Psi man mascot

Two Modes of Thought

In a little-noticed 1963 article, Ulric Neisser wrote, "Many writers have distinguished two types of mental processes." One type is well ordered, easy to describe, and controlled through conscious processes such as planning.

The other type is harder to describe and less easy to control. It is characterized by parallel processing (many things going on at once). Neisser added:

It cannot be a coincidence that the two corresponding modes have appeared as alternative possibilities in the design of "artificially intelligent" systems–that is, in programming computers. In that field, the two possibilities are often called "sequential" and "parallel."

What did Neisser point out in his 1963 article?

Neisser pointed out that sequential or serial processing corresponded to what another psychologist, Jerome Bruner, called analytic thought. Simultaneous or parallel processing corresponded to what Bruner called intuitive thought. Here is how Bruner (1960) described the two.

Analytic thinking character­istically proceeds a step at a time. Steps are explicit and usually can be accurately reported...

Such thinking proceeds with rela­tively full awareness of the infor­mation and operations involved. It may involve careful and deductive reasoning, often using mathematics or logic and an explicit plan of attack. Or it may involve a step-by-step process of induction and experiment...

Intuitive thinking character­istically does not advance in careful, well-planned steps. Indeed, it tends to involve maneuvers based seemingly on an implicit perception of the total problem. The thinker arrives at an answer, which may be right or wrong, with little if any awareness of the process by which he reached it. (Bruner, 1960, pp. 57-58)

What is analytic thought? Intuitive thought?

The analytic mode of thought occurs when we focus our attention and use our brains consciously to devise a strategy or plan of action. This is the type of con­trolled processing that Posner described as lighting up the anterior cingulate gyrus in the prefrontal areas of the brain.

By contrast, the intuitive mode of thought does not necessarily benefit from attention. In this mode, important processes must be allowed to happen, because there is too much going on at once for executive processes to guide it all. This type of activity involves parallel processing (what Bruner described as "implicit perception of the total problem").

In Chapter One (Psychology and Science) I suggested that science resembles other tools in amplifying existing functions of the brain. Science extends the human brain's natural ability to model or represent the universe.

Two different categories of scientific research enable two different kinds of powers in science. Observational research reveals correlations, allowing predictions. We might not know why a relationship exists, but we can use it to make a prediction. This resembles the function of intuition.

The other form of research is experi­mental research. It is consciously con­trolled as variables are manipulated to see their effects. That is like analytic thought. The emphasis is on identifying components of a system and how they interact, to understand how the system works. The two powers of science echo the two modes of thought.

The Two Modes Discovered Again

The two modes of thought were rediscovered again in the 1990s. Raichle (1994), writing in the Annual Reviews of Psychology, reviewed evidence from brain scanning experiments He concluded there were two distinct processing pathways in the brain.

One pathway tended to be active during the conscious phase of learning. Another dominated uncon­scious or automatic processing.

Raichle's conscious process corres­ponds to Neisser's serial process or Bruner's analytic process. Raichle's automatic process corresponds to Neisser's parallel process or Bruner's intuitive process. The names change, but the distinction remains the same.

How were "two modes of thought" discovered again in the 1990s? Which mode is "the head" and which is "the heart"?

Similarly, Epstein (1995) wrote in American Psychologist that evidence points toward "two...modes of information processing: a rational system and an emotionally driven experiential system." Epstein cited many studies suggesting a "conver­gence of a wide variety of theoretical positions" supporting the distinction.

Epstein said this is why we can experience "conflicts between the heart and the head." One system is active during conscious, controlled, or intellectual activity (the head), the other during unconscious, unwilled, emotional responses (the heart). The two do not always agree.

The perennial distinction surfaced again in 2011 with Daniel Kahne­man's Thinking: Fast and Slow. The book was a bestseller, described in the New York Times as "an aston­ishingly rich book: lucid, profound, full of intellectual surprises."

Kahneman described two cognitive systems. "System 1 is fast, intuitive, and emotional; System 2 is slower, more deliberative, and more logical." Sound familiar?

The two modes of consciousness can sometimes be doubly dissociated. In other words, brain injury can knock out one type of process while leaving the other undamaged, and vice versa.

This can happen with the syndrome called prosopagnosia or facial blindness. Some brain damaged patients cannot consciously recognize any face, including their own or that of the examiner.

However, these same patients show an unconscious emotional response to familiar faces, when monitored by an apparatus similar to a lie detector. The "heart" recognizes the familiar face, although the "head" does not.

Other brain-damaged patients show the exact opposite pattern. They show no emotional reaction to a familiar face, but they can consciously recognize and name familiar faces (Newcombe & Lie, 1995).

What is evidence of "two modes of processing" from the study of facial blindness?

Another example involves fear conditioning. When a person learns that a tone comes before an electric shock, normally there are two consequences occur:

1. The subject has conscious memory of the connection. The subject knows the tone comes before a shock.

2. The subject shows an emotional (autonomic) response to the tone, as measured on a device similar to lie detector. If the tone is sounded at a random time, the needles jump.

Bechara, Tranel, Damasio, Adolphs, Rockland, and Damasio (1995) found these responses, too, could be doubly dissociated. One case was a patient with damage near the amygdala.

The patient with amygdalar damage remembered the fact that a loud horn was sounded before each electric shock. But she showed no emotional reaction to the horn.

Another patient had damage to the hippocampus, an area involved in creating conscious memories. He showed an emotional reaction to the horn, but he was unable to recall it had been paired with a shock.

Finally, a third patient had damage to both the amygdala and hippocampus. That patient "acquired neither the [emotional] conditioning nor the facts."

What is evidence for two modes of processing from the study of fear conditioning?

If the two processes can be doubly dissociated, each eliminated without affecting the other, they represent different processes in the nervous system. This supports Neisser's insight (and everybody else's similar insight).


Bechara, A., Tranel, D., Damasio, H., Adolphs, R., Rockland, C., & Damasio, A.R. (1995) Double dissociation of conditioning and declarative knowledge relative to the amygdala and hippocampus in humans. Science, 269, 1115-1118.

Bruner, J. S. (1960). The Process of Education. Cambridge: Harvard University Press.

Kahneman, D. (2011) Thinking: Fast and Slow. New York: Farrar, Straus & Giroux.

. Newcombe, N. & Lie, E. (1995) Overt and covert recognition of faces in children and adults. Psychological Science, 6, 241-245.

Raichle, M. E. (1994). Images of the mind: Studies with modern imaging techniques. Annual Review of Psychology, 45, 333-356.

Write to Dr. Dewey at

Don't see what you need? Psych Web has over 1,000 pages, so it may be elsewhere on the site. Do a site-specific Google search using the box below.