July 7, 2003

Knowing about knowledge

By Eric Rangus

Larry Barsalou is a cognitive psychologist. To put it simply, he thinks about thinking. He designs experiments he hopes will help demonstrate and explain how the human mind works. Knowledge is a big deal to Barsalou. He thinks about it all the time.

“Brains don’t work like cameras. They have an attentional system that focuses on components of experience,” said Barsalou, whose lightning-quick speech and energetic delivery show he wears his passion for his work on his sleeve. “The brain doesn’t just record an image of an entire scene, instead it captures information about a scene's components such as a chair, a face, an eye, conversation, movement. If you are talking about a chair, it can focus on the overall shape, color or size. The brain is very flexible.”

In spring 2002, Barsalou, Winship Distinguished Research Professor of Psychology, earned a prestigious fellowship from the Guggenheim Foundation—one of just 184 people selected. He took a year’s sabbatical from teaching to work on The Human Conceptual System, a book that explores the history, different approaches, misconceptions and applications of the theories of human knowledge in conceptual systems.

Through a review of previously published research—a good deal of it his own—Barsalou’s book will discuss his theories on the subject, many of them originating outside the mainstream.

What Barsalou discusses so passionately are theories that even one of his own colleagues dismissed as crank and professional suicide 10 years ago. Now, though, many psychologists have begun to embrace them, along with scholars and professionals in other disciplines such as philosophy, literature and even business.

Since the 1950s, the majority of cognitive psychologists have held the view that the brain is similar to a computer—it accesses symbolic knowledge in different parts of the brain as if opening a file.

Dramatic advances in neuroscience over the past 20 years have helped drive Barsalou’s theories that knowledge and thinking involve sensory-motor mechanisms in the brain that simulate the experience one is thinking about—like watching a movie in the mind (although not necessarily consciously).

“If I am thinking about an elephant that’s not present, I’m running my visual system as if I were looking at one,” Barsalou said. “I’ve seen elephants, and my memory system has stored away the state that my visual system was in, and now if I’m thinking of one—the way that this theory goes—I simulate it. I re-enact my visual system partially, not exactly, as if an elephant were present. So I’m not hallucinating an elephant, but I’m getting a vague image of an elephant.”

Try it. Think about an elephant. What is the context? Is the elephant in a zoo? In a circus? Is it on television or a picture in a book? What is it doing? It is moving? What is the angle at which you see it?

The mind is re-creating the experience of seeing an elephant. The creature isn’t there, but the mind is pretending it is.

The elephant experiment is a nice party trick, but Barsalou has many formal avenues to support his theories as well. Through laboratory experiments, Barsalou and other researchers have provided empirical evidence for the view that sensory-motor systems represent knowledge through reenactments. For instance, one Barsalou experiment showed that subjects were able to process sensory properties of an object (if a person was thinking about "perfumed" for soap, she could quickly process "musty" for old books) faster than if she had to switch modalities (like "noisy" for television).

In other words, the brain had to switch gears, and that took just a few milliseconds more time. Knowledge relied on a sensory-motor function, just as he and his new school of cognitive psychologists have postulated.

Knowledge, of course, relies heavily on learning. And the experiences Barsalou investigates are largely subconscious—but nevertheless they unquestionably exist, as indicated through scientific experiments.

“One characteristic of expertise behavior is how automatic it is, like driving,” Barsalou said. “While you’re driving, you are listening to the radio or talking to someone; you don’t have a clue of what you’ve seen or done going down the road, but you’ve done it just fine.” The same is true of processing knowledge in most daily routines. “What you’ve done is mapped conditions in the world to the right responses so many times that when those conditions appear an unconscious part of your brain recognizes it and generates the right action.”

Barsalou is an intense, highly driven guy (instead of working 70–80 hours a week like he did when he was younger, Barsalou now works between 50–60 and he might take a vacation every year). Although he didn’t teach last year, Barsalou still was involved in roughly 40 collaborative projects, many of them involving experiments in his lab, and he spent a great deal of time writing. About one-third of his book is finished, and all of the primary research is complete. He just needs to pull together the narrative. Taking it easy doesn’t come easy for him, but Barsalou does find time to relax through meditation.

Barsalou has dabbled in meditation since he first began studying Buddhist teachings while growing up in the 1960s. “It’s a lot like developing a physical skill like playing tennis or the guitar,” Barsalou said. “Meditation is a particular state you have to train your mind to get into, where you are watching it rather than running along with it.”

Barsalou added that his meditative experiences as a young man helped drive his interest in cognitive psychology. “What I really liked about cognitive psychology is that it is not just a casual, subjective way of finding out how the mind works. You actually can design rigorous scientific experiments to verify what the mind is doing.”

Even when the subject is relaxation, for Barsalou, work is never too far away.