Dramaturgy of Desire:
An Analytical Approach to Dance Narratives.

Originally published in Choreoglogica Journal of the European Association of Dance Historians, Autumn 2006, Vol 2, no 1. 

Copyright Astrid Bernkopf 2006

Present-day dance research allows analytical investigations into various features and elements of the performance. However, a dogmatic and ideological attitude towards research topics and their treatment can be observed. Movement and choreography are the most prominent themes of research, followed by the body and issues relating to the performer. Such selective view has catered for one essential feature of theatre performances to be overlooked: namely, the stories that are told onstage. Hence, although dance analysis has developed research tools to study various elements, there is a lack of dance specific theory relating to danced narratives. Through this theoretical vacuum, the commonly accepted and traditionally upheld focus on movement and choreographic analysis has been further cemented in the discipline. 

The aim of this article is to introduce a method to investigate the narrative of the Romantic two-act ballet fantastique. This model of analysis may stand alone or may work in co-operation with other analytical tools. Theories on fairy tales and modes of storytelling in literature and theatre form the theoretical basis for the construction of this dance-orientated method of narrative analysis. Such approach stems from the erroneous perception/notion of the nineteenth-century ballet plot as a danced fairy tale. Moreover, the constant repetition of stories, and therefore actions, suggests that it may be treated in similar ways as the fairy tale. Consequently, a literary model of narrative analysis created for the fairy tale constitutes the basis of this method of inquiry. 

In his seminal work Morphology of the Folktale (2003), Russian scholar Vladimir Propp follows the general trend of Russian Formalism and investigates literary narratives. In Propp’s case, these narratives are magic tales, which, defined by seven distinct features, constitute a sub-category of the fairy tale. Magic tales display a supernatural bride or groom, a supernatural enemy, a marvellous task, supernatural helper, supernatural abilities or knowledge, magical tools and objects or any other supernatural or marvellous features, abilities and occurrences. A corpus of research was found by Propp (2003, p. 23 – 24) in one hundred magic tales presented in Afansiev’s collection that clearly belong to this category of the fairy tale. 

Concerning the structure of these tales, Propp states that the magic tale consists of thirty-one recurrent actions, which he defined as functions. These functions are main or key actions, which directly advance the story (ibid., p. 21). Minor details of these actions are discarded completely. Hence, it is important that the fairy tale’s hero receives a present and not what kind of gift this is. One significant feature of this method of structural investigation is, thus, that the functions are recurrent elements, which may change in their execution or appearance (ibid., p. 20). Therefore, an individual function is not bound to one or the other representation, but, although being a recurrent element of the fairy tale, it is at the same time in flux and variable. 

Nineteenth-century ballet tradition offers a corpus of research similar to the one Propp investigated. The ballet fantastique can be seen as close to the magic tale, since it confronts its audience with the same features that constitute the magic tale. Besides depicting supernatural occurrences and fairy tale-like elements, the ballet fantastique follows the standard plot treatment of the nineteenth-century ballet performance and displays the conventions predominant in ballet tradition. The time between 1830 and 1860 is generally considered as the age of the ballet fantastique, during which it grew and conquered the entire European continent, thus leading to national styles that nevertheless follow the most distinct conventions of the genre. Although works displaying the most significant features of the ballet fantastique can also be found outside this timeframe, such restricted period gives a compacter corpus of research that, despite its limitations, reveals changes in storytelling strategies during the three decades. A problem concerning these materials was, however, found in their composition in one, two or three acts. Such variety did not prove beneficial for a comparison of many ballet plots. Therefore, a focus on two-act productions only is followed throughout this study. A final limitation of source materials can be found in the complete reliance on the ballet scenario. As film or video recordings of these ballet performances are not at all available, the libretto stands in for the performance by giving a detailed account of the plot. Hence, reference to performance should be understood as relating to the performance as read and constructed from the libretto and not the live or recorded stage performance. Within financial and time constraints, it was possible to gather 70 scenarios following the criteria outlined above that constitute the main corpus of research. 

An analysis of these materials according to the rules Propp established resulted in the functions not following the order Propp has outlined. Moreover, some functions did not appear at all in any of the sources. On the other hand, however, other actions were constantly repeated in the ballet plot, therefore suggesting the presence of independent ballet-specific functions.

An investigation into the recurrent actions of the ballet plot has led to the definition of three types of functions. Firstly, a group of functions remains true to the definition Propp has attached to them and have merely been transcribed to fit their representation on the ballet stage. A second group has undergone slight changes, which are reflected in the differing definitions these received. However, the main idea or notion behind these functions still remained similar to that discovered by Propp for the fairy tale. Hence, their titles were kept, although their characterisation varies. The third group comprises functions that do not appear at all in Propp’s list. These functions were given definitions and are considered as ballet-specific functions. Although it could be argued that some of them might as well be present in the fairy tale and have been overlooked by Propp, this line of reasoning is not followed in a study focusing on the ballet narrative. All functions transcribed and defined for ballet tradition will from now on be labelled as ballet functions. 

One feature of the functions defined by Propp (2003, p. 21) that has not yet been mentioned is that any character may execute any of the functions. In a similar manner, the functions of ballet tradition are not connected to the characters of the ballet plot and may be found in the range of action of all characters. Additionally, their appearance is not bound to any particular enactment onstage. Therefore, like their colleagues of the fairy tale, the group of ballet functions can take on various guises. Moreover, they are not bound to any hierarchical presentation and may appear at any point within the plot. This is easily observable when considering the Wedding, which is the traditional end of the fairy tale and so appears last in Propp’s list. In ballet tradition, some productions set the Wedding at the very end of the plot as it can be seen in Der Kobold (librettist not stated, 1838), La Filleule des Fées (lib. Saint-Georges/Perrot, 1849) and Sacountala (lib. Gautier, 1858). Contrary to the trend set in these ballets, others such as La Fille du Danube (lib. F. Taglioni, 1836) and Orfa (librettist not stated, 1852) begin their course of action with the Wedding. The same applies for all other functions. These findings suggest that the venture into the recurrent actions of the ballet plot does not lead to a structural order of functions as the one Propp described for the fairy tale. 

So far, the storytelling strategies of the ballet fantastique have been compared to a literary form. Theatre has, in its course, developed its very own and distinct storytelling devices that also influenced the composition of the ballet plot. Theatre conventions can be found in all theatre arts and find their origins in Greek theatre tradition. In this sense, an influential treatise can be found in Aristotle’s Poetics (2005). Here, Aristotle describes the structure of a performance as consisting of an exposition, knotted middle section, péripétie and dénouement. Aristotle’s influence has not only left its imprints on storytelling and performance composition in drama and opera, but has also shaped the earliest attempts of independent storytelling in the field of dance. 

Following suit, French theorist and choreographer Jean-Georges Noverre states in his Lettres sur la Danse et sur les Ballets (1760) that a ballet must have “son exposition, son nœud & son dénouement” [its exposition, its entanglement & its resolution] (Noverre, 1760, p.20). This notion applies, according to Noverre not only to the entire plot, but should be the structural outline of each act and scene too. Italian choreographer Gasparo Angiolini also underlines the importance of Aristotle’s teachings by stating a ballet should have “un principo [sic], un mezzo, un fine” [a beginning, a middle, an end] (Angiolini, 1773, p. 14). Despite their querelle on who the true inventor of the ballet d’action was, in this particular point, an agreement between Angiolini and Noverre can be detected that shows how close the theatre arts were interrelated and influenced each other. Therefore, it can be said that these elements constitute theatre conventions, which regulate certain parts of the plot and provide a structural frame within which the story unfolds.

The exposition introduces the audience to the fictional world of the performance and displays the most important characters by conveying background information about them. Here, a whole set of functions is used to set up the future conflict between the characters. In the middle section, an intensification of the conflict is brought about, and the main characters are caught in a web of intrigues. Like the exposition, the middle section is composed of an entire range of functions. With the sudden and surprising reversal of action, the péripétie, the good fortunes of the main characters change through the open outburst of the conflict. At this point, all seems lost and the evil forces of the plot appear to win. Again, this part of the performance is presented to the audience through several functions. In some cases, a repetition of actions with increasing intensity takes place during the péripétie. This can be observed in Giselle (lib. Gautier/ Saint-Georges, 1841) where game-keeper Hilarion confronts Giselle three times with evidence that her sweetheart is an impostor. In the dénouement, the knot is untied, whereas the ending aims at either closure by seeking to provide a definite solution to the conflict or denies such sense of closure. Here, the last entrances take place, and in most cases all is restored to its proper place. The fictional social order that has been subverted in the course of the plot is re-established and the supernatural spirits return to their graves. 

The monologue is another theatrical feature that can be counted among theatre conventions. It is present in all theatre arts and carries very distinct messages. Information about the characters, their origins, plans and emotions is conveyed to help the audience become orientated within the plot and follow its progress. Again, as with the other conventions, the monologue does not display one particular function, but may include any function selected from the whole group. 

In addition to the monologue, narrative links provide connections between the single events of the plot. These bridge the acts and scenes and ensure a continuous flow of information. Usually, repetitions of functions serve as narrative links that establish connections by reminding the audience of events, facts or plans. In Giselle (lib. Gautier/Saint-Georges, 1841), the hunters retell the legend of the Wilis at the beginning of Act II, which, on the one hand, ties in to Giselle’s mother Berthe’s story in the first act and, on the other hand, indicates that they have entered the realm of the Wilis where the action of the ballet will continue.

Within the descriptions of the various theatrical elements one significant feature has been that not one single function constitutes one convention, but an entire range of functions. Moreover, repetitions of functions occur frequently and emphasise the problem of all functions appearing at any moment within the ballet plot. Furthermore, two of these conventions hold a specific position in this group. The monologue and narrative links are themselves not bound to a particular location within the plot as the other conventions are. Hence, they change through their differing appearance and location. 

A second group of stage conventions is represented by dramatic, operatic and choreographic conventions that are specific to each the theatre arts. These sub-conventions deal with storytelling aspects in ways particular to the genre and have their origins in the divergent nature of theatre arts and their development as independent art forms. In ballet tradition, the most apparent of these conventions is the presentation of the principal characters. As this convention is generally conveyed through a solo variation, all functions and narrative elements connected to the monologue can be found in it. 

Therefore, the interplay between mimed and danced scenes forms the first sub-convention in this category. In nineteenth-century ballet tradition, approximately the same amount of time was allocated to mime scenes and danced sequences (Poesio, 1999, p. 841; Smith, 2000, p. 175). Both conventions may convey the same narrative content and, therefore, it is a matter of choice whether one function is presented through danced or mimed movement. 

Another distinct element of the ballet performance are solo variations and sections for the corps de ballet. Generally, dances of soloists are seen as advancing the plot, whereas the divertissement engaging the entire corps de ballet is understood as causing a disruption of the ongoing action (Foster, 1986, p. 69). However, the materials suggest that divertissements and the dances of the corps de ballet carry narrative agencies that further the development of the plot and create suspense. The divertissement delays the outbreak of the conflict and is in many cases the enactment of a celebration. The audience is distracted by the dances, and the interference of the péripétie will have an even bigger impact. The solo variations, on the other hand, carry all narrative agencies of the monologue. 

One particularly well-known feature of Romantic ballet is the division of acts into one earthly and one otherworldly. With this tradition, the borders between reality and dream are blurred, and the plot is transferred to the realm of the subconscious. In this part of the performance, the dénouement traditionally takes place and the main character is tested. All this allows the plot to be untangled and a solution to the conflict may be reached. In most cases, this happens in an enchanted forest, an under water kingdom, fairyland or a fairy palace. 

All these sub-conventions represent a means through which the performance gains an individual appearance. As a result of their non-hierarchical status, they may be exchanged for each other, and no rule concerning which convention is to be preferred is detectable in the source materials. The same can be said about the functions occurring in these conventions. All functions may find their expression in danced as well as mimed form or may be used for solo variations and ensemble dances too, which once again overthrows any attempt of a linear explanation of the ballet plot. 

The final and perhaps most important element of a story are its characters. Romantic ballet tends to focus on a limited group of characters in ever-changing situations. So far, ballet characters have been divided into the “pagan dancer” (Gautier in Guest, 1986, p. 16) and “Christian dancer” (Gautier in Guest, 1986, p. 15) as Fanny Elssler and Marie Taglioni have been described. Another distinction would be that of the “robust, earthly, vivacious foreigner” (Foster, 1998, p. 200) which separates the characters according to their nationality. 

When focusing on the narrative and the narrative agency characters have in the ballet plot, a different distinction arises through five types of characters that are presented in differing combinations. Firstly, ballet shows the Hero, who yearns for a partner and has to prove himself worthy of his lover. The Unthreatening Woman displays the docile side of femininity and is the one and only correct choice for the Hero. She is a shy, industrious girl who does not indulge in sensual pleasures. Her counterpart, the Threatening Woman, lives an active and sensual life by hunting men on moonlit forest clearings. Her over-indulgence, activity and sexuality almost inevitably lead to her death; or at least to separation from the Hero. With the Parental Figures, the fictional society of the ballet narrative has the equivalent to parents who seek to protect their offspring from harm. Mothers, fathers, foster parents and fairy godmothers side with the lovers in their quest for a relationship. In most cases, all their attempts are in vain, and the lovers have to face their fate. However, contrary to this trend, in some ballets the Parental Figures are the ones who test the lovers and so ensure a happy end. The most important characters of the narrative are the Trickster Figures. These mischievous characters, be they male or female, are the ones who through their evil deeds and intrigues drive the lovers into the catastrophe. Thus, without the constant interferences of the Trickster Figures, the plot would lose momentum. A noteworthy feature is that the Trickster Figures do not only represent the antagonists of the lovers, but also show a counterweight to the Parental Figures whose aim is to unite the lovers.
These findings lead to the observation of ballet characters being close to those of Commedia tradition. Like the maschere of Commedia Dell’Arte, the most basic characterisation of these five types does not change from one narrative to the other. The difference between various narratives is the combination in which the characters are presented. Whereas in one ballet the Hero loves the Unthreatening Woman, he follows the Threatening Woman in the next. Furthermore, the personal descriptions of characters vary and so no Unthreatening Woman is exactly like the other due to different situations, background and personal stories. Moreover, it is observable that all functions may be executed by any of the characters. The actions are, thus, not part of the range of action of one or the other character. 

In regard to the model Vladimir Propp (2003) introduced for the fairy tale, it can be said that the actions of the ballet plot do not offer a linear structural solution. Conversely, the ballet narrative can be seen as composed of three very distinct layers. The first of these is constituted by the theatrical devices employed to convey the narrative. These have been identified in the group of theatre conventions and the sub-conventions of dramatic, operatic and choreographic traditions. Within this layer, a division between one macrostructure and one microstructure is discernable. The macrostructure’s elements form a framework for the microstructure. 

Macrostructural elements are the theatre conventions as represented by exposition, middle section, péripétie, dénouement and end. They appear in this fixed order and do not tend to shift place in nineteenth-century ballet tradition. As such, these features can be considered as constants of the ballet narrative. One feature of these constant elements that has not yet been considered is their duration. Whereas in Giselle (lib. Gautier/Saint-Georges, 1841) the exposition, middle section and péripétie take place during the first act, La Fille de Marbre (lib. Saint-Léon, 1847) presents these conventions until well into the second act. With such treatment an emphasis on either the first part of the narrative or the second part after the péripétie can be achieved. This also entails that the constants are not fixed and have to be considered variants at the same time. 

The elements of the theatrical microstructure can be found in the choreographic conventions and the monologue. They are not bound to any order and can appear at all points within the plot. Through these devices, the appearance of the acts or scenes is created from an interplay between the various conventions. Therefore, it is possible that one ballet presents a function as danced sequence, whereas another performance displays the same situation as mimed scene. As these conventions are another recurrent element of the ballet narrative, their simple presence results in them becoming constants, whereas their flexibility constitutes their variable nature. With such theatrical layout of the performance an individual sequence of conventions and their duration can be created for each ballet.
The next layer in this analytical model is formed of narrative conventions. These may again be divided into narrative macrostructure and microstructure. As literary studies have adopted Aristotle’s Poetics (2005) to investigate literature (Chatman, 1980, p. 47; Barry, 2002, p. 21 – 22; Jahn 2002, N2.1.4; Abbott, 2004, p. 53,), the conventions constituting the theatrical macrostructure are at the same time the features of the narrative macrostructure. Hence the double agency of this layer that keeps its constant and variant character for the narrative side of the performance. The narrative microstructure is formed by the arrangement of functions that gives each act or scene its individual appearance. As recurrent elements of the ballet narrative, the functions of Romantic ballet provide another set of constant elements. However, due to them not being bound to any hierarchy and thus appearing at any point in the plot, the functions are yet another set of constant variants.

Another implicit third narrative layer can be found in the characters and their conflict. The conflict as the underlying notion of the plot is created from the clashing interests of the characters. Each character strives to achieve or gain something. In the case of the nineteenth-century ballet plot, this generally is love and a relationship. However, through moral implications this goal is not always achieved. The conflict as the overarching notion can, thus, be considered as providing another macrostructure. The characters and their individual characteristics, emotions and desires can be seen as representing a microstructure each. The entanglements between the characters result in the individual presentation of a conflict in a particular ballet. As these conflicts deal with love and the trials of lovers, this thematic choice is another constant feature of the Romantic ballet plot. By following the example of Commedia Dell’Arte and presenting the characters in ever changing situations, the narrow focus on one thematic range nevertheless allows for variety and is, therefore, another shifting constant. The stock characters, on the other hand, are through their presence in the ballet plot constant elements, but in as far as they are portrayed in differing ways and combinations variants at the same time.

Having outlined the nineteenth-century ballet narrative as consisting of three distinct narrative layers composed through shifting constants, it remains to be said that with this analytical approach a first step towards narrative analysis in the field of dance analysis has taken place. The method does not represent an application of a pre-existing model, but has been conceived for theatre and its purposes. The single elements outlined in this article may be arranged in alternating ways as to suit the case study and account for the difference, variety and individuality found in the ballet scenario. Thus, it is possible to change the approach from one case study to the other without major problems. Additionally, one element may be left out or a focus on either narrative or theatrical structure can be followed. It is, therefore, that the Dramaturgy of Desire emerges through application.

Select Bibliography:

Abbott, H. [no further name given] Porter. The Cambridge Introduction to Narrative. Cambridge: Cambridge University Press, 2004.
Adshead, Janet (ed.). Dance Analysis. Theory and Practice. London: Dance Books, 1988.
Angiolini, Gasparo. Lettere di Gasparo Angiolini a Monsieur Noverre sopra i Balli Pantomimi. Milan: Gio. Battista Bianchi, 1773.
Aristotle. Aristotle’s Poetics. Accessed 09. May 2005.
Barry, Peter. Beginning Theory. An Introduction to Literary and Cultural Theory. Manchester/New York: Manchester University Press, 2002.
Beaumont, Cyril William. Complete Book of Ballets. London: Putnam, 1937.
______. The Ballet called Giselle. London: Dance Books, 1969.
Belsey, Catherine. Desire. Love Stories in Western Culture. Oxford/Cambridge: Blackwell, 1994.
Canton, Katia. The Fairy Tale Revisited. A Survey of the Evolution of the Tales, from Classical Literary Interpretations to Innovative Contemporary Dance-theater [sic.] Productions. Bern: Peter Lang, 1996.
Chatman, Seymour. Story and Discourse. Narrative Structure in Fiction and Film. New York: Cornell University Press, 1980.
Cobley, Paul. Narrative. The New Critical Idiom Series. London/New York: Routledge, 2004.
De Marinis, Marco. The Semiotics of Performance. Bloomington/Indiana: Indiana University Press, 1993.
Duchartre, Pierre Louis. The Italian Comedy. The Improvisation Scenarios Lives Attributes Portraits and Masks of the Illustrious Characters of the Commedia dell’ Arte. New York: Dover Publications, 1966.
Elam, Keir. The Semiotics of Theatre and Drama. London/New York: Routledge, 2002.
Foster, Susan Leigh. Reading Dancing: Bodies and Subjects in Contemporary American Dance. Berkeley/Los Angeles/London: University of California Press, 1986.
______. Choreography and Narrative. Ballet’s staging of Story and Desire. Bloomington: Indiana University Press, 1998.
Frye, Northrop. Anatomy of Criticism. London: Penguin, Books, 1990.
Freytag, Gustav. Die Technik des Dramas in Gesammelte Werke Band 2. Leipzig: Hirzel, [1910].
Garafola, Lynn. (ed.). Rethinking the Sylph. New Perspectives on the Romantic Ballet. Hanover/London: Wesleyan University Press, 1997.
Gozzi, Carlo. Fiabe teatrali. Milan: Mursai, 1998.
Guest, Ivor. The Romantic Ballet in Paris. London: The Pitman Press, 1966.
______. The Romantic Ballet in England. Its development, fulfilment and decline. London: Pitman Publishing, 1972.
______ (transl.). Gautier on Dance. London: Dance Books, 1986.
Horace. Ars Poetica. Translated by Leon Golden. Accessed 22 May 2005.
Jackson, Rosemary. Fantasy: The literature of subversion. London/New York: Routledge, 2003.
Jahn, Manfred. A Guide to the Theory of Drama. Accessed July 2002.
Jenkins, Keith. Re-thinking History. London/New York: Routledge, 2003.
Lämmert, Eberhard. Bauformen des Erzählens. Stuttgart: J. B. Metzlersche Verlagsbuchhandlung, 1968.
Lévi-Strauss, Claude. Die Struktur und die Form. Reflexionen über ein Werk von Vladimir Propp. in Propp, Vladimir Morphologie des Märchens. Germany [Frankfurt am Main]: Suhrkamp, 1975.
______. Structural Anthropology. Volume 1. London: Penguin Books, 1993.
Lüthi, Max. Das europäische Volksmärchen. Form und Wesen. Bern/München: A. Francke Verlag, 1960.
______. Es war einmal ... Vom Wesen des Volksmärchens. Göttingen: Vandenhoeck & Ruprecht, 1968.
Molinari, Cesare. Commedia dell’Arte. Mondadori: Milano, 1985.
Nicoll, Allardyce. The world of Harlequin. A Critical Study of the Commedia dell’Arte. Cambridge: Cambridge University Press, 1963.
Noverre, Jean-Georges. Lettres sur la Danse et sur les Ballets. Lyon: Aimé Delaroche, 1760.
Pfister, Manfred. Das Drama. Theorie und Analyse. München: Wilhelm Fink Verlag, 2001.
Poesio, Giannandrea. Dance Study Aids Part 8. Mime Acting in The Sleeping Beauty. In Dancing Times June 1999, Volume 1065, Number 89, p. 839 – 842.
______. Reviving the Gesture. In Dancing in the Millennium, An International Conference, Proceedings. Washington DC: George Washington University, July 19 – 23. 2000, p. 339 – 341.
Pontremoli, Alessandro. Drammaturgia della danza. Percorsi coreografici del secondo Novecento. Milano: Euresi Edizioni, 1997.
Praz, Mario. Liebe, Tod und Teufel. Die schwarze Romantik. Band 1. München: Deutscher Taschenbuch Verlag, Wissenschaftliche Reihe, 1970a.
______. Liebe, Tod und Teufel. Die schwarze Romantik. Band 2. München: Deutscher Taschenbuch Verlag, Wissenschaftliche Reihe, 1970b.
Propp, Vladimir. Die historischen Wurzeln des Zaubermärchens. München/Wien: Carl Hanser Verlag, 1987.
______. Morphology of the Folktale. Second revised edition. Austin: University of Texas Press, 2003.
Smith, Marian. Ballet and Opera in the Age of Giselle. Princeton/Oxford: Princeton University Press, 2000.
Stanzel, Franz K. Theorie des Erzählens. Göttingen: Vandenhoeck und Ruprecht, 2001.
Swann Jones, Steven. The Fairy Tale. The Magic Mirror of Imagination. New York/London: Routledge, 2002.
Todorov, Tzvetan. The Fantastic: a structural approach to a literary genre. New York: Cornell University Press, 1975.
Zimmermann, Bernhard. Die Griechische Komödie. Darmstadt: Wissenschaftliche Buchgesellschaft. 1998.
Zipes, Jack. Fairy Tales and the Art of Subversion. The Classical Genre for Children and the Process of civilization. London: Heinemann, 1983.