"Psychohistory" is one of three words (that I know of) that I get early-use credit for in The Oxford English Dictionary. The other two, for the record, are "positronic" and "robotics."

This is not at all unusual. Every science fiction writer makes up words and sometimes they actually penetrate the language (but then English is notoriously hospitable to neologisms-which is one of its strengths, in my opinion).

The more unimaginative and inevitable a word is, the more likely it is to be adopted, and I am not prone to making up words wildly. Thus, once the positron was discovered and named in 1935, and once "robot" became accepted as a term for a humaniform automaton in the 1920s, it was simply a matter of time before the words "positronic" and "robotics" appeared in print. That I seem to have been the first in each case is purely accidental.

In fact, when I first used the word "positronic" in print (in my story "Reason," which appeared in the April 1941 issue of Astounding Science Fiction) as a natural analogue of "electronic," I thought the word already existed. The same was true when I first used the word "robotics," in my story "Runaround," which appeared in the March 1942 issue of Astounding Science Fiction.

In the case of "psychohistory," however, I suspected that the word was not in common use, and might even never have been used before. (Actually, the O.E.D. cites one example of its use as early as 1934.) I first used it in my story, "Foundation," which appeared in the May 1942 issue of Astounding Science Fiction.

I came up with the word because John Campbell and I were discussing the course I was to take in the Foundation series once I came to him with my initial idea on the subject. I was quite frank in my intention of using Edward Gibbon's Decline and Fall of the Roman Empire as my model and as a basic guide for plot ideas, but I needed something that would make science fiction out of it. I couldn't simply call it the Galactic Empire and then just treat it as a hypertrophied Roman Empire.

So I suggested we add the fact that a mathematical treatment existed whereby the future could be predicted in a statistical fashion, and I called it "psychohistory." Actually, it was a poor word and did not represent what I truly meant. I should have called it "psychosociology" (a word which the O.E.D. lists as having first been used in 1928). However, I was so intent on history, thanks to Gibbon, that I could think of

nothing but psychohistory. In any case, Campbell was enthusiastic about the idea and we were off and running.

I modeled my concept of psychohistory on the kinetic theory of gases, which I had been beat over the head with in my physical chemistry classes. The molecules making up gases moved in an absolutely random fashion in any direction in three dimensions and in a wide range of speeds. Nevertheless, one could fairly describe what those motions would be on the average and work out the gas laws from those average motions with an enormous degree of precision.

In other words, although one couldn't possibly predict what a single molecule would do, one could accurately predict what umptillions of them would do.

So I applied that notion to human beings. Each individual human being might have "free will," but a huge mob of them should behave with some sort of predictability, and the analysis of "mob behavior" was my psychohistory.

There were two conditions that I had to set up in order to make it work, and they were not chosen carelessly. I picked them in order to make psychohistory more like kinetic theory. First, I had to deal with a large number of human beings, as kinetic theory worked with a large number of molecules. Neither would work for small numbers. It is for that reason that I had the Galactic Empire consist of twenty-five million worlds, each with an average population of four billion. That meant a total human population of one hundred quadrillion. (In my heart, I didn't think that was enough, but I didn't want to place any greater strain on the suspension of disbelief than I absolutely had to.)

Second, I had to retain the "randomness" factor. I couldn't expect human beings to behave as randomly as molecules, but they might approach such behavior if they had no idea as to what was expected of them. So it was necessary to suppose that human beings in general did not know what the predictions of psychohistory were and therefore would not tailor their activities to suit.

Much later in the game, I thought of a third condition that I didn't think of earlier simply because I had taken it so completely for granted. The kinetic theory assumes that gases are made up of nothing but molecules, and psychohistory will only work if the hosts of intelligence are made up of nothing but human beings. In other words, the presence of aliens with non-human intelligence might well bollix the works.

This situation may actually develop in future books of the Foundation series, but so far I have stayed clear of non-human intelligences in my Galactic Empire (partly because Campbell and I disagreed fundamentally on what their role would be if they existed and since neither of us would give in).

Eventually, I thought that my psycho history would fade out of human consciousness because the term came to be used by psychiatrists for the study of the psychiatric background of individuals (such as Woodrow Wilson, Sigmund Freud, or Adolf Hitler) who had some pronounced effect on history. Naturally, since I felt a proprietary interest in the term psychohistory as a predictive study of large faceless masses of human beings, I resented the new use of the word.

But then as time went on, I grew more philosophical. After all, it might well be that there could be no analogy drawn between molecules and human beings and that there could be no way of predicting human behavior. As mathematicians began to be interested in the details of what is now called "chaos," it seemed to me that human history might prove to be essentially "chaotic" so that there could be no psychohistory. Indeed, the question of whether psychohistory can be worked out or not lies at the center of the novel I have recently completed, Prelu de to Foundation, in which Hari Seldon (the founder of psychohistory) is portrayed as a young man who is in the process of trying to devise the science.

Imagine, then, how exciting it is for me to see that scientists are increasingly interested in my psychohistory, even though they may not be aware that that's what the study is called and may never have read any of my Foundation novels, and thus may not know of my involvement. (Who cares? The concept is more important than I am.)

Some months ago, a reader, Tom Wilsdon of Arden, North Carolina, sent me a clipping from the April 23, 1987, issue of Machine Design. It reads as follows, in full:

"A computer model originally intended to simulate liquid turbulence has been used to model group behavior. Researchers at Los Alamos National Laboratories have found that there is a similarity between group behavior and certain physical phenomena. To do the analysis, they assigned certain physical characteristics such as level of excitement, fear, and size of the crowd to model parameters. The interaction of the crowd closely paralleled the turbulent flow equations. Although the analysis cannot predict exactly what a group will do, it reportedly does help determine the most probable consequence of a given event."

Then, too, Roger N. Shepard, a professor of psychology at Stanford University, has published an article in the September 11, 1987 issue of Science entitled "Toward a Universal Law of Generalization for Psychological Science."

Unfortunately, although I made a valiant effort to read it, the mathematics was too tough for me and even the nonmathematical portions produced only a rather dim and hazy understanding within me. However, here is the summary of the article as given at the beginning:

"A psychological space is established for any set of stimuli by determining metric distances between the stimuli such that the probability that a response learned to any stimulus will generalize to any other is an invariant monotonic function of the distance between them. To a good approximation, this probability of generalization (i) decays exponentially with this distance, and (ii) does so in accordance with one of two metrics, depending on the relation between the dimensions along with the stimuli vary. These empirical regularities are mathematically derivable from universal principles of natural kinds and probabilistic geometry that may, through evolutionary internalization, tend to govern the behaviors of all sentient organisms."

As I said, I don't really understand this but I have the feeling that Hari Seldon would understand it without trouble. I am also concerned, suddenly, that psychohistory may be developed within the next century. I placed its development 20,000 years in the future. Is this going to be another case of my science- fictional imagination falling ludicrously short?