Regina about the "pious pictures," no mention is made of who created them; the narration appears guided by the idea and intention of what is portrayed rather than by an attempt to describe the visual composition in all its details. Mathis focuses on the spiritual aspect of this concert--and so does the music in which Hindemith sets this scene.

This "narrated portrayal" of the "Angelic Concert" is complemented in the scene that follows by an enactment combined with a narration of one of the two rear panels, "The Temptation of Saint Antony." The events presented on stage function on three levels. First, in the larger context of the operatic plot, Mathis's encounter with human tempters and monstrous tormentors appears like a bad dream--or a vision, given that he perceives himself as the Egyptian hermit Antony. Second, the verbal onslaught by the seven human tempters functions as a multi-layered interpretative embodiment of what is, beyond the reference to the pictorial representation in the altar panel, both the inner story of the temptations of Saint Antony and a dramatic portrayal of the plight in which Mathis is caught. Third, the physical attack by the monsters is at the same time a tableau vivant of Grünewald's depiction and its musical transmedialization: the choir does not only accompany with insults and spiteful interpretations the assault by hellish monsters to which Mathis/Antony is subjected in the center of the stage, but simultaneously narrates the scene as painted by Grünewald. All images evoked in this scene also reflect a deeper spiritual meaning since they can be understood as provocations, as torments that emerge from the victim's own doubting mind. They represent his spiritual nightmares and the internal enemies that haunt his soul.

The third scene in this sequence, entitled "The Visit of Saint Antony in the Hermitage of Saint Paul" after the Grünewald panel to which it relates, limits the enactment to the visual recreation: stage design, costumes, posture and position of the two actors. No narrative relates what we see and hear to the painting, thus allowing us to focus all the more on the symbolic significance of the scene.

The older hermit Paul ("embodied" by the operatic character, Cardinal Albrecht) acts as a spiritual adviser to Antony (= Mathis). While his verbal admonitions deal unequivocally with the reality of the artist in the time of the Lutheran Uprising and the Peasants' War, the scenic setting binds the conversation into the larger conflict of conscience that is, both literally and figuratively, through the ekphrasis of the altar panels, the subject matter of the opera. Hindemith's music adds a wealth of nuances that corroborate and enhance this interpretive layering.

Another fascinating case of such mediated enactment of a pictorial narrative exists in Stravinsky's opera The Rake's Progress. The stimulus here is a series of eight engravings etched in 1735 by William Hogarth, after an equal number of paintings he had completed a year earlier. Stravinsky, inspired by these prints, followed Aldous Huxley's recommendation to ask Auden for a libretto dramatizing the story told in the etchings, on which he could then base his opera. Auden accepted and, with the help of Chester Kallman, told the story of the rake in a style that aimed at once (I quote Williard Spiegelman) "to recapture the myths and language of an earlier, more optimistic world, and to examine that world from the perspective of our own.... the libretto is Auden's attempt to adapt certain poetic styles to the conditions of twentieth-century literary life, to imitate or parody older models in much the same way that Stravinsky's music casts new light on earlier operatic techniques."

However, what sets this case of musical enactment of a pictorial source apart from the two texted examples I will be examining as part of my study is the fact that the Auden/Kallman libretto is a literary ekphrasis in itself, of which Stravinsky's music then gives a setting. By contrast, the texts Honegger uses for his oratorio La danse des mort on Holbeinis Totentanz, while "authored" by Paul Claudel, are actually compiled from the Bible. As such they constitute something akin to a verbal embodiment of the common source that inspired the artist and the composer, rather than Claudel's ekphrastic reaction to Holbein's artistic rendering. The situation is very similar in the case of Janacek's composition on Josef Kresz-Mecina's panels The Lord's Prayer which the composer based on five tableaux vivants he had devised himself. Here, too, the text predates both the art and its musical transmedialization.

Compositions based on ekphrastic poems--poems that are themselves transformations of pictorial texts--while often charming, are least pertinent to the aims of my study. They are usually little more than mere settings. Francis Poulenc's songs on Guillaume Apollinaire's Le bestiaire, whose poems are in turn based on woodcuts by Raoul Dufy, fall into this category. So do Poulenc's settings of Paul Éluard's poems Travail du peintre, which verbally represent the style and characteristics of various contemporary painters, and Reynaldo Hahn's similarly motivated Portraits de peintres after poems by Marcel Proust. Poulenc also set Apollinaire's Calligrammes, which are not "poems on pictures" but rather "poems in the form of pictures"--a feature that is by necessity lost once the text, now used as "lyrics" for songs, is fitted between the staves of musical notation. At the other end of the spectrum, verbal ekphrasis may indeed stimulate further musical ekphrasis in an independent musical work. Thus Debussy's piano piece Clair de lune in Suite Bergamasque was apparently inspired by Paul Verlaine's ekphrastic poem (by the same title) after Antoine Watteau's painting, Fêtes galantes.

Finally, the composer's interest in a musical transformation of a given work of verbal or pictorial art may inspire a creative artist working in yet another medium to expand the ekphrastic process even further, adding yet another transformation. The cases of dual transmedialization I am most eager to examine are those involving three different media each: from the pictorial to the verbal and on to the musical, from the pictorial to the kinetic on to the musical, or from the poetic model to the musical transformation to the visual or kinetic interpretation. Detailed analyses exploring two examples for this kind of multimedial "fate" in interarts space of an original work will conclude my case studies.


 8 Central Questions in Research on Musical Ekphrasis

The central questions I will be asking regard the scope and nature of this interartistic, intersemiotic transmedialization and can be summed up as follows:



What choices do individual composers make in their quest musically to transmedialize a pictorial or literary representation?


Do the choices made by the composers of a certain historical and cultural context allow to distinguish and describe a newly emerging "convention" of intersemiotic transformation?


Does the range of stances adopted by composers to works of literature or visual art parallel those observed in ekphrastic poets?