PHILOSOPHICAL PROBLEMS OF DANCE CRITICISM

CHAPTER IV

THE IDENTITY OF WORKS OF ART IN DANCE

by Julie Charlotte Van Camp

Copyright Julie Charlotte Van Camp 1981

All Rights Reserved

NOTE: This dissertation may be downloaded, saved, printed, and reproduced for personal, educational, and scholarly uses, but only if this complete permissions notice and the full copyright notice are included with any such uses. This dissertation may not be sold or otherwise used for commercial purposes. For commercial use, please contact the author.

The page numbers from the original dissertation are included in the text here as follows: /p. x.

Endnotes from the original dissertation are included in the text here as follows: (x)

The dance world tolerates astonishing variation in the performance of dance works without loss of identity. Virtually every aspect of dance has been subject to considerable alteration; performances without major changes from previous ones of the same work are rare, although, significantly, increasingly less rare. Choreography is routinely altered, sometimes completely, (1) from other productions with the same name. (2) Traditional mime passages are shortened or eliminated, (3) plots altered, dramatic themes revised, (4) music inserted, (5) deleted, or re-arranged. (6) Audiences, critics, and artists alike usually find such practices quite acceptable, (7) in sharp contrast with the unthinkability of major (intentional) changes in, say, the score for Beethoven's Fifth Symphony. (8)

This chapter briefly surveys examples of the variation tolerated without loss of identity, and the differences among dance genre in the importance of different media (movement, music, etc.) in establishing identity and in the degree of variation tolerated. The inadequacy of identity theories relying primarily or exclusively on notational systems to explain identity in dance is discussed in terms of these practices. An alternative is proposed using the

test of infringement of copyright artworks, "substantial similarity," as determined by lay observers.

A. Actual Practices in Establishing Identity of Dance Works

the nineteenth-century full-evening "story-ballets," still predominant in the repertories of many companies, illustrate the great variation in performance tolerated by the dance world without loss of identity. Typical of these is Giselle. First produced in Paris in 1841, with an almost unbroken history of production, it is still in the repertory of many major ballet companies. (9) Numerous, substantial changes in choreography are well-documented, (10) and contemporary audiences demand identity only of certain well-known solos, pas de deux, and ensembles. (11) changes in the plot, manifested in different movement patterns, occurred in the nineteenth century, (12) but are not limited primarily to changes in the attitude and motivation of the major characters. (13) movements portraying dramatic details are still routinely altered, even in performances by the same company, (14) as is the traditional mime. (15) While strict identity of the choreography of much of the rhythmic movement is essential to the ballet, the non-dancing dramatic movement is not subject to such strict standards, but is considered an opportunity for dancers to interpret the roles by adding or substituting their own movement designs (normally, planned in advance of performance). As these dramatic movements (walking, etc.) can be as easily notated as pure, rhythmic dance movements,

the difference in essentiality does not reflect amenability to notation, but may reflect a view of dance as primarily movement, and only secondarily as a form of theater.

Alterations in the music of Giselle, as with other ballet classics, are considerable, despite the availability of musical notation. Major sections of the original score are routinely omitted or changed by repetition. (16) Music by another composer is usually added. (17) Music plays an important role, but one decidedly subservient to the choreography and the dramatic needs of a particular version, (18) helping to explain the liberal tolerance for variation.

As in other performing arts, considerable variation is acceptable in costumes, scenery, and lighting, as long as they are consistent with the dramatic theme of the ballet (e.g., peasant dresses in Act I and long white tutus for the Wilis in Act II). (19) Scenery (20) and props (21) are also varied, especially in authenticity and details, but remain consistent with the dramatic theme.

In sharp contrast are works choreographed by George Balanchine of the new York City Ballet, especially his neoclassical works which embody music in choreography. (22) The music is performed to meticulously follow the original composition by the composer, even in tempo. (23) The original choreography by Balanchine is adhered to precisely, (24) both specific movements and interpretive aspects designed by Balanchine. (25) Only Balanchine himself alters the choreography, something he often does to tailor roles to specific

dancers. (26) As most of his neo-classical ballets are plotless, this is not a factor, (27) but radical changes in costuming are common. (28) Balanchine ballets are the extreme case, but other contemporary or recent choreographic masterpieces receive similar deference. (29)

Notational systems are now widely available for dance, (30) and have been for some time, (31) but that availability does not explain the differences or even correlate with identity standards for various types of ballets or in different centuries. Although contemporary choreographers increasingly have their works recorded in written notation (32) or on film for archival and copyright purposes, many, including Balanchine, until very recently, have resisted that practice, partly because of the tremendous expense (33) and the continuing acceptability of handing ballets down from dancer to dancer. (34) Identity practices correlate more closely with conventions for appreciating and evaluating dance, than with the availability of notation. Identity standards for Balanchine ballets reflect the view that contemporary dance is primarily movement which expresses or embodies music. In the nineteenth century, the performances of particular dancers were more important than the choreography itself, and changes in the choreography were tolerated almost without limitation. In the twentieth century, with increasing emphasis on choreography more than individual performances, changes tolerated in the choreography, even in the story-ballets, are more limited. (35)

Notation reconstructor Rochelle Zide-Booth, after extolling the advantages of notation, raises what she treats as a wholly legitimate question, ". . . what . . . could or should be altered by the reconstructor?" (36) she treats sanguinely alterations in steps if beyond the capacity of the dancers. The aspect which must be preserved is the "quality" of the work, she says, (37) which is not a matter of strict adherence to notation. She also distinguishes works ". . . which must be reconstructed as closely as possible to their original productions in order to be valid, and . . . those which need some changes in order to make them acceptable to today's audiences," (38), including full-length classics from previous centuries.

Significantly, although a notational system for music has been available for centuries, identity requirements for the music in dance vary considerably. The strict identity of the concert hall is demanded for some, while identity with only general melodic themes is acceptable for others. Plots can be written down in words, to preserve with some precision the events to be portrayed, yet faithfulness to original plots has been limited. Since identity cannot be explained solely in terms of notational systems, other approaches are needed which account for and clarify the ways in which identity is actually established in dance.

B. Philosophical Theories of Identity in Dance

Philosophers have tried to account for identity in dance in a variety of ways, using appeals to notational systems,

production histories, cultural context, or some combination of those factors. Inadequacies of these theories can be summarized as either inadequate and inaccurate attention to the artform itself or as confusion about the purposes of identity theory.

Stephen Pepper, over 25 years ago, seemed to be unaware of the existence of written notation for dance when he wrote that ". . . there is no physical continuant of any importance like the written play in the drama or the score in music." (39) He seems instead to emphasize the importance of cultural agreements in identity, as when he says:

. . . lacking an ordinary physical continuant, that is replaced by a cultural continuant, the patterns of which are transmitted to perception through the medium of the dancers. The cultural continuant thus acts as a remote control upon the perceptual structure, just like a written play or a musical score except that its locus of existence is not on physical paper but in man's memories. (40)

Pepper could be claiming here that a notational scheme like musical notation could be constructed from the memories of dancers of particular works. But he could also be claiming that the memory of a dance is like the memory of a painting and is transmitted as completely as the frailty of human memory permits. Because he does not discuss or explain the role of notation as it does exist, it would be unfair and unfruitful to pursue detailed dissection of his observations.

Etienne Gilson recognized the potential of notation as

an encouragement to and reflection of increasing standardization of movement. Although noting that "the positions of the body and the movement of transition do not have the precision of a tone scale nor the exactness of musical duration," (41) he believes it possible that dance will develop " . . . definite works, stabilized through the resources of an appropriate score and as easily transmittable as are musical compositions today." (42) Gilson does not, however, explain how this evolution will occur nor does he pursue issues of compliance with such scores.

Joseph Margolis and Nelson Goodman, in more recent and detailed analyses, have explained philosophical identity in dance in terms of the increasing availability of notational systems, although they disagree sharply on whether such systems can wholly specify the identity of a work.

Margolis says that each performance is a token of a megatype fixed in a notational system, either an existing notation of a megatype or one constructed or constructable from observing performances. (43) "If we wish to assert that two performances are instances of the same megatype, we must be prepared to formulate a dance notation for which either performance will be a plausible token." (44) As we could construct a megatype for two performances which claimed to be The Dying Swan, yet failed in the same way to be so, formulation of the notation in itself does not show "authenticity." More importantly, Margolis does not say how much or what sort of compliance with notation is

required for a performance to be a "plausible" token of a megatype, nor who is best equipped to make such determinations, nor whether the degree of similarity must be the same for all types of dances, important questions taken up below.

He notes that prime notations which record generally accepted requirements for the identity of a work sometimes exist for dance (as they do for music), but prime notations might not always exist, for example, for folk dances. (45) Presumably, however, from among several competing megatypes which have been notated, one could become recognized as the "prime notation." Margolis does not say whether the version approved by the choreographer has a better claim to being considered the prime notation, but this is generally the case, as with ballets by Balanchine. He also does not say whether another fixation (such as a film or verbal description), either along or in conjunction with a symbolic notation, could establish generally-accepted requirements for identity, but it does in practice. He claims that dance (again like music) never has prime instances, which are critical instances (as of poems) which set standards for the other instances. A prime instance is ". . . a device for controlling the enumeration of tokens for a given megatype." (46) Margolis seems to deny that they exist for dance because of the difficulty distinguishing the megatype and the contributions of the individual dancer in a particular performance. (47) However, as he himself acknowledges in a

recent book, (48) devices which serve the function of prime instances do exist for dance, including definitive performances of a particular role (e.g., Anna Pavlova's performances of the Dying Swan, recorded on film; Vaslav Nijinsky's performances of Fokine ballets, recorded in the memories of living people, in extensive verbal descriptions, and expressive still photographs; (49) and Gelsey Kirkland's performance of Clara in The Nutcracker seen by millions in a recorded television program) or the world premiere performance of a new work by a dancer on whom it was choreographed. Such performances existed in time and, except for those on film, could never be seen again. Yet in practice they serve as important "devices for controlling the enumeration of tokens for a given megatype," if they are retained in the memory of dancers and audiences or if the performance was recorded on film, in photographs, or verbal descriptions. (50) it is difficult to see any difference between "prime instance" (defined in terms of its function) and something which serves that function.

Notation plays an important role in Margolis' views. First, he has said that, as notations are increasingly available, concert dances will be identified according to them. (51) to use his concepts, for a particular performance, a megatype can be formulated in notation; this notation could then be accepted as the prime notation. The importance of notation here is thus its potential as a standard, less fallible than human memory, against which particular

performances (tokens) can be compared for identity. But only a handful of people can actually read existing notational systems, and, because of their extreme complexity, this is not likely to change. Notational systems simply do not play any significant role in establishing identity in actual practice, as audiences and dancers alike continue to rely primarily on human memory to make identifications, with assistance from films and videotapes (which are not notations, but works of art themselves (52)), and verbal descriptions. Films of different dancers in the same work permit distinctions to be made between the choreographic design and the interpretations of individual dancers. The development and increasing availability of notational systems, understood as standards of identity, do not satisfactorily account for historical shifts in identity standards nor for present practices.

Notational systems are also important for Margolis because their existence reflects a standardization of the materials of an artform. (53) Sameness or identity can be established only when recognizable elements in different performances can be compared to determine identity with the megatype; the emergence of a notational system reflects that standardization of elements. However, the classical ballet vocabulary of standardized movements, with specific verbal descriptions (e.g., demi-plié in first position) existed long before contemporary notational systems were developed. To indicate time measures, dance has always had available

the notational system of music.

Ballet is noted for its constant innovations in the movement vocabulary, new steps, as well as new combinations of steps, but this does not mean it lacks the requisite standardization. The verbal characterizations long available can easily accommodate these new, nonstandard materials, as can notational systems which use locations of parts of the body. Notation was a very late arrival reflecting increasing standardization. It has not been absolutely essential to that process, nor the only evidence of it, and it has not added to or detracted from the continuing evolution of nonstandard movements.

A recent article by Margolis (54) stresses, not the importance of notational systems, but their inadequacy in accounting for identity. He acknowledges only that

Dance scores are primarily heuristic devices for recovering a minimal sense of the principal positions and movement of a given dance. (55)

He denies that such notation allows consideration of dance as an allographic art, as proposed by Goodman. (56) Along with the known divergence of dance performances, the reasons Margolis denies that dance could be allographic include, first, the discrepancy between ". . . the emphasis on visual recognition tout court in the notation and the requirements of actually generating dance movements in terms of the dynamics of motor activity controlled proprioceptively." (57) However, this reason seems to conflate the establishment of

identity of a work with the quite separate activity of learning how to produce the movements of a certain work or produce them in a particular way. Musical notation tells me nothing about how to blow or strike a musical instrument, but that does not render the notation "a mere heuristic device." A related conflation occurs later when he says:

There simply is not reliable correspondence between a dancer's performing a set of movements in accord with a mere notation and an audience's seizing the expressive qualities somehow conveyed by those movements. (58)

However, the former, performance, in accord with a notation, could be all that is needed for identity of the work, while the latter could be a separate matter of good performance of the work.

It is difficult to reconcile the various positions Margolis has put forth. Earlier, to assert that two performances are instances of the same megatype, we had to be able to construct a notation for which both are plausible tokens. Now it seems that construction of this notation is insufficient to establish sameness of the performances. The dancer must also know how to produce the movements. But why cannot the test of whether the dancer has succeeded in producing the movement be measured against that notation? Similarly, "expressiveness' depends on movements produced which could be notated. The same movements can be expressive of rather different things, depending on the interpretation of the dancer, but those variations need not and do not change

the identity of the work. In the black swan pas de deux of Act III of Swan Lake, some dancers express the same hardness of the evil magician while others, through their interpretation (the nuances of face, arms, phrasing, and manner of attack) express an underlying vulnerability akin to the white swan of the other acts. Yet audiences to not treat these as different works (assuming the same choreography) although they do treat them as different interpretations.

Margolis also denies the allographic character of dance because of the central role of uniquely expressive human bodies as instruments, (59) which ". . . cannot be satisfactorily represented by an extensional notation." (60) But this again confuses identity of a work with the evaluation of an individual dancer's interpretation of a work. An outraged choreographer, say, a Martha Graham, distressed by performance of one of her works by a dancer not trained or coached in her methods, might vehemently insist that it was not a performance of the Graham work at all. But Graham's likely behavior would betray this assertion of outrage. She might contemplate, for example, a lawsuit for infringement of the copyright or common law misappropriation of her work. A highly unorthodox performance of Swan Lake, resulting from an unusual and perhaps poorly suited body, results in several different assessments, which should be kept distinct: (1) the identity of the work is Swan Lake, but (2) the performance by a particular dancer is poor or inappropriate, perhaps because the body was not appropriate.

The increasing ease with which such distinctions can be made reflects the shift in understanding of the artform from one primarily concerned with the performances of individuals to one recognizing a collection of works performed in various ways by individuals. The outraged exclamation of choreographers (or critics) that a certain performance is not a performance of the work at all, must be carefully scrutinized. The overstatement could simply be an emphatic way of denying that the work had been well or acceptably performed, as opposed to a denial that it was the work at all.

Margolis may be correct in insisting that "the dance cannot be appreciated without some sense of how movements are actually generated," (61) but it does not necessarily follow that this knowledge is necessary to establish identity of a work. Margolis now believes this knowledge is necessary, when he rejects ". . . the notational constraint on the reidentification of a dance from performance to performance." (62) But Margolis has not shown that identity of a work necessarily depends on performance aspects which cannot be notated. Current practice suggests otherwise - we can and do distinguish between performances not of the same work and performances which are poor performances of the same work. Nor has he explained the minimal "heuristic" role which he says notation does play. If he simply means that most dancers still learn roles by watching other dancers instead of reading scores, I would agree. But some piano players still play "by ear" without reading music - that

does not show that notation has no role in establishing identity of dance.

Margolis' recent analysis is useful in highlighting the complex texture of the phenomenon of dance, but within this broad context the precise questions must be kept clear: identity of works, production of works, interpretation by dancers, evaluation of performances, and training of dancers to be able to perform works.

For Nelson Goodman, a notational system is also the sole measure of identity for dance works: "The function of a score is to specify the essential properties a performance must have to belong to the work . . . . All other variations are permitted; and the differences among performances of the same work, even in music, are enormous." (63) A notational system permits identification of ". . . a dance in its several performances, independently of any particular history of production." (64) Notation of any artform is developed ". . . in order to transcend the limitations of time and the individual." (65)

Like Margolis, Goodman argues that notational systems are possible only when there has been a necessary systematization of the materials of the artform, ". . . an antecedent classification of performances into works that is similarly independent of history of production," (66) which depends on there being distinctions ". . . between the constitutive and the contingent properties of a work." (67) He says that, although it is ". . . rough and tentative," (68) such a classi-

fication exists for a dance, a conclusion apparently reached because "prior to any notation, we make reasonably consistent judgments as to whether performances by different people are instances of the same dance." (69) Notational systems follow, but do not dictate, ". . . lines antecedently drawn by the informal classification of performances into works and by practical decisions as to what is prescribed and what is optional." (70) Goodman seems to consider the prerequisite classification as both standardization of particular movements and agreement about the identity of works considered in toto, while Margolis addresses only the former.

Goodman does not show that a notational system is or must be the sole device for establishing identity in dance. As he notes in discussing the prerequisites to a notational system, consistent judgment regarding identity of dance performances are already made independently of a notational system, although he does not indicate how such identifications are made. Further, it is simply not the case that the availability of notational systems has affected identity standards in dance so that no variations are permitted from the score and all other variations are permitted.

Certainly notation of dance is possible, and useful for theoretical analysis of an artform, but its availability does little to explain how audiences actually make comparisons for the purpose of establishing identity either before the emergence of notation nor once notation exists. A philosophical identity theory should explain what happens in

the artform. This should not be merely a descriptive account devoid of theory, but nor should it be a theory devoid of clarification of what actually happens.

Useful accounts of how the identity of dances is established in practice by audiences, choreographers, and dancers, have been developed recently by Adina Armelagos and Mary Sirridge. (71) However, in criticizing Goodman's proposals by showing the inadequacy of notational systems to provide the necessary and sufficient determinants for the identity of dance performances, they have failed to make certain important distinctions. First, they insist that performances must be characterized in terms of history of production to establish identity, but these characterizations are nothing but convenient descriptions which could be restated in terms of perceivable differences in performances and which could be notated or fixed in a tangible medium. Second, they do not recognize or examine the recent development of new ways of establishing identity to supplement long-established habits. These methods are being developed because of important new opportunities for copyright of choreographic works, (72) and because of the burgeoning growth of dance, the shortage of choreographers, and the resulting potential for infringement of copyrighted works.

Armelagos and Sirridge argue first that Goodman's theory is ". . . too weak, as it fails to include alleged 'incidentals' and rests on a thoroughly inadequate notion of dance movement." (73) they describe two types of these alleged

"incidentals." First, music, costuming, and lighting, factors excluded in Goodman's exclusive focus on human movement, can be critical in determining the identity of a work. This observation seems correct, although these "incidental" factors do not show the inadequacy of all notation, but only that multiple notation or other methods for the different media of dance may be needed. Further, these "incidentals" do not show that Goodman's approach is inadequate for identity of the movement and music. Second, they argue that the history of production is often essential in determining identity, as the use of dancers trained in a certain movement style (e.g., the Jose Limon school of modern dance) may be essential for the work to meet identity requirements. (74) However, although the training history of the dancers is admittedly a convenient shorthand for characterizing performances, it is not essential, for different training results in perceivable differences in performance which can be described in other ways. It is easier to describe those differences by reference to training than through notation, but this description has not been shown to be essential. In addition, like Margolis, they do not precisely reflect actual practices. An outraged choreographer may deny that a certain performance is "his work" - but by suing for infringement of copyrighted work he would betray the fact that he does believe it is his work, albeit poorly performed.

Armelagos and Sirridge argue next that Goodman's theory

is also "too strong," (75) as 'choreography, as it is in fact done and as it determines the identity of a work, will not produce the kind of compliants that Goodman (and the notator) want," (76) making the score inadequate not just for retrieval or production, but also for identity. The reasons for this inadequacy are not entirely clear. They suggest, but do not explicitly claim, that the inadequacy results because (1) "style, stage elements, and performers are sometimes allowed to vary widely from performance to performance," (77) and (2) because ". . . elements involved in producing a performance and the activity of the performance itself are invariably relevant to determining what work we are seeing." (78)

With regard to the first claim, variations in style are not necessarily problematic for a notational theory, as those aspects of performance could be simply nonessential. Stage elements (presumably, costuming, lighting, and so forth) do not show that notation of movement and music is inadequate for those aspects of a performance, but only that it is incomplete. The practice of substituting different performers at various performances suggests that dance identity can be free of the history of production, rather than necessarily tied to it. The problem with Goodman's approach is somewhat different. As few if any actual performances meet his standard of absolute compliance with a score, his theory is useless for understanding identity of actual works and performances without analysis of how much noncompliance

with movement and musical notation can be tolerated in practice without loss of identity.

Some criticism of the second reason why Goodman is "too strong" has been noted above and can now be further clarified. Armelagos and Sirridge say that a fundamental problem is ". . . that the common sense basic unit of movement construction, the step, is ill-defined," (79) and ". . . thus not interpretable independent of its history of production." (80) They acknowledge that notations can record ". . . a sequence of finely differentiated body positions," (81) but insist that this would still lack essential information about style and kinesthetic motivation. The "ill-definition" of a step does not seem to be unclearness as to position at a "slice of time," as for a note, but rather lack of clarity as to how much detail is needed about precise body locations at every moment. Again, however, it has not been shown that the perceivable differences resulting from "style, vocabulary, and kinesthetic motivation" could not be notated or that these causal factors are anything more than convenient ways of describing what can be seen and might be notated. The claim that scores are not adequate for producing a work ". . . if the style, its vocabulary, and its characteristic kinesthetic motivator and ideals are not antecedently known" (82) is irrelevant, as the adequacy of a score depends, not on how useful it is for producing a work, but rather on its usefulness for identification.

Armelagos and Sirridge note that in practice artists

are suspicious of notation, (83) but this could simply be the result of the availability of more familiar and convenient ways of transmitting dance works. However, circumstances in very recent years are making the use of notation (or fixation in a tangible medium such as film) much more attractive and necessary. With the well-known explosion in dance companies and activity, many companies are finding themselves unable to perform certain works, because a rehearsal master who knows the ballet is unavailable to teach it to the company. (84) The copyrightability of choreographic works, as of January 1, 1978, is encouraging choreographers to fix their works in a tangible medium, as the possibilities for infringement have increased substantially with the performance of works on television, the availability of home video-recorders, and the shortage of choreographers to create new works. How scores have traditionally been used proves less and less in view of the current shifts in attitude, necessity, and legal protection.

Armelagos and Sirridge (and the late article by Margolis) have not shown that all conceivable notational systems are inherently inadequate for establishing identity in dance, but nor has Goodman nor Margolis shown that or how notation can account for identity in practice. Goodman has not reconciled his demand for strict compliance with what is realistically possible in actually complying with this standard. Margolis has not shown what it means for notation to play a "heuristic" role without establishing identity. I

have tried to show that interpretations of individual dancers, while perhaps relevant to evaluation of the performance, can be and are kept separate from the establishment of identity, as can the method of production, even though this knowledge is necessary to produce the performance.

The issues remaining are, first, what notation or fixation is necessary to establish identity, in view of the unusual characteristics and recently changed circumstances of dance, and second, how much and what sort of compliance with those notations, fixations, or other standards is needed to constitute identity of a work in practice.

C. The Lay Observer Test of Substantial Similarity in Copyright Infringement

In attempting to understand how identity is established in practice, it is useful to consider the practices in the area of copyright infringement, an increasingly important application of identity theory. Determinations of copyright infringement use both notation to fix an absolute standard of identity and a lay observer test of substantial similarity to measure degrees of compliance with that notation. The proposal here is that an identity theory must include two things: a standard for identity, through notation, and a test for application of that standard, namely substantial similarity. Goodman sets up the standard but does not explain how it can be applied. Armelagos, Sirridge, and the later Margolis article reject the standard now winning growing acceptance, and instead apply a test of

history of production which confuses identity of works with quality of performances.

The central requirement for showing infringement of copyrighted artworks is "substantial similarity," as determined finally by lay observers. The analysis of similarity by experts, using notational analysis or any other methodology, is relevant and admissible, but not conclusive on the jury of lay observers. The lay observer test in the law (which also has limited experience with the artform of dance) is an important application of identity principles, especially since the availability of copyright protection promises to have far-reaching impact on identity in dance generally. Legal determinations of identity directly involve the issues of philosophical identity, and illustrate the viability of a two-pronged test of identity using both notation (or fixation) to set an ideal for absolute identity and a lay observer test to apply that ideal and specify actual identity of specific performances. How much and what sort of compliance with that fixation and how this compliance is determined is an important question in infringement theory, but one not addressed clearly in the context of philosophical identity.

Choreographic works were explicitly included in the subject matter of copyright (85) for the first time in the Copyright Act of 1976, (86) but many were eligible for protection under the previous law, primarily in the category of dramatic and dramatico-musical works; (87) common law protec-

tion was also available. (88) To prove infringement under the new statute, several matters preliminary to substantial similarity must be shown: (1) the copyrighted work must fall within the general subject matter of "choreographic work," (2) the work must be original, (3) the allegedly infringed aspects of the work must be expressions, not ideas, (89) and (4) the work must be fixed in a tangible medium. All of these elements are problematic in the law, because of the expanded subject matter of the new statute and also the scarcity of prior decisions involving choreographic works. This scarcity can be attributed, at least in part, to the failure of many choreographers to register their works under previous statutes, (90) and the practical difficulty of actually reproducing a work created by someone else. (91) Decisions under prior statutes mainly addressed the issue of whether a work could be protected under the category of dramatico-musical work, an emphasis reflecting both the limited subject matter of prior statutes and the nineteenth-century conception of dance as primarily a form of the theater.

The philosophical problem of definition arises in copyright consideration. "Choreographic work" is not defined in the new statute, although the legislative history says that the term is one of several with "fairly settled meaning," and that it is not "necessary to specify that 'choreographic works' do not include social dance steps and simple routines." (92) The precise meaning of "choreographic

works" is not clear, however, from prior statutes or case law. Nor is there any indication that Congress intended to limit "choreographic works" to those which were protected previously under the category of dramatico-musical work. (93) Congress clearly did intend to provide categories eligible for protection with "sufficient flexibility to free the courts from rigid or outmoded concepts of the scope of particular categories." (94)

In the absence of further guidance from the statute or case law, the meaning of "choreographic work" must be developed from common usage as the design of ballet or dance, which raises problems similar to those addressed in Chapter II. However, as separate copyright protection is available for musical and (non-movement) visual dimensions of dance (e.g., scenery and costumes), "choreographic work" for the purpose of the new statute is probably limited to the dimension of human movement. Thus, because of administrative considerations, "choreographic work" is defined more narrowly than is necessarily the case in ordinary usage by critics, audiences, and dancers. The comparison between copyright and philosophical identity is thus limited, but limited to a central problem of identity, that of movement.

To be eligible for copyright, a choreographic work must be original, (95) not in the sense of historic novelty, but rather that "the production is the result of independent labor. . . ." (96) this prerequisite of originality differs from the philosophical claim that identity of a work depends

on knowing the history of production. The showing of originality for copyright eligibility involves only one aspect of the history of production, viz., whether the protected work originated with the creator. A showing of infringement must include proof that the history of production of the allegedly infringing work includes access by the infringer to the protected work or, if such a showing is not possible, similarity so striking as to make creation without such access highly improbable. This inquiry into history of production is ordinarily severable from the analysis of similarity. Regardless of the results of the historical inquiry, two works could be found legally to have substantial similarity; the historical inquiry bears, instead, on the recovery for infringement, if such similarity were found. That is, infringement involves both similarity and access. The latter may be shown by striking similarity, but if lack of access can be proved, then there is no infringement, regardless of the degree of similarity.

In contrast, when Armelagos and Sirridge claim that the history of production must be known to establish identity, they mean that information about the way a production came about (e.g., through a particular training of the dancers) must be known to establish the actual similarity. They are not making a claim that history of production can be shown by striking similarity, nor that history of production and similarity are two separate factors which must be proven, but rather that similarity can and must be shown by history

of production.

Infringement of all types of artworks is determined according to the "ordinary observer" test, (97) which is simply whether an ordinary observer, perceiving two works, considers them to be "substantially similar." If so, and if the other requirements for copyright are met, one is an infringement of the other, protected work. This test, quite obviously, could accommodate the variations for identity standards in different types of genre, such as Giselle and works by Balanchine.

The law is not entirely clear as to who is an "ordinary observer." The classic definition of copy, "that which comes so near to the original as to give every person seeing it the idea created by the original," (98) is ambiguous, as it could refer either to (1) any person on the face of the earth, brought into court to view the work, or (2) every person who might, of his own volition, attend a performance or exhibition of the work. Some courts limit the ordinary observers on the jury to persons with some familiarity with the subject matter, (99) a very justifiable restriction for choreographic works. Although the burgeoning popularity of dance in this country is well-known, audiences still represent a small minority of the general public. Given the complete unfamiliarity of so many with this artform, only those who know how to look (e.g., those with experience as audience members of dance performances) should decide infringement. For persons with no familiarity with

the artform, comparing two dances could be like comparing poems in unknown foreign languages.

Courts have excluded from juries in infringement actions persons unable to perceive differences in the works in question. For example, the tone-deaf have been excluded from juries in music infringement. (100) Similarly, a case involving infringement of a French novel by three English dramas was withdrawn from the jury as it was not considered capable of making satisfactory comparisons of the works. (101) These exclusionary principle have not yet been applied in choreographic infringement, but clearly could be.

Significantly, expert testimony is not determinative of "substantial similarity," but it is relevant for "dissection and analysis" of the infringing passages to determine originality, (102) which is not left to the "lay observer" test. Expert testimony is also relevant for determination that a work is a "choreographic work," and thus subject to protection. (103) the emphasis of the lay observer test is clear: regardless of expert analysis, the ultimate test is always what lay observers think is substantially similar. The actual practice of typical audience members prevails over theory.

Infringement of a copyrighted choreographic work would almost always be a violation of the exclusive right "to perform the copyrighted work publicly" (104) or, with films of the work, the right "to display the copyrighted work publicly." (105) Substantial similarity is determined by comparing

the allegedly infringing performance (or film) with the protected work. The degree of similarity required varies according to proof of access by the infringer to the protected work, (106) a factor not relevant to philosophical identity, as noted.

The provision for degrees of similarity recognizes actual practice. There is no insistence that similarity be absolutely present or absent. Instead of saying that two works must be found either identical or not, juries may examine degrees of similarity. When proof of access is missing, the similarity must be found to be so striking that the possibility of independent creation of the infringing work is unlikely. (The ultimate issue in copyright law is copying, not similarity. Even though two works were identical, infringement would not be found if the creator of the allegedly infringing work could decisively prove lack of access, in which case the striking similarity would be overridden.) Infringement could be found for a lesser (but still substantial) degree of similarity, if access were decisively proven.

Although the jury (or a judge in a trial without jury, although he still might use an advisory jury) is the decision-maker on the factual question of whether substantial similarity exists in a particular case, the judge still has an important role in providing guidelines to the jury regarding the meaning of "substantial similarity." (The judge might also withdraw the decision from the jury if the

similarity is so trifling that no reasonable jury could find "substantial similarity.")

The extremes of similarity for choreographic works seem easy. Substantial similarity would be found in performances consisting of the identical movements by the same ensemble of dancers as provided in the fixation of the protected work, or the identical movements by the same ensemble of one or more discrete sections, movements, or "variations," as fixed in the notation or other medium.

At the other extreme, two performances would not be "substantially similar" if they shared only some of the same standardized steps, and not in any noticeably similar sequence (e.g., performances whose only similarity was the use of frequent fouettés and pirouettes, but not in the same sequence or combination). This would be analogous to two musical works using some of the same chords, but not in any noticeably similar sequence.

The vast middle ground between these extremes is problematic. Case law involving theatrical spectacles or pantomime provides little guidance, primarily the test of similarity in plot and characters. (107) Substantial similarity in literary works also rests primarily on detailed plot situations and characters. (108) As a choreographic work need not contain any dramatic situation, similarity for purposes of copyright infringement cannot depend primarily on these elements. (109) Even when choreographic works do include a plot or dramatic element, it is not normally the central or

distinctive characteristic of the work. Analogies with substantial similarity in musical works are more appropriate, (110) because of the non-verbal, temporal nature of the artforms, but no less complicated. Although it is widely believed that a certain number of bars of an identical melody is a decisive test of infringement, (111) similarity in music is not that simple and mechanical. It is at least clear that for all types of works, including dance, value is an irrelevant consideration. (112)

Similarity can be usefully assessed in terms of the elements of typical dance performances (identified in Chapter II):

(1) basic "steps," from either established or newly created movement vocabularies, which are

(2) combined in sequences of several steps

(3) for one or more dancers,

(4) in a performing area,

(5) to the accompaniment of music,

(6) for the purpose of telling a story and/or communicating or expressing human emotions or feelings,

(7) with the aid of mime, costumes, scenery, and lighting.

(1) The mere sharing of particular steps from established movement vocabularies (e.g., double pirouettes, grand jetés, etc.), would not make two works substantially similar. A newly created step, in isolation from any particular sequence, might be a distinctive motif, but would also not

make two works similar. (113) (Analogously, in music, use of some standard chords or one distinctive harmony would not make two works substantially similar.)

(2) a sequence of steps could show substantial similarity, but the obvious problems are (a) how long (in time and number of steps) the combination must be, and (b) how much similarity is required (absolute identity of the entire body throughout, absolute identity of arms and legs only, some identity of some steps, etc.) the shorter the sequence and the less absolute identity of particular parts of the body, the less likely would be substantial similarity. No simple rules could be developed somehow quantifying these questions for every type of dance.

(3) Choreography for more than one person could provide grounds for a finding of substantial similarity. (114) An ensemble might perform the same combination of steps (perhaps not unusual in themselves, or in combination), but at different times, to create a striking visual pattern of movement. Analysis of a movement design should not focus only on each dancer in isolation, any more than a symphonic musical work should be considered only one instrument at a time.

(4) The choice of performing space might be a distinctive and integral part of a work and grounds for substantial similarity. The choice of location for movement is certainly a deliberate part of the design, especially when it is as unique as ramps running across the audience or the steps

leading to a public monument. For copyright infringement, however, the choice of performing space alone would probably be excluded from protection as a "procedure," although the pattern of movement designed for that space could be protected.

(5) The music used to accompany dance movements is subject to copyright under a different category, (115) but the choice of a particular musical accompaniment for certain movements might be a distinctive contribution, providing grounds for substantial similarity, as well as originality. Performing the Black Swan pas de deux from Swan Lake to a Sousa march instead of the traditional Tchaikovsky score might not only be distinctive and original, but historically novel. It is not clear whether this choice would be generally recognized within the meaning of "choreographic work" for purposes of copyright, however.

(6) Because of the absence of words, the dramatic element of dance (plot or story; emotions or feelings) is necessarily so generalized as to provide questionable grounds for substantial similarity (as well as failing the requirement of originality). Plots in themselves, apart from the movement, also might not be considered part of a "choreographic work" (116) for copyright purposes. Words with a distinctly original plot (e.g., avant-garde works using spoken dialogue or new "story-ballets") might still be eligible for registration as dramatico-musical works, (117) to ensure that protection is obtained for the dramatic element.

as well as the design of the movement.

(7) costumes, scenery, and lighting would probably be eligible for protection as "pictorial, graphic, and sculptural works." (118) As with the choice of music, the choice of costumes for use with certain movements might be historically novel and strong grounds for substantial similarity (e.g., performing the Dying Swan in a scarlet unitard), but would not clearly be part of a choreographic work in the sense of movement design. (119)

Several aspects of typical dance performances are subject to copyright under different categories of protection (music, scenery, costumes), but regardless of whether those aspects of a particular work have actually been copyrighted, they do not affect the substantial similarity of the movements themselves. (120) It could be argued that in infringement cases, especially if an ordinary lay jury were to make the final determination, performances for comparison purposes of the works in question should be presented with identical costumes, sets, and music. Even if all such elements are not protected by copyright, they affect the overall perceptual impact of the aspects of the work which are protected. (121)

With human movement only now clearly central under the copyright law in the new category of "choreographic works," a basic understanding of substantial similarity between movement designs has yet to be developed. A few basic principles for copyright infringement would clearly apply (such

as disallowance of only insignificant alterations, which could be exploited by pirates to evade copyright), but decisions for now must rely on the determination of the fact-finders without well-developed guidance from the court. Given the diversity and complexity in contemporary dance, it is unlikely that any simple tests of similarity will or should be developed.

A crucial comparison between substantial similarity in copyright infringement and philosophical identity is the role of notation or, more broadly, fixation in a tangible medium. To be eligible for copyright protection, works must be "fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device." (122) The legislative history of the Copyright Act emphasizes the flexibility available in the choice of medium, "to avoid the artificial and largely unjustifiable distinction . . . under which statutory copyrightability in certain cases has been made to depend upon the form or medium in which the work is fixed." (123) However, there is no hint in the legislative history that Congress specifically considered the use of film and notation in dance.

Choreographic works can be fixed through films, videotape, or a written notational system. Detailed verbal descriptions using standard ballet vocabulary would also seem acceptable. Fixation through film or videotape records

every element of a performance, including the interpretation of particular dancers, (124) resulting in a copyrighted work considerably more detailed than written notation could provide. The resulting problems do not seem to have been recognized by the drafters of the new copyright law.

First, the interpretations of individual performers which would be recorded on film or videotape have traditionally been excluded from protection, (125) although some courts have included these interpretations in the protected work if recorded in some way. (126) It could thus be argued that all interpretive aspects of a performance fixed on film or videotape are protected. If such interpretive elements are not included in the protected work, it would often be impossible to identify which aspects were part of the choreographic work and which were the interpretive contributions of the performers. Many interpretive elements could conceivably be the work of either the performer or the choreographer (e.g., a certain turn of the head, facial expression or phrasing of the steps). Another problem is whether the choreographer should be considered the "author" (and thus eligible for copyright protection) of those interpretive elements recorded on film, but contributed by the performer. For simplicity, there could be a presumption that the choreographer has contributed all protected elements of the work, including "interpretive" elements, with the burden on a challenger to show otherwise. It might also be argued that the dancers and the choreographer should be

considered joint authors.

Another problem resulting from the possibility of multiple fixation is determination of what constitutes the choreographic work protected by copyright when there are discrepancies between the visual recording and written notation, when both fixations have been deposited with the Copyright Office. Should the protected work consist only of those elements shred by both forms of fixation? Should the film be seen as a supplement to the more skeletal written notation? The distinctive elements captured only on film might be precisely the characteristics best showing substantial similarity or originality in an infringement suit. It could be argued, however, that the discrepancies constitute, prima facie, the interpretive and unprotected elements of the work, that is, the nonessential characteristics for te purpose of identity.

Another issue is whether common law protection would be available under state law for those aspects of works which are not fixed. (127) If a work is revised considerably, by the choreographer, from the fixed copy, it could be argued that the revisions, if unfixed, should constitute a new work protected at common law. (128)

Another issue concerns the actual performances to be compared in determining infringement. If a written notation was deposited for the copyrighted work, a performance would have to be reconstructed from the notation for the court, since so few people can read such notations. It is

uncertain whether this reconstructed performance should be supervised by the copyright owner or a neutral party to ensure reconstruction of only those notated aspects. The choice of party to oversee this reconstruction could critically affect the outcome of the case. It is also not clear how to compare an allegedly infringing performance, which had not been recorded, with a protected work.

Despite these numerous significant issues involving the method of fixation, the initial response under the new statute by choreographers seems motivated mainly by economic considerations, with preference for substantially less expensive videotape over notation. (129)

In summary, infringement and identity theories share several important characteristics. In both, the aesthetic, economic, or other value of a work is irrelevant in determinations of similarity. Thus, in both, care must be taken in analyzing statements that "This is not Work X" to be certain that this is not really an elliptical claim that "this is not a good performance of Work X." Both involve a search for tests of sameness. In both the choreographer makes decisions about what constitutes the work. For example, in copyright, the choreographer must decide what notation and/or film is submitted to the copyright office. For performances, he or she must decide what work is presented, initially and often in reconstructions of the work. In both spheres, the reconstructions of others might also be found similar to the choreographer's intended version. (130)

Infringement and identity theories also differ in significant ways. Only in infringement theory is actual copying and access relevant. Even so, this does not affect a finding of substantial similarity, but only the degree of similarity which must be proven. That is, the history of production (whether or not there was access) is not relevant to the finding of substantial similarity by a jury. In contrast, some identity theories claim that history of production is essential in establishing identity, but that has been challenged here.

Another important difference is that infringement is found where there is similarity of the whole or of some major part of the work, while identity pertains only to the work as an entirety. Infringement could be a major part of a movement design or it could be of the movement design but not of the other media (music, scenery, etc.) protected under different categories of copyright. Infringement of a significant part of a work is not the same as identity of a work as an entirety. Identity theory thus involves additional factors but the method (notation plus lay observer test, e.g.,) need not necessarily be different.

Another difference is the importance of degrees of similarity. Infringement involves findings of various degrees of similarity (depending on the degree of access shown), while identity theory is interested in absolute similarity. It is suggested here that because of the complexity of dance and the uncertainty of identity standards, philo-

sophers could fruitfully look at degrees of similarity to understand how identity is established.

Another significant difference is that infringement theory ignores trivial or minor differences in works, not because of the belief that such differences are necessarily irrelevant to a showing of similarity, but rather because recognizing such differences as reducing similarity would invite and tolerate plagiarists and pirates. This motivation is not present in identity theory, but it could incorporate the same lack of interest in minor differences on the grounds that they do not affect the perceptions of ordinary observers, a secondary motivation in infringement theory.

Another difference is the absence in identity theory of various statutory prerequisites and exemptions to a showing of infringement, including originality, the "public domain," and the "fair use" exemption, but these are not relevant to the test of substantial similarity. Copyright is concerned primarily with protection of the economic right of use of intellectual property; substantial similarity is only one element of that right.

Identity and infringement theories differ in their underlying purpose and motivation. Copyright rests on the assumption that economic motivation will encourage creativity. If similarity is sufficiently substantial that potential consumers of the creation will be misled into purchasing the infringing work, that economic benefit will be

denied to the true author. The test thus rests squarely on the perceptions of ordinary observers and potential consumers. No such motivation exists explicitly in philosophical identity theory, but it is important to try to specify purposes for identity theory. It is possible that identity standards could differ depending on the purpose. For the purpose of criticism, works must be identified so they can be compared in different performances by different artists. As critical evaluation is, in part, evaluation of the quality of performance of a given work, that work must be identified to make such comparisons. Criticism also involves comparison if the aesthetic value of different choreographic designs. In order for such comparisons to be meaningful for a reader, the work must be identifiable by the interested readership. Identity theory also strives to lay the theoretical groundwork and understanding which would explain the development of notational systems and such practical applications as the copyright system.

Conclusions

The proposal here is that identity in dance is best understood as a matter of both (1) a written notation which provides an ideal or absolute benchmark for identity, and (2) agreements among ordinary observers of dance at a particular time and with regard to particular types of works regarding the necessary sort of compliance with the notation. These agreements are like the guidelines for ordinary observers in infringement cases. For example, the mere

presence of standardized movements, in itself, does not constitute similarity or identity, but a recognizable movement phrase, analogous to a melody line in music, does contribute to finding similarity. A trivial difference such as a different placement of the head does not detract from a finding of identity. These guidelines can be identified and verbalized by looking at actual practice in the dance world. The task of identifying these guidelines is a continuing responsibility of critics, copyright lawyers, choreographers, and other members of the dance world.

Notation plays an important, but not the only, role in establishing identity. It provides a standard against which acceptable variations can be assessed. It reflects growing standardization in dance with regard to characterizing movements, although the standard for an entire work consists of both a notation (either actual or constructable) against which deviations can be assessed following ordinary observer guidelines.

This two-pronged proposal explains why, although both Giselle and a Balanchine ballet can be notated, the variation tolerated from those notations is so different. The acceptable variation is a reflection of different ordinary observer guidelines. The guidelines for nineteenth-century classics emphasize consistency with dramatic mood and plot with great tolerance for variations in steps. Those for a work by Balanchine emphasize close adherence to the notated steps. There is a trend toward the latter, reflecting

growing interest in the creative aspects of choreographic designs as opposed to the interpretive aspects of individual performers. But many of these genres still co-exist in dance, with different identity standards.

This dual test of identity accommodates a built-in limit on absolute identity in dance, namely, the nonuniformity of the human instrument in the artform. It also accommodates the numerous variations possible because the many media in dance can be combined with different emphasis on particular media in different dance genre.

The approach of the law to infringement of copyright does not provide any easy answers, but it does suggest basic guidelines for better understanding identity. First, the ultimate test and thus the final object of analysis should be the reaction of lay observers to actual performances, including how much and what kind of similarity is needed to establish identity. Notational analysis is useful to the extent that it explicates and clarifies those practices. Second, a broader sense of "fixation" rather than notation is a more generous tool which serves the purpose of removing the record of dance from fallible human memory, while recording interpretive nuances not included in the notation and for which there is not a clear distinction between essential and non-essential performance aspects.

Theories which explain identity in other artforms can be extended to dance only with great caution, because of the complexity of the artform, the unique role of the human

body as instrument, and the very recent development of dance as an artform. But philosophical analysis can be used quite productively in understanding the way in which people actually establish identity.

NOTES

(1) The Nutcracker is an extreme example of a ballet produced by numerous companies, productions sharing only the Tchaikovsky score and the scenario. See, e.g., Alan M. Kriegsman, "'Nutcracker' Theme and Variations," Washington Post, December 23, 1975. George Balanchine claims that the numerous versions are "about the same" in "story and action," as well, of course, as the music. George Balanchine and Francis Mason, Balanchine's Complete Stories of the Great Ballets (Garden City, N.J.: Doubleday & Co., Inc., 1977), p. 413. This extreme case presents the question whether audiences really treat all these versions as the same ballet or as different versions of the same ballet. Or as different ballets sharing only story and music. Although the answer is not clear, the latter characterization is probably most accurate. Return to text

(2) "Tinkering with the traditional versions of the nineteenth-century ballets occurs all the time everywhere." Walter Sorell, The Dancers' Image: Points and Counterpoints (New York: Columbia University Press, 1971), p. 65

Changes are also made in new ballets. A very recent example is The Four Seasons, created for the New York City Ballet in 1979. Jerome Robbins choreographed two entirely different solos to different music for Peter Martins and Mikhail Baryshnikov, alternating as the lead male dancers. Critic Arlene Croce refers to these as "two versions" of the work. "Other Verdi Variations," New Yorker, February 5, 1979, p. 114. Their partners, Patricia McBride and Suzanne Farrell, performed solos similar choreographically, except that McBride ". . . does chaines piques instead of the string of double soutenu turns that Farrell knocks off." Ibid., p. 116. Performances seen by this writer, New York State Theater, with Mikhail Baryshnikov, January 18, 1979; Kennedy Center, with Mikhail Baryshnikov, February 22, 1979; with peter martins, February 24, 1979.

Names are sometimes the only thing changed, as with the change of Balanchine's Le Palais de Cristal to Symphony in C. Jack Anderson, "Ballet: City's 'Symphony in C' Is a Crystal Palace," New York Times, January 9, 1979. Return to text

(3) The nineteenth-century story-ballets normally included much mime to convey the dramatic action. The mime in those ballets surviving in current repertories is almost always reduced or eliminated.

For example, in Balanchine's version of Coppelia, first performed in 1974, "The mime passages have been quickened, compressed, and in some instances threaded directly into the dancing." Alan M. Kriegsman, "A Sparkling 'Coppelia,'" Washington Post, February 21, 1976. Alexandra Danilova, who assisted with the production is reported to have said:

"'Some things, like the mime, I speeded up.' . . . 'Some things must be made acceptable to our time.'" Hubert Saal, "Saratoga Smiles," Newsweek, July 29, 1974.

In the Coppelia by American Ballet Theatre, first presented in 1968, guest Paolo ". . . Bortoluzzi tends to substitute naturalistic gestures for the conventional mime required for 19^th-century ballet." Anna Kisselgoff, "2 Stars From Italy Join In a Spirited 'Coppelia,'" New York Times, January 12, 1976.

Productions of Swan Lake also illustrate the wide variance in the use of mime. See, e.g., Arlene Croce, "The Royal Line," New Yorker, May 17, 1975, p. 162; Clive Barnes, "Ballet: 2 Remarkable Local Debuts," New York Times, May 12, 1975. Return to text

(4) For example, Erik Bruhn's version of Swan Lake for the National Ballet of Canada, which retains most of the original Petipa choreography, though adding some new passages, develops a very unusual psychological, Freudian view of the Prince. The evil Rothbart is portrayed by a woman resembling the Prince's mother. Anna Kisselgoff, "Ballet: Canadians Dance Their Own 'Swan Lake,'" New York Times, July 28, 1978; Tobi Tobias, "'I Am Not Finished as a Dancer,'" New York Times, June 29, 1975; Clive Barnes, "Nadia Potts Is a Stylish and Distinctive Odette-Odile," New York Times, August 5, 1975. Mikhail Baryshnikov, Baryshnikov at Work, Charles Engell France, ed. (New York: Alfred A. Knopf, 1976), pp. 153-5.

". . . though all productions of [the nineteenth-century ballet Coppelia] tell the same tale, they can look different because of the various interpretations choreographers may give to the events of the scenario. Some "Coppelia's" are farcical. Some are macabre. Others are fables about the folly of infatuation. Still others emphasize Swanilda's wit and cleverness - and even her touch of bossiness. . . ." Jack Anderson, "Cuban Ballet Performs an Unfamiliar 'Coppelia,'" New York Times, June 17, 1978.

Baryshnikov's performance of Balanchine's Apollo has been described as ". . . the most Dionysian Apollo we have seen . . . Very obviously, Mr. Baryshnikov does not see the ballet as an allegory about virtues of classical art. Its usual serenity has been broken by the realism he has imposed in angst-laden acting upon the stylized movement." Anna Kisselgoff, "Dance: Baryshnikov Star of Festival," New York Times, June 12, 1978; see also, Anna Kisselgoff, "Dance: Festival shows Baryshnikov at Work," New York Times, June 13, 1978.

Rudolf Nureyev's production of Romeo and Juliet for London Festival Ballet, a ballet produced by many choreographers and companies, emphasizes the theme of fate throughout the work. Anna Kisselgoff, "Ballet: Nureyev's Interpretation of 'Romeo,'" New York Times, July 20, 1978; Alan M.

Kriegsman, "Stalking Rudolf Nureyev," Washington Post, August 6, 1978; Deborah Jowitt, "Slugging It Out in Verona," Village Voice, August 14, 1978. Kisselgoff also says the production ". . . is really more about Romeo and his friends than about Romeo and Juliet." "A New Juliet by Londoners At the Met," New York Times, July 22, 1978. Kriegsman characterizes the production as "more starkly medieval than wantonly Elizabethan. . . ." "Nureyev Is 'Romeo,'" Washington Post, August 2, 1978.

Oscar Araiz' version for the Joffrey Ballet uses three different dancers as Juliet to develop the various facets of the heroine. Jennifer Dunning, "'Romeo and Juliet' by Araiz Is Danced by the Joffrey Ballet," New York Times, November 12, 1978. Return to text

(5) For example, in Balanchine's revival of Coppelia, Delibes' traditional score is ". . . restored to uncut proportions, with some interpolations from other Delibes ballets. . . ." Alan M. Kriegsman, "A Sparkling 'Coppelia,'" Washington Post, February 21, 1976; see also, Arlene Croce, "I have Made You and You Are Beautiful," New Yorker, August 5, 1974, pp. 75-7. Return to text

(6) For example, the order and length of Tchaikovsky's Swan Lake has been altered frequently. See Humphrey Searle, Ballet Music (New York: Dover Publications, Inc., 1973), pp. 69-73.

A comment by Alan Kriegsman is typical of critical acceptance of such musical changes: "The fracturing and rearrangement of the music [Prokofiev's Romeo and Juliet], for instance - a choreographer uses music to serve the ends of the dance, and what's valid is what works." "The Esthetic Gimmickry of Araiz's 'Romeo and Juliet,'" Washington Post, August 4, 1978. Return to text

(7) Arlene Croce reports "some outcry" from the audience about choreographic changes by Gelsey Kirkland in Swan Lake (e.g., "Her entrance in the coda eliminated the fouetté in the half turn into arabesque, and so another Odette cameo was partially erased"), but Croce says ". . . the only sensible objection to altered text in 'the classics' is to wrongness of effect. I'd have preferred some bright substitution for the standard relevé-passé/entrechat sequence, which, with Kirkland's thin thighs and calves, has no dazzle. But even though this Odette abjured dazzle, the overwhelming impression she left of originality and power was not damages." "Beyond Ballet Theatre," New Yorker, July 4, 1977.

Clive Barnes also speaks approvingly of certain changes: "In the classics the Bolshoi Ballet has traditionally permitted its chief interpreters considerable leeway in individual characterizations. It is a principle, sometimes

followed in the West, that has a lot to comment it, for dancers instead of feeling themselves encased with the unbreakable confines of a production, are given a little, precious room to move, the change to interpret a role themselves, rather than merely to execute someone else's interpretation." "Ballet: the Bolshoi Scores With 'Giselle,'" New York Times, May 2, 1975.

On rare occasions, critics hint at a limit to changes beyond which a work ceases to be the "same" work. E.g., Marcia Siegel says: "I would like to see some choreographer restudy the whole of Swan Lake, unify all its elements, throw out the pantomime and audience-wooing and old-time hokum, and create a clean, dramatically believable story ballet out of the material. But then, of course, it wouldn't be Swan Lake." At the Vanishing Point (New York: Saturday Review Press, 1972), pp. 56-7. Interestingly, she does not hint at loss of identity as Swan Lake in an earlier discussion of a radically different version: "George Balanchine's version of Swan Lake, consisting of the second act only, is a more acceptable solution. Without sacrificing the dance experience, he dispenses with superfluities of plot, spectacle, divertissement, and stereotyped character and concentrates on the encounter between the Prince and the Swan Queen . . . . Pantomime is employed sparingly, and he lets his dancers make use of the expressive possibilities within the ballet technique." Ibid., p. 56.

Another rare example of changes in Swan Lake so radical that it was considered by some to have lost its identity was Mikhail Mordkin's 1938 version for Ballet Theatre, a one-act version retaining only the music, with a cast including The Ideal, the Poet, and the Protective Fates (the Black Maidens). In 1940, he added the traditional first act pas de trois, danced by cygnets and a huntsman, instead of the usual party guests. Charles Payne characterizes this "tone-poem version" as a "different" ballet. American Ballet Theatre (New York: Alfred A. Knopf, 1978), p. 39. Return to text

(8) Such changes are apparently as unacceptable in opera as in the concert hall, despite opera's use of unique human voices (similar in uniqueness to the role of human bodies in dance). At a recent performance of Don Carlo, mezzo-soprano Marilyn Horne was booed for singing "O don fatale," in the key of F instead of A, although Horne insisted that Verdi had originally planned the aria in F. "Marilyn Horne Booed for Doing Her Homework," New York Times, February 8, 1979. Return to text

(9) See, David Blair, Program Notes, Giselle, American Ballet Theatre, Kennedy Center, April 4, 1976, p. 22B; George Balanchine and Francis Mason, Balanchine's Complete Stories of the Great Ballets, pp. 280-292. Return to text

(10) It is well-established that the French choreographer Marius Petipa made many revision in the choreography during his career in St. Petersburg in the latter half of the nineteenth century, as have numerous others. See, e.g., David Blair, "Historical Background of Giselle," in Kennedy Center, April, 1976, p. 22B; Erik Bruhn, "Restaging the Classics," in Payne, American Ballet Theatre, pp. 326-7. Today, many changes in the choreography are still common in surviving productions. These range from doing a series of leaps and whirls instead of a usual walk to replacing a familiar series of assembles with a different jump. Clive Barnes notes of Vladimir Vasiliev, as Albrecht in the Bolshoi Ballet's production: "After his entrance he does not merely walk to Giselle's house, as is the custom, but leaps over in a couple of grand jetes, and whirls back in a series of chaine turns. Such untoward fireworks might have been totally inappropriate, but here they appeared just right." "Ballet: The Bolshoi Scores With 'Giselle,'" New York Times, May 2, 1975. Anna Kisselgoff says of a performance by Eleanor D'Antuono with American Ballet Theatre: "[She] was in error when she destroyed the traditional symmetry of Albrecht's and Giselle's doing assembles together. Instead she inserted a pointless diamond-shaped jump." "Ballet: Basic 'Giselle,'" New York Times, December 28, 1975. But Kisselgoff does not hint at any loss of identity for this "error."

In the Cuban production, "Even when there is new choreography by Miss Alonso, it has logic. The peasant pas de deux of Act I is now danced by six girls and four boys and these girls are Giselle's usual six friends." Anna Kisselgoff, "Dance: Miss Alonso Stars in 'Giselle,'" New York Times, June 15, 1978. Many other changes in the Cuban version are described by Alonso herself in "Performing Giselle," in Payne, American Ballet Theatre, pp. 341-2.

The new production by London Festival Ballet restores many choreographic elements of the original version changed in productions over the last century. As described by Anna Kisselgoff: "Essentially, it is a version that acknowledges a romantic trademark in ballet - the expressive power of movement. The mime is not subordinate, it is in fact more frequent and clearer than in many productions. But there are also dance passages restored and additional choreography by Miss [Mary] Skeaping, with changes in sequence. In Act I, she has moved up the usual peasant Pas de Deux, to the traditional Borgmuller music, and shortened it by a solo for each dancer. Giselle's solo, also considered an interpolation, now follows this pas de deux and is part of the entertainment for the aristocratic hunting party. By contrast, the vendange dance has been lengthened, using the restored Adolphe Adam music for a new pas de deux, including new solos for Giselle, the peasant girl courted by Albrecht, the count in disguise. Act II . . . contains the most dramatic restorations. The misty opening forest scene

not only shows a group of gamekeepers and Hilarion . . . stunned by the sudden shadowy appearance of wilis, but it also presages the highly effective and ghostly new scene in which the men are literally surrounding by these menacing maidens and a Myrtha leaping among all. Much has been written about a passage called the wilis fugue, in which the wilis attack Giselle and Albrecht as they stand protected by the cross at her grave. Miss Skeaping has restored this section, giving it a whirlwind quality, interesting but not as dramatic as expected." "London Festival Ballet in Skeaping 'Giselle,'" New York Times, July 23, 1978. See also, David Vaughan, "Revivals, 1976," Ballet Review, VI (No. 1, 1977-78), 33-4; Alan M. Kriegsman, "Standing Ovation for A Superb 'Giselle,'" Washington Post, August 10, 1978. Return to text

(11) For example, the lilies pas de deux, Albrecht's solo, and the ensembles for the Wilis in Act II are rarely altered. Return to text

(12) One striking change was the ending of the ballet, when Giselle returns to her grave and Albrecht departs. In the earliest performances, his fiancee Bathilde appeared at the close of the ballet and forgave him, but this was dropped in later nineteenth-century productions. See, e.g., Sorell, Dancer's Image, p. 421.

Another variation is the cause of Giselle's death in the first act. In the original version, she had a heart condition, went made upon learning of Albrecht's deception, and died from the weak heart. See, e.g., Sorell, Dancer's Image, p. 67. In American Ballet Theatre's production, Giselle is portrayed simply as sickly, and she dies of "a broken heart." Program Notes, Kennedy Center, April, 1976, p. 22A. In other performances, including that of the English Lynn Seymour, Giselle dies by stabbing herself to death. "Reverting to ballet tradition, Miss Seymour's Giselle stabbed herself in Act I. To expire of a broken heart would have been too pallid for this dramatic heroine so willing to die for love." Anna Kisselgoff, "Ballet: 'Giselle' Via Lynn Seymour," New York Times, June 12, 1976. Return to text

(13) In current productions, there is still great variation in the departure of Albrecht from the grave. Ivan Nagy pauses briefly to watch Giselle re-enter her grave, then turns his back, picks up his cape, and walks across the stage. In sharp contrast, Mikhail Baryshnikov, stunned and remorseful, stares intensely at the grave as he impulsively picks up an armful of lilies on the ground. He then walks slowly backward across stage, dropping lilies on the ground unknowingly as he continues to stare in horror at the grave. He finally looks down at the one remaining lily in his hand, looks again at the grave, and throws himself on the floor. These differences are consistent with Nagy's interpretation

of Albrecht as something of a cad with a short-lived touch of remorse and Baryshnikov's interpretation of Albrecht as truly falling in love with Giselle and feeling deep anguish over his original deception. Performances seen by this writer, American Ballet Theatre, Kennedy Center, April 4, matinee and evening.

Indeed, there are almost as many endings as there have been Albrecht's Christopher Lyall of the English Ballet Rambert wandered in a daze all over the stage and then threw himself on the grave. Giselle: Act II, Ballet Rambert, 1959 (film). Anton Dolin gave a very melodramatic ending, clutching at himself and finally throwing himself on the ground. Giselle: Excerpts, Ballet Theatre [n.d.] (film). Films seen by this writer at the Jerome S. Robbins Film Archives, Dance Research collection, New York City Public Library, Lincoln Center.

The interpretation of Vladimir Vasiliev of the Bolshoi is described by Clive Barnes: "His Albrecht seems quite unlike anyone else's. He does not play the role as a caddish noble seducer, but rather as a man caught by circumstances, hopelessly in love with the peasant girl Giselle, and forced into a deception that fills him with guilt. In the second act he approaches Giselle's grave like a man in a trance. He even abstractedly drops a flower and stoops to pick it up, hardly knowing or caring where he is or what he is doing." "Ballet: The Bolshoi Scores With 'Giselle,'" New York Times, May 2, 1975.

At the end of Act I in some versions, Albrecht is pulled away from the dead Giselle by townspeople as an outsider; in others, he stays by her side as the curtain comes down. See, e.g., Anna Kisselgoff, "Ballet: Basic 'Giselle,'" New York Times, December 28, 1975. Erik Bruhn explains the difference as that between a cad who allows himself to be pulled away and one truly in love with Giselle, who refuses to leave her side as she dies. "Restaging the Classics," in Payne, American Ballet Theater, p. 325. See also, Mikhail Baryshnikov's rationale for the latter interpretation, Baryshnikov at Work, pp. 26-34.

Clive Barnes also notes that Hilarion, the suitor rejected by Giselle for Albrecht, used to be played as a villain, but now is usually presented as "likable." "Ballet: Fine Baryshnikov and Makarova in 'Giselle,'" New York Times, January 7, 1975; see also, Alan M. Kriegsman, "Ballet Nacional de Cuba," Washington Post, June 12, 1978: "Unlike most Hilarions one sees, he emphasized not a spiteful, bruised ego, but adoration for Giselle and heartbreak over her collapse - he was so compelling one wondered why Giselle ever turned him down." Return to text

(14) Quite common are changes in small dramatic details, such as the way Albrecht first gets Giselle's attention, how Giselle hides from her mother, or who pours the wine for the royal party. It is especially instructive to note that many

differences in performances by the same company. In performances seen by this writer on April 4, 1975, of American Ballet Theatre's production of Giselle, Marianna Tcherkassky and Mikhail Baryshnikov led the cast at the matinee, and Natalia Makarova and Ivay Nagy performed in the evening, with numerous differences in detail. Nagy first got Giselle's attention by making kissing noises, while Baryshnikov clapped his hands. Makarova hid from her mother by vanishing into the crowd of peasants, while Tcherkassky follows the most common practice of crouching behind Baryshnikov. In Makarova's performance, the wine for the royal party in the first act was poured and served by her mother, while Tcherkassky did this herself. Makarova knelt to kiss the hem of the dress worn by the Princess, while Tcherkassky rubbed her cheek against the garment near the waist. Such small differences are quite common in different performances of the ballet, but they are especially striking when they occur in performances by the same company on the same day with almost identical casts, except for the lead roles. Return to text

(15) For example, the recent production by the National Ballet of Cuba actually reinserts some old time by Giselle's mother telling of Giselle's possible doom. Anna Kisselgoff, "Dance: Miss Alonso Stars in 'Giselle,'" New York Times, June 15, 1978; see also, Erik Bruhn, "Restaging the Classics," in Payne, American Ballet Theatre, p. 326. Return to text

(16) Contemporary companies commonly alter the original score by repeating various passages. The current ABT production, for example, repeats such passages as the first act arrival of the royal party and the Wilis theme in the second act during the famous passage in which the corps moves across stage in a cantilevered arabesque position. Comparisons of recording of "Adam's own orchestration" by Vienna Philharmonic, Herbert von Karajan, conducting (London: CS-6251) and performances seen by this writer of ABT's production of Giselle at the Kennedy Center and the Uris Theatre, New York, New York, 1975-81. Return to text

(17) Music by another composer, Burgmuller, was inserted for the peasant pas de deux in the first act, although this is not used in all productions. Humphrey Searle, Ballet Music (New York: Dover Publications, Inc., 1973), p. 59. Return to text

(18) "Such music is of course ineffective in the concert hall - but that is not its place. It is intended to be listened to as an adjunct to the dance, and it supports the dancing by adding its rhythmical, emotional, atmospheric, or dramatic qualities." Searle, Ballet Music, p. 58. Return to text

(19) More radical innovations have drawn severe criticism. Mikhail Mordkin's 1939 version for Ballet Theatre featured ". . . horrendous costumes . . . for the second act - the Wilis were swathed in cellophane . . . and their bodies and tulle skirts were sprinkled with sparkling pailletties. . . " payne, American Ballet Theatre, p. 39. In the 1946 production, the Wilis were dressed in black tulle and blue underskirts, making them "vengeful spirits," not "the airy, sylph-like creatures to which audiences had grown accustomed." Ibid., p. 148. The new production was ". . . only tolerated by the public and the critics. . . ." Ibid., p. 150. Return to text

(20) The second act of Giselle has been performed with as little as a single cross on stage and as much as an authentic-looking forest, complete with trees, logs, and grass around Giselle's grave. Return to text

(21) E.g., critic Edwin Denby notes that Carlotta Grisi, who was the original Giselle in 1841, used wires in the second act to "amplify" her leaps. He also claimed in 1943 that these wires were still being used in productions in Paris, London, and Leningrad. Edwin Denby, Looking at the Dance (New York: Horizon Press, 1949), p. 28. Whether or not the latter is true, the use of wires, acceptable in the nineteenth century, is unheard of in current productions. Return to text

(22) See Chapter III, notes 86-88. Return to text

(23) Interview with Robert Irvine, Musical Director, New York City Ballet, on the Public Broadcasting System telecast of the NYCB's Coppelia, "Live from Lincoln Center," January 31, 1978. Return to text

(24) "Tudor and Balanchine themselves may constantly fuss with their ballets: Tudor may assign one ballerina a double role in "Undertow," while Balanchine may redo the mime and scrap the decor and even the very title of "Ballet Imperial." But as the authors of the works, their changes are comparable to a poet's revisions. No one else, though, would dream of touching these ballets." Jack Anderson, "Idealists, Materialists, and the Thirty-Two Fouettés," Ballet Review, V (No. 1, 1975), 15. Return to text

(25) Balanchine's rehearsal mistress, Rosemary Dunleavy, explains the importance to her of following those interpretative aspects: "'It's not just doing the right steps in the right time. It's giving those dance phrases the energy, the impetus, the particular rhythms that Balanchine wants them done with . . . . I believe that those of us who are

working with Mr. Balanchine, who've seen and heard how he wants his ballets done, can help to preserve that,'" Tobi Tobias, "She Knows Balanchine's Ballets by Heart," New York Times, January 21, 1979.

Suzanne Farrell disputes the implication, however, that this allegiance to Balanchine's interpretive aspects leaves dancers with absolutely no freedom to add their own touches. Discussing her role in Tzigane, she has said, "The whole first solo is pretty much mine. I do like to suggest and sometimes I improvise on stage, but honestly and with no tricks, not to be rude or assertive but because things happen on stage. And you use whatever happens. I never want to turn in the same performance." Jennifer Dunning, "The Ballet Odyssey of Suzanne Farrell," New York Times, February 11, 1979. Return to text

(26) "Balanchine treats even his landmark achievements as organic, evolving material, from time to time redesigning elements in them to reflect a given performer's gifts (or conceal his or her weaknesses), and to embody his own changing vision." Tobi Tobias, "She Knows Balanchine's Ballets by Heart."

In his Nutcracker, ". . . Mr. Balanchine apparently permits the male star to show himself off . . . thus, in one such passage, Mr. Baryshnikov danced choreography featuring turns in place, while at the matinee the choreography danced by Helgi Tomasson featured beats." Jack Anderson, "Dance: New Baryshnikov," New York Times, December 12, 1978.

Baryshnikov's solos in Balanchine's Coppelia have also been altered from Balanchine's other versions. Critic Arlene Croce has said she thinks the solos were not changed enough to make them appropriately spectacular for Baryshnikov! "Fashion at the Ballet," New Yorker, December 4, 1978, p. 220.

Balanchine has also recently "reconceived" some of his ballets for television. For Chaconne, ". . . the style of a Baroque court-ballet has given way, progressively to a romantic reverie." Anna Kisselgoff, "Dance on TV-A New Phase," New York Times, December 10, 1978.

See also, Don Daniels, "Academy: The New World of Serenade," Ballet Review, V (1975-6), 1-12. Return to text

(27) Although a few Balanchine ballets have plots (e.g., Coppelia, Nutcracker, Prodigal Son, Pulcinella), the vast majority are plotless. See, Nancy Reynolds, Repertory in Review: 40 Years of the New York City Ballet (New York: The Dial Press, 1977) Return to text

(28) Chaconne, once performed in practice clothes, is now presented in classical Greek-like costumes by Karinska. Reynolds, Repertory in Review, pp. 330-2.

NYCB also has presented Fokine's Les Sylphides, with the original title of Chopiniana, in white practice clothes instead of the traditional romantic tutus. Reynolds, Repertory in Review, pp. 281-2. Return to text

(29) Robert Joffrey, currently reviving several twentieth-century masterpieces by Leonide Massine, Agnes de Mille, and Frederick Ashton, ". . . is committed to redoing the works as they were presented at their debuts. 'If you can't do them properly, the way the creators intended, then you shouldn't do them at all,' he said." Ken Sandler, "'I'm Too Busy to Choreograph,'" New York Times, November 12, 1978. Return to text

(30) Labanotation has been available since 1928 and Benesh Notation is also widely available. "Benesh notators say their techniques are faster and easier to read than Labanotation. The adherents of Labanotation say theirs are more accurate, more detailed and more responsive to the choreographer's intentions." Still, ". . . only a few dancers can read a Labanotation or Benesh score, and no prominent choreographer writes one." Earl Ubell, "Dance Notation Steps Into A New Era," New York Times, October 24, 1976. Return to text

(31) Arnold L. Haskell notes the early use of verbal descriptions: "In the early days, when the main feature of ballet was its geometric pattern, it was comparatively simple to fix the movements in a system of script. Even when the technique of ballet became far more extended but ballet meant two or three persons actually performing with the corps de ballet as a decorative background, the problem was an easy one, especially as the music was simple in rhythm. The movements of classical ballet were restricted to a framework that it is possible to describe in words so that a dance could practically be transmitted by post." Ballet (Harmondsworth, England: Penguin Books, 1938), pp. 49-50.

Edmund Morris also notes early ways of recording dance, but finds film preferable. "From the earliest days of cave-painting, men have attempted to record the ephemeral beauty of the dance. Choreography was formally committed to paper as long ago as the Renaissance, and countless prints and photographs make it possible to reconstruct, albeit jerkily, the flow of romantic ballet. In this century, of course, we have been fortunate in that the invention of the movie camera, not to mention videotape, has made it possible to capture at least a trillionth of the mimetic movements into which the human body is constantly breaking." Nowhere in his lengthy article on preserving dance does he even mention contemporary notational systems. "Where Dance Memories Leap To Life," New York Times, August 10, 1975.

In the sixteenth century, Thoinot Arbeau developed a "system of dance notation." Anatole Chujoy, The Dance

Encyclopedia (New York: A. S. Barnes and Company, 1949), p. 16.

Gregorio Lambranzi published, in 1716, New and Curious School of Theatrical Dancing, which ". . . gives the theme and air for each [dance], suggestions for steps, and, by means of an engraving, conveys the style-atmosphere and setting of the dance, and the costume to be worn." the system is not a strict rendering of the dances, but ". . . a rough outline . . . leaving the dancer to fill in the details. . . ." Derra de Moroda, trans., Cyril W. Beaumont, ed. (New York: Dance Horizons, 1966), p. 8. Return to text

(32) Ann Hutchinson cites the 1938 written notation of Kurt Jooss' The Green Table and the 1949 use by Ballet Theatre of the score of Billy the Kid for a revival as examples of the use of notation for purposes of reconstructing ballet. Other works notated, although not yet needed for such purposes, are Balanchine's Symphony in C, Orpheus, Theme and Variations, and Bouree Fantasque, and modern works by Doris Humphrey and Charles Weidman. "The Preservation of the Dance Score through Notation," in Sorell, Many Faces, pp. 154-5. Return to text

(33) The cost of notating in Labanotation Jerome Robbins' Les Noces, a 25-minute work for 26 dancers, was $4,000 for 1000 hours of work, resulting in a 365-page score. Recently, the technology has been developed for using a computer to considerably speed up this process, which may eventually help reduce the cost. Earl Ubell, "Dance Notation Steps Into A New Era."

Arnold L. Haskell believes films vastly superior to written notation, regardless of cost: "Although there are a variety of systems of script, not one of them is of much practical use today. . . . the film alone could come to the rescue, a film that could analyze in slow motion, and already the cost of ballet is prohibitive." Ballet, p. 50. Return to text

(34) See, e.g., the profile of Rosemary Dunleavy, Balanchine's assistant, who ". . . is responsible for teaching and maintaining the work of the corps de ballet in the prodigious Balanchine repertory - some 35 ballets in performance this season and several times as many retrievable upon demand. . . . Most of this choreography is recorded in Miss Dunleavy's head." Tobias, "She Knows Balanchine's Ballets by Heart." Return to text

(35) Dance critic Jack Anderson explains this difference between centuries as one between Idealists, before the twentieth century, and Materialists, in this century: "Idealist dance-goers regard a dance as the incarnation in movement of ideas or effects; typically, Idealists may not mind that in

different productions of what is ostensibly the same ballet steps are changed, provided that the alterations express the same idea, produce the same effect, or illumine the work's central concept. The Materialist, in contrast, regards a dance as an assemblage of specific steps . . . from which ideas or effects may be derived." "Idealists, Materialists, and the Thirty-Two Fouettés." Return to text

(36) "Reconstructing a Dance through Notation - And Making It Live!" (Paper presented at the 38^th Annual Meeting of the American Society for Aesthetics, Milwaukee, Wisconsin, October 23, 1980), p. 3. Return to text

(37) Ibid. Return to text

(38) Ibid. Return to text

(39) "The Aesthetic Work of Art," in Art and Philosophy, ed. by W. E. Kennick (New York: St. Martin's Press, 1964), p. 122. Return to text

(40) Ibid. Return to text

(41) Forms and Substances in the Arts, trans. By Salvator Attanasio (New York: Charles Scribner's Sons, 1966), p. 201. Return to text

(42) Ibid., pp. 201-2. Return to text

(43) Joseph Margolis, "The Identity of a Work of Art," Mind, LXVIII (January, 1959), 50. See also, Joseph Margolis, Art and Philosophy (Atlantic Highlands, N.J.: Humanities Press, 1980), p. 63). Return to text

(44) "The Identity of a Work of Art," 48; Art and Philosophy, p. 63. Return to text

(45) "The Identity of a Work of Art," 48; Art and Philosophy, p. 62. Return to text

(46) "The Identity of a Work of Art," 41; Art and Philosophy, p. 55. Return to text

(47) "The Identity of a Work of Art," 46-7; Art and Philosophy, p. 61. Return to text

(48) Art and Philosophy. Return to text

(49) Many current revivals of roles made famous by Vaslav Nijinsky in the early twentieth century are invariably evaluated by comparisons with Nijinsky, of whom there is absolutely no film footage, although there is a wealth of still photographs. For example, "Nureyev was disappointing in 'Spectre.' . . . His poses and arm movements were undoubtedly modelled after photographs and descriptions of Nijinsky, but they emerge in the wrong keys. . . ." Alan M. Kriegsman, Washington Post, August 9, 1978. See also, Arlene Croce, "Nureyev as Nijinsky, Babilée as Babilée," New Yorker, March 26, 1979, p. 110; Hubert Saal, "Homage to Nijinsky," Newsweek, March 19, 1979, p. 67; Anna Kisselgoff, "Nureyev in Nijinsky's 'Faune,'" New York Times, March 25, 1979.

Anna Kisselgoff suggests that Nureyev's deviation from the original disqualifies his performance as one of Le Spectre de la Rose, although this may be only a neatly turned phrase that does not mean as much as it seems to say: "He danced magnificently in 'Le Spectre de la Rose' but he did not dance 'Le Spectre de la Rose.' The serpentine arms, overly rotated, were clearly wrong. If one goes by the version remembered by Andre Eglevsky, who was coached for a month in this role by [the choreographer of 'Spectre, Michel] Fokine, many of the steps were missing here and the virtuosity included was exactly what Fokine said he did not want." "Nureyev Dances Nijinsky Role," New York Times, July 26, 1978. Return to text

(50) These devices are, admittedly, not unproblematic, especially since memory may conflict with other records of these definitive performances. Beryl Grey has reported the problems reviving Nijinsky's famous role Le Spectre de la Rose for London Festival Ballet: "The company's revival . . . brought down upon it the wrath of the critics because of a chair used that appeared in early designs and critical accounts but was later discarded. Even photographs, Miss Grey finds, are not dependable. 'People forget that when dancers go to photo calls - which are absolute hell, a bit like packing up for the country - they forget their bits and bobs and often have to put on someone else's trousers.'" Jennifer Dunning, "She runs the London Festival Ballet," New York Times, July 18, 1978. Return to text

(51) Margolis, "The Identity of a Work of Art," 49; Art and Philosophy, p. 63. Return to text

(52) "The Identity of a Work of Art," 45; Art and Philosophy, p. 64. Return to text

(53) "The Identity of a Work of Art," 44-5; Art and Philosophy, p. 59. Return to text

(54) "The Autographic Nature of the Dance" (paper presented at "Illuminating Dance" Conference, Temple University, Philadelphia, Pa., May 5, 1979). Return to text

(55) Ibid., p. 1. Return to text

(56) Infra, note 63 et seq. Return to text

(57) "The Autographic Nature of the Dance," p. 2. Return to text

(58) Ibid., p. 5. Return to text

(59) Ibid., pp. 6-8. Return to text

(60) Ibid., p. 8. Return to text

(61) Ibid., p. 11. Return to text

(62) Ibid., p. 13. Return to text

(63) Languages of Art (Indianapolis: The Bobbs-Merrill Company, Inc., 1968), p. 212. Return to text

(64) Ibid. Return to text

(65) Ibid., p. 121. Return to text

(66) Ibid., p. 212. Return to text

(67) Ibid., p. 121. Return to text

(68) Ibid., p. 212. Return to text

(69) Ibid., p. 213. Return to text

(70) Ibid., p. 121. Return to text

(71) Adina Armelagos and Mary Sirridge, "The Identity Crisis in Dance," JAAC, XXXVII (Winter, 1978), 129-39. Return to text

(72) Although the Copyright Office is unable at this time to report the extent of registration under the new copyright Act, there are indications that interest is growing rapidly among choreographers in the new protections available. "Special Problems of Choreographic works," 1980 Committee Reports-Section of Patent, Trademark and Copyright Law (Chicago, Illinois: American Bar Association, 1980), p. 158. Return to text

(73) Armelagos and Sirridge, "The Identity Crisis in Dance," 129. Return to text

(74) Ibid., 131. Return to text

(75) Ibid., 129. Return to text

(76) Ibid., 133. Return to text

(77) Ibid., 132. Return to text

(78) Ibid. Return to text

(79) Ibid., 133. Return to text

(80) Ibid. Return to text

(81) Ibid., 133-4. Return to text

(82) Ibid., 133. Return to text

(83) Ibid., 134. Return to text

(84) For example, the Washington Ballet was unable to present Balanchine's Concerto Barocco in March 1979, as originally scheduled, because of the unavailability of anyone from the New York City Ballet to teach the work to the company. Conversation by this writer with Peter Grigsby, company manager. Return to text

(85) 17 U.S.C. 102(a)(4). Return to text

(86) 17 U.S.C. 101 et seq. (1976). Return to text

(87) Registration of Claims to Copyright, 37 C.F.R. 202.7 (1976). Return to text

(88) Common law protection under state law, available for unpublished works, should have been attractive to choreographers, since few published notations or films of their works, McMahon, Choreography and Copyright, ART & LAW, January, 1978, at 1-2, and since performance of a work does not constitute publication. Ferris v. Frohman, 223 U.S. 424 (1912). However, no reported cases exist for common law copyright of choreography. 1 M. NIMMER, NIMMER ON COPYRIGHT 123 n. 525 (1975). Indeed, without any fixation of the work, it would be extremely difficult to provide infringement by another performance. B. Varner, Copyright in Choreographic Works, in SENATE COMM. ON THE JUDICIARY 86^th CONG., 2D SESS., STUDIES ON THE COPYRIGHT LAW REVISION 94, 103 (Comm. Print 1961) [hereinafter cited as Copyright in Choreographic Works].

This protection continues to be available in very limited circumstances for works not yet fixed in a tangible form, although there are not yet any reported decisions on this question since passage of the new Act. "Special Problems of Choreographic Works," p. 157. Return to text

(89) Protection is excluded by the statute for "any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work." 17 U.S.C. 102(b). The legislative history makes clear that the provision is intended to codify, without change, the previously recognized "dichotomy between expression and idea.' S. REP. NO. 473, 94^th cong., 1^st Sess. 54 (1975); H.R. REP. NO. 1476, 94^th Cong., 2^nd Sess. 57 (1976). Both the House and the Senate Reports also characterize the things excluded as "intellectual concepts." S. REP. NO. 473, 94^th cong., 1^st Sess. 54; H.R. REP. NO. 1476, 94^th Cong., 2^nd Sess. 56 (1976). Obviously excluded from protection would be such concepts or ideas as "exploration of the emotion of grief," "stylization of a wedding ceremony with one couple and a corps de ballet," "examination of good and evil through an allegory involving swans," and "abstract, angular movement to embody music by Stravinsky."

Not subject to the exclusion would be non-verbal "ideas," in the sense sometimes used by writers on dance that movement itself is "the message," or idea. Lincoln Kirstein, for example, has said, "Increasingly, ballets fail to tell stories. They are about the dance itself, just as symphonic music is about sound." L. Kirstein, Comments and Views Submitted to the Copyright Office on Copyright in Choreographic Works, in SENATE COMM. ON THE JUDICIARY, 86^th CONG., 2D SESS., STUDIES ON COPYRIGHT LAW REVISION 113 (Comm. Print 1961) [hereinafter cited as Comments and Views]. Such movement patterns, if they meet other requirements for copyrightability, should have protection, as this non-verbal sense of "idea" does not seem to be the intent of Congress.

The decisive test for an excluded "idea" would seem to be whether it could be described verbally in conceptual, as opposed to strictly descriptive, terms.

Another obvious exclusion, as a "procedure" or "system," would be any element of a standard dance vocabulary, such as a plié or a pirouette. The Copyright Office would not register, under previous law, "a mere dance step or variation as such, apart from a developed choreographic work in which it appears." COPYRIGHT OFFICE, CIRCULAR 41: CHOREOGRAPHIC WORKS 1 (April 1977). Less obvious, but almost certainly also excluded, would be the development of a strikingly novel step (such as the split tour jete introduced by Mikhail Baryshnikov) or novel movement vocabularies (such as that of Martha Graham). These would be excluded for the same reason that a new atonal musical scale or a new genre, such as jazz music, would be excluded. Kirstein suggests, for example, that "certain choreographers are inventive and do originate certain combinations, but so do musicians invent sound and rhythms, but they cannot reserve them for their unique use, except in the context of a given use, as in a symphony." L. Kirstein, Comments and Views 114. Since many "steps" actually involve several "building blocks" (e.g., preparation-turn-landing), the primary problem here is drawing a distinction between a new step, excluded from protection, and an original combination of steps, probably eligible for protection as part of a complete choreographic work. Return to text

(90) Such prominent choreographers as George Balanchine, Jerome Robbins, and Alvin Ailey commonly made videotapes for such purposes as rehearsal aids and archival collections, but did not submit them to the Copyright Office for registration. Kibbee, Copyright Protection for Choreography, ART & LAW, January, 1976, at 4. Return to text

(91) Kirstein observes: "To my knowledge, there are not half a dozen choreographers in the world actually capable of reproducing works which they have not themselves designed. It is true that some systems of notation have been advanced but the knowledge taken to read them and put them to practical use is another inhibition against plagiarism or theft direct." L. Kirstein, Comments and Views 113. Return to text

(92) S. REP. NO. 473, 94^th Cong., 1^st Sess. 52 (1975); H.R. REP. NO. 1476, 94^th Cong., 2^nd Sess. 53-54 (1976). The House Report alone refers to the Act's "explicit recognition of all forms of choreography." H.R. REP. NO. 1476, 94^th Cong., 2^nd Sess. 52 (1976) (emphasis added). Return to text

(93) The Copyright Office defined "choreographic work" under the previous statute as "a ballet or similar theatri-

cal work that tells a story, develops a character, or expresses a theme or emotion by means of specific dance movements and physical actions." COPYRIGHT OFFICE, CIRCULAR 41: CHOREOGRAPHIC WORKS 1 (April 1977); accord, 37 C.F.R. 202.7 (1976). However, as works need no longer be registered as dramatico-musical works, this definition is not necessarily applicable to the new category of "choreographic work." The creation of the new category suggests that Congress intended to create a broader class of protection. It could be argued that the new category was added solely to resolve any uncertainty that choreography could be protect under the dramatico-musical category. However, if this were the sole aim of Congress, any such doubt could easily have been resolved by clarification in the legislative history of the new act, rather than the creation of an entirely new category.

Notably, the new statute omits any requirement for fixation in writing of any dramatic element present in the work, as would have been required under several early revision bill. B. Varner, Copyright in Choreographic Works 99. Return to text

(94) S. REP. NO. 473, 94^th Cong., 1^st Sess. 52 (1975); H.R. REP. NO. 1476, 94^th Cong., 2^nd Sess. 53-54 (1976). Both reports, using identical language, also stress that the list of categories "is 'illustrative and not limitative,' and that the seven categories do not necessarily exhaust the scope of 'original works of authorship,'" thus further emphasizing the intended flexibility of the categories. Id. See also, "Special Problems of Choreographic Works," pp. 155-6. Return to text

(95) 17 U.S.C. 102(a). Return to text