1905 - 1967

Dr. C. George Boeree

George Kelly was teaching physiological psychology at Fort Hays Kansas State College in 1931. It was the time of the dust bowl and the Depression. Recognizing the pains and sorrows of the farming families of this part of west-central Kansas, he decided to do something a little more humanitarian with his life: He decided to develop a rural clinical service.

Mind you, this was hardly a money-making operation. Many of his clients had no money. Some couldn't come to him, and so he and his students would travel, sometimes for hours, to them.

At first, Kelly used the standard Freudian training that every psychology Ph.D. received in those days. He had these folks lie down on a couch, free associate, and tell him their dreams. When he saw resistances or symbols of sexual and aggressive needs, he would patiently convey his impressions to them. It was surprising, he thought, how readily these relatively unsophisticated people took to these explanations of their problems. Surely, given their culture, the standard Freudian interpretations should seem terribly bizarre? Apparently, they placed their faith in him, the professional.

Kelly himself, however, wasn't so sure about these standard Freudian explanations. He found them a bit far-fetched at times, not quite appropriate to the lives of Kansan farm families. So, as time went by, he noticed that his interpretations of dreams and such were becoming increasingly unorthodox. In fact, he began "making up" explanations! His clients listened as carefully as before, believed in him as much as ever, and improved at the same slow but steady pace.

It began to occur to him that what truly mattered to these people was that they had an explanation of their difficulties, that they had a way of understanding them. What mattered was that the "chaos" of their lives developed some order. And he discovered that, while just about any order and understanding that came from an authority was accepted gladly, order and understanding that came out of their own lives, their own culture, was even better.

Out of these insights, Kelly developed his theory and philosophy. The theory we'll get to in a while. The philosophy he called constructive alternativism. Constructive alternativism is the idea that, while there is only one true reality, reality is always experienced from one or another perspective, or alternative construction. I have a construction, you have one, a person on the other side of the planet has one, someone living long ago had one, a primitive person has one, a modern scientist has one, every child has one, even someone who is seriously mentally ill has one.

Some constructions are better than others. Mine, I hope, is better than that of someone who is seriously mentally ill. My physician's construction of my ills is better, I trust, than the construction of the local faith healer. Yet no-one's construction is ever complete -- the world is just too complicated, too big, for anyone to have the perfect perspective. And no-one's perspective is ever to be completely ignored. Each perspective is, in fact, a perspective on the ultimate reality, and has some value to that person in that time and place.

In fact, Kelly says, there are an infinite number of alternative constructions one may take towards the world, and if ours is not doing a very good job, we can take another!


George Kelly was born on April 28, 1905, on a farm near Perth, Kansas. He was the only child of Theodore and Elfleda Kelly. His father was originally a presbyterian minister who had taken up farming on his doctor's advice. His mother was a former school teacher.

George's schooling was erratic at best. His family moved, by covered wagon, to Colorado when George was young, but they were forced to return to Kansas when water became scarce. From then on, George attended mostly one room schools. Fortunately, both his parents took part in his education. When he was thirteen, he was finally sent off to boarding school in Wichita.

After high school, Kelly was a good example of someone who was both interested in everything and basically directionless. He received a bachelor's degree in 1926 in physics and math from Park College, followed with a master's in sociology from the University of Kansas. Moving to Minnesota, he taught public speaking to labor organizers and bankers and citizenship classes to immigrants.

He moved to Sheldon, Iowa, where he taught and coached drama at a junior college, and met his wife-to-be, Gladys Thompson. After a few short-term jobs, he received a fellowship to go to the University of Edinburgh, where he received a bachelor of education degree in psychology. In 1931, he received his Ph.D. in psychology from the State University of Iowa.

Then, during the depression, he worked at Fort Hays Kansas State College, where he developed his theory and clinical techniques. During World War II, Kelly served as an aviation psychologist with the Navy, followed by a stint at the University of Maryland.

In 1946, he left for Ohio State University, the year after Carl Rogers left, and became the director of its clinical program. It was here that his theory matured, where he wrote his two-volume work, The Psychology of Personal Constructs, and where he influenced a number of graduate students.

In 1965, he began a research position at Brandeis University, where Maslow was working. Sadly, he died soon afterward, on March 6, 1967.


Kelly's theory begins with what he called his "fruitful metaphor." He had noticed long before that scientists, and therapists, often displayed a peculiar attitude towards people: While they thought quite well of themselves, they tended to look down on their subjects or clients. While they saw themselves as engaged in the fine arts of reason and empiricism, they tended to see ordinary people as the victims of their sexual energies or conditioning histories. But Kelly, with his experience with Kansan students and farm people, noted that these ordinary people, too, were engaged in science; they, too, were trying to understand what was going on.

So people -- ordinary people -- are scientists, too. The have constructions of their reality, like scientists have theories. They have anticipations or expectations, like scientists have hypotheses. They engage in behaviors that test those expectations, like scientists do experiments. They improve their understandings of reality on the bases of their experiences, like scientists adjust their theories to fit the facts. From this metaphor comes Kelly's entire theory.

The fundamental postulate

Kelly organized his theory into a fundamental postulate and 11 corollaries. His fundamental postulate says this: "A person's processes are psychologically channelized by the ways in which he anticipates events." (This and all subsequent quotations are from Kelly's 1955 The Psychology of Personal Constructs. ) This is the central movement in the scientific process: from hypothesis to experiment or observation, i.e. from anticipation to experience and behavior.

By processes, Kelly means your experiences, thoughts, feelings, behaviors, and whatever might be left over. All these things are determined, not just by the reality out there, but by your efforts to anticipate the world, other people, and yourself, from moment to moment as well as day-to-day and year-to-year.

So, when I look out of my window to find the source of some high-pitched noises, I don't just see exactly and completely what is out there. I see that which is in keeping with my expectations. I am ready for birds, perhaps, or children laughing and playing. I am not prepared for a bulldozer that operates with a squeal rather than the usual rumbling, or for a flying saucer landing in my yard. If a UFO were in fact the source of the high-pitched noises, I would not truly perceive it at first. I'd perceive something. I'd be confused and frightened. I'd try to figure out what I'm looking at. I'd engage in all sorts of behaviors to help me figure it out, or to get me away from the source of my anxiety! Only after a bit would I be able to find the right anticipation, the right hypothesis: "Oh my God, it's a UFO!"

If, of course, UFO's were a common place occurrence in my world, upon hearing high-pitched noises I would anticipate birds, kids, or a UFO, an anticipation that could then be quickly refined with a glance out of the window.

The construction corollary

"A person anticipates events by construing their replications."

That is, we construct our anticipations using our past experience. We are fundamentally conservative creatures; we expect things to happen as they've happened before. We look for the patterns, the consistencies, in our experiences. If I set my alarm clock, I expect it to ring at the right time, as it has done since time immemorial. If I behave nicely to someone, I expect them to behave nicely back.

This is the step from theory to hypothesis, i.e. from construction system (knowledge, understanding) to anticipation.

The experience corollary

"A person's construction system varies as he successively construes the replication of events."

When things don't happen the way they have in the past, we have to adapt, to reconstruct. This new experience alters our future anticipations. We learn.

This is the step from experiment and observation to validation or reconstruction: Based on the results of our experiment -- the behaviors we engage in -- or our observation -- the experiences we have -- we either continue our faith in our theory of reality, or we change the theory.

The dichotomy corollary

"A person's construction system is composed of a finite number of dichotomous constructs."

We store our experience in the form of constructs, which he also referred to as "useful concepts," "convenient fictions," and "transparent templates." You "place" these "templates" on the world, and they guide your perceptions and behaviors.

He often calls them personal constructs, emphasizing the fact that they are yours and yours alone, unique to you and no-one else. A construct is not some label or pigeon-hole or dimension I, as a psychologist, lay on you, the "ordinary" person. It is a small bit of how you see the world.

He also calls them bipolar constructs, to emphasize their dichotomous nature. They have two ends, or poles: Where there is thin, there must be fat, where there is tall, there must be short, where there is up, there must be down, and so on. If everyone were fat, then fat would become meaningless, or identical in meaning to "everyone." Some people must be skinny in order for fat to have any meaning, and vice versa!

This is actually a very old insight. In ancient China, for example, philosophers made much of yin and yang, the opposites that together make the whole. More recently, Carl Jung talks about it a great deal. Linguists and anthropologists accept it as a given part of language and culture.

A number of psychologists, most notably Gestalt psychologists, have pointed out that we don't so much associate separate things as differentiate things out of a more-or-less whole background. First you see a lot of undifferentiated "stuff" going on (a "buzzing, blooming confusion," as William James called it). Then you learn to pick out of that "stuff" the things that are important, that make a difference, that have meaning for you. The young child doesn't care if you are fat or thin, black or white, rich or poor, Jew or Gentile; Only when the people around him or her convey their prejudices, does the child begin to notice these things.

Many constructs have names or are easily nameable: good-bad, happy-sad, introvert-extravert, flourescent-incandescent.... But they need not! They can be unnamed. Babies, even animals, have constructs: food-I-like vs. food-I-spit-out, danger vs. safety, Mommy vs. stranger.

Probably, most of our constructs are non-verbal. Think of all the habits that you have that you don't name, such as the detailed movements involved in driving a car. Think about the things you recognize but don't name, such as the formation just beneath your nose? (It's called a philtrum.) Or think about all the subtleties of a feeling like "falling in love."

This is as close as Kelly comes to distinguishing a conscious and an unconscious mind: Constructs with names are more easily thought about. They are certainly more easily talked about! It's as if a name is a handle by which you can grab onto a construct, move it around, show it to others, and so on. And yet a construct that has no name is still "there," and can have every bit as great an effect on your life!

Sometimes, although a construct has names, we pretend to ourselves that one pole doesn't really refer to anything or anybody. For example, a person might say that there aren't any truly bad people in the world. Kelly would say that he or she has submerged this pole -- something similar to repression.

It might be, you see, that for this person to acknowledge the meaningfulness of "bad" would require them to acknowledge a lot more: Perhaps mom would have to be labeled bad, or dad, or me! Rather than admit something like this, he or she would rather stop using the construct. Sadly, the construct is still there, and shows up in the person's behaviors and feelings.

One more differentiation Kelly makes in regards to constructs is between peripheral and core constructs.  Peripheral constructs are most constructs about the world, others, and even one's self.  Core constructs, on the other hand, are the constructs that are most significant to you, that to one extent or another actually define who you are.  Write down the first 10 or 20 adjectives that occur to you about yourself -- these may very well represent core constructs.  Core constructs is the closest Kelly comes to talking about a self.

The organization corollary

"Each person characteristically evolves, for his convenience in anticipating events, a construction system embracing ordinal relationships between constructs."

Constructs are not just floating around unconnected. If they were, you wouldn't be able to use one piece of information to get to another -- you wouldn't be able to anticipate! When you are talked into a blind date, and your friend spends a great deal of energy trying to convince you that the person you will be going out with has a great personality, you know, you just know, that they will turn out to look like Quasimodo. How do you get from "great personality" to "Quasimodo?" Organization!

Some constructs are subordinate to, or "under," other constructs. There are two versions of this. First, there's a taxonomic kind of subordination, like the "trees" of animal or plant life you learned in high school biology. There are living things vs. non-living things, for example; subordinate to living things are, say, plants vs. animals; under plants, there might be trees vs. flowers; under trees, there might be conifers vs. deciduous trees; and so on.

Mind you, these are personal constructs, not scientific constructs, and so this is a personal taxonomy as well. It may be the same as the scientific one in your biology textbook, or it might not be. I still tend to have a species of conifer called Christmas trees.

There is also a definitional kind of subordination, called constellation. This involves stacks of constructs, with all their poles aligned. For example, beneath the construct conifers vs. deciduous trees, we may find soft-wood vs. hard-wood, needle-bearing vs. leaf-bearing, cone-bearing vs. flower-bearing, and so on.

      conifers -- deciduous
          |           |
     soft-wood -- hard-wood
          |           |
needle-bearing -- leaf-bearing
          |           |
  cone-bearing -- flower-bearing

This is also the basis for stereotyping: "We" are good, clean, smart, moral, etc., while "they" are bad, dirty, dumb, immoral, etc.

Many constructs, of course, are independent of each other. Plants-animals is independent of flourescent-incandescent, to give an obvious example.

Sometimes, the relationship between two constructs is very tight. If one construct is consistently used to predict another, you have tight construction. Prejudice would be an example: As soon as you have a label for someone, you automatically assume other things about that person as well. You "jump to conclusions."

When we "do" science, we need to use tight construction. We call this "rigorous thinking," and it is a good thing. Who, after all, would want an engineer to build bridges using scientific rules that only maybe work. People who think of themselves as realistic often prefer tight construction.

But it is a small step from rigorous and realistic to rigid. And this rigidity can become pathological, so that an obsessive-compulsive person has to do things "just so" or break out in anxiety.

On the other hand, sometimes the relationship between constructs is left loose: There is a connection, but it is not absolute, not quite necessary. Loose construction is a more flexible way of using constructs. When we go to another country, for example, with some preconceptions about the people. These preconceptions would be prejudicial stereotypes, if we construed them tightly. But if we use them loosely, they merely help us to behave more appropriately in their culture.

We use loose construction when we fantasize and dream, when anticipations are broken freely and odd combinations are permitted. However, if we use loose construction too often and inappropriately, we appear flaky rather than flexible. Taken far enough, loose construction will land you in an institution.

The creativity cycle makes use of these ideas. When we are being creative, we first loosen our constructions -- fantasizing and brainstorming alternative constructions. When we find a novel construction that looks like it has some potential, we focus on it and tighten it up. We use the creativity cycle (obviously) in the arts. First we loosen up and get creative in the simplest sense; then tighten things up and give our creations substance. We conceive the idea, then give it form.

We use the creativity cycle in therapy, too. We let go of our unsuccessful models of reality, let our constructs drift, find a novel configuration, pull it into more rigorous shape, and try it out! We'll get back to this later.

The range corollary

"A construct is convenient for the anticipation of a finite range of events only."

No construct is useful for everything. The gender construct (male-female) is, for most of us, something of importance only with people and a few higher animals such as our pets and cattle. Few of us care what sex flies are, or lizards, or even armadillos. And no-one, I think, applies gender to geological formations or political parties. These things are beyond the range of convenience of the gender construct.

Some constructs are very comprehensive, or broad in application. Good-bad is perhaps the most comprehensive construct of all, being applicable to nearly anything. Other constructs are very incidental, or narrow. Flourescent-incandescent is fairly narrow, applicable only to light bulbs.

But notice that what is relatively narrow for you might be relatively broad for me. A biologist will be interested in the gender of flies, lizards, armadillos, apple trees, philodendra, and so on. Or a philosopher may restrict his or her use of good-bad to specifically moral behaviors, rather than to all kinds of things, people, or beliefs.

The modulation corollary

"The variation in a person's construction system is limited by the permeability of the constructs within whose range of convenience the variants lie."

Some constructs are "springy," they "modulate," they are permeable, which means that they are open to increased range. Other constructs are relatively impermeable.

For example, good-bad is generally quite permeable for most of us. We are always adding new elements: We may never have seen a computer before, or a CD player, or a fax machine, but as soon as we have, we want to know the best brand to buy. Likewise, a person who will look around for a rock if a hammer is not available uses the construct concerning "things to hammer with" in a permeable fashion.

On the other hand, flourescent-incandescent is relatively impermeable: It can be used for lighting, but little else is likely to ever be admitted. And people who won't let you sit on tables are keeping their sit-upon constructs quite impermeable.

In case this seems like another way of talking about incidental vs. comprehensive constructs, note that you can have comprehensive but impermeable constructs, such as the one expressed by the person who says "Whatever happened to the good old days? There just don't seem to be any honest people around anymore." In other words, honesty, though broad, is now closed. And there are incidental constructs used permeably, such as when you say "my, but you're looking incandescent today!" Permeability is the very soul of poetry!

When there is no more "stretch," no more "give" in the range of the constructs you are using, you may have to resort to more drastic measures. Dilation is when you broaden the range of your constructs. Let's say you don't believe in ESP. You walk into a party and suddenly you hear a voice in your head and notice someone smiling knowingly at you from across the room! You would have to rather quickly stretch the range of the constructs involving ESP, which had been filled, up to now, with nothing but a few hoaxes.

On the other hand, sometimes events force you to narrow the range of your constructs equally dramatically. This is called constriction. An example might be when, after a lifetime of believing that people were moral creatures, you experience the realities of war. The construct including "moral" may shrink out of existence.

Notice that dilation and constriction are rather emotional things. You can easily understand depression and manic states this way. The manic person has dilated a set of constructs about his or her happiness enormously, and shouts "I've never imagined that life could be like this before!" Someone who is depressed, on the other hand, has taken the constructs that relate to life and good things to do with it and constricted them down to sitting alone in the dark.

The choice corollary

"A person chooses for himself that alternative in a dichotomized construct through which he anticipates the greater possibility for extension and definition of his system."

With all these constructs, and all these poles, how do we chose our behaviors? Kelly says that we will choose to do what we anticipate will most likely elaborate our construction system, that is, improve our understanding, our ability to anticipate. Reality places limits on what we can experience or do, but we choose how to construe, or interpret, that reality. And we choose to interpret that reality in whatever way we believe will help us the most.

Commonly, our choices are between an adventurous alternative and a safe one. We could try to extend our understanding of, say, human heterosexual interaction (partying) by making the adventurous choice of going to more parties, getting to know more people, developing more relationships, and so on.

On the other hand, we might prefer to define our understanding by making the security choice: staying home, pondering what might have gone wrong with that last unsuccessful relationship, or getting to know one person better. Which one you choose will depend on which one you think you need.

With all this choosing going on, you might expect that Kelly has had something to say about free will vs. determinism. He has, and what he has to say is very interesting: He sees freedom as being a relative concept. We are not "free" or "unfree;" Some of us are free-er than others; We are free-er in some situations than in others; We are free-er from some forces than from others; And we are free-er under some constructions than under others.

The individuality corollary

"Persons differ from each other in their construction of events."

Since everyone has different experiences, everyone's construction of reality is different. Remember, he calls his theory the theory of personal constructs. Kelly does not approve of classification systems, personality types, or personality tests. His own famous "rep test," as you will see, is not a test in the traditional sense at all.

The commonality corollary

"To the extent that one person employs a construction of experience which is similar to that employed by another, his psychological processes are similar to the other person."

Just because we are all different doesn't mean we can't be similar. If our construction system -- our understanding of reality -- is similar, so will be our experiences, our behaviors, and our feelings. For example, if we share the same culture, we'll see things in a similar way, and the closer we are, the more similar we'll be.

In fact, Kelly says that we spend a great deal of our time seeking validation from other people. A man sitting himself down at the local bar and sighing "women!" does so with the expectation that his neighbor at the bar will respond with the support of his world view he is at that moment desperately in need of: "Yeah, women! You can't live with 'em and you can't live without 'em." The same scenario applies, with appropriate alterations, to women. And similar scenarios apply as well to kindergarten children, adolescent gangs, the klan, political parties, scientific conferences, and so on. We look for support from those who are similar to ourselves. Only they can know how we truly feel!

The fragmentation corollary

"A person may successively employ a variety of construction subsystems which are inferentially incompatible with each other."

The fragmentation corollary says that we can be inconsistent within ourselves. It is, in fact, a rare person who "has it all together" and functions, at all times in all places, as a unified personality. Nearly all of us, for example, have different roles that we play in life: I am a man, a husband, a father, a son, a professor; I am someone with certain ethnic, religious, political, and philosophical identifications; sometimes I'm a patient, or a guest, or a host, or a customer. And I am not quite the same in these various roles.

Often the roles are separated by circumstances. A man might be a cop at night, and act tough, authoritarian, efficient. But in the daytime, he might be a father, and act gentle, tender, affectionate. Since the circumstances are kept apart, the roles don't come into conflict. But heaven forbid the man finds himself in the situation of having to arrest his own child! Or a parent may be seen treating a child like an adult one minute, scolding her the next, and hugging her like a baby the following minute. An observer might frown at the inconsistency. Yet, for most people, these inconsistencies are integrated at higher levels: The parent may be in each case expressing his or her love and concern for the child's well-being.

Some of Kelly's followers have reintroduced an old idea to the study of personality, that each of us is a community of selves, rather than just one simple self. This may be true. However, other theorists would suggest that a more unified personality might be healthier, and a "community of selves" is a little too close to multiple personalities for comfort!

The sociality corollary

"To the extent that one person construes the construction processes of another, he may play a role in a social process involving the other person."

Even if you are not really similar to another person, you can still relate to them. You can, in fact, "construe how another construes," "psych him out," "get inside her head," "see where he's coming from," and "know what she means." In other words, I can set aside a portion of myself (made possible through the fragmentation corollary) to "be" someone else.

This is an important part of role playing, because, whenever you play a role, you play it to or with someone, someone you need to understand in order to relate to. Kelly thought this was so important he almost called his theory role theory, except that the name had already been taken. These ideas, in fact, came from the school of thought in sociology founded by George Herbert Mead.


The theory so far presented may sound very cognitive, with all its emphasis on constructs and constructions, and many people have said so as their primary criticism of Kelly's theory. In fact, Kelly disliked being called a cognitive theorist. He felt that his "professional constructs" included the more traditional ideas of perception, behavior, and emotion, as well as cognition. So to say he doesn't talk about emotions, for example, is to miss the point altogether.

What you and I would call emotions (or affect, or feelings) Kelly called constructs of transition, because they refer to the experiences we have when we move from one way of looking at the world or ourselves to another.

When you are suddenly aware that your constructs aren't functioning well, you feel anxiety. You are (as Kelly said) "caught with your constructs down." It can be anything from your checkbook not balancing, to forgetting someone's name during introductions, to an unexpected hallucinogenic trip, to forgetting your own name. When anticipations fail, you feel anxiety. If you've taken a social psychology course, you might recognize the concept as being very similar to cognitive dissonance.

When the anxiety involves anticipations of great changes coming to your core constructs -- the ones of greatest importance to you -- it becomes a threat. For example, you are not feeling well. You think it might be something serious. You go to the doctor. He looks. He shakes his head. He looks again. He gets solemn. He calls in a colleague.... This is "threat." We also feel it when we graduate, get married, become parents for the first time, when roller coasters leave the track, and during therapy.

When you do things that are not in keeping with your core constructs -- with your idea of who you are and how you should behave -- you feel guilt. This is a novel and useful definition of guilt, because it includes situations that people know to be guilt-ridden and yet don't meet the usual criterion of being in some way immoral. If your child falls into a manhole, it may not be your fault, but you will feel guilty, because it violates your belief that it is your duty as a parent to prevent accidents like this. Similarly, children often feel guilty when a parent gets sick, or when parents divorce. And when a criminal does something out of character, something the rest of the world might consider good, he feels guilty about it!

We have talked a lot about adapting to the world when our constructs don't match up with reality, but there is another way: You can try to make reality match up with your constructs. Kelly calls this aggression. It includes aggression proper: If someone insults my tie, I can punch his lights out, in which case I can wear my tie in peace. But it also includes things we might today prefer to call assertiveness: Sometimes things are not as they should be, and we should change them to fit our ideals. Without assertiveness, there would be no social progress!

Again, when our core constructs are on the line, aggression may become hostility. Hostility is a matter of insisting that your constructs are valid, despite overwhelming evidence to the contrary. Examples might include an elderly boxer still claiming to be "the greatest," a nerd who truly believes he's a Don Juan, or a person in therapy who desperately resists acknowledging that there even is a problem.

Psychopathology and Therapy

This brings us nicely to Kelly's definition of a psychological disorder: "Any personal construction which is used repeatedly in spite of consistent invalidation." The behaviors and thoughts of neurosis, depression, paranoia, schizophrenia, etc., are all examples. So are patterns of violence, bigotry, criminality, greed, addiction, and so on. The person can no longer anticipate well, yet can't seem to learn new ways of relating to the world. He or she is loaded with anxiety and hostility, is unhappy and is making everyone else unhappy, too.

If a person's problem is poor construction, then the solution should be reconstruction, a term Kelly was tempted to use for his style of therapy. Psychotherapy involves getting the client to reconstrue, to see things in a different way, from a new perspective, one that allows the choices that lead to elaboration.

Kellian therapists essentially ask their clients to join them in a series of experiments concerning the clients' life styles. They may ask their clients to loosen their constructs, to slip them around, to test them, to tighten them up again, to "try them on for size." The intent is to encourage movement, essential for any progress.

Kelly, with his background in drama, liked to use role-playing (or enactment) to encourage movement. He might take the part of your mother and have you express your feelings. After a while, he might ask you to reverse roles with him -- you be your mother, and he'll be you! In this way, you become aware of your own construction of your relationship and your mother's construction. Perhaps you will begin to understand her, or see ways in which you might adapt. You may come to a compromise, or discover an entirely new perspective that rises above both.

Kelly's therapy often involves home-work, things he would ask you to do outside the therapy situation. His best known technique is called fixed-role therapy. First, he asks you for a description of yourself, a couple of pages in the third person, which he calls the character sketch. Then he constructs, perhaps with the help of a colleague, another description, called the fixed-role sketch, of a pretend person.

He writes this sketch by examining your original sketch carefully and using constructs that are "at right angles" to the constructs you used. This means that the new constructs are independent of the original ones, but they are used in a similar way, that is, they refer to the same range of elements.

If, for example, I use genius-idiot as a construct in dealing with people, I don't give them a lot of room to be somewhere in between, and I don't allow much for change. And, since we use the same constructs on ourselves as we use for others, I don't give myself much slack either. On a really good day, I might call myself a genius. On most days, I'd have no choice, if I used such a dramatic construct, but to call myself an idiot. And idiots stay idiots; they don't turn into geniuses. So, I'd be setting myself up for depression, not to mention for a life with very few friends.

Kelly might write a fixed-role sketch with a construct like skilled-unskilled. This is a much more "humane" construct than genius-idiot. It is much less judgmental: A person can, after all, be skilled in one area, yet unskilled in another. And it allows for change: If I find that I am unskilled in some area of importance, I can, with a little effort, become skilled.

Anyway, Kelly would then ask his client to be the person described in the fixed-role sketch for a week or two. Mind you, this is a full time commitment: He wants you to be this person 24 hours a day, at work, at home, even when you're alone. Kelly found that most people are quite good at this, and even enjoy it. After all, this person is usually much healthier than they are!

Should the client come back and say "Thank you, doc! I believe I'm cured. All I need to do now is be "Dave" instead of "George" for the rest of my life," Kelly would have a surprise in store: He might ask that person to play another fixed-role for a couple of weeks, one that might not be so positive. That's because the intent of this play-acting is not that the therapist give you a new personality. That would quickly come to nothing. The idea is to show you that you do, in fact, have the power to change, to "choose yourself."

Kellian therapy has, as its goal, opening people up to alternatives, helping them to discover their freedom, allowing them to live up to their potentials. For this reason, and many others, Kelly fits most appropriately among the humanistic psychologists.


Perhaps the thing most associated with George Kelly is his role construct repertory test, which most people now call the rep grid. Not a test in the traditional sense at all, it is a diagnostic, self-discovery, and research tool that has actually become more famous than the rest of his theory.

First, the client names a set of ten to twenty people, called elements, likely to be of some importance to the person's life. In therapy, these people are named in response to certain suggestive categories, such as "past lover" and "someone you pity," and would naturally include yourself, your mother and father, and so on.

The therapist or researcher then picks out three of these at a time, and asks you to tell him or her which of the three are similar, and which one is different. And he asks you to give him something to call the similarity and the difference. The similarity label is called the similarity pole, and the difference one is called the contrast pole, and together they make up one of the constructs you use in social relations. If, for example, you say that you and your present lover are both nervous people, but your former lover was very calm, then nervous is the similarity pole and calm the contrast pole of the construct nervous-calm.

You continue in this fashion, with different combinations of three, until you get about twenty contrasts listed. By eyeballing the list, or by performing certain statistical operations on a completed chart, the list might be narrowed down to ten or so contrasts by eliminating overlaps: Often, our constructs , even though they have different words attached to them, are used in the same way. Nervous-calm, for example, may be used exactly like you use neurotic-healthy or jittery-passive.

In diagnosis and self-discovery uses, you are, of course, encouraged to use constructs that refer to people's behaviors and personalities. But in research uses, you may be asked to give any kind of constructs at all, and you may be asked to give them in response to all sorts of elements. In industrial psychology, for example, people have been asked to compare and contrast various products (for marketing analyses), good and bad examples of a product (for quality control analyses), or different leadership styles. You can find your musical style constructs this way, or your constructs about political figures, or the constructs you use to understand personality theories.

In therapy, the rep grid gives the therapist and the client a picture of the client's view of reality that can be discussed and worked with. In marriage therapy, two people can work on the grid with the same set of elements, and their constructs compared and discussed. It isn't sacred: The rep grid is rare among "tests" in that the client is invited to change his or her mind about it at any time. Neither is it assumed to be a complete picture of a person's mental state. It is what it is: a diagnostic tool.

In research, we can take advantage of a number of computer programs that allow for a "measurement" of the distances between constructs or between elements. We get a picture, created by the people themselves, of their world-views. We can compare the views of several people (as long as they use the same elements). We can compare a person's world-view before and after training, or therapy. It is an exciting tool, an unusual combination of the subjective and objective side of personality research.

For more information on the Rep Grid, see The Qualitative Methods Workbook (Part Five)!


Kelly published The Psychology of Personal Constructs in 1955. After a brief flurry of interest (and considerable criticism), he and his theory were pretty much forgotten, except by a few loyal students, most of whom were involved more in their clinical practices than in the advancement of the psychology of personality. Curiously, his theory continued to have a modest notoriety in England, particularly among industrial psychologists.

The reasons for this lack of attention are not hard to fathom: The "science" branch of psychology was at that time still rather mired in a behaviorist approach to psychology that had little patience with the subjective side of things; And the clinical side of psychology found people like Carl Rogers much easier to follow. Kelly was a good 20 years ahead of his time. Only recently, with the so-called "cognitive revolution," are people really ready to understand him.

It is ironic that George Kelly, always true to his philosophy of constructive alternativism, felt that, if his theory were still around in ten or twenty years, in a form significantly like the original, there would be cause for concern. Theories, like our individual views of reality, should change, not remain static.

There are legitimate criticisms. First, although Kelly is a very good writer, he chose to reinvent psychology from the ground up, introducing a new set of terms and a new set of metaphors and images. And he went out of his way to avoid being associated with other approaches to the field. This inevitably alienated him from the mainstream.

In a more positive vein, some of the words he invented are now firmly fixed in mainstream psychology (although many still think of them as "trendy!"): Anticipation has been made popular by the famous cognitive psychologist Ulric Neisser; Construct, construction, construal, and all its variations can be found in books and articles right alongside of words like perception and behavior. Sadly, Kelly, just like other innovators, seldom gets any credit for for his innovations, mostly because psychologists are rarely trained to pay much attention to where ideas come from.

The "rep grid" has also become quite popular, especially since computers have made it much easier to use. As I mentioned before, it is a nice blend of the qualitative and the introspective that even critics of Kelly's overall theory have a hard time finding fault with.


Much of Personal Construct Theory is phenomenological. Kelly acknowledged his sympathies with the phenomenological theories of Carl Rogers, Donald Snygg and Arthur Combs, and the "self-theorists" Prescott Lecky and Victor Raimy. But he was skeptical of phenomenology per se. Like so many people, he assumed that phenomenology was some kind of introspective idealism. As we shall see in later chapters, that is a mistaken assumption.

But a phenomenologist would find much of Kelly's theory quite congenial. For example, Kelly believes that to understand behavior you need to understand how the person construes reality -- i.e. how he or she understands it, perceives it -- more than what that reality truly is. In fact, he points out that everyone's view -- even the hard-core scientist's -- is just that: a view. And yet he also notes, emphatically, that there is no danger here of solipsism (the idea that the world is only my idea), because the view has to be of something. This is exactly the meaning of the phenomenologist's basic principle, known as intentionality.

On the other hand, there are aspects of Kelly's theory that are not so congenial to phenomenology. First, he was a true theory-builder, and the technical detail of his theory shows it. Phenomenologists, on the other hand, tend to avoid theory. Second, he had high hopes for a rigorous methodology for psychology -- even using the experimental scientist as his "fruitful metaphor." Most phenomenologists are much more skeptical about experimentation.

The emphasis on theory-building, fine detail, and the hope for a rigorous methodology do make Kelly very appealing to modern cognitive psychologists. Time will tell whether Kelly will be remembered as a phenomenologist or a cognitivist!


The basic reference for George Kelly is the two volume Psychology of Personal Constructs (1955). The first three chapters are available in paperback as A Theory of Personality (1963). Another paperback, written especially for the "layperson," is Bannister and Fransella's Inquiring Man: The Theory of Personal Constructs (1971).

Kelly wrote a number of very interesting articles as well. Most of them are collected into Clinical Psychology and Personality: Selected papers of George Kelly, edited by Brendan Maher (1969). There are other collections of works, by Kelly and his followers, available. Look especially for collections edited by Don Bannister.

Lastly, there is a Kellian journal, called The Journal of Personal Construct Psychology. It includes theoretical and research articles by Kellians and psychologists with similar orientations.

Copyright 1997, 2006  C. George Boeree