It used to be every novelist's dream to see his book purchased by Hollywood, then projected 25 feet high at the local bijou, with impossibly beautiful people playing the parts that had begun as fractured slivers of the poor schmo's id. He was so happy and rich, he didn't even notice they ruined it.

That still happens, and they still ruin it, as per Tom Clancy et al. In fact, just looking at the newspaper movie section, we see that a good many of the films in release originated as works of fiction. Besides "The Sum of All Fears," from the Clancy canon, there's "Minority Report" (from a Philip K. Dick short story), "The Bourne Identity" (from a Robert Ludlum novel), "About a Boy" (from Nick Hornby's novel) and "Divine Secrets of the Ya-Ya Sisterhood" (from a bestseller by Rebecca Wells).

The novel or short story -- that is, the work of fictional prose with an organic story, a set of recognizable characters and motives, and something wrong with it -- has provided material to films for almost a century. The earliest may have been a silent "Frankenstein." I saw it sometime back at the National Institutes of Health, and in it, all the pathologies of the literary adaptation were on display: They took a great 600-page novel and reduced it to seven minutes. Okay, so it was made in 1904.

But still: "It's not as good as the book." Think how many more times you've heard that than "It's much better than the book." That suggests a relentless reality: It may be that the novel isn't the ideal form for translation into the feature film.

A few adaptations, of course, have been masterpieces, most of those exceptions falling into two categories.

In the first, filmmakers simply ignore or abandon the original novel and use it as a platform to explore similar themes. The great "Dr. Strangelove" is one; the novel, "Red Alert" by Peter George, was largely a straightforward docudrama based on George's excellent knowledge of the inner workings of strategic-defense culture.

Stanley Kubrick and his collaborator, the freaky genius Terry Southern, and George himself reinvented the movie as dark farce on the subject of the end of the world. See it back-to-back with, say, "Fail-Safe," which plays the same story straight (and was itself adapted from a novel), and you see why one has lasted while the other slipped into oblivion.

The other consistently well-done literary adaptation takes off from a peculiar kind of novel: one so short and of a piece that it's almost a movie already. "The Maltese Falcon" is one, "The Postman Always Rings Twice" another. The best example of this is Lewis Milestone's memorable version of "All Quiet on the Western Front." Or think of Martin Ritt's stunning version of "The Spy Who Came In From the Cold" with the magnificently bleached-dry Richard Burton. Yes, it does happen. But not often.

Perhaps the best form to carry the complexity, the density, the drama and the humor of a big novel is the television miniseries. In most cases, you need a dozen hours or so to accommodate the best of prose fiction. Anything less truncates, shortchanges and destroys all those things that somehow are the novel: the milieu, the delightful smaller characters, the odd subplot, the subtle manipulations of point of view, the on-schedule arrival of epiphany.

The masterpiece of novel adaptation on television comes, as does the above "Came In From the Cold," from the complex John Le Carre, whose classic "Tinker, Tailor, Soldier, Spy" was a beloved BBC series; later, an independent British television network did a wonderful job on "Smiley's People." On our side of the Atlantic, the fabulous CBS version of "Lonesome Dove," Larry McMurtry's epic novel of a cattle drive, is generally conceded to be top-drawer.

Novel-into-feature-film, though, usually doesn't work reliably. It never has. It never will. But they do it anyway, the dopes. And they will always do it. Now there's a new kind of novel from which Hollywood draws: the graphic novel, such as "Road to Perdition," which just arrived in a movie version. It's by no means the first, but it may be the biggest and most ambitious.

What is a graphic novel? Well, it's not exactly a comic book. If you call it one, I am sure you will be e-bombed into oblivion by its many aggressive adherents. In fact it is so far removed, thematically and artistically, from the comic book, it might even be seen as an anti-comic book, almost a reaction against the comic's arbitrary limits on complexity and maturity of theme. The comic book -- from Porky Pig to Superman to Sgt. Rock -- has pretty much been a settled issue since the '30s: a short, punchy story expressed over no more than 12 pages, with all dialogue contained in word bubbles and all sentences ending in exclamation points! The characters, super-powered or not, are constants, and the formula is precise. On top of those formal obligations are moral ones, as stated by an oppressively authoritarian comics code (a reaction against so-called "bad" comics in the '50s) that mandates peppy tastefulness.

Most baby boom kids saw through comic books by the time they reached adolescence -- the same time they were noticing how phony movies and TV were. The books had their own kind of ridiculousness to them. Take a late-'50s hero, Sgt. Rock of Easy Company, in the DC Comics series. Rock never got hit and never panicked and always triumphed, shooting down Messerschmitts with his Tommy gun and throwing that grenade so well it always went down the Tiger tank's suspiciously open hatch. The gun went "Budda-budda!," the explosion went "Ka-BLANG!" and at the end, Rock always said, "Okay, Easy, let's move out." The strip was usually drawn by a great comic artist named Joe Kubert from scripts by Robert Kanigher; it was always amazing how much power and poignancy these stories managed. But you could feel, toward the end of the run (I read 'em all!), Kubert and Kanigher's disenchantment with the format. It was so limiting.

When change came -- in the late '60s, that epoch of change -- it came in the form of the graphic novel, imported from the Europe that always took popular art more seriously than did Americans.

This is no time or place for a history of the graphic novel, for two reasons: First, it would take too long, and second, I don't know it. What I do know is that, simultaneous with tear gas in the streets, loud music on the radio and love-ins on campuses, a magazine called Heavy Metal appeared on newsstands, and it pretty much blew the minds of whoever wandered between its pages.

It was adult comic art, and the stories were dark, troubled, violent, almost obscene, rendered with painfully vivid detail. They were long and complex, usually set in a sci-fi context. The high champion of this genre was the French comic artist calling himself Moebius, whose dystopian vision was mind-boggling, to say the least, and wildly influential. ("Blade Runner," ostensibly based on another Dick short story, showed clear Moebius influences.)

The profusion of graphic novels that have been adapted for screenplays -- the "Batman" graphic novels (not the comic books), "From Hell," "Ghost World" and on and on -- makes a salient point: Possibly the graphic novel is a better story vehicle than the novel for a feature film.

That's because a graphic novel isn't really a novel; what it is, really, is a movie in graphics. It tells stories visually, finding images to express emotions that would somehow be beyond a prose artist's ability to convey.

Let's look at a specific image sequence from "Road to Perdition," which was written by Max Allan Collins and drawn by Richard Piers Rayner. It's a seemingly nothing little twist of action, yet the comic team expresses the subtlety, the emotional vividness, with an eloquence that is probably unmatchable in prose.

In this sequence the narrator, young Michael Sullivan Jr., has just learned that his beloved father is really a contract killer for the Looney mob, which runs the Tri-Cities area of Rock Island and Moline, Ill., and Davenport, Iowa, in cahoots with the Capone mob in Chicago. He's seen his father machine-gun a batch of men. But now they are home, and Michael's mom wants to know where he was, what happened. Which parent does Michael please? The answer is that he loves his father still, no matter what, and promises in his mind to never give the secret away. His mother demands an explanation, thinking that he's run away and gotten in some kind of trouble.

Dad replies, "I've spoken to the boy. He won't do it [run away] again. No more questions."

"But . . ." she insists.

"We'll not speak of it again."

They go upstairs and the narrator informs us: "A covenant was formed between my father and me."

Now how to express this idea? Here's how a novelist writing in Michael Jr.'s voice might make the point:

"My father's dry, large hand brushed against mine as we headed upstairs. Whether it was intended or accidental, I'd never know; but there was something in the touch freighted with communication: It was that we had connected in a bond that sealed us off from the world of women and commerce, the world of school and factory. We had bonded and it was permanent and that was that. Blah blah, and then, furthermore, blah blah blah. And, finally, I had to conclude: blah blah and blah again."

Another kind of writer, of the Raymond Carver school of minimalism, might offer something more mysterious:

"His hand brushed against mine."

We'd have to figure that one out ourselves, supply our own interpretation. Maybe we get it, maybe we don't. Maybe we don't even notice it, so flat and unemphatic is it. But the director of a film and the artist of a graphic novel can use the close-up, which makes the distinction; it emphasizes that touch without overdoing it. The artist Rayner conveys all in a single image, spare, unclotted with sentimentality or the wretchedness of powerful thoughts clumsily expressed: We see the two hands brush. That's enough, that's everything.

The artist, who has seen movies, knows the power of the camera to fragment images into small parts, and that those parts carry significance. Thus it is his decision to become the camera. He closes in and finds a way to express the connection between father and son with immediacy and precision, and apparently without conscious symbolism: It's just two hands, after all. But each hand carries a larger meaning of the being to which it's attached, and we know the man is troubled and the boy is adoring and feels privileged to join this private intimacy with the father.

Henry James couldn't get that; nor could Joyce or Hemingway or Faulkner. But Rayner does. Moreover, we read the secret message being communicated, even if we've never opened a graphic novel before, because we've been educated to the vocabulary by movies. We accept the conceit of the close-up without a second thought; we appreciate its symbolic meaning without needing to grasp that it's symbolic.

So for a screenwriter or an adapter, the heavy lifting is already done. It's particularly the case for action sequences, where the profusion of images, the dynamic fluidity with which one becomes another, where clarity is less important than momentum, where the individual components of close-quarters battle -- the flash of guns, the jingle of spent cartridges, the whup of a bullet striking and penetrating flesh, the drift of gun smoke -- are best communicated in a montage, be it a series of semiabstract drawings or a film sequence edited with stunning brilliance.

The poor prose guy -- how far behind the curve is he? Struggling and lost in his snare of motives and psychological realities, it's almost impossible for him to create in words the sensation of action that either a movie or a graphic novel can. And I'd venture a bet that no thriller writer could stay with Rayner, who gives us a brilliant evocation of the gunfights of the '30s.

And that leads to a big question: If the DreamWorks studio goes to the trouble of buying a graphic novel and producing a movie based on it, why oh why did the filmmakers then all but abandon it?

"Road to Perdition" the novel is a black-and-white affair, and it's gritty, violent, Catholic and unbelievably fast-moving. It's a homage not merely to the crime films of the '30s but also to the Japanese samurai sensibility. It is said to be inspired by Kazuo Koike and Goseki Kojima's graphic novel series "Lone Wolf and Cub," about a masterless samurai on the road with his small son. The father's tenderness toward his son contrasts dynamically with his spasms of violence, as the best swordsman in Japan. The violent sequences, like Rayner's, are slick, abstract and compelling.

Collins and Rayner pick up on that. They give their Mike Sullivan a batch of shootouts where he outguns multiple antagonists, and the gimmick is usually that he is unarmed, he nabs a gun from somebody threatening him, and he goes to town, shooting faster and more accurately. But he's more than a dervish, he's a freakin' ninja: He spins, he rolls, he ducks, he darts, he never misses. So when Sullivan goes to speak to his betraying boss after his wife and child have been killed, he blows away 11 henchmen. A few pages later, visiting Al Capone's headquarters, he blows away 12 more.

Tom Hanks, as director Sam Mendes' Mike Sullivan, does none of this. It's not that he's not good with a gun (he is) but the violence has been Occidentalized: made smaller, less outrageous, less bloody, less absurdist, yet also more stylized. The many shootouts are staged like scenes from a highlight reel of Greek tragedy. A good fix? Well, possibly; after all, who wants to see Tom Hanks kill dozens of people if they aren't German soldiers?

On the other hand, when the novel's Michael Sullivan kills a bunch of Capone guys, it makes the war between him and Capone's mob, which fills the book and the movie's second half, dramatically believable. It flows from action; it is a consequence, directly, of Sullivan's own actions and he must deal with it.

In the movie, without that Capone gang massacre, the last half feels unrooted in cause and effect; it is motiveless, arbitrary. You cannot figure out why the Capone mob is going so nuts after this guy. As a courtesy to a downstate crime lord? That motive doesn't really wash; so the movie, in this and many other ways, begins to feel subtly incoherent, almost disassociative. They've fixed the story so much, they've broken it real good.

Indeed, director Mendes has largely abandoned the pictorial brilliance of Rayner and Collins's conception and substituted one of his own.

It's sparser, more formally aestheticized, less spontaneous, neater, less vibrant. It's wet (full of rain all the damned time), which, one critic notes, is less a climatic than a moral condition. It hasn't an ounce of the zest, the gritty vitality, the streety violence of the original. It's not a crime melodrama anymore, but some solemnized uber-myth, as stylized and stultified as Kabuki or Noh.

So the new paradigm isn't all that different from the old one. You write/draw a graphic novel, Hollywood buys it, and you go to the local bijou and see it 25 feet tall with beautiful people playing characters that began as slivers of your id. You're now rich, so you don't notice how they ruined it.

In other words, the picture book was better than the movie.