Orson Card - Speaker for the Dead

What I discovered then--the spring of 1983--was that the book was unwritable. In order to make the Ender Wiggin of Speaker make any kind of sense, I had to have this really long, kind of boring opening chapter that brought him from the end of the Bugger War to the beginning of the story of Speaker some three thousand years later! It was outrageous. I couldn't write it.

When Compute!, the publisher I was working for as a book editor, sent me along to the American Booksellers Association convention in Dallas, I noticed that Tom Doherty himself was at the Tor Books exhibit. I greeted him, and then on impulse asked him if I could talk to him. I had no well-formed plan in mind, and I was a little frightened when he said, "Sure," and set an appointment not long after. Our meeting consisted of walking through the crowds as I explained to him the problem I was having writing Speaker . The only solution I could think of, I said, was to write a novel version of Ender's Game, so I could put all that material about how Ender became a Speaker for the Dead at the end of that book, thus allowing Speaker to begin at its true beginning.

Once I proposed the idea (having only thought of it a short while before) it seemed so obvious that I wondered why I hadn't tried to sell a novel version of Ender's Game years before. (Only later did I realise that it wasn't until I was working on Speaker that the character of Ender Wiggin grew enough to be able to sustain a novel.) Still, Tom agreed with me that a novel version of Ender's Game was a good idea. "Let's do it" he said. "Same terms as Speaker ?".

"Sure," I said, hardly believing that the decision could be made so easily--I hadn't talked to him more than five minutes.

"Fine. We'll send a contract to Barbara as soon as I get back to New York."

Lo! It happened exactly as he said! This was something I had never seen before--a publisher making a decision instantly and then having everything he said turn out to be true! I still marvel at it--a publisher who is not only an honest man but also loves (and reads ) books, makes decisions quickly, and then can sell the books he publishes!

Gratefully I set aside Speaker and began plotting Ender's Game. By the time I quit my job at Compute! that fall, after only nine months in the position (I'm not cut out for corporate life anymore, I'm afraid), I was raring to go. I began Ender's Game before Christmas that year, took a break to go to Utah to promote my novel Saints , and then returned home and finished the book in a couple more weeks.

Then I turned to Speaker and the real suffering began. By now, of course, the title had changed from Speaker of Death to Speaker for the Dead , as the concept had clarified at the end of Ender's Game. By now, the character of Ender had developed so much that my original draft of the opening of Speaker was almost laughable. I had begun (except for the "introductory chapter") with Ender's arrival on the planet Lusitania. Just in time to speak the death of an old lout named Marcão. But it was hollow and empty and it just wasn't working. So I went back to the drawing board and began all over again.

I began the book several more times, each time getting a little farther, but each time being blocked because it still wasn't right. I didn't know what "right" was, of course--but I did have several hundred pages of "wrong." (During this struggle with Speaker I wrote the novel Wyrms , which in some ways was a tryout of the scientific ideas in Speaker and, eventually, Xenocide —using a semisentient molecule that adapts itself easily to alien species in order to take them over and control them.)

Finally I knew I had to begin with the character of Novinha, who hadn't even existed in the original outline. And the characters of Pipo and Libo had also emerged, along with Pipo's death, pretty much as they happen in the first few chapters of the book you now hold in your hands. But I still wasn't done. It still wasn't enough. I was about 200 pages deep and the book was dead in my hands and I didn't know what to do.

It happened that a good friend of mine. Gregg Keizer was working for Compute! In fact I was the one who had recruited him away from his job as a junior high school English teacher (for which I think he has forgiven me) and brought him out to North Carolina. I had met Gregg when he became my student at a science fiction writing class I taught in the University of Utah's evening school program back in the seventies. He was one of those frustrating students who are simply brilliant when they walk in the door, so the teacher can't take the slightest credit for anything they do. He was also one of the most decent human beings I know, which makes me very nervous around him--so nervous, in fact that the only times I have ever gotten thoroughly and stupidly lost have been while he was in the car with me and I was supposed to know where I was going. Some teacher!

(I once was so certain that a story of Gregg's would sell that I made a wager with my class--if it didn't sell within one year, I would run naked through the corridors of Orson Spencer Hall on the U of U campus, which is where our class met. The story didn't sell in a year--a pox on editors!--and, perhaps out of an exaggerated commitment to aesthetics, I reneged on the bet. Since the story did sell a short while afterward, Gregg has never demanded that I make good, but he does have the debt hanging over my head.)

Anyway, right during the time that I was stymied on Speaker, Gregg and I decided to go to New York for the 1985 Nebula weekend. Ender's Game had only-just been published and neither of us had anything on the ballot. We just wanted to go to New York and to the Nebulas, so why not? I brought along the manuscript of Speaker for him to read--or perhaps I gave it to him in advance--I don't remember now. I do remember, though, sitting at the foot of his bed while he lay there and explained the problems he saw in Speaker .

He had many good ideas. Of course, most of them dealt with small fixes for problems in the manuscript as it now stood. One comment he made, however, illuminated everything for me. "I couldn't tell Novinha's kids apart," he said. "I couldn't remember which was which."

I had enough experience by then to know exactly what this meant. He couldn't tell Novinha's kids apart because they weren't characters. They were nothing but placeholders. At first I toyed with the idea of simply cutting them out. In my novel Saints , Ihad run into a problem with a younger sister of my protagonist--I kept forgetting she existed and completely neglecting her for hundreds of pages at a time. The solution was to eliminate the character; callously, I had her die in infancy. But excision wasn't the right move in this case. Because I wanted Novinha to be voluntarily isolated, I had to have her be otherwise acceptable to her neighbors. In a Catholic colony like Lusitania this meant Novinha needed to have a bunch of kids.

Yet I had no idea who they were or what they would do in the story. Once you've read Speaker, of course, you'll wonder what the story would be without Novinha's children, and the answer is, It wouldn't be much! But at the time I hadn't developed their role in the story; yet there was something in the story that led Gregg to want them to amount to something more--that made him want to be able to tell them apart.

It meant throwing out all but the first couple of chapters of what I had written so far (and, in fact, I ended up completely writing the novel from the beginning), but it soon dawned on me that it was worth doing, for this was the final idea, the one that would pull me through the whole book. I had observed before that one thing wrong with science fiction as a whole was that almost all the heroes seemed to spring fully-grown from the head of Zeus--no one had families. If there was a mention of parents at all, it was to tell us they were dead, or such miserable specimens of humanity that the hero could hardly wait to get out of town.