Someone once asked Isaac Newton how he managed to reach solutions to problems that others found impenetrable. He answered, "By thinking and thinking and thinking about it."

I don't know what other answer people can possibly expect. There is the romantic notion that there is such a thing as "inspiration," that a heavenly Muse comes down and plunks her harp over your head and, presto, the job is done. Like all romantic notions, however, this is just a romantic notion.

Some people may be better at solving problems and getting ideas than others are; they may have a livelier imagination, a more efficient way of grasping at distant consequences; but it all comes down to thinking in the end. What counts is how well you can think, and even more, how long and persistently you can think without breaking down. There are brilliant people, I imagine, who produce little, if anything, because their attention span to their own thoughts is so short; and there are less brilliant people who can plug away at their thoughts until they wrench something out of them.

All this comes up in my mind now because a friend of mine, a science fiction writer whose work I admire enormously, in the course of a conversation asked, in a very embarrassed manner, "How do you get your ideas?"

I could see what the problem was. He had been having a little trouble coming up with something and he thought that perhaps he had lost the knack of getting ideas, or had never really had it, and he turned to me. After all, I write so much that I must have no trouble getting ideas and I might even have some special system that others could use, too.

But I answered, very earnestly, "How do I get my ideas? By thinking and thinking and thinking till I'm ready to jump out the window."

"You, too?" he said, quite obviously relieved.

"Of course," I said. "If you're having trouble, all it means is that you're one of us. After all, if getting ideas were easy, everyone in the world would be writing."

After that, I put some serious thought into the matter of getting ideas. Was there any way I could spot my own system? Was there, in fact, any system at all, or did one simply think at random?

I went back over what happened in my mind before I wrote my most recent novel, Nemesis, which Doubleday published in October 1989, and I thought it might be helpful to aspiring writers, or even just to readers, if I described the preliminary thinking that went into the novel.

It started when my Doubleday editor, Jennifer Brehl, said to me, "I'd like your next novel not to be part of a series, Isaac. I don't want it to be a Foundation novel or a Robot novel or an Empire novel. Write one that's completely independent."

So I started thinking, and this is the way it went, in brief. (I'll cut out all the false starts and dead ends and mooning about and try to trace a sensible pathway through it all.)

The Foundation novels, Robot novels, and Empire novels are all interconnected and all deal with a background in which interstellar travel at super luminal speeds is well established. Of my previous independent novels, The End of Eternity deals with time-travel; The Gods Themselves with communication between universes; and Fantastic Voyage II with miniaturization. In none of these is there interstellar travel.

Very well, then, let me have a new novel which exploits an entirely new background. Let it deal with the establishment of interstellar travel, with the first interstellar voyages. Immediately I imagined a settled solar system, an Earth in decay, large numbers of space settlements in lunar orbit and in the asteroids. I imagined the space settlements as hostile to Earth and vice versa.

That gave me a reason for the drive to develop interstellar travel. Naturally, technological advances may be made for their own sake (as mountains are climbed "because they're there") but it helps to have a less exalted reason. A settlement might want to get away from the solar system to create a completely new society, profiting by past experience to avoid some of humanity's earlier mistakes.

Good, but where do they go? If they have true interstellar flight, as in my Foundation novels, they can go anywhere, but that's too much freedom. It introduces too many possibilities and not enough difficulties. If humanity is just developing interstellar flight, it might not be a very efficient process at first and a settlement trying to escape might find itself with a very limited range.

Now where do they go? The logical place is Alpha Centauri, the nearest star, but that is so logical that there's no fun to it. Well, then, what if there's another star only half as far as Alpha Centauri? That would be easier to reach.

But why haven't we seen it, if it exists? Well, it's a red-dwarf star and very dim, and besides there's a patch of interstellar dust between it and ourselves and that dims it further so that it just hasn't been noticed.

At that point, I remembered that a few years ago there was some speculation that the Sun might have a very distant red-dwarf companion that once in every revolution penetrated the comet cloud and sent some comets whizzing into the inner solar system where one or two might occasionally collide with Earth and produce the periodic waves of life-extinction. The red dwarf was called Nemesis.

The suggestion seems to have died down, but I made use of it. My characters would go to the nearby red dwarf, which I would call Nemesis, and then use that as the name for my novel. Of course, you can't very well have a habitable planet circling a red dwarf star, but I wanted one. It would give me greater flexibility than simply to have the settlement go into orbit about the red dwarf. That meant I had to think up a set of conditions that (if you don't question things too closely) would make it sound as though a habitable planet could exist. For that I had to invent a gas giant, with an Earth-sized satellite, and it would be the satellite that would be habitable.

Now I needed a problem. The obvious one would be that Nemesis was circling the Sun and would eventually pass through the comet cloud. I rejected that because it had been well discussed in the media and I wanted something a little less expected. So I decided that Nemesis was an independent star that happened to be en route to a relatively near miss of the solar system, with possibly dangerous gravitational effects.

That was a good problem, but I needed a plausible solution. That took some time but I finally thought one up. (Sorry, I won't tell you what it is. For that you'll have to read the book.)

What I needed next was a good character that would serve as the spinal column of the book, around whom everything would revolve. I chose a fourteen-year-old girl, with certain characteristics that I thought would make her interesting.

Then I needed a place to start the book. I would begin with my main character and have her do or say something that starts the chain of events that will take up the rest of the book. I made the choice and then waited no longer. I sat down and started the book.

But, you might point out that I didn't yet have the novel. All I had was the social framework, a problem, a solution, a character and a beginning. When do I make up all the details that go into the characteristically involved plot of one of my novels (and Nemesis is quite involved).

I'm afraid that I make that up as I go along, but not without thought. Having worked out the first scene, I find that by the time I've finished that, I have the second scene in mind, at the conclusion of which I have the third scene, and so on all the way through to the ninety-fifth scene or so, which ends the novel.

To do that, I have to keep on thinking, on a smaller and more detailed scale all the time that I'm doing the book (which takes me nine months, perhaps). I do it at the cost of lots of lost sleep and lots of lack of attention to people and things about me (including an occasional blank stare even at my dear wife, Janet, who never fails to get the alarmed notion that "something's wrong" each time I go into a spasm of thought).

But then isn't it possible that two-thirds of the way through the book I realize that toward the beginning I made a wrong turn and am now beating my way down a blind alley. It is possible, but it's never happened to me yet, and I don't expect it to. I always build the next scenes on whatever it is I have already done and never consider any possible alternatives. I simply have no time to start over again.

However, I don't mean to make the process sound simpler than it really is. You must take into account, in the first place, that I have a natural aptitude for this sort of thing, and, also, that I have been doing it for over half a century now, and experience counts.

Anyway, this is the closest I can come to explaining where I get my ideas.