The 13th Warrior, the medieval adventure movie with Antonio Banderas as an Arab, opens at theaters everywhere today. Since this movie provides an entertaining, attention-holding two hours at the theater, with a rare appearance by Omar Sharif, and since I am aware of the difficulty of making any kind of successful movie, let alone a relatively ambitious costume/action spectacular, I can recommend it as an acceptable way to spend eight or nine dollars.

However, I expect that critics will not be kind to this film, and would be hard-pressed to really defend it against unkind reviews. The public will find its own way as usual.

But the easiest criticism of The 13th Warrior - that it is hampered by silly and improbable action scenes - will not go very far in explaining what it is that the film fails to achieve. This movie commits the sin (a mortal sin, really, when you're talking about a period picture) of not savoring its own atmosphere. All of its flaws can be traced to a lack of faith in period moviemaking. I have not read Eaters of the Dead, the Michael Crichton book on which the movie is based, nor have I followed the pre-release news reportage on the film's production. But there are enough disconnected elements here - settings used only briefly, plotlines that dwindle away - to indicate that this movie is another victim of post-focus-group recutting. At some point, screening audiences reacted negatively, and large chunks of the film were lopped off. There's nothing wrong with that (Many anecdotes exist of how The Blair Witch Project was reshaped into its successful final form through post-screening re-edits), but what is lacking in this movie is something no preview audience could provide - a willingness to indulge its own setting.

It's certainly presumptuous for a movie piker to issue such nebulous critical directions, but successful period setting is easy enough to recognize when you see it. The unjustly maligned Amistad features a wonderful scene in which a boatload of escaped Africans comes ashore in the woods of early 18th-century Connecticut, and the first sign that they are in America and not - as they expected - in their homeland is when a man rides by on a velocipede - one of those proto-bicycles that the rider propels by pushing his feet on the ground, Fred Flintstone-style. The scene communicates important information, and surprises the audience while doing so. A successful period production understands that the old days were not simply cruder than our own times, but also that the people were more sophisticated than their circumstances would indicate. Period detail involves appreciating how people once lived their lives, and moviemaking involves turning that appreciation into entertainment. Michael Mann's 1992 version of Last of the Mohicans remains the benchmark for this kind of thing, turning its creaky frontier tale into an intense, improbable and swashbuckling adventure - and not by introducing new elements but by burrowing deeper into the story's 18th-century texture.

It's hard to imagine a more promising premise than The 13th Warrior's. The movie's fish-out-of-water, Arab-among-Vikings premise should be good for at least half an hour of intriguing screen time. The movie can barely milk it for five minutes. This is particularly frustrating because Antonio Banderas achieves the precise note of fussbudget sophistication that is needed to make this work. He registers disgust at seeing Norsemen drink, spit and wash in the same bowl of water. But there's nothing here beyond a grossout reflex; it doesn't require a twentieth century viewer to make a concession to history or culture. Later, at a meeting with the Beowulfian King Hrothgar, a big, dirty dog is allowed to mope around the throne room, but Banderas's Ahmed Ibn Fahlan shows no disapproval - despite the fact that even in contemporary Arab culture dogs are treated with disdain and almost never allowed indoors. It's philosophical contrasts like these that create character, and a sense of place; filmmakers glide over them at their own risk.

Innumerable opportunities to play off the Arab's relative sophistication are also missed. Couldn't we have had at least one scene where Ibn Fahlan employs algebra to get his comrades out of a jam, or cures snakebite with some fancy Middle Eastern medical know-how? When Banderas explains that an enemy with Darth Maul face tattoos is a man - and not, as previously assumed, a monster - everybody takes him at his word. Why? Did he in his Baghdad days meet up with face-painters from Africa or the South Seas? We don't know, and the movie is too stingy to offer even such tarted-up flashes of color. Over time, it becomes hard to take pleasure in a movie that takes so little pleasure in itself.

Even the film's inspired bits are served before their time. The scene in which Banderas learns the Norse language uses a simple and inspired movie trick. He listens to his allies' conversation, and at first, all he and we hear is an uninterrupted flow of gibberish. Then a stray word of English sneaks in, then two or three words, then half a phrase, and so on, until the language becomes recognizable. It's a neat effect, and in its slick way it reflects the mental process of learning a foreign language. But the movie doesn't have the patience to work this effect into the flow of the action, or acknowledge that even under the best conditions this process takes months. Instead, the entire transformation takes place at once; if you're not paying close attention you might think Ibn Fahlan has learned the entire language in one night. The problem isn't that the scene is too Hollywood, or too improbable. It's just too easy.