One hundred years ago next month, the Austrian physician published the first edition of his book The Interpretation of Dreams. Many psychoanalysts still follow him faithfully, probing dreams to understand unhappiness.

Some scientists even say that modern research techniques give new support to Freud's century-old idea. But others think that Freud slipped up.

"Neuroscience evidence does show that dreams are motivated by instinctual drives when we sleep," said Mark Solms, a neuroscientist and psychoanalyst at the Royal London Hospital and the Anna Freud Center in London.

Harvard neuroscientist Robert Stickgold disagrees.

"He was 50 percent right and 100 percent wrong," Dr. Stickgold said of Freud. "He had good observations, but he built this interpretive structure that there's no evidence for."

When Freud wrote his book in 1899, scientists didn't know much about how the brain works. The shape and structure of the brain was known, but little was understood about which parts of the brain govern which behaviors.

While Freud tried to understand the brain by listening to his patients and analyzing his own dreams, scientists now have many sophisticated tools at their disposal. Researchers routinely use machines to measure how active different parts of the brain are, and know, at least in general, which parts of the brain control which aspects of behavior.

Although some neuroscientists question whether it's even possible to test Freud's ideas, psychoanalysts today are trying to learn more about dreaming and other unconscious thoughts using modern technology. A new science journal, Neuro-Psychoanalysis, has even been created to help psychoanalysts and brain biologists share their findings.

Before Freud's writings on dream interpretation, dreams were generally thought to be either meaningless, or to predict the future, said Dr. Edward Nersessian, a psychoanalyst in New York City and co-editor of the journal. But Freud, after analyzing his own dreams, proposed that dreams have meaning. He also maintained that the purpose of dreaming was to allow people to sleep, despite thoughts or physical discomfort that would otherwise wake them.

For example, Freud thought that without dreams, feeling thirsty might wake a person. A dream about drinking a tall, cool glass of water would temporarily satisfy the thirst, letting the person continue to sleep.

Likewise, disturbing events of the day also threaten sleep.

"If your mom is angry because you didn't call her, and that was on your mind, you may have a dream that when interpreted would indicate" those feelings, Dr. Nersessian said.

For instance, he said, a dream that your mother buys you a present could be interpreted as the brain's effort to calm your anxieties, allowing you to sleep.

Freud said further that most dreams were strings of thought that represented the fulfillment of unfulfilled wishes. A dream of water fulfilled the wish to drink; the dream about the present fulfilled the wish to make peace with mom. The overriding wish, according to Freud, is to preserve sleep, Dr. Nersessian said.

Freud explained dreams that don't make sense or seem completely out of the blue as necessary distortions of dreams that would otherwise seem unacceptable.

"So a wish to kill your sister is a dream that some other woman dies," Dr. Nersessian said. A dream of actually killing your sister might be disturbing enough to wake you up - and would be what some psychoanalysts consider a failed dream.

But how do psychoanalysts know whether a dream of a woman's death is a sign of hostility to a sister or something else?

Freud himself is said to have been asked the question by a student who wanted to know why he always smoked a cigar. The implication was that the cigar was a phallic symbol. Freud is said to have dismissed the symbolism, replying, "Sometimes a cigar is just a cigar."

But that argument doesn't sit well with some researchers.

It's hard to test whether Freud's theory of hidden meanings is valid "as long as he gets to decide what the cigar is," said Harvard Medical School's Dr. Stickgold.

Dr. Stickgold said he favors the idea that dreaming is the brain's way of processing the events of the day. He and his colleagues have done a study, still unpublished in a scientific journal, of people who were told to play the computer game Tetris. The goal in Tetris is to rotate falling geometric pieces so they fit together neatly.

As many Tetris junkies may know, dreams of the falling pieces are common. Dr. Stickgold instructed 22 people to play the computer game for two to three hours, and then woke them up during the first hour of the night. About 75 percent of the people reported Tetris dreams.

In some ways, Dr. Stickgold said, the Tetris dreams fit with Freud's theory that dreams are based on the events of the day.

However, he says: "This isn't wish fulfillment. This isn't deep trauma that they are playing [out] in disguised form.

"This is the brain that's saying I had a lot of activity today, and I should try to form connections."

It's more likely, he suggested, that the Tetris dreams were just the brain's way of sealing the skills of the game in the person's memory.

Dr. Stickgold also found that people who had played Tetris before the study dreamed of the version they had learned on years earlier, not the version used in the study.

"So there's the brain using this time to go and look for associations between recent memories and old memories," he said.

Dreams can seem to have hidden meaning, Dr. Stickgold argued, because once in a while, just by chance, they really fit with a person's view of life. Dr. Stickgold likened symbolism in dreams to finding significance in a computer spell checker's suggestions. What if, he said, the spell checker suggested substituting "grinch" for "Gingrich?"

If you're of a certain political bent, he said, you could react by thinking "that really fits, that's incredible."

But that's bad reasoning.

"Your spell checker doesn't have a political ax to grind," he said.

Dr. Stickgold said he doesn't discount the idea that the way people react to dreams can shed light on their lives.

"Do they have meaning? Yes, as soon as you wake up and create the meaning," he said.

On that much, psychoanalysts agree. The interpretation of dreams has changed since Freud, said Dr. Leon Hoffman, a New York City psychoanalyst. Whereas Freud did most of the dream interpretation himself, today psychoanalysts and patients work together on dream interpretation. Dreams, and other behaviors, are interpreted in the context of events going on in the patient's life.

But Freudian psychoanalysts do believe that unconscious thoughts influence someone's life, Dr. Hoffman said, and that dreams are a window into the unconscious.

"You're trying to understand in the patient what factors are outside of his or her awareness that motivate his or her actions," Dr. Hoffman said.

While psychoanalysts contend that talking to patients is a good way to learn about the function and meaning of dreams, biologists prefer to go straight to the source - the brain itself.

In the 1950s, '60s and '70s, a popular theory developed that dreams occurred during a special phase of sleep called rapid-eye movement, or REM. The part of the brain that triggers REM is a region near the nape of the neck, called the pons. During REM sleep, the pons was thought to trigger regions in the front of the brain, where the dream images were formed.

Some scientists took this to mean that the pons was haphazardly triggering memories in the brain, like a CD player programmed to choose songs at random. Under this scenario, dreams had no hidden meaning.

Scientists now know that REM sleep and dreams don't always go hand in hand. They also know that during dreams, areas of the brain connected with instinctual urges, such as eating and sex, are active.

Those areas, the ventromedial frontal lobes, seem to be crucial for dreaming, Dr. Solms said. He has studied nine patients with damage to those areas; none of them dream. Studies in animals have shown that the equivalent area drives the basic instinct to satisfy thirst, hunger, warmth and sexual desire. Because if the ventromedial frontal lobes are damaged, dreaming can't occur, Dr. Solms concludes that the lobes are key to dream formation.

The dreams about Tetris don't negate the idea that dreams are driven by instinctual urges, he argues. It may be, he said, that dreaming about the game is a manifestation of a basic, competitive urge.

There are limits, Dr. Solms said, to what biologists can say about Freud's theory that dreams have hidden meanings.

"There's no neuroscience evidence that's relevant either way to support that," he said.

Psychoanalysts agree, saying the only evidence they have comes from sessions with patients.

Just because Freud's theories aren't fully testable by science, Dr. Nersessian said, doesn't mean they're not worth thinking about.