"EVERYBODY'S right, and everybody's wrong!" Thus spake Alexandra Danilova, the celebrated ex-ballerina and associate of George Balanchine at the New York City Ballet. Her pronouncement came during a panel session at the four-day Dance Critics Association conference, which ended in New York last Sunday. It was intended to sum up her observations on the restaging of ballets, but it was also a perfect epitaph for the entire discussion of the conference theme -- "Reconstruction and Revival" -- which, in turn, is an issue of burning concern in the dance world today.

The critics' conference was planned to coincide with a bevy of performances by leading dance companies -- a conjunction of forces extraordinary even for New York, the world's acknowledged dance mecca. Ongoing simultaneously with the conference was the New York City Ballet's eight-day Stravinsky Festival at the New York State Theater, honoring the composer who was Balanchine's chief collaborator on the occasion of the centennial of his birth; a three-week season by the Martha Graham Dance Company at City Center, including three key revivals and a pair of new works (today marks the final performance); and a two-week visit by the Royal Danish Ballet, fresh from their Kennedy Center triumphs, at the Metropolitan Opera House, along with the usual variegated slew of smaller-scale events throughout the city.

These three major troupes, as it happens, offer a striking illustration of the contemporary pertinence of the conference theme. The Danish company rests largely on its perpetuation of the legacy of August Bournonville, the 19-century ballet master who provided the company its technical and esthetic basis, as well as a major portion of its repertoire. Both the New York City Ballet and the Graham troupe are, primarily, embodiments of the genius of single creative figures whose careers have lasted already for more than 50 years. Balanchine is 78 and has had heart surgery in recent years; Graham is 10 years older and is presently suffering from shingles. As it is, their companies are continually involved in processes of revival and reconstruction, and dance lovers cannot help but ruminate on what will happen when they are no longer among us -- obviously, posterity will be utterly dependent on our ability to keep their works alive beyond their earthly presence.

This is no easy matter. As the critics' conference underscored in many ways, it is hard enough to ensure that a given piece of choreography retains its identity and character even when its creator is alive, active and involved. The first law of dance is its ephemerality, greater in degree than for any other art. John Taras, one of the NYC Ballet's choreograhers and a conservator of the Balanchine repertoire, noted at one meeting that no dance is ever the same twice: The same ballet with an identical cast is inevitably different -- significantly so -- from one performance to the next.

This is where Danilova's parting epigram comes in, for this seeming tautology expresses the profoundest truth about dance history and the conditions under which masterworks are transmitted from one generation to another. Everybody is wrong -- to extrapolate from Danilova's remark -- in the sense that the premier performance of a new work is unique; all subsequent renditions of the same material are irreversibly bound to deviate from the original, sometimes by intention but in any case by the very nature of the dance experience. In fact, as was often noted at the dance critics' sessions, any dance performance that is not a premiere is ipso facto a reconstruction or a revival. Everybody is right -- again in line with Danilova's meaning -- in the sense that, since no absolute authority is possible, anyone is "entitled" to determine what the true substance and nature of a dance work amounts to. In a way, it is only our desire to be faithful to the intentions or spirit of a choreographer that stands between the accepted conventions of an era, on the one hand, and esthetic anarchy, on the other.

Both the Stravinsky Festival and the Graham season pointed up many of the thornier aspects of these matters as they affect actual production and performance. It was clear from the outset that the new offerings of the Stravinsky commemoration had scant chance of becoming more than minor achievements, for the simple reason that the composer's storehouse has already been picked clean.

Balanchine has been choreographing to Stravinsky's music since 1925, and the company's 1972 Stravinsky Festival engendered at least four ballet masterworks by him; apart from those scores Balanchine has deemed unsuitable for choreographic treatment -- "Le Sacre du Printemps" and "Les Noces" among them -- he has already used all the major compositions with the exception of the 1940 Symphony in C (which John Clifford choreographed for the company in the late '60s).

Of the new things Balanchine attempted for the new festival, two were exceedingly brief -- "Tango" and "Elegie" -- and two were less dance works than theatrical tableaux -- "The Flood" and "Persephone" (Balanchine himself said of the latter that it is "not a ballet"); the remaining piece, "Variations," he was unable to finish in time for the festival. (Balanchine confirmed, by the way, that the company will bring its new Stravinsky acquisitions to the Kennedy Center this fall, except "The Flood" and "Persephone" which are too unwieldy for travel; "Variations" may come, he said, "if it's good.")

There have also been a small number of substantial contributions from other NYC Ballet choreographers (altogether, five were represented, plus guest Lew Christensen) including Jerome Robbins and Peter Martins. But inevitably, the programs found their true glory in the great Balanchine-Stravinsky creations of past years, of which 11 were staged.

Among those I saw were the "Divertimento from 'Le Baiser de la Fee'" and "Apollo," both of which exemplified problems that pertain generally to revivals. In the case of "Divertimento," what was demonstrated was the power of the dancer to transform the dance: Though Helgi Tomasson was as stirring as ever in his original role and young Katrina Killian danced beautifully, the latter missed almost entirely the fey pathos that was the hallmark of Patricia McBride's poignant interpretation, with the consequence that the ballet itself seemed not just diminished but altered, though not a step had been changed.

The performance of "Apollo" in its recently revised form, stripped not only of the decor and costumes but also the entire Prologue depicting the birth of Apollo, evoked the subtitle of the critics' conference -- "Which Dance Is This, Anyway?" Aside from interpretive colorations attributable to Peter Martins as Apollo and the rest of the cast, the changes in this case were Balanchine's own deliberate alterations. Certainly it's the prerogative of the choreographer to make whatever changes he wants, and backstage during the festival Balanchine left no doubt that he strongly prefers the truncated version. What's more, he said he may go even further in the future in eliminating narrative vestiges from the ballet to come still closer to his ideal of unadorned classical abstraction.

"The Prologue is passe," he said. "What's good about the ballet is Apollo confronting the Muses; it's a ballet about dancing, about classical dancing, and there's still one variation I'd like to change some day, to take out that awkwardness of Apollo at the start, as if he's still learning to walk and dance -- maybe I'll do it, if I'm not too tired, lazy or old." Balanchine is free to revise as he chooses, but we are free to favor one version or another. What, though, is history to make of this, assuming alternative versions survive? Which "Apollo" deserves to become the paradigm for future generations, the last one, or the one that seems in the long run the most esthetically satisfying?

The lesson of the Graham season was that sometimes it is easier to preserve the letter than the spirit, when what is involved is a dance work that had its origin in a period temperamentally far removed from the present. The one work I caught, "Dances of the Golden Hall," is in itself a throwback; Graham has called it her Denishawn piece, and though the brassy Halston loincloths are distinctly of today, the grandiose mythic conception and vague ethnicity make the piece look tawdry and outlandish to contemporary eyes.

The most interesting item was the revival of the stark, ritualistic masterpiece of 1931, "Primitive Mysteries," whose strengths, both structural and poetic, somehow assert themselves powerfully despite the attritions time has imposed. The problem here is that the dancers of Graham's present company (apprentices drilled by Yuriko made up the ensemble; Takako Asakawa was the virginal celebrant) seem to lack inner connection with both the vision and the visceral impulses that were the work's original source.

The performance had depth, conviction and sense of weight, but the ultimate qualities of belief and earthy brawn evaded capture -- the dance looked studied rather than lived in. Dorothy Bird Villard, who danced in the premiere of "Primitive Mysteries" had spoke of it at the conference, had high praise for the company's present effort: "This time [compared with the 1977 revival] I felt as if I were doing it, it was so close to what it had been." But she added, "What I didn't feel was the thing behind it . . . the sense of love in it." Other revivals of the season, and current performances of older work, were similarly afflicted -- at a gala one evening, Aaron Copland led the orchestra in an incandescent reading of his "Appalachian Spring," but the dancing, devoid of the idealistic fervor than can rejuvenate this Graham classic of 1944, turned the piece into a kind of frontier soap opera.

The conference dealt with all these questions in considerable depth. There were panels on Balanchine, on Graham and Cunningham, on Bournonville, and on the methodologies and dilemmas of reconstruction and revival; most of the panelists were people, like Taras, and Irina Nijinska (daughter of Bronislava Nijinska) and Bertram Ross (choreographer, formerly a longtime principal and partner of Graham) who cope professionally with the problems in a constant way.

There were also movement sessions on mime (led by Niels Bjorn Larsen of the Royal Danish Ballet) and early modern dance, as well as film and video showings comparing verions of masterworks of diverse style. The highlights included a performance of Fanny Elssler's "Cracovienne," the celebrated character dance that Bournonville so admired he jotted it down in his own dance shorthand, here painstakingly reconstructed by Danish scholar Knud Arne Jurgensen and captivatingly danced by RDB principal Lis Jeppesen; and a brief, tantalizing film demonstrating the work of Millicent Hodson, an American scholar who is intrepidly trying to retrieve the possibly irretrievable -- Vaslav Nijinsky's legendary, riot-provoking choreography for "Le Sacre du Printemps" of 1913. No fundamental enigmas were resolved; the bottom line, as far as conclusions were concerned, was Danilova's "everybody's right and everybody's wrong." But the central issues got a thorough airing, and what did emerge was a sense of common resolve to help put dance in touch with its own elusive past, both distant and immediate.