Siglind Bruhn’s Personal Home Page











A Concert of Paintings:
“Musical Ekphrasis” in the Twentieth Century

Siglind Bruhn 

Abstract: Not only poets may respond to a work of visual art with a creative act in their own medium, transposing the style and structure, the message and metaphors from the visual to the verbal. Composers, more and more frequently in our century, are also exploring this interartistic mode of transfer. Although the musical medium is reputedly abstract, composers, just like poets, can respond in many different ways to a visual representation. They may transpose aspects of both structure and content; they may supplement, interpret, respond with associations, problematize, or play with some of the suggestive elements of the original image. This article begins with some methodological considerations regarding the musical equivalent of what literary scholars know as ekphrasis (see Spitzer 1955, Hagstrum 1958, Krieger 1967, 1992, Lund 1992 [1982], Clüver 1989, 1997, Scott 1991, 1995, Mitchell 1992, 1994, Heffernan 1993, Yacobi 1995, 1998). Central questions concern the definition of musical ekphrasis in relation to "program music," and music's ability to narrate or portray extra-musical realities, i.e., to relate to them by way of mimesis or reference. In a second section, I attempt to position musical ekphrasis within the grid of interartistic interactions laid out by Hans Lund, and address some central issues of terminology in research on musical ekphrasis. Next I draw on three groups of symphonic compositions and two hybrid works (one pairing music with dance, the other with a biblical text), all composed in response to works of visual art, in order to attempt an assessment of the possible scope (and limitations) of the undertaking to reflect images in tones. To concretize the idea of correspondence between pictorial and musical configurations, I conclude with a comparative case study. My overall aim in this essay is thus threefold: to survey the range of artistic expression that can be made available for such transmedializations, to examine the degree to which the composers' creative responses draw on a body of shared cultural conventions, and to develop some first steps towards a methodology of "musical ekphrasis."


1 Music about Works of Art and Literature?

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. The literary topos through which a poem (or any other verbal text) addresses itself to the visual arts has received much attention in recent years and been subjected to intense scrutiny. As Leo Spitzer reminds us in his beautiful study of Keats's Ode to a Grecian Urn, ekphrasis has been "known to Occidental literature from Homer to Theocritus to the Parnassians and Rilke, [as] the poetic description of a pictorial or cultural work of art, which description implies, in the words of Théophile Gautier, 'une transposition d'art,' the reproduction, through the medium of words, of sensuously perceptible objets d'art" (Spitzer 1962 [1955]: 72). Jean Hagstrum, in a seminal book that may well have triggered the enthusiastic revival and the cluster of treatments of the topic in the literary debate of the past two decades, delineates the term ekphrasis in the context of what he refers to as "iconic poetry," as relating "to that special quality of giving voice and language to the otherwise mute art object" (Hagstrum 1958: 17-29 [18]). Hagstrum builds not only on Spitzer but also on Julius Schlosser-Magnino (1935: 12) and George Saintsbury's definition of ekphrasis as "a set description intended to bring person, place, picture, &c., vividly before the mind's eye" (Saintsbury 1902, I: 491). From the latter, Hagstrum takes his crucial emphasis on the concept of enargeia, "pictorial vividness," which Clüver has since taken up (Clüver 1998). Another angle was emphasized by Murray Krieger, who speaks of "the imitation in literature of a work of plastic art" and of poems that "in imitating a plastic object in language and time, make that object in its spatial simultaneity a true emblem of itself" (Krieger 1992 [1967]: 265, 267). Following this direction of thought, Peter Wagner points out that ekphrasis, both as a poetical and rhetorical device and as a literary genre, is Janus-faced; "as a form of mimesis, it stages a paradoxical performance, promising to give voice to the allegedly silent image even while attempting to overcome the power of the image by transforming and inscribing it" (Wagner 1996: 11, 13). For his discussion of mimesis in art, particularly concerning the relation between representation, illusion, and fiction, he draws on Kendall Walton (1990; Walton's work, while not explicitly dealing with the notion of ekphrasis, is germane to my concern here, not least because he has since turned to explorations of the representational nature of music: see Walton 1988, 1994). Finally regarding ekphrasis as a literary genre, Grant Scott (1991), Tom Mitchell (1992), and James Heffernan (1993) define ekphrasis as "the verbal representation of visual representation," a wording that Claus Clüver later expands to "the verbal representation of a real or fictitious text composed in a non-verbal sign system" (Clüver 1997: 26).

The phenomenon of ekphrasis is typically traced back to Homer's description of the shield of Achilles in the Iliad. As a rhetorical device, Peter Wagner (1996: 12) traces the term to writings attributed to Dionysius of Halicarnassus, in whose wake it was established as a school exercise in rhetoric. In the sense of the interart genre that interests us here, Murray Krieger dates the birth of a reflection on ekphrasis to the third century. As he reminds us (Krieger 1992: 7-8), Imagines of Philostratus the Elder was a series of descriptions of pictorial works of art which Philostratus the Younger, in his own derivative series of similar descriptions, called "ekphrasis."

In contrast to ekphrasis proper, with its history of more than two thousand years, the musical equivalent of ekphrasis is a very recent phenomenon. Moreover, the first examples of the budding new genre, written in the last years of the nineteenth century, were mostly not distinguished from the broader category of "program music." (See Schoenberg's Pelleas und Melisande on Maurice Maeterlinck's Symbolist puppet play, to my mind the first true example of a musical mimesis of drama, and the comments by Susanne Langer 1957: 225-226, referred to below.) Musical compositions with explicit reference- whether verbal, in titles and accompanying notes, or onomatopoeic-have existed for much of the history of Western music (from the many keyboard works about balmy or inclement weather or battle scenes, collected in Victorian drawing-room anthologies, through Beethoven's Pastoral Symphony to Richard Strauss's Symphonia domestica1) ; yet, I claim, musical ekphrasis is a distinct and as such much more recent phenomenon. An important task in approaching the subject matter of this study, then, is that of formulating the criteria by which we can differentiate between musical ekphrasis and what is generally known as "program music." The two genres belong to the same species: both involve purely instrumental music that has its raison d'être in its response to a definite literary or pictorial source; both have variously been described as "illustrative" or "representative" music (Newman 1905: particularly 103-188; 125-126). While it is possible to employ "program music" as an 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 distinguish it from the more generic expression.

In literature, the equivalent distinction is that between, on the one hand, ekphrasis as defined by Spitzer 1955, Hagstrum 1958, Krieger 1967, 1992, Lund 1992 [1982], Clüver 1989, 1997, Scott 1991, 1995, Mitchell 1992, 1994, Heffernan 1993, Yacobi 1995, 1998, and, on the other hand, "Beschreibungsliteratur"---a descriptive epic in which characters are reduced to objects (Buch 1972, also discussed in Lund 1992 [1982]: 13-15)2---or "iconic projection" (Lund 1992 [1982]: 63-201). One way of locating the difference is to ask whose idea of reality is being represented. "Program music" narrates or paints, suggests or represents scenes or stories (and, by extension, events or characters) that enter the music from the composer's mind. Musical ekphrasis, by contrast, narrates or paints stories or scenes created by an artist other than the composer of the music, and in another artistic medium. Furthermore, musical ekphrasis typically relates not only to the content of the poetically or pictorially conveyed source text, but usually also to one of the aspects distinguishing the mode of primary representation-its style, its form, its mood, a conspicuous arrangement of details, etc. The distinction may emerge fully when it is rephrased in analogy to one made in literary theory: that between Tamar Yacobi's modes of "re-presentation" and "representation." Yacobi (1995: 611-612) cites two examples from Tolstoy's Anna Karenina, where Anna's portrait functions as a (narrative) ekphrasis whereas the description of Kitty, "travelling to her estate, framed by the window of her carriage," is merely "picturelike." Program music represents, while musical ekphrasis re-presents.

For the purpose of this essay, the primary medium transmedialized by a composer will always be that of the visual arts. Yet as I have demonstrated in detail elsewhere (Bruhn 2000a), music can also create sonic representations of verbal representations. I recognize this variant of ekphrasis particularly in compositions of non-vocal music that recreate a literary text-as opposed to setting it (as in art song or Literaturoper) and thus including the primary medium in the re-presentation.

Yet even among compositions whose titles seem to point to a prior representation in the visual or verbal medium, there are a number that dwell on the borderline between the general category of program music and that of musical ekphrasis. Three well-known examples may bring out this hybrid state: Liszt's Hamlet (1858), Mussorgsky's Pictures at an Exhibition (1874), and Liszt's Hunnenschlacht (1857).

At first inspection, Liszt's Hamlet appears to be a musical impression of the character that was verbally depicted by Shakespeare. However, Hamlet and the legend around him have existed prior to the poet who made him famous, and Liszt's music, which does not attempt to relate to the play's plot or structure, linguistic or stylistic features, speech patterns or philosophical content, cannot, in my opinion, be said to function as a transformation of the play. Mussorgsky's Pictures at an Exhibition presents a slightly more complex case. Originally a piano suite that became known as a piece in the repertoire of symphonic poems thanks to Ravel's ingenious orchestration, it was allegedly created in response to a memorial exhibition of architectural drawings, stage designs, and watercolors in honor of the composer's recently deceased friend, Victor Hartman. In the absence of full details about the individual exhibits- "only six of the designs and illustrations that directly relate to Pictures may be identified with certainty" (Russ 1992: 16)-I find it difficult to appraise to what extent the composition constitutes a series of transmedializations into music of Hartmann's pictorially created world. It is far more likely that they represent general impressions, distilled from the composer's experiences of the exhibition and the fond memories of his friend's character. Still, even assuming that some of the individual "pictures" re-presented in music have indeed not had pictorial counterparts in the Hartman exhibition and therefore do no satisfy an understanding of ekphrasis that privileges one-to-one equivalences, Mussorgsky's composite work, with its series of (musical) tableaux linked by the "promenade" of the narrator within the musical fiction, may be read as an example of what Tamar Yacobi defines as an "ekphrastic model [...] as distinct from a unique art-work" (1998: 23), a "pictorial model, a common denominator, a generalized visual image" (1995: 601). Mussorgsky's composition may thus re-present not so much items from the historical and unique Hartman exhibition that served as a stimulus, but the art theme of the "picture gallery," complete with viewer(s).

The third borderline case I wish to address is Liszt's Hunnenschlacht, composed in response to Kaulbach's fresco by the same title. Judged by the fairly narrow definition of ekphrasis by Kibédi Varga (1989: 44), who requires "an exact description meant, to a certain degree, to evoke and substitute for the painting itself," Liszt's work would fall outside the genre. The musical rendering transcends the historic specificity as well as the particular pictorial source, painting a musical picture of a battle, with no equivalence to structural or stylistic features in Kaulbach's fresco. On the composer's account (quoted in Moore 1966: 142-143), even more influential than the mural itself were the thoughts his friend Kaulbach shared with him while conceiving the pictorial representation. It is as if the visual artist chose to depict one half of these thoughts, while the composer portrayed the other. The painter ultimately concentrated on the continuation of the combat among the souls of the slain seen in the mist that floated upon the surface of the lake during sunset. By contrast, the composer, according to his own testimony, chose to focus on the battle between a ferocious barbarian and "the personification of Celestial succor." Where the painter, fascinated by the mingling of bodies, portrayed violent rage without end, the pious composer depicted the ultimate victory of "divine truth, universal charity, the progress of humanity, and the hope of the world," which "sheds over all things a radiant, transfiguring, and eternal light." Yet, as Tamar Yacobi (1995: 604-605) has shown in discussing Lessing's injunctions regarding the statue of Laocoön and Virgil's poetic mimesis, a "medium-sensitive" attitude "sets free the re-presentational device [...]: ekphrasis may develop (rather than at best parallel, at worst attenuate) the original image." Within such a broader definition of ekphrasis, Liszt's Hunnenschlacht may constitute an example of ekphrasis that is intriguingly different from the majority of those I have thus far explored; it would also push the beginning of the genre back by about forty years, from around 1895 to 1857.

These borderline cases notwithstanding, the general distinction between musical ekphrasis and program music seems nonetheless crucial. A conflation of the two is detrimental to composers as well as listeners and scholars. Composers, particularly at the beginning of the twentieth century, when program music was gaining a bad reputation, often concealed their full intent in the hope of being taken seriously, having their compositions featured in well-respected concert venues and favorably reviewed by the influential art critics of the day-a privilege largely reserved to authors of "absolute" music. Such concealment occurred not only with regard to programs of the more general kind-one is reminded of Mahler's withdrawing his poetic outlines for his symphonies (Müller 1988: 216)-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 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 (Schoenberg 1950, Bailey 1984). 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 musical ekphrasis and its proper recognition.

The backdrop for this uneasiness with the concept of musical representation (and re-presentation) is to be found in the debate about "absolute" music made notorious by Eduard Hanslick. Drawing on a term that Richard Wagner first used with a pejorative slant in 1846 (Wagner 1871-1883, ix: 123) and on an assessment of "autonomous music" already formulated half a century earlier by Adam Smith (1980 [1795]), Hanslick propagates the idea that only music that is entirely independent of extra-musical influences, that, as Smith (205) had phrased it, "may be said to be complete in itself, and to require no interpreters to explain it" is "pure" music: that the value of music lies in its formal relations and not in its expressiveness. In opposition to a musical aesthetics of affect, imitation, and emotion (Wirkungs-, Nachahmungs-, Empfindungsästhetik), Hanslick declared the classical work of instrumental music, particularly that of Mozart, to be the truest realization of the potential of music since it pretends to be nothing but itself (Hanslick 1991 [1854]). With his famous definition that the theme of a musical composition is its proper content (95-104) and that the best of music constitutes, in its essence, "tönend bewegte Formen" (32), he took a stand against both the most successful portion of contemporary composition, particularly works of Berlioz, Wagner, Liszt, and Verdi, and against the very influential poetic descriptions and "readings" of instrumental music by such literary interpreters of music as E.T.A. Hoffmann, Wackenroder, Tieck, and Jean Paul. According to his notion of "absolute" music, Hanslick called all music that referred to extra-musical programs "relative," derivative, and hence of lesser artistic value.

Almost a century after Hanslick, the influential American philosopher, Susanne Langer, developed her theory of musical meaning by extension of and in contradistinction to Hanslick (Langer 1957 [1942]). While conceding that all modes of human understanding are forms of "symbolic transformation" (xiv, 26-52), and that music serves as a paradigm of a symbolic system, Langer confirmed Hanslick's position in defining such symbols as presentational rather than discursive (79-102), as having meaning but not asserting anything in particular. On her account, works of music can be "credited [...] with significance, although (by reason of the moral censorship which distorts the appearance of basic desires) we can never say what they signify" (207; emphasis in the original). Although Langer seems to endorse Hanslick and his school when she asserts that music "is preëminently non-representative" (209), she grants that insofar as music has "significance" (by which she seems to mean what contemporary scholars term "signification"), it is semantic (218), and that its aim is "not self-expression but formulation and representation of emotions, moods, mental tensions and resolutions-a 'logical picture' of sentient, responsive life, a source of insight, not a plea for sympathy"(222). While she has only scorn for program music, described as "the recognition of natural sounds in musical effects," a music that "deliberately imitates the clatter and cries of the market place, hoofbeats, clanging hammers, running brooks, nightingales and bells and the inevitable cuckoo" (220), she does allow for content-expressing music provided the correct "psychical distance" is maintained. Her defense of music's exhibiting formal characteristics that are analogous to whatever it purports to symbolize-"if it represented anything, e.g., an event, a passion, a dramatic action, it would have to exhibit a logical form which that object could also take [...] the musical figure which we recognize as such must be a figuration under which we could apprehend the thing referred to" (225-226)-could even be taken to endorse the complex responses of ekphrastic music to its stimuli.

I will have to leave for a separate study the question why musical ekphrasis seems to have found its form only in the late nineteenth century. It may have emerged in reaction against Hanslick's condemnation of all music that is not "absolute" as subjective and emotional, hence of lesser value; a detailed, serious, and imaginative musical interpretation or mimesis of a work of plastic art could hope to hold its ground against this judgment. Another conceivable reason may be its affinity to contemporary developments of Modernism, as Mack Smith's study of ekphrasis in literary prose suggests in effect: "The Symbolists [...], in eschewing realist externality, strove to portray internal states that concrete detail and referential language are ineffective in describing. The language of symbolism is evocatively ambiguous so as to suggest almost ineffable internal states. Music is the closest parallel to the symbolists' aims, so music became the emblem of their aspirations" (Smith 1995: 246-247).

More recently, the idea of instrumental music exhibiting a semantic content has been taken up by scholars of musical semiotics, musical hermeneutics, and musical narratology. Yet while in the literary discussion, Claus Clüver had early on developed a theoretical approach to ekphrasis which draws on Roman Jacobson's semiotic theories, exploring ekphrasis as a mode of "intersemiotic transposition" or even "intersemiotic translation" (Clüver 1989: 58-62), semioticians of music like Vladimir Karbusicky (1986, 1987, 1990), Kofi Agawu (1991), Raymond Monelle (1992, 1995), and Robert Hatten (1994, 1995) have resisted the advance from a recognition of music's semantic potential to a tracing of actual intersemiotic correspondences. This remains true even when a brilliantly perceptive philosopher of music like Naomi Cumming, entrenched in semiotic thinking, touches on the known relation between Peter Sculthorpe's orchestral composition Mangrove and Ian Fairweather's painting by the same title: while the fact of a visual stimulus is observed and beautifully documented by her, no attempt is made to develop thoughts that might lead to a theory of correspondence, mimesis, or transformation (Cumming 1998). My own reading of musical compositions in the light of the extra-musical stimulus that brought them into being has benefitted specifically from Carl Dahlhaus's reflections on musical hermeneutics (1978) as well as from the music-hermeneutic works of Constantin Floros (1977-1985, 1980, 1981, 1989), Hermann Danuser (1975), and Anthony Newcomb (1984, 1998). None of these investigations, however, touches on music's ability to re-present subject matter that pre-exists as a representation in another artistic medium.

The present study aims at a first answer to the question what it may mean when composers claim to respond to a poem or painting, a drama or sculpture, by transforming that artwork's features and message into their own medium: musical language. In pursuing this question, I start from 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 painting to its musical rendering; in fact I maintain that they correspond to a degree that justifies adapting the terminology of ekphrasis developed in the adjacent literary field. In view of this wider application, I propose to extend Claus Clüver's (1997: 26) definition of ekphrasis so that it might extend to nonverbal modes of re-presentation.. 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 (literary) ekphrasis is a three-tiered structure of reality and its artistic transformation:


a real or fictitious "text" functioning as a source for artistic representation;


a primary representation of that "text" in visual form (as a painting, drawing, photograph, carving, sculpture, etc.) or, for that matter, in film or dance (see Raftis 1989), i.e., in any mode that reaches us primarily through our visual perception; and


a re-presentation in poetic language of that first representation.

The poetic rendering can do, and usually does, more than merely enumerate the details of the visual image and their spatial position within the work of art. 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 "musical ekphrasis" is


a real or fictitious "text" functioning as a source for artistic representation;


a primary representation of that "text" in visual or verbal form; and


a re-presentation in musical language of that first (visual or verbal) representation.

That music represents reality does not go without saying. In recent years, the application to music of the term "representation" has become more accepted and consierable extended in its range to music that would previously have been classified as "absolute"; see particularly Walton 1988, 1994, and Scruton 1980, 1997: 118-139. Earlier in the twentieth century, however, philosophers, music theorists, and general art critics were divided on this issue. Some went back as far as Schopenhauer, who had firmly objected to the notion of musical imitations of "phenomena of the world of perception" (Schopenhauer 1969, I: 264). Tovey most prominently, but with him many musicologists of the time, maintains that programmatic elements in "serious" music are irrelevant to its value as music. Arguing that musical programs are incidentals that the listener can safely ignore while concentrating on the "musical" significance of the sounds, he claims specifically that "not a bar of the 'Pastoral' Symphony would be otherwise if its 'program' had never been thought of" (Tovey 1956: 168). Schopenhauer believes that music has a very limited mimetic potential, while Tovey declares any musical representation undesirable. This was not always the dominant view. Under the heading "imitation" in his Dictionnaire de musique (1975 [1768]: 198-199), Jean-Jacques Rousseau 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 mimesis than its sister arts. Rousseau regarded music, and within it specifically melody, as the last vestige of a form of utterance suffused with emotional warmth, a warmth that verbal language once also knew but has since lost and that, he claimed (1760: xiv), corresponds to a natural expression of the emotional to such a degree that one wants a special dictionary to understand it.

Conventions established between the parties engaging in communication through representation need not, and in fact do not, end with verbal language. While Susanne Langer (1967 [1953]: 31) is, of course, correct when she states that "the elements of music are not words," I disagree with her further conclusion that words alone function as "independent associative symbols with a reference fixed by convention." Musical language has developed a highly sophisticated catalogue of signifiers that are understood, within the conventions of our cultural tradition, as referring to non-musical objects. Among the best known are


the figures of musical rhetoric developed in the fifteenth and sixteenth (Tinctoris 1475, Luther 1538, Figulus 1575, etc.) and encoded in the seventeenth and eighteenth centuries (Burmeister 1601, Muffat 1698, Mattheson 1739, etc.); the illustrative devices developed in the madrigalism of the late sixteenth and early seventeenth centuries; the affective connotations linked with keys and tonalities; the "affective types," developed by the eighteenth-century music theorist Johannes Kirnberger (1773) as an extension of the rhetoric-of-music tradition; and the influential system of categorizing the connotations of intervals; 3


the retracing of a visual object (like the Cross) in the pitch outline; thus the French composer Olivier Messiaen, in a piano piece about the Cross looking upon the infant Jesus (in Vingt regards sur l'Enfant-Jésus), invents a theme whose initial notes, A-Ab-Bb-G, trace the characteristic shape of a cross with an extended vertical beam, tilted at roughly the angle at which one imagines Jesus carrying it during his walk to Calvary (see Bruhn 1997b: 47-50);


the acrostic (letter-name) representation of, or allusion to, persons---from Bach's famous pitch signature ("BACH," which sounds B-flat/A/C/B-natural after the German fashion of naming notes), the related one of Schoenberg (who, with the four letters A/E-flat/C/B-natural for A[rnold] SCH[oenberg], clearly attempted to establish a proximity to Bach), and those of Shostakovich, Schumann, Berg, Webern, etc., to the acrostic bows of reverence to a patron (Schumann's A/B-flat/E/G/G, used as a theme in a famous set of piano variations to honor one Mr. Abegg) or to a lover (Berg's HF, used throughout his Lyrical Suite to wink behind his wife's back to Hanna Fuchs), and other cryptographic messages, Does the range of stances adopted by composers to works of literature or visual art parallel those observed in ekphrastic poets?


the suggestive power of rhythmic and metric signifiers (lilting triple time for lullabies and romance songs, double-dotted notes to suggest military prowess, etc.), and


the semantic interpretation of brief musical units or timbrally distinct utterances as "gestures" on the basis of their kinesthetic shape (Lidov 1987). 4

These basic signifiers constitute intrinsically different ways of music's "referring to" (as opposed to "depicting") non-musical objects; correspondingly different is the degree to which each presupposes specialized knowledge. Rhetorical figures, modeled after (verbal) oratory, are readily understood by those familiar with them since they function almost like a linguistic 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)5 concepts of "high" and "low" to what is heard as faster or slower vibration; and letter-name allusions are decodable only after translation of the musically received message into its notational equivalent and its basically arbitrary, though conventionally prescribed alphabetic signifiers.

Yet even the latter two cases, which initially communicate only through a mental concept arbitrarily connected with them in our culture, will turn into convention by force of repeated association. The listeners' experience of a correlation between certain musical figures and implied meanings develops from unexpected recognition---or the recognition of unexpectedness---via repeated exposure to anticipation, thus establishing a set of conventions that may gradually come to bypass the original mental link, even develop into forms where the link is actually inaccessible. Similarly, the Germanic naming of pitches (with B and H as well as the suffix-inflected Fis for F# and Es--pronounced like, and used for, "S"--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, as it were, if lovers of Western music across language barriers recognize that D-Eb stands for Dmitri Shostakovich, on the basis of the Germanic spelling of the letters DS, used in the absence of a note-name equivalent to the Cyrillic letter for "Š."

Composers using musical figures 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) along with their syntactic organization, vertical texture, horizontal structure, tonal organization, and timbral coloring, are invested with communicative value. 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. Finally, countable units---from notes to beats, measures, or sections---invite play with numerical symbols both traditional and innovative. These latter cases venture ever further into the realm of what I have called "the joy of literacy": not only do such significations elude the uninitiated, we no longer expect them to be accessible even to experienced listeners, but only to skilled readers of the score.

Furthermore, music, as I hope to demonstrate, is capable of a kind of descriptive effect that Wendy Steiner (1982: 43-46), 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 achieve this without compromising its intrinsic logic. The reason for this greater flexibility is that music, while resembling verbal texts in that it develops in time, simultaneously "paints." Like the media of visual art, it conveys to its audience the 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. Lessing, in the "Kollektaneen zur Litteratur" published within his "Nachlaß," quotes Bach as praising Telemann as a "great painter" who, in a certain aria, had "so inimitably expressed the marvel and horror at the appearance of a ghost that one could, without the words, which are highly miserable, immediately hear what the music wanted" [so, daß man auch ohne die Worte, welche höchst elend sind, gleich hören könne, was die Musik wolle]. But, Lessing (quoting Bach) adds, Telemann often "exaggerates his painting to the degree of absurdity when he paints things that music should not paint."(Lessing 1955, 6: 223; translation mine). Regrettably, Bach seems not to have provided an example---or if he did, Lessing does not cite it.

But can non-vocal music also narrate? In his work on Mahler, Anthony Newcomb (1998) 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. Musical plot characters (or "agents") are defined by the composer and identified by the listener through attributes that "may be located in various musical elements---for example, in instrumentation, tempo, texture, interval vocabulary, metric design, rhythmic motive or style, or harmonic support," and "these actors may then be understood as actors in the unfolding of a plausible series of actions and events" (Newcomb 1998: 61-62). The difference between such musical agents and agents in representational painting, literature, or film is that these genres are forced to attach attributes to specific human figures and to flesh out these figures with many ancillary details. Music has no need to do so; just as it can present thematic generality (in what Newcomb calls "archetypal plots") without having to attach it to any specific situations or settings, so it can present shifting constellations of attributes to construct plausible agencies---among them human---left undefined. "The actions and events can [...] remain purely musical ones and still imply referential narrative patterns, stories, or plots" (ibid: 63).

Also, Carolyn Abbate (1991: xi-xii) reminds us of 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." Such distance, she believes, is achieved through "moments of diegesis," in which musical voices speak across the sensual matter we are hearing. However, she cautions, such "moments of diegesis" are far from normal or universal in non-texted instrumental music. Since both Newcomb and Abbate are here referring to music that does not, by its title or genre, claim to be a representation of an extra-musical reality, the allowance for "narrative acts of music" is exceedingly encouraging. Non-texted, non-allusive music may not be able to convey the fine details of a plotline because it cannot establish the extra-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 (1994) confirms, mere titles often suffice to provide this essential factual skeleton and make music patently representational---and even narrative.

Thus far I have argued that music, like visual art and literature, is capable of depicting and referring to objects in a world outside its own sonic realm (like a person called Abegg or a cross carried by a man called Jesus), and that what is represented in a pictorial, literary, or musical medium may be an image or a story, a design or a narrative. I now turn to the more specific question of how music may represent something that does not belong to the primary reality "out there in the world" (say, memories of a day in the country-side or a vacation in Italy) or "in here in the soul" (say, a love story or a learning experience), but has previously been represented in a work of visual art or literature.


2 Musical Ekphrasis as a Case of Interart Transfer

The preceding argument leads me to another perspective of what musical ekphrasis is and is not. The distinction is usefully captured when one asks whether what is given in a particular case can be described as "poems or paintings and music?," "poems or paintings in music?," or "poems or paintings into music?" In this section of my essay, then, I wish to inquire 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, generally, to deal with the musical material in a systematic way and, specifically, to know what to exclude and why, I turn to the already established methodologies in (literary) ekphrasis.

In this context, the approach of the Scandinavian interart researcher Hans Lund (1992 [1982]) proves helpful. In his chapter on "The Picture in the Poem: A Theoretical Discussion," Lund provides a scheme intended to define what stance the author of the secondary representation (here, a poet; in our case, a composer) adopts towards the work of art (a painting or, in our case, a painting or poem or 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 two of them.) Here are Lund's definitions one by one, and my own adaptations for the field of 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. (Lund 1992 [1982]: 8)

What Lund sketches here corresponds, it seems to me, to two somewhat different genres in the case of musical composition: setting and collaboration. Both answer the questions put at the outset of this section with "Poems or paintings and music." Among collaborations involving music as one of the key components are works like Parade (by Cocteau + Satie + Massine + Picasso) and L'Histoire du soldat (Stravinsky + Ramuz), to name only two outstanding examples. Such collaborations 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. Neither type of collaboration is ekphrastic, for two reasons. First, it is usually unclear which sign system among those involved should be considered primary and which constitutes the re-presentation. Second, one may assume that the myriad aspects of communication, which would otherwise be expressed within a single artistic text, are 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." 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 visual artist, nevertheless did not, to my knowledge, create any work in which his dual talent engendered a single overarching artistic message. The closest analogue in recent music is probably Erik Satie. Many of his piano scores (see, e.g., Sports et Divertissements, published in facsimile) tread a fine line between musical score and visual artwork. The brief pieces are prefaced with drawings by Charles Martin and may have been intended, or so Satie scholars believe (Whiting 1999: 400-404), to be looked at as much as performed. From the time when emblematic writing itself flourished, at least one composition seems to function as a musical analogue. In the early seventeenth century, Michael Maier (1568-1622) created a work entitled Atalanta fugiens which consists of fifty musical settings in a fugued style, i.e. with voices imitating one another, accompanied by emblems and epigrams. Also known as "Michael Maier's alchemical emblem book," Atalanta fugiens is specifically intended to be appreciated "per oculis et intellectui."

Yet the field of music encompasses compositions that manifest a much rarer combination of talents than a dual aptitude for poetry and painting (like that of William Blake), music and painting (like that of Arnold Schoenberg), or music and poetry (like, more recently, that of Kurt Schwitters). This unique combination is synaesthesia. In correspondence with some painters who claim to be putting on canvas the hues communicated to them in musical sounds, composers endowed with the gift of seeing colors when hearing pitches or chords may purport to be creating a work consisting of sound and color. In the case of a composer who, like Olivier Messiaen, expects his audience to "see" with their inner eye the hues expressed in his chords (Messiaen 1944, Griffiths 1978), 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 on trust. By contrast, in compositions like Alexander Scriabin's Prometheus, 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 (Mirka 1998).

The second musical equivalent to Lund's "combination" are settings of one text in another medium. While often intriguing in themselves, these also constitute a hybrid form in comparison to the more particular phenomenon of musical ekphrasis. 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 affected, 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.

This is not to argue that vocal music can never include ekphrastic means. However, the genres---opera, oratorium, and the lied mainly---are more characteristically perceived as instances in which a message conveyed in a verbal text is supported and perhaps enhanced in the musical parameters chosen to realize and accompany the verbal utterance. Works in which the music is made to convey a subtext that considerably transcends or slants that communicated in the words it sets is as rare as it is fascinating; for examples, see my studies of Alban Berg's Altenberg-Lieder (Bruhn 1998b) and of Paul Hindemith's two divergent interpretations of one of Rilke's poetry cycles (Bruhn 2000b). Yet these exceptions notwithstanding, I agree with Roger Scruton that it is "necessary to distinguish music which purports to carry its narrative meaning within itself from music which is attached to a narrative arising independently, whether through the words of a song or through the action of a dramatic work" (Scruton 1980: 285).

The case is somewhat more complex when a choreographer chooses a piece of music to which to compose a ballet. The possibilities range from cases in which the music is used primarily as an aesthetically satisfying vehicle for the choreography, to cases in which it actually inspires a conceptual interpretation. In order to distinguish with confidence between the two, one would ideally need to create an artificial situation in which one could focus on choreographies in a silent performance---as, for instance, on video recordings with the sound turned off. The question would then be whether such a purely kinetic work could be experienced 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 type of relation.

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 hourglass and the like in the pattern poems of baroque poetry, as well as Apollinaire's Calligrammes and the concrete poetry of Modernism. (Lund 1992: 8-9)

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 content. Musical notation would not be in existence without the content 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 here, there are several other cases that would require answering our initial question about the relationship of music to its stimulus with "the visual or the verbal in music." Music knows the equivalent to "stanzas in the shape of a goblet," integrating a strong visual element. The most obvious examples can be found in compositions written in graphic notation. This system, in which a composer specifies (or, more often, merely suggests) aspects of an intended performance, developed from the verbal directions found in earlier scores. These are now expanded and, in part or in toto, replaced by imaginative symbols that aim to activate the performer's creative participation. Known at least since the middle of this century (see Morton Feldman's Projections of 1950-51), this notational practice has moved more and more into the area of a non-specific analogy of sign and intended content, inviting the performer's creative response rather than a translation of signs regulated by conventions. However, I doubt that we are generally dealing here with a "piece of visual art" even at the simplest level of the term. Notation, in all cases, is graphic in nature. And while an explicitly graphic notation of music that claims to do without any kind of "alphabet" or transliteration of clearly delineated phenomena often takes the idea into interesting territory, I would hesitate to count such scores among the "integrations of music and picture." 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 that are visually arranged in such a way as to suggest depicted objects. (In this sense, the scores of Sylvano Bussotti could be compared with concrete poetry, in that the visual aspect of the written form conveys a message of its own.) 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., see above) 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 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 S'fod, 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 camp authorities would hardly have recognized this, but the ones for whom the message was intended did: it reads "Mazkeret lejom hashana harishon vehu acharon begalut Terezín"---In remembrance of the first and also last anniversary of the Terezín exile (see Karas 1985: 80-81).

Furthermore, 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 in that case would not concern us here. Illustrations in musical manuscripts, however, form a category of their own. Prior to Satie's sketches in the early twentieth century, they were known primarily from manuscripts of late medieval and Renaissance music. An example is the famous Chansonnier Cordiforme, the "heart-shaped chansonnier." More fanciful than useful for music making, the manuscript shows a kind of troubadour song written into a preciously illuminated heart. 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. Thus in a four-part untexted canon by Bartolomé 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 colored the deep blue of the sky. 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 reveals 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. When composers (like Hindemith in the case of Mathis der Maler and The Harmony of the World) prepare symphonic excerpts of their operas, they conspicuously rely almost exclusively on segments that were purely instrumental music in the first place: overtures and orchestral interludes (Bruhn 1998a: 336-340). Conversely, original libretti, i.e. dramatic texts written explicitly as textbooks for operatic setting, notoriously lack the subtleties a theatrical audience would expect, and the few attempts to perform an operatic libretto on the stage of a lyrical theater have yielded rather disappointing results (Gier 1998). Yet the constituent parts of opera---the libretto on the one hand and the "pure" music on the other---are also capable of functioning independently to a greater degree than is the case in the pattern poems Lund mentions. While the visual element in an hour-glass poem is nothing but an empty outline (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 a crude if not unpoetic kind when taken as dramatic works; but, as the now established term "Literaturoper" indicates, there are a number of literary works that originated as dramas and continue to exist as such, before and after their transformation into an opera. Yet these are exceptional cases and certainly not 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 type.

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. (Lund 1992: 9)

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 about painting, they are referred to as ekphrasis. In music, such ekphrasis can take as its object a work of literature (as in Ravel's piano composition, Gaspard de la nuit, whose three pieces are musical transformations of three poems by Aloysius Bertrand; see Bruhn 1997a: 181-228) or a work of visual art, on which I focus here. Among compositions re-presenting literary texts there are symphonic works 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 drama, Pelleas und Melisande) as well as musical works about poems short and long (Schoenberg's sextet Verklärte Nacht on a poem by Richard Dehmel and Elliott Carter's transmedialization, in Concerto for Orchestra, of Saint-John Perse's epic poem Vents). Musical re-presentations of paintings include musical "triptychs" that twentieth-century composers base on works of quattrocento artists (see, e.g., Ottorino Respighi's Trittico botticelliano and Bohuslav Martinu's Les Fresques de Piero della Francesca), transmedializations of Romantic paintings (e.g., Serge Rachmaninov and Max Reger on Böcklin's Isle of the Dead), three musical ekphrases on an early modern work by Paul Klee (see Peter Maxwell Davies's, Gunther Schuller's, and Giselher Klebe's compositions, all bearing the title, "The Twittering Machine"), or the responses, by two contemporary Danish composers, to drawings by M.C. Escher (Per Nørgård's "Ant Fugue" from Prelude and Ant Fugue [with Crab Canon]: Hommage a M.C. Escher, and Hans Abraham's Three Worlds).

Compositions that function as musical ekphrasis may be identified by their titles, which occasionally include explicit reference to the artist on whose work the composition is based (Trittico botticelliano), perhaps even together with a generic allusion to the transmedialized artwork (Les Fresques de Piero della Francesca, The Chagall Windows) or specifically identified (The Twelve Jerusalem Chagall Windows). In other cases, composers may consider the artwork's titling so unique as to serve as an unambiguous marker (Isle of the Dead, Twittering Machine, La mort de Tintagile, Pelleas und Melisande). Occasionally there may be what seems like an overabundance of verbal markers, for example when Hindemith entitles an instrumental composition "Hérodiade (de Stéphane Mallarmé)" and, in addition, places half-sentences from the poetic text above the staves at every significant structural juncture of the music, making it clear that the music follows the poem chronologically (a procedure which is by no means common; contrast, for example, Ravel's re-presentations of Bertrand's poems). At the other end of the spectrum there are cases in which the title of the musical work is not sufficient to convey the source. Among the many reasons why this may be so, three stand out, along with what they disclose about the very different attitudes adopted by composers toward the re-presented artwork.


First, the source text may be little known. Non-French audiences will often not be familiar with Bertrand's poetic cycle Gaspard de la nuit; so Ravel convinced his publisher to reprint the three poems in full, each facing the beginning of the musical re-presentation in the score, thus guaranteeing access to the source works for anybody who cares to know.


Second, a title may provide insufficient guidance when it is based on a pun or a sophisticated quotation from a third-party. This is the case with Nørgård's Prelude and Ant Fugue: Hommage à M. C. Escher, which plays ingeniously not only with one of the lesser-known among Escher's works, a 1953 print entitled Ant Fugue, but also with two chapter headings in Douglas R. Hofstadter's famous book, Gödel, Escher, Bach (1999 [1979]). In Hofstadter's book, technical chapters (with correspondingly technical titles) alternate with poetic ones, which bear musical titles. Among the latter, there are two that appear as if linked by suggestive elisions. The first musical chapter in Part II is entitled "Prelude . . ."; it is followed, after a chapter elaborating on "Levels of Description, and Computer Systems," by ". . . Ant Fugue." In this latter chapter, Hofstadter reproduces the Escher print that also graces Nørgård's score. The "Prelude . . ." chapter is illustrated with Escher's 1963 woodcut, Möbius Strip II, in which ants crawl over the unending surface of the famous twisted loop. Nørgård's own prelude, thus linked in more than one way to the following fugue, is a play on the first prelude from the high point of all compositions of "preludes and fugues," Bach's Well-Tempered Clavier, which is associatively interpreted here in something like an endless loop.


Third, the name given to a composition may even by deliberately misleading. Such seems to be the case with Carter's titling, Symphony for Three Orchestras, which challenges audiences to imagine (or listen for and try to identify) the alleged three instrumental bodies pitted against one another, concerto-style. With listeners (and students of the score) thus distracted by alleged intra-musical questions, the composition would be unrecognizable as a work of musical ekphrasis were it not for the extensive preface, in which the composer names and introduces the poet, identifies the work on which he drew, Vents, and reprints several significant excerpts from the poetic text.

Finally with regard to allusive markers, a particularly intriguing case is that in which the composer transmedializing a visual artwork points at the specific source text by way of one or several musical signifiers. An educated audience listening to a performance of Respighi's Trittico botticelliano, either in the concert hall without access to a program that would give the movements' headings, or on the radio without the benefit of some introductory remarks, will smile with recognition at the opening sounds of the initial movement. The first sixteen measures constitute a (modified but immediately identifiable) quotation from one of the most memorable phrases in the first of Vivaldi's Four Seasons violin concertos, "Spring." Considered in the context of Meir Sternberg's comprehensive theory of "quotation" as mimesis of discourse, where the represented object "is itself a subject or manifestation of subjective experience: speech, thought, and otherwise expressive behavior" (Sternberg 1982: 107), Respighi thus presents us with a multiple ekphrasis. By means of something like the musical equivalent to indirect speech, the composer draws our attention to the season of blossoming flowers (and love) as the suggested object of representation. Moreover, the extraordinary amount in this movement of what musicians refer to as "timbral coloring," achieved above all with the aid of tuned percussion instruments like glockenspiel, celesta, xylophone, etc., recommends the movement as one in which music "paints." Finally, the movement comprises two additional musical "quotations" (as Sternberg employs the term): an allusion---through characteristic rhythmic and melodic figures---to the (musical) genre of a "Renaissance dance," and a troubadour hymn to spring, the latter a signifier linking the time of blossoming even more unambiguously with idealized love. For the (admittedly rare) listener capable of identifying even these more esoteric mimetic inserts, the imagined listening experience, unencumbered by verbal information, might yield sufficient hints to identify the source text. As a painting in Renaissance style about spring and idealized love, Botticelli's --- may well be the first that comes to mind. Attuned to the "universals of quotation" (Sternberg), this compound of second-order mimeses thus serves as an intralinguistic (i.e., musical within the language of music) marker alerting the audience to the (extra-musical) object of re-presentation, a specific work of visual art.

When transformation of an artwork is brought onto the theatrical stage and blended with the miming aspect of that genre, the result is a case of enactment. I know of at least three compositions based on serial paintings that can be shown to contain distinct elements of enactment: Igor Stravinsky's The Rake's Progress, Paul Hindemith's Mathis der Maler, and Arthur Honegger's La danse des morts. Hindemith's opera Mathis der Maler integrates and musically interprets several panels from the Isenheim Altarpiece, the masterpiece of the operatic protagonist's historical model, Mathis [or Matthias] Grünewald (see Bruhn 1998a: 340-350). Stravinsky's opera The Rake's Progress, the stimulus of which is a series of eight engravings etched in 1735 by William Hogarth, is written after a libretto in which W. H. Auden, with the help of Chester Kallman, dramatized the story told in the etchings. The fact that the Auden/Kallman libretto is a literary ekphrasis in itself, which Stravinsky then casts into musical form, sets this case of musical enactment of a pictorial source apart from the third example. For his oratorio La danse des mort, based on Hans Holbein's Totentanz, Honegger uses texts that, while "authored" by Paul Claudel, are actually compiled from the Bible and assorted folk-songs. As such they constitute something akin to a verbal embodiment of the common source that inspired both 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 Leoš Janácek's composition The Lord's Prayer in response to Josef Kresz-Mecina's panels, a work the composer based on five tableaux vivants he had devised himself (Fink 1988: 122-124). Here, too, the text predates both the visual artwork and its musical transmedialization.

Compositions rendering ekphrastic poetry---i.e., poems that are themselves transformations of pictorial texts---usually constitute not cases of musical ekphrasis but rather mere settings (see my distinction above). 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 inspired 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" (so-called "picture poems")---a form that is necessarily 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 musical ekphrasis in a musical work that then presents, as it were, a third-level transmedialization. Thus Debussy's piano piece Clair de lune in Suite Bergamasque is based on Paul Verlaine's ekphrastic poem (by the same title) after Antoine Watteau's painting, Fêtes galantes.

Finally, the composer's musically transmedializing a work of verbal or pictorial art may inspire a creative artist working in a third medium to extend the ekphrastic process even further, adding yet another level of re-presentation. The cases of two-phase transmedialization in which I have been most interested are those involving three different media each: from the pictorial to the verbal and on to the musical (as in Honegger's oratorio, La danse des morts, written---with the mediation of a text that Claudel compiled from the Bible and other sources---on Hans Holbein's Totentanz series), from the pictorial to the kinetic on to the musical (as in Hindemith's Nobilissima Visione---on Giotto frescoes of Saint Francis of Assisi, in the Bardi Chapel of Santa Croce in Florence, via Léonide Massine's choreography), or from the poetic model to the musical transformation to the visual or kinetic interpretation (as in the two ballets written on Claude Debussy's and Hindemith's Mallarmé-based symphonic compositions: Vaslav Nijinsky's Après-midi d'un faune and Martha Graham's Hérodiade; see Bruhn 2000a).

The central questions I have been asking regard the scope and nature of this interart, 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 these choices, situated within a certain historical and cultural context, allow us to ascertain and describe a new "convention" of intersemiotic transformation?


Does the range of ekphrastic stances adopted by composers toward works of verbal or visual art (mimesis, supplementation, association, interpretation or recontextualization, playful response, etc.) parallel those observed in ekphrastic poets?

3 Means of Musical Transmedialization

In the following examination of the possible scope (and limitations) of the undertaking to reflect images in tones, I will draw on three groups of symphonic compositions based on visual representations: a musical "triptych" in three paintings by Sandro Botticelli, two cyclical compositions on Marc Chagall's Jerusalem Windows, and three short works by unrelated composers on Paul Klee's The Twittering Machine (see Table). In passing I will also mention two hybrids: Hindemith's Nobilissima Visione, a ballet music on Giotto's Saint Francis frescoes in the Bardi chapel at Santa Croce in Florence, and Honegger's La danse des morts, an oratorio for which Claudel compiled a text mainly from biblical verses. In these two cases, the choreographer and the poet respectively share the response to the work of art with the composer.

Table 1: Examples of musical ekphrasis





Sandro Botticelli

L'adorazione dei Magi
La nascita di Venere

}Trittico botticelliano

Ottorino Respighi (1897-1936)

Marc Chagall

stained-glass windows,
synagogue, Jerusalem, Hadassah-Hebrew Medical Center

(1) The Twelve Jerusalem Chagall Windows
(2) The Chagall Windows

Jacob Gilboa (1920-)

John McCabe (1939-)

Paul Klee

Die Zwitschermaschine

(1) "The Twittering Machine" from
7 Studies on Themes of Paul Klee
(2) Die Zwitschermaschine
(3) "The Twittering Machine" from
Five Klee Pictures

Gunther Schuller (1925-)

Giselher Klebe (1925-)

Peter Maxwell Davies (1934-)

Hans Holbein

Simulachres et Historiées Faces de la Mort, wood cuts

La danse des morts, oratorio

Arthur Honegger (1892-1955)
with poet Paul Claudel


7 frescoes on St. Francis of Assisi, Bardi Chapel, Santa Croce, Florence

Nobilissima Visione, ballet
with choreographer Léonide Massine

Paul Hindemith (1895-1963)

One way of exploring the ability of music to transmedialize works of visual art is to arrange the devices that a composer chooses in a single work of ekphrasis on an imagined scale between, at the one end, the iconic or "natural" and, at the other end, the allusive or "arbitrary." Ottorino Respighi's instrumental interpretation of Botticelli's three paintings is a case in point. The Italian composer exploits the entire spectrum of depictive means available to the musical language. There are even moments in his work that come very close to onomatopoeia. Respighi uses an (extant or convincingly imitated) Renaissance dance ("dance" here refers to a musical genre, not a choreography) to make us hear the music to which the three Graces in Botticelli's Primavera --- sway. In another musical section, he quotes one of the most famous troubadour songs---which happens to be a hymn to spring and is thus doubly suitable for a transmedialization of the painting Primavera---to signify the concept of idealized love that, according to the Neoplatonic views informing Botticelli's allegorical paintings, was epitomized by Venus. When the composer colors the refrain of the troubadour song with harmonics (high, ethereal-sounding overtones) in the violins, he relies on our visual-conceptual comprehension of this musical device as a representation of loftiness; we are invited to understand that he is pointing to the "elevated" nature of the love to which the Florentine artist dedicated his painting about Venus.

While music is no doubt "a language," we all know that it cannot label or describe directly; it cannot simply say or show "red" or "green", "behind" or "in front," "apple" or "chair." In order for a music listener to understand how music responds to a work of art, it is even more necessary than in the case of an ekphrastic poem that the beholder be acquainted with the stimulus. The significance of the listener's familiarity with the primary work of art increases in proportion to the degree to which a composer establishes original links between musical means and the extra-musical content of that primary work. Musical gestures of military prowess as suggested in trumpet fanfares and drum rolls, or the depiction of narrowness in a melody with a strangely restricted pitch range carried by a leading voice, convey distinct pictures of triumphant, potentially threatening force and (physical or emotional) constraint respectively, whether or not an audience is capable of associating the agents in the suggested scenarios with fictional characters in the primary artwork. By contrast, a harmonic is just an overtone. A harmonics-inflected tune from the troubadour repertoire may be assumed to signify elevated love whether or not a listener identifies the genre and retraces the connection to idealized adoration. However, it requires familiarity with the painting being referred to, as well as knowledge of the general humanistic ideas informing depictions of mythological content in the era of Neoplatonism, if an audience is to follow Respighi as he represents, through this symbolic attribute, Venus the goddess of divine love as painted by Botticelli.

Intersecting with such an imagined scale from the iconic to the allusive is another, one defined by the nature of what is being transmedialized. Composers may set themselves the task to represent in their music not only a content imagined as a scene or story, but beyond that the artistic language in which this content was first expressed, its form, and even stylistic or otherwise idiosyncratic details of the primary representation.

Music can take as its point of departure the medium employed in the original representation. A composer may aim to capture the multiple fragments in identical or similar hues by which stained-glass windows are distinguished from other visual representations, stringing together "shards" of music by way of multiple repetition, including even an emulation of the variations in color owing to thinning intensity close to the edges; all this and more can be found in Gilboa's The Twelve Jerusalem Chagall Windows. Or, in cases where a medium becomes part of the content, a composer may draw his or her inspiration from the element suggested within the primary representation. Thus McCabe, in responding to the visual representation of water that Chagall incorporated in his portrayal of some of the tribes of Israel, uses the shimmering effect of an aleatory texture with multiple layers to create musical images of water in the corresponding sections of his composition. Music can also translate an implied component of the image, as Davies did when he organized some of the basic musical entities in his "Twittering Machine" in a precisely overlapping fashion that imitates the---invisible but no doubt necessary---cogwheels in Klee's contraption. All three cases constitute "iconic" manners of musical depiction.

A musical transformation can tangibly point to the time in which a scene or story is placed. The time may be that of Saint Francis of Assisi's twelfth century (conveyed through the use of a trouvère song from the era, as in Hindemith's Nobilissima Visione based on Giotto frescoes) or that of twentieth-century technology, which informs "twittering machines" and other contraptions (suggested through the use of serialized pitch organization in Schuller's composition or of motoric rhythm in Klebe's piece).

A composer can invite the listener to draw for his understanding of the ekphrastic pointer on a shared pool of musical signifiers, through the citation of existing musical material. Respighi's composition begins with a rather obvious paraphrase of the first concerto in Vivaldi's Four Seasons. Since that concerto is, of course, entitled Primavera, he thus creates by quotation an allusion to the "spring" of Botticelli. Later in the same work, Respighi suggests the orchestral palette which his Russian teacher, Rimsky-Korsakov, had employed in his composition Shéhérazade, and by doing so places his musical representation of the "Adoration of the Magi" as an event occurring in the Orient.

Last but not least, as the fortuitous cases of several composers responding to the same work of art demonstrate, the range of possible approaches is open not only with regard to the question how something will be re-presented or re-told, but just as much by the choice of what aspect of the primary artwork, or what angle of its interpretation, is chosen to be foregrounded in the musical transmedialization. While Davies and Schuller focus on the joint fate of the four pathetic bird heads attached to Klee's crank, albeit allowing for very different degrees of rebellion versus helpless compliance, Klebe sees in the same painting four individuals taking very different stands under shared duress. And while Gilboa feels inspired by a combination of the words heard in Jacob's blessing and the radiant, translucent play of light through glass seen in Chagall's windows, McCabe concerns himself with the dynamics prevailing among the twelve brothers and tribal ancestors as both the biblical patriarch and the artist presented them.

As my close reading of these case studies has convinced me, there is no parameter of the musical language that cannot be used in the service of depicting or referring to an extra-musical reality in general and, more specifically, an extra-musical work of art with its form and content. At best, preferences can be observed in the means chosen by individual composers. Almost all facets of rhythmic structuring have at one time or another been employed to convey affect. Double-dotted attacks characteristically stand for allusions to "fate," while the play with syncopations and other rhythmic irregularities often serves to capture ambivalence. Rhythmic flexibility is often employed to epitomize the concept of freedom, especially the freedom from the constraints of conventions. By extension, rhythmic flexibility may stand for a cliché characterization of a cultural other---be it the King from the Orient at the beginning of Respighi's "Adoration of the Magi" or the faun in a composition by Debussy: both are musically represented in rhythmically very elastic woodwind melodies. By contrast, rhythmic monotony evokes different associations, depending on the context in which it is heard. In connection with extensive note repetition and percussive sounds, a listener will hear monotony as combative energy (as in the "Judah" of McCabe's Chagall Windows). In a largely homophonic setting, rhythmic monotony reminds one of machine-generated noises; this plays a role in the first movement of Klebe's Twittering Machine.

Pitches, the smallest units of diastematic material, are used as signifiers both individually and collectively. (To give an example from a musical ekphrasis of a verbal text: Debussy in his Prélude à l'après-midi d'un faune, based on Mallarmé's poem, employs two pitches as tonal centers throughout the composition to represent the two competing aspects in the faun's perception of what happened to him---the thwarted sexual adventure and the invention of the flute made from reed, or the sensual versus the aesthetic.) The connection of two or more pitches into intervals or chains thereof plays a significant role in musical representation in general. Drawing on the signification of certain intervals established in the context of the rhetoric-of-music tradition of the sixteenth and seventeenth centuries, mentioned above, composers may employ tritones and semitones as a contrast to perfect fourths and fifths to distinguish emotional insecurity from self-assurance. Schoenberg thus employs the semitone, the smallest interval in the traditional scale of Western music, and the allegedly unsingable tritone, that artificial interval created through the bisection of the octave which was long dubbed diabolus in musica, in the material characterizing the women in his musical re-presentation of Dehmel's poem, Verklärte Nacht (Transfigured Night), and of Maeterlinck's drama, Pelléas et Mélisande. Semitone and tritone stand for human smallness and the not-quite-accepted, respectively, while the "perfect" intervals, as emblems of consonance, express conformity with conventional views and are reserved, in Schoenberg's composition, for the depiction of Dehmel's self-assured "man" and for the somewhat formal and rigid Golaud in Maeterlinck's play.

Pitches joined to form recurring motifs may assume signifying power either through the associations composers and audiences share about the message implicit in certain contours, or through context-specific relationships. McCabe's composition The Chagall Windows, in which recurring pitch contours are used as leitmotifs for qualities such as strength of character, recklessness, etc., provides particularly convincing examples of the former. In other cases, a motif that has been firmly established in its original contour may later recur with changes arrived at through processes that in themselves imply distinct meanings with regard to the extra-musical agent signified. Here, the details of the resulting shapes may communicate inflections with regard to the emotional or spiritual state of the entity referred to. This is superbly achieved in Hindemith's Nobilissima Visione, where the composer invites his listeners to follow Saint Francis's psychological and spiritual experiences in the musical form of the unexpected transformations, interruptions, distortions, and reaffirmations to which he subjects the trouvère song.

Next to rhythms and pitches, the timbres of the various instruments lend themselves particularly well to characterizations of suggested dramatis personae. The link between a certain sound color and a human agent to be imagined may be based on conventional attributes or introduced in a composition to suggest traits not necessarily anticipated by the audience. In the first category we find the faun's flute, the horn of the huntsman, and the drum of the soldier. These instrumental emblems, identifying generic rather than unique agents, function as "ekphrastic models" in the sense proposed by Yacobi (1995, 1998); they are sonic signifiers that present, as a kind of synecdoche, the attribute associated with their bearers in countless works of literature and visual art. In a related sense, bells of various kinds, including the glockenspiel, are employed the way they have long been, as sounds indicating a quality of the religious ritual. The organ as the prototypical instrument of church ceremony is linked to the Christian Mass. In La danse des morts, Honegger uses the instrument sparingly, reserving it for moments when he intends to make tangible the loving Father communicating with His children. The other aspect of God, the terrible Yahweh who appears to frightened humans in fire or thunder, the judge of human failure and weakness, is timbrally represented by the trombone, an instrument with a long history of suggesting divine judgment and punishment. In religious music, the instrument was often employed as a latter-day emblem of the angelic instruments announcing the Last Judgment, which Luther, in his translation of the Bible, had rendered as trombones (see the trombone statements with religious connotations in Mozart's edition of Handel's Messiah as well as in his own Requiem, in the final movement of Beethoven's Ninth Symphony, Berlioz's Grande messe des morts, etc.) In addition to its characteristic use in church music, the trombone was also featured as a timbral signifier of divine pronouncements in opera: see the oracle scene of Gluck's Alceste, the "sacrifice" scene of Mozart's Idomeneo, the judgment-at-supper scene in Mozart's Don Giovanni, and many others.

Other instrumental colors, while not necessarily conventionally pre-defined, nevertheless serve as intuitively comprehensible timbral signifiers. This applies to the sounds of the trumpet for the dashing young fellow and of the whip for brutal disposition. The bassoon, whose often mocking and always nasal sound quality associates it with charming humor and bucolic playfulness, offers itself as a very appropriate signifier, especially for pastoral scenes.

Finally, regarding the means available for the musical representation of extra-musical content in general and the transmedialization of works of visual art in particular, all musical parameters (pitch, intervals, harmony, rhythm, meter, tempo, timbre, texture, structure) can be used in quotations of existing musical material or allusions to known musical genres, to music-related circumstances, or to other meaning-carrying content in music. In terms of the relation of the musical composition to the primary work of art, a musical entity then functions not as a symbol but as a signal, in the sense defined by Susanne Langer (1967 [1953]: 26): "a signal is comprehended if it serves to make us notice the object or situation it bespeaks. A symbol is understood when we conceive the idea it presents."

Allusions used to signal an "ekphrastic model" are often marked by timbres with conventionally fixed associations, as in the examples listed above (the hunting call employed to typify a characters, etc.) We find another interesting instance in the orientalizing features of the "Adoration of the Magi" in Respighi's Trittico botticelliano: the music thereby suggests not a trait pertaining to a person, but a geographic locale---actually one quite different from the solidly Italian ambience painted by Botticelli. This example can be understood as a case of one "model" (an orientalizing melisma as a generic attribute of the peoples from the Eastern shore of the Mediterranean) referring to another (the "adoration of the Magi" as a pictorial model, of which Botticelli's painting is one specific realization). The same movement also contains another fine example of a musically represented ekphrastic model. The composer suggests the topos of dirge or elegy through the combination of the lilting siciliano rhythm with the mournful color of a slow and legato-playing bassoon, and recreates in his melopoeia the impression given by the chants of the oriental liturgy.

When Respighi organizes some of his material in the way familiar from the music accompanying Renaissance dances, he reduces the full orchestral palette to three soloistic wind instruments. These, embodying with their slim and sober sound the aesthetics of the neoplatonically inclined society of fifteenth-century Italy, intone the typical irregular meters of the dance music known to have been played at the Florentine court during Botticelli's time. This is most certainly an eloquent allusion. It may even be a specific quotation (as opposed to the generic ones mentioned above), but if so, the composer must have been using a source not generally available to today's researchers. Another passage that similarly comes close to such quotation is the chorale used by Arthur Honegger as a recurring theme in La danse des morts. It sounds so tantalizingly like a hymn that "must exist" that this researcher, for one, spent weeks trying to discover a source that, alas, remained elusive.

Such pointing quotations---insertions of verifiable pre-existing musical material into a new musical frame-work---may be anything from subtle to conspicuous. In what follows, I will give examples representing various points on an imagined scale from the obvious to the concealed. Most typical among the materials lifted from another musical context are syntactic entities like themes and motifs or even entire song stanzas. They may be inserted into their new frames---in our case, the transmedializing composition within which they are made to serve as a musical signifier to a non-musical artwork (or an aspect thereof)---with or without the specific textures and timbres that originally pertained to them. With few exceptions, such quotations are taken from one of three sources: another composer's composition, the folk repertoire of either the composer or the country in which the story is placed, and sacred music with its hymns and chants.

Easiest to identify among the musical quotations found in the works I investigated are Respighi's two borrowings, in the first and third movements of Trittico botticelliano, of conspicuous figures from Vivaldi's Four Seasons. Strictly speaking, these are paraphrases, but the combination of an identical key and octave range together with the imitation of the original timbre and ornamentation captures the atmosphere so well that the effect is one of having heard a bit of Vivaldi. Equally accessible is the quotation, in Honegger's oratorio, of the "Dies irae" melody, which is superimposed over the remaining strands of the texture in very slow notes and thus gives listeners ample time for identification. Respighi's rather straightforward way of quotating from Vivaldi's "Spring" to point at Botticelli's painting have already been commented upon; Honegger's device for evoking "that day of wrath" in a composition that interprets Holbein's Totentanz woodcuts in light of God's judgment of the people of Israel is equally accessible and persuasive. But while no ardent music lover will miss the well-known plain-chant or the snippets from the much-beloved violin concerto, and hardly anyone will fail to make the connection implied in the words or the title respectively, the same cannot be taken for granted in the case of Respighi's quotation, in the second movement of his Botticelli Triptych, of the late-Gregorian hymn melody, "Veni veni Emmanuel." It is a specific quotation, and yet, since the melodic contour is not that with which we are familiar from the Christmas carol sung to the English translation of the words, "O come, come Emmanuel," only the initiated---scholars and lovers of Gregorian hymns---will recognize the inset and thus benefit from the hint by hearing a further musical signal pointing to Botticelli's Adorazione dei Magi.

Hindemith (in Nobilissima Visione after Giotto's frescoes on Saint Francis of Assisi) and Respighi (in the "Spring" movement of Trittico botticelliano) each quote a troubadour melody. Both melodies are famous examples within their genre, but not necessarily familiar to a general concert audience. And yet, notwithstanding the discrepancy between the composer's intention and his listeners' lacuna in "literacy," which in this case hinders the identification of both the musical contours and the words assigned to them, quotations such as these can ideally be recognized by thoughtful music appreciators and thus signal interpretative aspects within the transmedialized primary text. Saint Francis, then, appears primarily not as the founder of an order of mendicant little brothers, or as one who cares for the sick and loves even lepers, or as one who receives the stigmata, but, in line with one of his self-characterizations, as "God's minstrel" (Jørgensen 1955: 127), the poet-composer of the Canticle of Brother Sun. Interestingly though, while Respighi's and Hindemith's troubadour songs both express, by virtue of this genre, the trope of idealized love for an inaccessible lady, they differ in the way these quotations are used. In both cases, the material is quoted instrumentally; but even the melody alone, constituting one component of an original word-music entity, invites the question who is pictured as singing. In Trittico botticelliano, no specific singer is suggested. The quotation thus serves exclusively to point to the main character in Botticelli's painting, Venus, and the assumed gist of the message intended in the painting as a whole: the representation of a personification of ideal love. By contrast, in Hindemith's Giotto-inspired ballet music, the troubadour song serves as the signature tune of Saint Francis and thus characterizes not the presumed addressee of the song (the Virgin Mary) but the protagonist of the story, Saint Francis himself.

A musical quotation may thus embody anything from a timbrally, rhythmically, or otherwise characterized "model" to an (ideally) identifiable larger syntactic entity. The object of the montage may function as a musical marker that helps define the object of the transmedialization ("the music must be referring to a representation of spring, since Respighi gives us a snippet of Vivaldi's Spring") or as a window on an unexpected aspect of the primary artwork (showing Saint Francis as, above all, a singer---the one thing he is not in Giotto's frescoes in the Bardi chapel). In either case, such insets, speaking autonomously within the superimposed frame, add a unique means by which an ekphrastic composition communicates the viewpoint adopted in the re-presentation.


4 Variations of Ekphrastic Stance

In the 417-page "theory" prefacing his three-volume annotated anthology of ekphrastic poetry, Gisbert Kranz (1981-87) organizes the attitudes a poet may assume towards a work of visual art into a number of categories. His complex grid of cross-classification groups ekphrastic poems in terms of (1) what they achieve with regard to the visual depiction (transposition, supplementation, association, interpretation, provocation, play, or concretization); (2) what their intention is with regard to the work and its artist (to describe, praise, critique, moralize about; to use as a didactic, political, or sociocritical tool; to express delight, etc.); (3) what speech attitude the poet adopts (allocutive, monological, dialogical, apostrophizing, or epic; genetic, meditative, or cyclical); (4) what order of reality pertains to the picture represented in the poem (real, fictitious, or cumulative); (5) what occasioned the poem (including the situation of the author, of the pictorial work, and of the artist). For my present purpose. I will focus specifically on five of the categories within the first group, i.e., ekphrastic transposition, supplementation, association, interpretation, and playfulness, with a view to showing that the distinction made in the sister domain is equally relevant in the field of musical ekphrasis. (Kranz's understanding of ekphrastic "transposition" as one of many ekphrastic procedures seems confusing in light of the work of scholars, from Tardieu 1944 through Spitzer 1955 to Clüver 1989, who use it as a general designator, as in "transposition d'art" or "intersemiotic transposition").

In ekphrastic "transposition" thus defined, a poet recreates not only the content of the primary work of art but also, and significantly, pertinent aspects of its form or its arrangement of details. A beautiful musical equivalent to this kind of poetic transposition exists in Walter Steffens's 1966 flute-and-piano composition, La Femme-Fleur, after Pablo Picasso's 1947 oil painting by that title. As Monika Fink (1988: 105) has shown in her analysis of the piece, the composer achieves a stunning translation of the pictorial structures. Picasso's painting traces a process of abstraction that takes as its point of departure a naturalistically rendered female nude and converts it, in a series of consecutive retouchings, into a sunflower with a woman's face. In correspondence with this pictorial transmutation, Steffens uses serial technique to transmute an original melodic idea into a pointillist abstraction of itself. Just as the contours in the portrait are distorted and then restored to form new shapes, so is the serially developed melodic line broken up and regrouped into new musical entities. Fink argues convincingly that the conclusion of Steffens's piece, in which the flute returns to the elegant melodic phrase of the beginning, is a musical interpretation of the fact that the gaze of the appreciator, having taken in Picasso's transmutation of woman into flower, ends by returning to the woman when it discerns her face among the petals.

In ekphrastic supplementation, a poet adds to the visual representation any of the innumerable nonspatial dimensions that a painter or sculptor may imply but cannot realize directly: sensory experiences, for instance, as when words describe sound, smell, taste, and touch. Supplementation may also allow a poet to read postures as arrested gestures and infer a possible "before" and "after" that would enclose the captured moment. Finally, words can attribute complex chains of thoughts and nuances of feelings to the depicted characters, where the painter's brush has to limit itself to more general suggestions. Among the musical works I have examined, one relates to its model in ways corresponding to these poems: Hindemith's Nobilissima Visione on Giotto's frescoes at Santa Croce in Florence.

In Hindemith's composition, one is struck by the degree to which the composer uses means and aspects not available to the visual artist to create a work that Giotto would have recognized as true to his spirit. The fact that he can, without in the least straining our image of Saint Francis, add the sonic component as a means of expression is particularly fortuitous. The composer's representation of the little brother of Assisi relies primarily on the saint's devotion to joyous singing, expressed in his much-documented love for the songs of the French troubadours and trouvères and his most beautiful artistic creation, the Canticle of Brother Sun. Giotto's frescoes of Saint Francis touch their beholder deeply as they capture his modest yet blissfully radiant personality. Yet the devout artist has no means of giving us the songs through which Francis expresses his deep inner joy, the heavenly music he hears and to which he dances with his improvised violin, and his praise of Creation, which he leaves as his poetic testament. On all this, the music dwells; here it excels.

Ekphrasis by association does not render exactly what the eyes see (in a visual image or on the printed page of a literary work) but, inspired by the primary work of art, spins off new thoughts or familiar mental or emotional connections. Two musical works that associate without replicating details of content or form: are Resphighi's Botticelli Triptych and Honegger's La danse des morts.

When Respighi renders in music an imaginary triptych of Botticelli paintings, he transmedializes not so much the scene or story visible in the three depictions but the associations that he himself has and that he hopes in turn to induce in his audience. Thus, under the title of the first of the chosen canvases, ---, the mythological scene with Venus in the center evokes thoughts of Ficino's theories about allegory (Ferruolo 1955). In 1478, a few years before the painting was created, Marsilio Ficino had attributed a dual meaning to Venus: she represents Humanity; her soul and mind are Love and Charity. Respighi musically epitomizes these thoughts in a hymn to spring and idealized love. The backdrop of the depicted scene and the playful characters of Zephyr and Cloris to the right may imply a bucolic context, which the composer renders in the form of a bassoon theme of distinctly pastoral tone and humor. The movement of the three Graces, seen in a circular arrangement with fingers interlaced, evokes in him the association of Renaissance dances. In Botticelli's Birth of Venus, the canvas that is re-presented in the third movement of the composition---the other "wing" of the triptych, as it were---Respighi notes the recurrence of two of the characters, Venus and Zephyr, and responds by recalling conspicuous features of the Renaissance music he introduced in his "Primavera" movement. In response to Botticelli's Adoration of the Magi, as already mentioned, he calls to mind a late-Gregorian hymn that hails the coming of the Savior, as well as suggesting a popular Italian Christmas song. Furthermore, and perhaps most importantly, the very arrangement of the three independent paintings into a Trittico botticelliano carries its own associations, which create an overarching message that neither the painter, nor his contemporaries, nor possibly any of today's art experts would have perceived were it not for this composer's musically couched interpretation.

The borderline between this kind of association and ekphrastic interpretation is permeable. In the latter, the transmedializing artist uses not associations of a personal nature, but rather implications known to and shared by all three parties: the creator of the original work, the responding artist, and the community of appreciators. These familiar implications include knowledge regarding the historical, legendary, or literary background of the scene or story, which would typically form part of a cultural framework within which all participating parties are embedded. Of the symphonic compositions I chose for my study, three can be said to rely on interpretation: McCabe's and Gilboa's musical transformations of Chagall's Jerusalem Windows and Honegger and Claudel's joint transmedialization of Holbein's woodcut series about the "dance of the dead"-itself a particular instance of an "ekphrastic model" in the sense introduced by Yacobi (1995, 1998). In all three compositions, the familiar associations upon which both the visual and musical representations build and on which the communication with beholders rely are taken from the Hebrew Bible (or the Christian Old Testament, respectively). In the case of Chagall's representations of the twelve tribes of Israel, the Bible provides an obvious source of association and reservoir for interpretive angles; in Holbein's work, the connection with specific biblical passages is not self-evident and therefore all the more intriguing.

Paul Claudel, asked to provide a verbal text that would enable Honegger to transmedialize Holbein's artwork into an oratorium, associates specific sections of the Bible, especially from the books of Ezekiel and Job, with the message he finds expressed in the "religious" images of the woodcut series. Just like tableaux nos. 1-5 and 40-41 in the visual representation, these passages function as a verbal frame based on biblical imagery from the Creation to the Last Judgment that encloses an array of encounters between death and his clients. Holbein's 1526 Simulachres et Historiées Faces de la Mort, usually referred to in modern German as "Totentanz" (thus a "dance of the dead" rather than a "dance of death"), might also be read as a specific instances of an age-old folk tradition, one that was no doubt largely innocent of the visions and spiritual struggles of the prophets Claudel evoked. Having made the primary association between this particular rendering of the Totentanz with biblical prophecies and injunctions, the two authors of the twentieth-century ekphrasis then take the artist's depictions as a point of departure for their idiosyncratic reading. In the text and the music of the oratorio, Holbein's less-than-pious sequence of encounters with the beckoning skeleton is rendered, in a single movement, as a dance of the reassembled bones provocatively singing French folk melodies and revolutionary songs, their melodies piled one on top of the other with open disrespect for any shared meter or tonality. This riotous and unsynchronized cacophany seems to collapse the sequential imagery of the thirty-four central "client-of-death" woodcuts into a simultaneous, many-voiced musical representation of defiance.

Yet these dancers will soon be truly dead; Honegger's music states this as unambiguously as do Holbein's vignettes. The artist conveys this by way of the hourglass, another "ekphrastic model" found throughout the art-historical canon. Hidden in different corners of twenty-four of the woodcuts, the memento mori reminds beholders that no matter how various the forms of the "dance" depicted here, all of them ultimately lead to the grave. Claudel creates a verbal re-presentation of the artistic model in his collage, strung together from various biblical verses (Gen 3:19, and the Song of Songs 8:6, Matth 6:25), which reads "Souviens-toi, homme, que tu es esprit et la chair est plus que le vêtement et l'esprit est plus que la chair et l'œil est plus que le visage et l'amour est plus que la mort" (Remember, man, that you are spirit and the flesh is more than the garment and the spirit is more than the flesh and the eye is more than the face and love is more than death.) Honegger sets this phrase by means of three musical devices that add up to a striking transmedialization. In casting the phrase with multiple repetitions, as a bass below a texture of ever-changing upper voices, the composer shapes a passacaglia. This musical form, which dates back to the seventeenth century, has been deployed by recent composers (e.g., Benjamin Britten's Peter Grimes, Alban Berg's Wozzeck, Paul Hindemith's Marienleben) with a view to its figurative connotations, suggesting a "persistent foundation" or "basic idea." In stringing the phrase into a seamless twelve-fold reiteration and thus securing its continuous presence beneath roughly two thirds of the boisterous singing in this movement, Honegger takes up Holbein's two-times-twelve-fold inset of the hourglass that apprears in more than two thirds of the "client-of-death" representations. In presenting the phrase recto tono (i.e., in a rhythmicized but monotone style of recitation on a pitch repeated throughout the entire phrase), he achieves an effect that lessens the listeners' attention to the individual words, so much so that they are likely to miss, the first few times at least, that this is not at all the oft-heard admonition regarding man being dust and returning to dust. The thought that these revelers should be spirit appears at first quite contrary to what is expressed in Holbein's genre etchings, at least when one focuses on the entertaining images that make up the large central section. It is only the biblical "frame," established in tableaux 1-5 and 40-41, that invites beholders to interpret the popular humor shown in the "client-of-death" vignettes as a mere veneer covering the profound question of what humans are, perishable flesh or immortal spirit.

Expanding on this angle of reading Holbein, Honegger (and Claudel) reserve the remaining movements, and thus by far the larger portion of the composition, for various kinds of reflections about death and, most prominently, for Man's direct conversation with God. This encounter, not emphasized in the images but owed to Claudel and Honegger's interpretation, prompts the composer to base his work on musical signifiers such as timbres---the thunder of God's presence, the angels' trombones, the solo voice accompanied by a violin obbligato in the style of a Bach aria---that shift the focus from a (spirited, often even facetious) representation of the common fear of untimely death to that of a deeply spiritual struggle.

Finally, as in ekphrastic poetry, the playful stance is easiest to recognize in musical transmedializations. In a mode similar to Gerhard Rühm's numeric "poem" on Brancusi's Endless Column and Paul de Vree's syllabic recreation of Jean Tinguely's honky-tonk contraptions, Peter Maxwell Davies and Gunther Schuller focus on the humor in Klee's depiction and respond to it with wit and (musical) humor.

This case, exceptional in that three composers respond to the same painting---two in a playful way, one rather seriously---provides an opportunity to witness in miniature the transmedializing eloquence of instrumental music. I will therefore now turn to this case study and explore it in some detail.


5 Three Ways of Listening to Birds on a Crank:
Musical Interpretations of Paul Klee's Witty Criticism of Modern Culture

Paul Klee is known to many musicians as the artist with the fermatas and the many musically suggestive titles ("The Pianist in Distress," "The Literary Piano," "The Order of High C," "Old Sound," "A Master Must Pass Through a Bad Orchestra," "Fugue in Red," "Drawing in Two Voices," "The Canon of Color Totality," and scores of wordings reminiscent of music-theoretical books). And he may well be the envy of many a musicologist on account of his "Graphic Translation of a Three-part Passage from J. S. Bach."

Among the many composers who have been inspired by Klee's very "musical" paintings, the three most prominent---the American Gunther Schuller, the Englishman Peter Maxwell Davies, and the German Giselher Klebe---have all gained particular popularity with just these ekphrastic works, and specifically with the rendering of the Swiss artist's vision of a mechanized bird concert. Klebe's Die Zwitschermaschine created a stir when it was premiered at the first postwar Donaueschingen Festival for New Music in 1950; Davies's Five Klee Pictures (1959-62) alerted the musical establishment in England to the then young grammar-school music teacher; and Schuller's biographers continue, despite his amazingly prolific output during the forty years since, to speak of his Seven Studies on Themes of Paul Klee (1959) as his most widely received work. To judge from the literature, the three contemporaries never crossed paths; most certainly, their compositions originated independently of each other.

Klee created Die Zwitschermaschine in 1922, developing an idea he had first expressed a year earlier in the ink drawing, Konzert auf dem Zweig (Concert on the Twig). The miniature, deceptively childlike and innocently witty at first glance, can be interpreted on many levels, as the art-historical literature growing around it confirms. Even H. W. Janson's brief description in History of Art (1962: 527) suggests something of the depth expressed in Klee's sparse design. "With a few simple lines," Janson writes, "Klee has created a ghostly mechanism that imitates the sound of birds, simultaneously mocking our faith in the miracles of the machine age and our sentimental appreciation of bird song. The little contraption (which is not without its sinister aspect: the heads of the four sham birds look like fishermen's lures, as if they might entrap real birds) thus condenses into one striking invention a complex of ideas about present-day civilization."

The four birds, along with the slightly wavy twig on which their thin legs are perched, are almost identical with those Klee drew in Concert on the Twig. Each consists of a stylized head with an open beak and a single eye, a stick-figure body and corresponding single-line legs. Weird as their appearance may be, one can distinguish four different poses. Are these different temperaments? characters? reactions to the mechanism that supposedly drives them? From their beaks protrude variously shaped caricatures of a tongue, easily interpreted as symbols for the different nature of their utterances.

Beginning from the left, the first bird, whose body is stretched tall with the head cocked backwards, emits a vertical form that could be mistaken for the handle of a spoon, were it not for the precursor drawing in which a round black dot, placed into the corner of the beak and thinly connected with the "handle," reveals an exclamation point. With this punctuation mark and the overall body posture, complete with the open eye and the neatly feathered crest, the bird seems self-possessed and assertive. It is also the only one among the four that has a tail, albeit of wispy hair-lines only, which is seen floating in the air in a rather relaxed manner. The second bird from the left comes to little more than half the size of the first. This is partly due to the fact that its legs are spread wide, as if in a desperate attempt not to lose its footing (a fear caused, presumably, by the whirling motion to which the twig is subjected). While the first bird is assertive, this one is despondent. It hangs its head straight down, with the crest falling over its open beak. The utterance emerging from its throat is depicted as a curled shape, limp and listless, as if wanting to coil itself back into the head. The third bird, taller again, looks sideways and slightly down, away from the crank that threatens all of them. The black thread of its "song" is curved upward, combining the cunning of a fish hook with a lack of direction. Its crest spirals downward, surrounding the vertical stick that represents its body as if the bird were hugging itself to give itself courage. Where the eyes in the other three birds are small, thin black circles, this creature shows a hugely oversized eye, white with a dark, anxiety-stricken pupil. The fourth bird is portrayed with yet another character. As Maurice Shapiro (1968: 68) describes it so well, it "faces forward, to the right, advancing. His head is taut and compact; his crest streams out behind in rays; his eye is bright and directed. From the tightly drawn mouth there extends a needle-sharp barb, two-ply, machined like a ratch, which is aimed in the direction of the threatening dark mass that moves in from the upper right of the picture; faced by this formidable creature, the clouds open and a space clears on the extreme right. If this demoniacal woodpecker sings, the notched barb in his throat will sound more rattle than twitter."

If these are the birds that are to produce the twittering, what exactly is the machine and how does it activate them? Klee limits himself to the merest hints: a crank at the right-hand side, with a handle large enough to appear commanding, if not threatening, attached to a thin horizontal thread that intersects at various points with the curved twig but ends in a wheel fastened to nothing that the viewer can detect. Below this arrangement, there is a rectangular shape that, to this beholder, seems to double as a rudimentary stage for the bird concert and as a safety net should one of them fall. The strange object seen to the left of the group of birds will represent a music stand when perceived in connection with the stage, or else appear as a device pertaining to the machine, both supporting the thread that operates the wheeling motion and, with its harpoon-like points, keeping the birds confined from any sideward escape.

There are, then, at least three rather different ways of reading Klee's painting. If one chooses not to regard the stick-figures as deliberately pitiful, one could see a witty drawing of a machine that attempts to use birds' voices for a concerted action, driven by a crank, and thus controlling the speed and perhaps the volume of the twittering. This reading need not be entirely pessimistic; there would be nothing that prevents the individual birds from varying their tunes as they are used to doing; nor would there be any reason why the birds, once they have recovered from their shock after the forced motion has stopped, might not resume their singing without any input of the "machine." In another interpretation, the denatured birds, reduced to little else than the beaks that are to produce predictable sounds, can be read as Klee's critique of a technological age that has sacrificed nature and its original music to the cracking sounds of soulless machines. This reading might include a sociopolitical component; most likely it would imagine the birds as deprived by an anonymous power of their liberty and self-determination. Finally, one could focus on the four differently characterized creatures, seeing them as epitomes of four ways of reacting to the threat of nonsensical automatization. In this scenario, the crank, or the fact that what Klee draws is really a manually operated contraption and as such subject to human whim, failure, overexcitement, tiredness, etc., may not be so central as the possible attitudes towards any kind of impersonal interference with basic freedom.

Peter Maxwell Davies conceived his orchestral composition, Five Klee Pictures, for the students of Cirencester Grammar School in Gloucestershire, whose director of music he was from 1959-1962; he later revised and expanded particularly the third movement, "The Twittering Machine." With an instrumentation now adjusted to the possibilities of professional orchestras, the revised work was first performed in 1976 (see Davies 1978).

Three features are immediately striking in this piece. In terms of texture, one distinguishes three levels: a primary ostinato (a figure that stubbornly repeats itself over and over) which is heard throughout without pause, a pair of voices adding two secondary ostinati, present in all but the very last measures, and twenty further voices. In terms of notation, improvising parts exist on each of the three levels side by side with parts in fixed notation. In terms of emotional structure, the sixty-four measures of the composition are laid out as a protracted increase in volume and speed during the initial three quarters, followed by a sudden hush when the original tempo and volume is restored, and concluding ever more slowly, softly, and thinned out.

The primary ostinato is presented by trombones and low strings. Before a background of ascending triads in regular quarter-notes, the solo cello alternates with the first trombone in presenting a curved figure of eight eighth-notes. Each figure ascends from its first note through flattened steps to the diminished fifth and redescends through two raised pitches, with the result that it covers all seven semitones within the tritone. Since this is later complemented, in each of the two partner instruments, by the cluster between the tritone and the octave, the primary ostinato voices really make use of all twelve semitones. Such "all-encompassing" circular shapes seem to paint sonic images of wheels. This impression is reinforced when it becomes clear that each of these figures, exceeding a single three-four measure and concluding on the first beat of the next, overlaps with the beginning of the complementary figure---passing on the motion as one cog-wheel to another, an ingenious musical image of the mechanical wheel in the contraption depicted by Klee.

The main ostinato in low strings and trombones is soon joined by a secondary ostinato presented in the piano and four horns. However, no sooner have listeners understood that these are ostinati, i.e., unchanging, stubbornly repeated figures, than the piano, soon followed by the strings and the trombone, begins improvising on its figure; only the horn remains "obstinate" in its repetition. The same happens when the ostinato groups are joined, in irregular but well-spaced succession, by the remaining strings and winds as well as by five percussion instruments: half of them barely present a basic form of their utterance before they begin exploring joyously free and ever different renditions. During the eight measures leading up to the fff climax in m. 48, twelve voices are engaged in repeating patterns with variation of either rhythm or pitch or contour, fifteen reiterate figures without changes, and one (the bass drum) plays developmental changes specified in the notation.

After this climactic moment, the music returns to the original tempo and volume level, while the voices added to the ostinato figures are limited to new, relaxed three-note gestures or repeated trills. Soon, however, these "voices" drop out along with those of the secondary ostinato, leaving only the primary patterns to fade towards what is nevertheless heard as a sudden cessation of the mechanical motion.

One cannot speak about the musical impression of this organized chaos is, of course, given that every performance, by the same players but even more so by a different ensemble, will result in differences on many levels. Yet whatever the details in a specific interpretation, the general idea of the portrayal remains. The "twittering" we hear has all the variety and simultaneous fundamental monotony of the simple birds Klee draws. The repertoire of the voices that are added to the basic mechanism of the ostinato patterns consists of just a few pitches, repeated innumerable times with little room for variation. Heard in relation to Klee's picture, the basic ostinato patterns seem to epitomize the mechanical contraption that, in this image of modern denaturation, launches the bird calls. Once the cog-wheels are turning, they start gathering one bird call after another, gain ever more momentum and intensity, and gradually draw all voices around them into a whirling motion.

No doubt it is the machine that, with the precision of its movements holding a rein on all it initiates, keeps all bird calls metrically aligned. One wonders whether Davies was only being practical (not wanting to tax the powers of the intended first performers---and listeners---beyond their limits) when he decided to keep the composition in very traditional three-four time throughout, without any of the polymetric juxtapositions his contemporary, Olivier Messiaen, was using in the same decade for his many bird pieces. But there may well be a statement here, one not far from what Klee may have intended: that man-made machines deprive nature of its glorious freedom of expression, confining it into the narrow boxes of that which the human mind can grasp without exerting itself.

Gunther Schuller wrote his musical response to pictures by Klee in 1959, at the same time as his British colleague (Schuller 1962). Like Davies, he places "The Twittering Machine" in the center of a cycle, which he entitled Seven Studies on Themes of Paul Klee. The work was commissioned by the Minneapolis Symphony Orchestra and premiered on 27 November 1959 under the baton of Anal. Dared.

Like Davies, Schuller makes use of musical devices that suggest themselves for transmedializing Klee's vision of a contraption with wires that make bird heads sing. Ostinato figures again represent the mechanical apparatus with its ever-recurring circular motion, and the inflections of tempo---a gradual increase at the beginning, matched by a slackening and return to a more comfortable tempo for the final section---epitomize the part played by the human hand operating the crank.

But here the similarities end. Schuller's ostinato does not continue throughout the piece but rather punctuates it. The first statement spans eight measures, the second eight beats, and the third eight sixteenth-notes; in the four-four time of the piece this translates as a twofold diminution to a quarter. The three participants in the ostinato differ from one another only minutely. While the speed of their sonic attacks differs, they all play with a four-note chromatic cluster in such a way that all four pitches are heard simultaneously at all times.

During the second half of the initial ostinato passage, the representation of the mechanical device is colored with the first "twittering" noises: individual attacks in the highest register of two piccolos, flute, oboe, clarinet, and solo violins I and II, soon joined by almost equally high-pitched notes in trumpet, English horn, and solo violas. Later in the piece, once the ostinato-carrying instruments have fallen silent, these timbres will be supplemented by even stranger avian sounds, produced by a bassoon playing in the highest register and a solo violoncello playing in harmonics. The rhythm of these twitters is extremely complex, and while the volume is different for each attack, the general level increases as the ostinato recedes in diminuendo and thinned-out participation.

This twittering, we must not forget, is generated by the turn of a crank and the wires that activate (presumably artificial) avian heads. As befits music supposedly created with the help of such a contraption, the pitches are rigidly ordered along the lines of serial composition.

A "theme with two variations" built from row transformations is followed by an eight-bar phrase that is held together by a long ritardando, a slackening down to one quarter of the main tempo. During this dramatic deceleration, the entire body of musical voices descends through more than three octaves. The effect is further enhanced in terms of timbre: the brightest voices disappear and are substituted, if only for a few notes, by bass clarinet, contra bassoon, trombone, and solo bass. This plunge into the low register uses none of the tonal models introduced so far. Instead, maintaining a continuation of the complex rhythms characteristic of the bird calls, many of the individual voices now present not individual pitches but falling semitones. It is interesting that the drastically decelerating descent includes, at its slowest, falling quarter-tone segments and ends with bass notes that, according to the composer's marking, droop in little glissandi. This fact seems to suggest that Schuller has in mind a very tangible image: the twittering-machine mechanism, winding down as the hand that turned the crank is getting tired or bored, is reaching a degree of slowness where it no longer guarantees accurate action; the effect is like the out-of-tune flattening of notes on a record not keeping its turning speed.

After a brief general pause, the ostinato figure embarks on its much shorter second statement. All is contracted here; especially, the crescendo that launches the cluster figures is compressed into a single beat. During the equally abbreviated diminuendo that follows, the "bird" voices, juxtaposed with a powerful crescendo, reverse their earlier plunge and ascend rapidly, suggesting a powerful rewinding of the "bird-call" mechanism. The bird calls in their high register resume where they had left off, in such a way that the piece works its way backward from varied phrases to the initial unadorned one. The piece closes---or the machine screeches to a final halt---on what may be heard as a strongly misshapen C-major chord.

Of the three composers inspired by Klee's picture, Giselher Klebe is the only one who focuses on the individuality of the birds and their reactions as they are subjected to forced motion and enunciation. He explains his goal in the preface of the autograph (Klebe 1959): "The musical concept adopted the pictorial layout of the picture by Paul Klee and thus mounted the four 'twittering components' onto a 'machine'-like rack." Klebe's composition must thus be understood first and foremost as a four-part character sketch. The common thread that runs through the single components is certainly full of repetitive rhythmic gestures, as is to be expected in any musical portrayal of machines. However, it is not so much the contraption itself that is depicted here as the way the four creatures caught in it perceive their fate. This effect conveys itself very powerfully to a listening audience. There is a thread that holds together the "four birds" in their four symphonic movements, binding sections that differ in vivacity and mood in much the same way as the shared bondage in the teeth of the "machine" unites Klee's birds with their different temperament.

The idiosyncratic portrayal of the four characters in the composition's four sections reveals Klebe's interpretation of Klee's depiction. The first section---like the bird to the left with its erect posture and beak pointed heavenward---is assertive and self-assured. In fact, misleading as such simplistic correspondences generally are, everyone listening to the music with Klee's miniature before their eyes will be tempted to identify the slashing full-orchestra sf strokes heard throughout the Allegro as instances of the bird's exclamation-mark enunciation. The refrain-like material with its irregularly interrupted sixteenth-notes in the strings evolves a picture of anger. Its strident sounds are the result of a juxtaposition of incompatible semitones, a repeated D# in the second violins and an E-based figure in the first violins and cellos. Alternating with these straightforward calls of defiance, the motivic material shows no sign of the paucity of expression expected from beings deprived of their freedom. The first singer presents eight distinct melodic gestures and one powerful rhythmic ostinato. The motifs differ in texture from solo to homorhythmic duet and chordal setting, involve all the instruments of the orchestra, and are tonally cast in a free form of dodecaphony which, although it often reaches the full set of twelve semitones, is not confined by rules of serial sequencing. The rhythmic ostinato, presented in the percussion, is free in its own way, observing neither metric preferences nor regular spacing between recurrences. Much as this bird may be bound into the machinery of percussive sixteenths that pervade the section, it manages to maintain a large repertoire of self-expression.

In the Andante section, Klebe portrays the dejected, literally crestfallen posture of the second bird in a pattern based on a dotted rhythm, juxtaposed horizontally and often also vertically with triplets. Double basses in weary-sounding pizzicato pulsations underlie solo entries that develop into complicated details, as if the anxious bird found itself entangled in just too many wires. The duet of two trumpets with cup-mutes confirms this character portrayal in the realm of tone color as well.

The third section is deceptive in its verbal indications. Although the tempo marking suggests a moderate pace after the preceding slower one, the actual impression of the music is for long stretches one of dread-filled stasis. Here the rhythm often seems frozen, and a very limited number of harmonic soundscapes keeps recurring. When, as a contrast, the motion picks up, the sudden vivacity seems to suggest madness and is far from any joy. This is powerfully conveyed by means of a rather unusual timbral combination: an interplay of brief staccato utterings in the piano with the xylophone and the drum and another very dry percussion instrument. The dramatically enlarged eye in Klee's third bird and its strangely shredded crest find an eerie musical equivalent here.

The recapitulation of material from the first section, implied by the "tempo I" in the heading of the final section, is compressed (just as the bird on the far right is much shorter and more compact than the craning one on the far left). This reprise includes, in its first half, reminiscences from the two intermittent sections, as if this bird was briefly trying on the full array of possible reactions developed by its siblings in response to the manipulative situation. Having thus gathered strength and, apparently, worked up a righteous indignation (with a wild stretto of the initial motif and frequently renewed exclamation-mark strokes in sff), this bird converts assertiveness into raw aggression. The music concludes in an unambiguous percussive fury, as if hurling arrows into any man-made clouds that may dare to darken the skies of avian freedom.

As the brief description of these compositions by Davies, Schuller, and Klebe confirm, the three composers differ substantially in their approach to the painting and are thus ideal representatives of the various ways in which Klee's miniature may be interpreted. Davies sees a satire in which the birds never give up their efforts to outsmart the machine. True, they are bound into a wheeling motion of ever faster rotation. But, this composer is convinced, nothing, not even an attempt at mechanization, will ultimately be able to stifle their expressive freedom. Instead, as the tempo increases, an ever greater number of birds takes part in the joyous ride, contributing ever-varied flourishes or syncopated variations of their respective calls. Klebe does not seem to focus on the idea behind the painting and the question who comes out victorious, birds or human machinery. Instead, he translates the structure and concrete components, the four individualized characters as he finds them depicted so charmingly by Klee. Finally, Schuller's movement strikes me as a brilliant piece of musically couched cultural criticism. A completely mechanized set of avian squeaks is subjected to a series of transformations that only the human obsession with abstraction could invent. The musical picture is one of detached heads on wires, jerked this way and that, with a result that is as mesmerizing as so many of the high-tech contrivances characteristic of our time can be. But the birds' voices are no longer their own, as becomes sadly obvious when they go "out of tune" with the slowing down of the mechanism.





There has been ample discussion by scholars of program music whether the presence of a title is sufficient to bring a piece under the rubric, insofar as this usage seems to confuse the distinction between "a piece that expresses some emotion suggested by the title from another that either evokes its subject or (in some more concrete sense) actually attempts to describe it" (Scruton 1980). Since the authorities in the field (see, e.g., Newman 1905, Tovey 1937, 1956, Schering 1941, Lockspeiser 1973, Orrey 1975, and Scruton 1997) differ on the issue itself as well as on the allocation of individual compositions, the following list includes examples of both categories. Stock items in a catalog of programmatic compositions thus widely understood would include works from all countries of continental Europe. Musical amateurs have at their disposition sizeable collections of keyboard pieces from the English sixteenth and seventeenth centuries, anthologized for the drawing room and graced with fanciful titles; see, e.g., William Byrd's suite, Battel, whose fifteen pieces present a sequence of tableaux from idealized war, and the Fitzwilliam Virginal Book with titles like "Faire weather," "Calm whether[sic]," "Lightening," "Thunder," and "Clear day." In Germany, H. I. F. von Biber (1644[?]-1704) wrote "Mystery or Rosary Sonatas" for violin, each depicting one of the fifteen Mysteries of the Rosary, as well as many tone paintings, the most well-known being Battalia for violin solo, strings, and basso continuo, and Johann Kuhnau (1660[?]-1722) composed six "Bible Sonatas" which, he claims in the preface, depict "the fight between David and Goliath," "the melancholy of Saul being dissipated by music," "the marriage of Jacob," etc. In France, François Couperin (1668-1733) composed a musical portrait of "The Pilgrims" while Jean Philippe Rameau (1683-1764) undertook to render emotional snapshots in pieces entitled "Sighs," "Tender Plaints," and "The Joyous Girl." In Italy, Antonio Vivaldi (1678-1741) composed large canvases of the seasons in Le quattro stagioni (accompanied by four sonnets, possibly his own) as well as many other works carrying programmatic titles, such as the chamber concertos La tempesta di mare, La caccia, Sonata al Santo Sepolchro, and Per la Solennità di San Lorenzo. Half a century later, the Austrian Karl D. v. Dittersdorf (1739-99) wrote twelve programmatic symphonies on topics like Ovid's Metamorphoses and The War of the Human Passions. Several famous compositions from the period of Viennese Classicism first put programmatic works on the map of serious music, notably Haydn's Creation, Beethoven's Pastoral Symphony and Wellingtons Sieg (the "Battle Symphony"), and his Egmont, Leonora, and Coriolan overtures. However, as Wagner pointed out in an essay on Liszt's symphonic poems (Wagner 1871-1883, v), while these compositions aimed at sketching a character or telling a story in musical texture, their form was still determined almost entirely by the laws of absolute music. This apparent contradiction of content and form changed only very gradually. Beethoven's legacy determined the developments in the nineteenth and well into the twentieth century. Early on, the programmatic orchestral compositions by Hector Berlioz (1803-69), Symphonie fantastique and Harold en Italie, took off from the impressionistic example of Beethoven's Sixth Symphony; they led to many further examples of musical "nature canvases," particularly in Eastern Europe---Modest Mussorgsky (1839-81) with Ivanova noch'na lisoy gore (St. John's Night on the Bare Mountain) and Bedrich Smetana (1824-84) with Vltava (Moldau) come to mind. The "tone poems" of Richard Strauss (1864-1949) include autobiographically inspired pieces, notably Aus Italien (From Italy) and Ein Heldenleben (A Hero's Life), along with compositions based on tales of pranksters and eccentric knights (Till Eulenspiegel, Don Quixote) as well as a tone poem for which he himself wrote out a detailed synopsis that he then exactly illustrated in his music: Tod und Verklärung (Death and Transfiguration). To comment on just one: Don Quixote: Fantastic Variations on a Theme of Knightly Character, cast as a sinfonia concertante with viola and violoncello as solo instruments (roughly for Don Quixote and Sancho Panza), is a masterpiece of musical pictorialism, with hilariously realistic windmills, sheep, and flying horses, which paved the way for much of the musical pictorialism in the twentieth century. The twentieth century continues the tradition with successful compositions from Claude Debussy's La Mer (The Ocean, 1905) to Murray Schafer's Son of Heldenleben (1968) and Peter Maxwell Davies's Runes from a Holy Island (1977).

Incidentally, Buch in his study on "Beschreibungsliteratur" does not address the question of ekphrasis, i.e. of literary texts re-presenting visual texts. Like the music-lovers who deem the term program music quite enough for covering all compositions that represent something, be it the program part of the world out there or part of another artist's creative work, Buch considers "Malereigedichte" a category subordinated to the larger genre of "descriptive literature."

Catalogues of such figures of musical rhetoric comprise up to 160 distinct forms (Unger 1969 [1941]). The simplest overview suggested (Buelow 1980: 795) lists (1) figures of melodic repetition, (2) figures based on fugal imitation, (3) figures formed by dissonance structures, (4) interval figures, (5) hypotyposis figures, (6) sound figures, and (7) figures formed by silence. To give one example: the group hypotyposis comprises the figures anabasis, catabasis, circulation, fuga, hyperbole, metabasis, passaggio, transgressus, and variatio (Buelow 1980: 798). For some of the treatises containing original definitions, see Johannes Tinctoris, Complexus effectuum musices (ca. 1475), Martin Luther, Encomion musices (1538), and W. Figulus, Cantionum sacrum (1575); for just three of the many compilations undertaken in the subsequent music-historical period see Joachim Burmeister, Musica autoschediastiké (1601), Gottlieb Muffat, Florilegium (1698), Johann Mattheson, Der vollkommene Capellmeister (1739). Johann Philipp Kirnberger's treatise, "Die wahren Grundsätze zum Gebrauch der Harmonie ... als ein Zusatz zu der Kunst des reinen Satzes in der Musik was first published in 1773 (Berlin and Königsberg). An English translation by D.W. Beach and J. Thym, "The True Principles for the Practice of Harmony," can be found in Journal of Music Theory 23/2: 163-225. Connotations linked with keys and tonalities range from the simplistic "major means happy, minor means sad" to the more complex allusions said to reside in the Greek modes. To give one example: Boethius (480?-?524) tells in the prologue of his de musica (see Strunk 1965 [1950]: 82) that "the Phrygian tone, that is, the third, sung to a musical instrument, aroused one young man listening, the suitor of a certain girl, and provoked him to such rashness that he wanted to break into the girl's room at once, by force. And when the Phrygian tone was changed to Hypophrygian, that is, the third to the fourth tone, the young man calmed down, appeased by the gentleness of the tone."

David Lidov (1989) and, more recently, Robert Hatten (study in progress) explore the creation of semantic content in instrumental music through representations of the body. They believe that gestures may exploit a listener's identification with motor activity (think of the difference, when watching and hearing a violinist, of the effect created by serene bowing that flows from an undulating upper body, as against, e.g., furiously hacking attacks executed with tense shoulders and neck). In extension to purely kinetic features, a specific timbral quality created on an instrument may be linked---quasi-metaphorically---with a particular vocal grain, as if answering questions like "what kind of feeling would be expressed if this timbre were that of a human voice?"

For the purpose of my present argument, I am using "metaphor" as describing both placement and motion in auditory space and nuances of affective content. For a lucid investigation of the fuzzy boundaries and extremely varied landscapes within the territory of "musical metaphor," see Cumming 1994: 3-28.

See Leonard Meyer's definition that "Musical meaning arises when our expectant habit responses are delayed or blocked---when the normal course of stylistic mental events is disturbed by some form of deviation" (Meyer 1967: 10).

The work is listed alternatively with the explanatory subtitles hoc est, emblemata nova de secretis naturae chymicae and the longer Secretioris naturae secretorum scrutinium chymicum per oculis et intellectui: accurate accommodata, figuris cupro, emblemata, epigrammata, illustratum, opusculum ingeniis alterioribus. The music is for three unspecified voices; the emblems are engravings in copper.

This is a unique example of musical notation---albeit one appearing out of the usual context, i.e. on the title page of a composition and thus not intended for performance---that simultaneously functions as verbal text. Unfortunately, as my Israeli friends tell me, not all parts of the intended Hebrew message are entirely clear; possibly, like many European-born and bred Jews, the composer may not have known the language intimately enough. Given the historical situation, however, the suggested transliteration seems like a safe guess at Pavel Haas's intention.


Earlier versions of the reflections contained in this essay were presented at various conferences: sections 1-2 at the International Symposium on Word & Music Interactions (Bloomington, IN, February 1998), sections 3-4 at the Fifth World Conference on Word & Image Studies (Claremont, CA, March 1999), and the case study in the final segment at the Conference "Cultural Functions of Interart Poetics and Practice" (Lund, Sweden, May 2000). A developed version will appear in vol 22:3 of Poetics Today (fall 2001). The thoughts that inform this essay also appear in my monograph, Musical Ekphrasis: Composers Responding to Poetry and Painting (Hillsdale, NY: Pendragon Press, 2000), where especially the case study is analyzed in much greater detail.



Abbate, Carolyn


Unsung Voices: Opera and Musical Narrativity in the Nineteenth Century (Princeton, NJ: Princeton University Press).

Agawu, Kofi


Playing with Signs: A Semiotic Interpretation of Classic Music (Princeton, NJ: Princeton University Press).

Bailey, Walter B.


Programmatic Elements in the Works of Schoenberg (Ann Arbor: UMI Research Press).

Boethius, Anicius Manlius Severinus


De musica, modern edition in Strunk 1965 [1950]

Bruhn, Siglind


Images and Ideas in Modern French Piano Music: The Extra-musical Subtext in Piano Works by Ravel, Debussy and Messiaen (Stuyvesant, NY: Pendragon Press).


Musikalische Symbolik in Olivier Messiaens Weihnachtsvignetten (Frankfurt: Peter Lang).


The Temptation of Paul Hindemith: Mathis der Maler as a Spiritual Testimony (Stuyvesant, NY: Pendragon Press).


"Symbolism and Self-Quotation in Berg's Picture Postcard Songs," in Encrypted Messages in Alban Berg's Music, edited by Siglind Bruhn, 157-90 (New York: Garland).


Musical Ekphrasis: Composers Responding to Poetry and Painting (Hillsdale, NY: Pendragon Press).


Musical Ekphrasis in Rilke's MARIEN-LEBEN (Amsterdam: Rodopi).

Buch, Hans Christoph


Ut pictura poesis: die Beschreibungsliteratur und ihre Kritiker von Lessing bis Lukács (Munich: Hanser).

Buelow, George J.


"Rhetoric and music," in The New Grove Dictionary of Music and Musicians, edited by Stanley Sadie, vol. 15: 793-803 (London: Macmillan).

Burmeister, Joachim


Musica autoschediastiké; English translation by Benito V. Rivera (New Haven: Yale University Press, 1993).

Clüver, Claus


"On Intersemiotic Transposition," Poetics Today 10: 55-90.


"Ekphrasis Reconsidered: On Verbal Representations of Non-Verbal Texts," in Interart Poetics: Essays on the Interrelations of the Arts and Media, edited by Ulla-Britta Lagerroth, Hans Lund, and Erik Hedling , 19-34 (Amsterdam: Rodopi).


"Quotation, Enargeia, and the Functions of Ekphrasis," in Pictures Into Words: Theoretical and Descriptive Approaches to Ekphrasis, edited by Valerie Robillard and Els Jongeneel, 35-52 (Amsterdam: VU University Press).

Cumming, Naomi


"Metaphor in Roger Scruton's aesthetics of music," in Theory, analysis and meaning in music, edited by Anthony Pople , 3-28 (Cambridge: Cambridge University Press).


"Encountering Mangrove: An Essay in Signification," in Signs in Musical Hermeneutics. Special issue, Semiotics 13/1-4, edited by Siglind Bruhn, 61-102.

Dahlhaus, Carl


"Vorwort," Beiträge zur musikalischen Hermeneutik (Kassel: Bärenreiter), pp. 7-10.

Danuser, Hermann


Musikalische Prosa (Regensburg: Gustav Bosse).

Davies, Peter Maxwell


Five Klee Pictures (London: Boosey & Hawkes).

Ferruolo, Arnolfo B.


"Botticelli's Mythologies, Ficino's 'de Amore', Poliziano's 'Stanze per la Giostra': Their Circle of Love," The Art Bulletin XXXVII /1: 17-25.

Figulus, Wolfgang


Cantionum sacrum (Frankfurt a.d. Oder)

Fink, Monika


Musik nach Bildern: Programmbezogenes Komponieren im 19. und 20. Jahrhundert (Innsbruck: Edition Helbing).

Floros, Constantin


Gustav Mahler, 3 vols. (Wiesbaden: Breitkopf & Härtel).


Brahms und Bruckner: Studien zur musikalischen Exegetik (Wiesbaden: Breitkopf & Härtel).


"Das esoterische Programm der Lyrischen Suite von Alban Berg," Musik-Konzepte 4: 5-48.


Musik als Botschaft (Wiesbaden: Breitkopf & Härtel).

Gier, Albert


Das Libretto: Theorie und Geschichte einer musikoliterarischen Gattung (Darmstadt: Wissenschaftliche. Buchgesellschaft).

Griffiths, Paul


"Catalogue des Couleurs: Notes on Messiaen's Tone Colours on his 70th Birthday," Music Theory cxix: 1035.

Hagstrum, Jean H.


The Sister Arts: The Tradition of Literary Pictorialism and English Poetry from Dryden to Gray (Chicago: The University of Chicago Press).

Hanslick, Eduard


Vom Musikalisch-Schönen: Ein Beitrag zur Revision der Ästhetik der Tonkunst (Darmstadt: Wissenschaftliche Buchgesellschaft; reprint 1991).

Hatten, Robert


Musical Meaning in Beethoven (Bloomington, IN: Indiana University Press).


"Metaphor in Music," in Musical Signification: Essays in the Semiotic Theory and Analysis of Music, edited by Eero Tarasti, 373-391 (Bloomington, IN: Indiana University Press).

Heffernan, James


Museum of Words: The Poetics of Ekphrasis from Homer to Ashbery (Chicago: University of Chicago Press).

Hofstadter, Douglas R.


Gödel, Escher, Bach: An Eternal Golden Braid (New York : Basic Books; 1999).

Janson, H. W.


History of Art (Englewood Cliffs, NJ: Prentice Hall).

Jørgensen, Johannes


"God's Singer," in Saint Francis of Assisi (Garden City, NY: Image Books), pp. 127-217.

Karas, Joza


Music in Terezín 1941-1945 (New York: Beaufort Books).

Karbusicky, Vladimir


Grundriß der musikalischen Semantik (Darmstadt: Wissenschaftliche Buchgesellschaft).


"Signification in Music: A Metaphor?," in The Semiotic Web, edited by Thomas A. Sebeok and Jean Umiker-Sebeok, 430-444 (Berlin/New York: Mouton de Gruyter).


Sinn und Bedeutung in der Musik: Texte zur Entwicklung des musiksemiotischen Denkens (Darmstadt: Wissenschaftliche Buchgesellschaft).

Kibédi Varga, Alexander


"Criteria for Describing Word-and-Image Relations," Poetics Today 10: 31-53.

Kirnberger, Johann Philipp


"Die wahren Grundsätze zum Gebrauch der Harmonie ... als ein Zusatz zu der Kunst des reinen Satzes in der Musik" (Berlin and Königsberg); an English translation by D.W. Beach and J. Thym, "The True Principles for the Practice of Harmony," can be found in Journal of Music Theory 23/2: 163-225.

Klebe, Giselher


Die Zwitschermaschine: Metamorphose über ein gleichnamiges Bild von Paul Klee, für Orchester, op. 7 (Berlin and Wiesbaden: Bote & Bock)

Kranz, Gisbert


Das Bildgedicht: Theorie, Lexikon, Bibliographie, 3 vols. (Cologne: Böhlau).

Krieger, Murray


"Ekphrasis and the Still Movement of Poetry, or Laokoon Revisited," in The Poet as Critic, edited by Frederick P. W. McDowell, 3-26 (Evanston, IL: Northwestern University Press).


Ekphrasis: The Illusion of the Natural Sign (Baltimore: The Johns Hopkins University Press).

Langer, Susanne K.


Philosophy in a New Key: A Study in the Symbolism of Reason, Rite, and Art (Cambridge, MA: Harvard University Press; reprint 1957).


Feeling and Form: A Theory of Art Developed from Philosophy in a New Key (London: Routledge & Kegan; reprint 1967 ).

Lessing, Gotthold Ephraim


Ausgewählte Werke. Vol. 6 Religiöse Streitschriften, Aus dem Nachlaß, Briefe, Würdigungen (Hamburg: Standard-Verlag).

Lidov, David


"Mind and Body in Music," Semiotica 66/1: 69-97.

Lockspeiser, Edward


Music and Painting: A Study in Comparative Ideas from Turner to Schoenberg (London: Cassell).

Lund, Hans


Text as Picture: Studies in the Literary Transformations of Pictures, translated by Kacke Götrick (Lewiston, NY: Edwin Mellen). First published in Swedish as Texten som tavla: Studier i litterär bildtransformation [Lund: LiberFörlag, 1982]).

Luther, Martin


Encomion musices (Wittenberg; for a modern reprint see, e.g., Walter Blankenburg, "Überlieferung und Textgeschichte von Martin Luthers Encomium musices," Luther-Jahrbuch xxxix (1972): 80

Mattheson, Johann


Der vollkommene Capellmeister (Hamburg; facsimile reprint Kassel : Bärenreiter, 1954).

Messiaen, Olivier


La Technique de mon langage musical (Paris: Alphonse Leduc).

Meyer, Leonard B.


Emotion and Meaning in Music (Chicago and London: The University of Chicago Press).


Music, the Arts, and Ideas: Patterns and Predictions in Twentieth-Century Culture (Chicago, IL: University of Chicago Press).

Mirka, Danuta


"Colors of a Mystic Fire: Light and Sound in Scriabin's Prometheus," in Signs in Musical Hermeneutics, Special issue of The American Journal of Semiotics [Vol. 13, nos. 1-4]: 227-248, edited by Siglind Bruhn (New York: Garland).

Mitchell, W. J. T.


"Ekphrasis and the Other," South Atlantic Quarterly 91: 695-719.


Picture Theory: Essays on Verbal and Visual Representation (Chicago, IL: University of Chicago Press).

Monelle, Raymond


Linguistics and Semiotics in Music (Chur: Harwood Academic).


"Music and Semantics," in Musical Signification: Essays in the Semiotic Theory and Analysis of Music, edited by Eero Tarasti (Bloomington, IN: Indiana University Press): 91-107.

Moore, Earl V.


The Symphony and the Symphonic Poem (Ann Arbor: Ulrich's Books).

Müller, Karl-Josef


Mahler: Leben. Werke. Dokumente (Mainz: Schott / München: Piper).

Muffat, Gottlieb


Florilegium; modern edition by Heinrich Rietsch (Vienna: Universal Edition, 1959).

Newcomb, Anthony


"Once More: 'Between Absolute and Programme Music': Schumann's Second Symphony," 19th Century Music 7: 233-50.


"Agencies/ Actors in Instrumental Music: The Example of Mahler's Ninth Symphony, Second Move-ment," in Musik als Text: Kongreßbericht Gesellschaft für Musik Freiburg 1993, vol. 1: 55-61 (Kassel: Bärenreiter).

Newman, Ernest


"Programme Music," in Musical Studies (London: John Lane), pp. 103-188.

Orrey, Leslie


Programme Music: A Brief Survey from the Sixteenth century to the Present Day (London: Davis-Paynter).

Raftis, Alkis, ed.,


Dance and poetry: International Anthology of poems on dance (Princeton: Princeton Book Co.).

Rousseau, Jean-Jacques


Essai sur l'origine des langues, où il est parlé de la mélodie et de l'imitation musicale (Paris)


Dictionnaire de musique (first published Paris: Chez la veuve Duchesne); English translation by William Waring (New York: AMS Press, 2nd edition 1975).

Russ, Michael


Musorgsky: Pictures at an Exhibition (Cambridge, UK: Cambridge University Press).

Saintsbury, George


A History of Criticism and Literary Taste in Europe (New York 1902); quoted from Hagstrum 1958: 18.

Schering, Arnold


Das Symbol in der Musik (Leipzig: Köhler & Amelang).

Schlosser-Magnino, Julius


La letteratura artistica: Manuale delle fonti della storia dell'arte moderna (Florence, Nuova Italia).

Schoenberg, Arnold


"Analysis: Pelleas and Melisande," album notes (Capitol recording).

Schopenhauer, Arthur


The World as Will and Representation, trans. E. F. J. Payne (New York: Dover).

Schuller, Gunther


Seven Studies on Themes of Paul Klee (London: Universal Edition; facsimile study score).

Scott, Grant F.


"The Rhetoric of Dilation: Ekphrasis and Ideology," Word & Image 7: 301-10.


The Sculpted Word: Keats and Ekphrasis (New York: University Press of New England).

Scruton, Roger


"Programme Music," in The New Grove Dictionary of Music and Musicians, edited by Stanley Sadie, vol. 15: 283-7 (London: Macmillan).


The Aesthetics of Music (Oxford: Clarendon Press).

Shapiro, Maurice


"Klee's Twittering Machine," The Art Bulletin 50/1: 66-69.

Smith, Adam


Essays on Philosophical Subjects (Oxford: Clarendon Press; reprint 1980).

Smith, Mack


Literary Realism and the Ekphrastic Tradition (University Park: The Pennsylvania State University Press).

Spitzer, Leo


"The Ode on a Grecian Urn, or Content vs. Metagrammar," Comparative Literature, 7 204-25.


"The Ode on a Grecian Urn, or Content vs. Metagrammar," in Essays on English and American Literature, edited by Anna Hatcher, 67-97 (Princeton: Princeton University Press).

Steiner, Wendy


The Colors of Rhetoric: Problems in the Relation between Modern Literature and Painting (Chicago: University of Chicago Press).

Sternberg, Meir


"Proteus in Quotation-Land: Mimesis and the Forms of Reported Discourse," Poetics Today 3/2: 107-56.

Strunk, Oliver


Source Readings in Music History (New York: Norton; 1965).

Tardieu, Jean


Figures (Paris: Gallimard).

Tinctoris, Johannes


Complexus effectuum musices; English translation by Albert Seay in Corpus scriptorum de musica 22 (American Institute of Musicology, 1975).

Tovey, Donald Francis


Essays in Musical Analysis, vol. iv: Illustrative Music (London: Oxford University Press).


"Programme Music," in The Forms of Music (New York: Meridian Books), pp. 167-174.

Unger, Hans-Heinrich


Die Beziehungen zwischen Musik und Rhetorik im 16.-18. Jahrhundert (Hildesheim: Olms; 1969).

Wagner, Peter


Icons - Texts - Iconotexts: Essays on Ekphrasis and Intermediality (Berlin and New York: Walter de Gruyter).

Wagner, Richard

1871-73, 1883

Gesammelte Schriften und Dichtungen (Leipzig: Fritzsch).

Walton, Kendall


"What is Abstract About the Art of Music," The Journal of Aesthetics and Art Criticism, XLVI/3: 351-364.


Mimesis as Make-Believe: On the Foundations of the Representational Arts (Cambridge, MA: Harvard University Press).


"Listening with Imagination: Is Music Representational?" The Journal of Aesthetics and Art Criticism 52/1: 47-61.

Whiting, Steven Moore


Satie the Bohemian: From Cabaret to Concert Hall (Oxford : Oxford University Press).

Yacobi, Tamar


"Pictorial Models and Narrative Ekphrasis," Poetics Today 16: 599-649.


"The Ekphrastic Model: Forms and Functions," in Pictures Into Words: Theoretical and Descriptive Approaches to Ekphrasis, edited by Valerie Robillard and Els Jongeneel, 21-34 (Amsterdam: VU University Press).