You can observe these characteristics only by their outward manifestations—by the words, actions, gestures, and subtler mannerisms of people.
When your subconscious is stocked with such well-filed material—when your concretes are filed under the proper abstractions and your abstractions are amply illustrated by concretes—then you can approach an assignment such as “present a characterization of Roark.” And then, if you tell yourself that he is independent, honest, and just, your subconscious will throw at you the kind of concretes that make you feel, while writing a scene: “Yes, Roark would say this, but he would not say that.”
The best, most natural dialogue is usually written as if the writer is listening to dictation. You might get stuck on any particular point and have to question yourself; but normally, dialogue writes itself. You have an idea of the scene, and when you write, the dialogue “just comes” to you—exactly as, in a conversation, your own answers come to you. That is, you speak from your premises, knowledge, and estimate of the situation.
In writing dialogue, you must react on two or more premises. As Roark, you speak from a certain premise; as Keating, you say something else. Your mind must know the connection between certain abstractions and their concrete expressions so well that you can write for three or five or any number of people, constantly switching premises in your mind. You cannot do this by conscious intention. You must reach the stage where the process feels “instinctive”—where, the moment when you speak for Roark, you have a sense of what he would say, and when Keating has to answer, you have a sense of what he would say.
This sudden “feel” of a character is not a mystical talent. In the process of writing, you feel that you “just know” what Roark or Keating would say; but this feeling means only that your understanding of the premises involved has become automatic.
When I wrote the Roark-Keating scene, I did not think consciously of those implications of each line that I explained earlier. But when I write a line inspirationally, I can tell myself why it is in character, and why another line would be out of character. To judge the objective validity of what you write, you must be able afterward to tell yourself why a given line is right for one character (what it conveys) and why something else is right for another character (what it conveys). After the writing, you must be able to do the kind of analysis I did of the Roark-Keating scene.
At first you should do this kind of analysis every time you write something, in order to train yourself in the process. Later, all your rational justifications will be in order and available to your conscious mind, but you will not have to check on them each time. You will know by a lightning-like sum what kind of touch is right, and why it is right.
Also, when you begin writing, write only as much as you are sure of. Do not force your characters into artificial behavior; do not say arbitrarily: “I don’t know what he’d say, so I’ll put in the first line available.” If you do not know what a character would do or say, you simply have to give it some more thought.
When I create a character, I find it helpful to project him visually. This gives me a concrete focus so that the character does not float in my mind as a mere collection of abstract virtues or vices. Seeing his appearance is like having a physical body on which I can hang the abstractions.
That is how Roark was created. I did not base him on any particular human being; but the start of the character in my mind was the image of a redheaded man with long legs and gaunt cheekbones. I formed as clear an image of his figure as I could, and this became the focus for all the abstract characteristics I had to give him. I have done the same for all of my heroes.
In regard to villains and characters who are neither particularly good nor bad, I find it helpful to focus on some acquaintance or public figure—not on the details of this person, but only on the essence. In the case of Toohey, I had in mind four living journalists and writers. I did not think of any one of them in specific detail, nor did I study their writings or lives. But my total impression of them