Robert Schumann, who was just as fine a critic as he was a composer, wrote eloquently about the Schubert Ninth Symphony -- which he happens to have discovered and introduced.

He made a case for size, for rigor, even at the cost of rough edges -- with the implication that some works even benefit from their rough edges. The lesson here is that size and occasional disproportion are valid indeed, and are not necessarily just efforts to make splashes.

The Schubert, alas, is not being played here this summer. But another work, one similarly flawed and magnificent, is drawing audiences: Eugene O'Neill's "The Iceman Cometh," in its American National Theater production at the Kennedy Center. As with the Schubert piece, the imperfections are as unquestionable as the greatness. Do not expect the perfectly gauged graces of a "Hamlet" or a "Cosi fan tutte."

Both the Schubert and the O'Neill seem almost prophetic in their relevance to some of the plagues of our gnarled, disorderly times. The ebb and flow of history make certain works of art more important for certain periods, less so for others (the heroics of the Beethoven Fifth during World War II, for instance). And both the Ninth and "Iceman" stretch almost to the limit what many of us have come to accept as proper pacing in art -- and, by analogy, life.

We have long been into -- and are perhaps beginning to emerge from -- a McLuhan-esque age in which imperatives of the marketplace have too often taken precedence over artistic imagination in shaping our art -- whether it is Warhol or "The Waltons." But neither Schubert nor O'Neill ever heard of fast food, extended-payment deals, ultra-supersaver plans and the rest. Lucky men.

It is an essential element of the genius of both artists that they were not restricted by such patterns. They were not frozen into arbitrary, technologically-dictated time and space frames -- everything from the 30-second television spot to the 60-minute long-playing record. They could be truly daring. They could go to the very essence of what freedom in art is all about -- to discover new insights, new experiences. And their notions are a salutary foil to the way the rigidity of contemporary attention spans may distort the way we listen, the way we read.

"The television audiences and the theater audiences are not the same," says Kennedy Center Chairman Roger Stevens. But few theatergoers are not exposed to television. And who, really, is immune from its influence?

O'Neill wrote "Iceman" in four acts -- the way Shakespeare normally wrote in five acts. Such creators do not ordinarily extend their works to such lengths just because they want to keep you sitting there. ANT Director Peter Sellars says, "Sure, there are moments in the theater when you get impatient. After all, sometimes in a Shakespeare play you will think, well, you've heard the exposition in the first three acts and you know how the last act is going to turn out, so can't you take a break and skip the fourth act? The answer is no. Because in great drama, the accumulation of material always adds to the payoff at the end. And that's one of the things that makes it great."

Granted, the conventions of the five-act play or the four-movement symphony may themselves impose certain restrictions on artists, but nobody ever said plays or symphonies absolutely had to be done that way. And within such forms, there is enormously more artistic flexibility and range than in, say, the 30-minute sitcom. (In fact this "Iceman" actually has about 10 minutes cut, but Sellars maintains it is the most complete version ever given. There was consideration of eliminating one of the three intermissions, but director Jose Quintero strongly opposed it.)

Obviously, Robert Schumann (who rescued the Ninth from total obscurity after coming upon an incredible cache of unperformed Schubert manuscripts owned by the composer's brother) could not have anticipated the way we think more than 150 years later. But Schumann's ideas on the Ninth are much to the point in the case of the O'Neill play:

There is, wrote Schumann, "the heavenly length of the symphony, like that of one of Jean Paul's romances in four thick volumes, never able to come to an end, for the very best reason -- in order to leave the reader able to go on romancing for himself. How refreshing is this feeling of overflowing wealth."

Schumann sounds almost apologetic -- as if he saw from the beginning what would be the problem with this astounding symphony.

Neither length nor ungainliness, though, detract from its greatness. The Ninth, I believe, is Schubert's most formidable instrumental work, along with the String Quintet, but after Schubert wrote it, in his last year (1828), few saw "heaven" in its enormous dimensions.

Musicians simply refused to play it. The symphony was too hard; it was too unidiomatic. They rationalized that it was bad because it had some awkward moments -- indisputable even today. How, for instance, should any violinist be expected to play those 200-odd measures of successive triplets in the last movement? They seemed unnecessary, and they really are not all necessary -- but they are there, part and parcel of one of the mightiest symphonies ever created. Sometimes artists simply have to have their way. If they are going to think big, they have to have latitude to be wrong on details.

In fact, when Mendelssohn directed the London premiere in 1842, the violinists fell into titters.

I cannot be sure, but I have a hunch that both Schubert and O'Neill were positively defiant of the artistic time frames of their day.

In "Iceman's" craggy, sometimes slow, occasionally exasperating 4 3/4 hours, there is an uncompromising willingness to be difficult as if O'Neill thinks that's necessary to make his points. Suavity is not a particular strength of either "Iceman" or the Schubert Ninth; it is as if each artist sensed that the ordeal of bringing off what they were trying to do demanded such intensity of concentration that they didn't dare spare their intellectual energy to observe the conventional graces. For instance, Cherubini's captivating Symphony in D, written at roughly the same time as the Schubert, is considerably more "well made," in the sense that an Ibsen play is. But just as in those amazing unfinished Michelangelo slave sculptures, imperfection may sometimes actually enhance power.

O'Neill must have been well aware that he was severely testing the patience of his audiences when he let "Iceman" unfold at this pace (even "Long Day's Journey Into Night" is considerably less long a theatrical journey). There are 19 neurotic, possessed characters compressed into the claustrophobic confines of "Iceman's" shady little barroom hangout -- and each is developed with such dramatic and syntactical richness that there really isn't any bit part. (I spend more time at the opera than the theater, and I can't really think of an opera with more major characters, short of Wagner's "Ring.")

One critic has referred to the play's "operatic grandeur," and indeed the most eloquent moments of the play are either monologues or duets. It reminds one of early Verdi, with all those fairly conventional set forms. And, for all its brilliance, "Iceman" does largely lack that genius for compression -- a kind of emotional shorthand -- that Verdi brought to those last glories of his long creative life, "Otello" and "Falstaff."

Jason Robards, who plays Hickey in "Iceman," referred in an interview to his crazed final monologue (more than 20 minutes long) as "an aria" that is "sung." The splendid Robards is too modest. This is a "mad scene," the kind of dramatic creation with which a Callas could take even a sometimes-creaky old bel canto vehicle like Donizetti's "Lucia di Lammermoor" and make it into an unforgettable experience. Robards' performance does not suffer greatly by comparison with Callas'. And unlike her, he doesn't have a conductor to help out. Also, he has to do it almost every night.

There are many things that make both the Schubert Ninth and "Iceman" so superior to lesser works, such as "Lucia." But most important is this: the academic phrase for it is "artistic vision," but the more precise word, the one that most characterizes the two personalities in the human sense -- of both Schubert and O'Neill -- is uncompromising "stubbornness" in the service of art.