Small Changes (54 page)

By the time they went to bed Saturday night she was tired, what with the week’s shopping, laundry, making supper and cleaning up. They made love but she could not concentrate enough to come before he had finished. Then she could not drift off. The longer she lay beside his sleeping body, the wider awake she became, the more uselessly alert. Finally she let her mind play over some of the technical materials she had been reading and pondering.

About three she hopped out of bed and tiptoed across the hall to her study—bleak, unfinished room she must paint eventually. She sat down and started to write, as quickly as her hand could move on the pad. Her brain was working in leaps, in great encompassing bolts of new material, and she could not transcribe her thoughts quickly enough.

About five she went back to bed. When the alarm went off three hours later (Neil liked to rise early even on Sunday) she felt she had been kicked in the head. Her good humor returned with the memory of her beautiful idea. Away with the missile nonsense, she would settle down to working on it today and get it whipped into shape. Yes, Abe had asked her just last week if she wanted to give a seminar soon, and she had said she would think about it—feeling small desire to haul out her neglected thesis. Unfinished business made her queasy in her conscience. But here was an idea that fed out of that material but was light-years past it and exciting enough to stun them all. She would get Abe to let her work on it. They could carry her on overhead for a while till they could get a new project funded. She did not mind working alone, although that could never be as exciting as the give-and-take of a really creative group. However, it would be preferable beyond comparison to being stuck on Dick’s project.

Everything about the missile contract was political. She
had wasted a month generating ideas about machine design only to have it finally sunk in to her that since the contract for the hardware had already been let to UNIVAC, there was an unspoken pressure operating against any ideas Logical might produce. That is, UNIVAC was going to implement the machine design and was deeply unfriendly to ideas that didn’t go with the predilections of their already existing hardware and their technical people. Logical’s far-out suggestions for machine design were just a boondoggle, a baroque embellishment on the contract, and never intended to be implemented by anybody.

The technical people they had to deal with at Bell Labs kept changing. The old faces would disappear and new people turn up to deal with them who had no idea what they were doing and what had happened before. The same thing kept happening with their government contract interface, where people kept finishing their tours of duty and going off, leaving important pieces of paper lost in the works and requests unheeded and interim reports in limbo.

She hated the whole clanking apparatus and she hated the intended product. Now she had a way out. She would tell Abe Monday afternoon that she wanted to be scheduled for a seminar. There she would present her new idea and they would all be turned on. They had to: it was a breakthrough: a descriptive method that would be adequate for describing
both
artificial languages and computers. This it would do in such a way that an algorithm could translate automatically—without further human intervention—a program written in the artificial language into machine language.

This was the so-called compiler-compiler problem and it had never been solved. But she saw a way in. She could not guess—no way of knowing at this point—if her idea would work, but what a breakthrough if it did. Even if it didn’t, she was sure the effort would generate interesting materials. What she saw was an intelligent method for partitioning up the possibilities—a pattern recognition approach which she hoped and believed would produce useful partitionings in each artificial language and which would then explore those possibilities.

What usually went wrong with such schemes was that a language developed for describing the syntax of other languages—like artificial or computer languages—lacked a semantic bridge between the two structures. Invariably that
bridge had to be spelled out by a programer for each language-computer pairing.

But she was conceiving a super compiler-compiler that could simply be handed descriptions of five languages, five machines, and could handle specifications of which language was to be linked with which machine. Then the super compiler-compiler would accept programing and knock out machine code for the required case. The combinatorics involved made it a hairy problem. What she thought she saw was a method for making the partitionings, the choices, not blindly but heuristically: it was gorgeous if far out. But it just had to grab Abe. This would get her back onto something exciting. She had not joined Logical to do ugly boring nonsense mired in politics. Truly it made her sick. Even if she had been a right-wing nut she would have hated the project, since the probabilities of the anti-missile missile system working at all were abominably low. She had known that for months.

She was ashamed of her work. At odd moments she imagined trying to explain it to Phil, running into old friends, políticos, and bumbling through an answer to what she was doing. The contract stank, but she was beginning to realize everything around was sticky. Even the pure research she had been working on with Neu was funded by the Department of Defense, through A.R.P.A. In fact all the most interesting projects in computer science seemed funded by the Pentagon, when she came down to it. Only they could afford big machines and only they were interested in the techniques on the level of sophistication where she wanted to work. Thus to say something was at the cutting edge of the science was almost for sure to say that the military was paying for it, since nobody else could afford to.

Neil did not make those connections. He thought of the whole thing as what Phil would have called a vast ripoff: that the government was paying him for playing his elaborate and beautiful games. He saw no results in the newspaper from the intricate webs of logic he spun out. She shied away from confronting him. After all, she did not know what to propose as an alternate source of funding. Neil was a good person, a kind person, but he had not been trained to make connections in terms of social consequences. And if she were less blind, what difference did that make?

Well, the first step was getting out of the project she loathed.
Then she would gradually try to move herself into some other area of expertise and research—a branch of computer science not quite so dependent on fancy, vast, and expensive equipment, so that somebody besides the D.O.D. could afford to use it. She could not turn to Neil for help. He knew the ins and outs of the computer field on a more intimate level than she did, but her attempts to talk with him always ended up either as technical arguments about the worth of somebody’s ideas or as anecdotal histories on his part of brilliant systems men. Nor could she tell him exactly what had soured her. She would have had to tell him about her last meeting with Wilhelm Graben and that she had no intention of doing. That would upset him. It was meaningless. The last gasp of a dying way of life.

The day of the seminar came. She had not been talking half an hour when she had a strong physical sensation of being bombarded by waves of hostility. She could not even complete setting out her idea before the attacks began. She seemed to have hit a bare nerve in almost everyone. Neil of course disqualified himself: the only person who might have understood the importance of the idea and followed her argument. No, she was being paranoid. Jaime could follow, Abe could follow, Fred and Ted should be able to follow easily. Yet Fred was hardly able to be polite. The idea seemed to affront him personally. He sounded angry as he argued that her compiler would make improper partitionings and never be able to escape. It would forever continue to try for matches between the languages in a nightmare loop straggling into infinity. He made it sound like a bad trip: a memory of years before with Phil on acid, being caught in a small self-enclosed universe where the same words repeated endlessly and eternally, and she knew that never would she be able to escape forward.

Ted sounded just as mad, though he spoke with arch sarcasm. He said how he thought it was wonderfully fascinating to presume to solve in half an hour what was in its essence an insoluble problem. One that a great many individuals of vast reputation and experience had spent much time and energy on. They all did enjoy a good science fiction tale now and then, so why not a sci fi seminar? The junior members of the research teams could always be depended on for a bit of flying sauceritis. Perhaps every technical person was
entitled to an ego trip sometimes, but most of them would prefer doing it on their own time.

His sarcasm entered her like shrapnel. Her hands sweated into the chalk, dust compacting on her damp palms. She could not tell if he was angry because she had
presumed
to attack such a problem, or because
she
had presumed.

Jaime was the only one who seemed to want to grapple with the idea. His objections were in terms of whether the descriptive techniques were really adequate to all cases. She answered him as best she could and then the rout continued. When it was finally done she walked out quickly, shaking, and shut the door of her office. She had never seen anyone attacked that way. She could not believe the response. She had been sure the idea was exciting, but that excitement had been fury. Yet perhaps she was being defensive. There were large holes. She had been hoping by presenting the idea early to get criticism that would spur her on to rapid progress. Mostly, of course, she had hoped to get off the missile project.

She must confront Abe directly. He had not attacked her, he had only asked whether the number of possibilities could really be sufficiently reduced by the pattern-recognition approach. That was a serious question. She felt she had handled it well. She would appeal to him. Only he decided what they would take on as projects and what they would not.

Friday afternoon she got an appointment to see him. With the name Abe one would have expected a fifty-year-old Jewish businessman, a pudgy grandfather: but Abe was from Lincoln country and had been named by his father, a self-made agribusinessman. How he had ended up in computers was anyone’s guess. He was a big-boned man with sandy hair and perennially squinted eyes of in-between color, a firm handshake and a just-folks manner that disappeared in a flash when he was annoyed or dealing with people he wanted to impress. He had a firm reputation and had been in analysis since before Neil had met him. It was said in the company that Abe had chosen Boston as a site for Logical, not so much because it was a fertile field for computer contracts as because that was where his analyst practiced.

Abe had a large family, the oldest in prep schools and the youngest still a toddler, all by the wife whom Miriam had seen only twice, dressed up for some occasion with the wives of other company presidents in a manner which cost a lot to no effect Miriam could figure out—beige linen dresses
that went into the wallpaper with the gaunt body. He lived the farthest from the office, up almost into New Hampshire in a farmhouse supposed to date from 1830, on which he was always financing renovations. He had several acres and a palomino. Abe held aloof from the rest of the company, although all three directors had known each other well before they formed it. He was opaque to her. He hated to fire technical staff—leaving that, when it must be done, to Dick—yet indifferent to them as people. He was proud of how good an organization he had built: his stable, perhaps.

From time to time he picked favorites known to be up and coming. Ted had been especially favored for six to eight months, because he had developed an idea they had incorporated in a proprietary product: a program they hoped to market to other companies. Lately Ted had fallen from grace because the idea did not work after they sold it. The program did not do what it was supposed to—true of a great many programs, perhaps most of them—but Ted had been caught because the way it didn’t work had become visible too quickly. She had never been a favorite son in the company, but she would not mind that position at all.

Abe heard her out patiently enough, though she had the feeling his mind was only partially engaged. Where was the rest of him? He sat behind his neat desk with his long legs spread and his feet propped on a lower drawer pulled out, fiddling with his pipe. Half the men at Logical smoked pipes. This was a meerschaum. The tobacco had a sweetish smell that tickled her nose. It was close in the office and she wished that the windows opened. Outside it was raining lightly but persistently. The sky looked low, sagging just over the expressway and about to fall in long filmy strands over the cars beginning to clog it for rush hour.

“People weren’t exactly unanimous behind this idea. I like to feel my technical people are agreed that an idea has merit before I would go pulling someone off a project he’s already on.”

“But with any idea that has a certain amount of risk—that involves taking a giant step—you can’t have agreement beforehand.”

“It is an exciting idea, Miriam, and I’d be the last to deny that, and we certainly got some excitement out of the boys on Wednesday. But you know we’re not a big outfit with a lot of extra capital to bankroll our ideas. Even if this idea
does pan out, chances are it’d be too costly to have much commercial application. Now”—he held up his hand to head her off at the pass—“you know and I know we’re here to do research. But we aren’t in a financial position right at this moment to fund our own pure research off the cuff.”

“It wouldn’t take long for me to reach the point where we could see if it works or not. It really would be a breakthrough if I’m right. If I’m wrong, it wouldn’t take long.”

“I know that you feel strongly, and I wish I could give you the green light. Tell you what we can do. I don’t want you to give up. I want you to go ahead and work on your idea on the side while you’re on the ABM contract. Then in six months or so we can review your progress. I think we might have a reserve for developing our own technical ideas by then. You’ll have your idea roughed out and filled in. I like to feel the staff is behind any idea we finance on the cuff, because it’s a risk to all of us. But you’re not to feel discouraged. Work on it, keep at it, and we’ll see if we can’t come up with something in six months, nine months. You don’t want people to feel you can get supported on a project of your own because Neil threw his weight around.”