Version Control

by Dexter Palmer

  Champions of interferometry back then would have dismissed cryogenic bar research as “undead science.” That is, it was science that continued to thrive in its own small corner of the culture, even though the community as a whole decided to pretend that it didn’t exist. Cryogenic bar research was an edge case, but there are examples of undead science that are more blatant. At least one automobile company continued to fund cold fusion research into the mid-1990s, even though the idea had been widely discredited in 1989. Even august Princeton University had hosted a laboratory dedicated to paranormal research as late as 2007. Next to that DeWitt’s research wasn’t embarrassing at all. It had merely acquired the arguably unearned whiff of the faintly dubious, mostly in circles populated by those who had a likelihood of indirectly profiting if it failed.

  I did not grasp all this at twenty-three. If some elder statesman of the field had told me all this to my face, I would have warmed myself under his cold gaze by the light of my ideals, and assumed with certainty that I would succeed where all those before me had failed.

  Still other things were lost on me in my idealistic youth.

  I was not aware of the extent to which, when one gets to the cutting edges of the field, problems of physics become intertwined with problems of sociology. It is not enough to see the truth. You must also convince others that you have seen it, which becomes increasingly difficult as science gets farther away from direct observation of phenomena. Modern physics involves looking at data and making a series of inferences about it, and those inferences tell us things about the nature of the world that the naked eye cannot.

  In the matter of gravitational waves, agreeing on the validity of the chain of inferences that connected the data to the claim was no light thing. This is the case in any subfield of modern physics to some degree. Our areas of expertise have become so highly specialized, and those who are even just slightly out of our subdiscipline, not to mention the distant controllers of purse strings, have little idea how to evaluate the quality of the inferences that experts make. They rely on intangibles and externalities. They ask around about a given scientist’s reputation, or decide if he’s a “good guy,” or they “go with their gut.” So presentation matters. The data can’t make its own case, and only a fool believes otherwise.

  I tell the post-docs and grad students who spend their days with the CVD that though they have chosen to work on an extremely difficult project, they are, in a sense, quite lucky. When we get positive results, it won’t be a matter of convincing others that we’ve picked up a barely detectable signal. The results will be quite clear, as clear as an apple falling or a match bursting into flame.

  Unless

  —

  The last and most important thing I discovered at twenty-three, and as obvious as it is I still need to remind myself of it daily, is that science is done by human beings, who abhor failure above almost all things. But Science depends on the constant failure of its practitioners in order to thrive: the wrong paths that some scientists take in their careers indicate the correct paths for others.

  Science does not have an ego. Scientists do, though, and in the backs of their minds, they are all well aware of the high likelihood of failure in their chosen profession. If even a raft of positive results sometimes earns a researcher little more than a few lines in an article or a citation in a footnote, then negative or null results, even if they advance the field, are almost certain to be forgotten.

  Put these two things together, then. Consider the need not just to see the truth, but to be recognized by others as seeing it. Consider the tendency to view only positive statements in science as valuable. It is easy for someone to conclude, perhaps not even consciously, that if a scientist’s career does well when he produces positive results that are verified by others, then Science as a whole progresses when those who practice it are recognized by others to produce positive results. A statement that is both true and wrong.

  I was attending the physics department’s open house at Nicolls College. I was twenty-three, and a vessel waiting to be filled.

  The open house is when graduate students meet their potential advisers. On the one hand, faculty want to attract the very best students to their labs, since they have limited financial and intellectual resources and don’t want to waste them on second-stringers. On the other hand, if they attract few or no students at all, their grant money dries up as the holders of purses begin to suspect that it’s time to prune one of Science’s dying branches. Meanwhile the students attempt, based on what little they know, to size up their competition, to measure their own potential worth to the department and the field. It does you no good to land in the lab of a storied professor if you can never catch his eye once you’re there. You will have a better chance in the long term if you settle for someone distinctly second tier who will listen when you speak.

  DeWitt must have seen me as an easy mark. Back then I had this idea of the scientist as an iconoclast who forged his own path. While the other grad students at the open house dressed and groomed themselves in ways that were vaguely reminiscent of the portraits of physicists that were taken during the 1950s, with starched button-down shirts and tweed jackets and even the occasional buzzcuts and horn-rimmed glasses, I’d taken my fashion cues from Apple Computer’s new CEO: jeans and a black turtleneck.

  DeWitt’s gaze lit upon me as I wandered the crowded conference room, with the faculty members standing against the walls as if they were coyly waiting to fill their dance cards. He was, I remember, brazen. He stretched out his arm toward me, curled his fingers, and pulled his hand back to himself, as if he held an invisible rope that was tied to my heart. And I turned and I came to him, easy as that.

  He wasn’t dressed like the other guys. If you had given the fifteen-year-old version of me a pencil and told me to draw a scientist I would have drawn him. Not a geek in a lab coat with Coke-bottle glasses, but someone at ease with himself, who gave the studied impression that he was only tapping a tenth of the coolness he kept in reserve. He had a mane of silver hair and a salt-and-pepper beard to match. No glasses to obscure his sky-blue eyes. Lanky and rangy, with stringy muscles that suggested regular fifty-mile bike rides. Faded jeans, Chuck Taylor hi-tops, and a Pink Floyd tour shirt that still looked new even though the show it advertised was five years back.

  “I don’t take you for a Floyd fan. I take you for a Rush fan, you have that look,” he said, and maybe he focused on the way my eyes lit up instead of the studied, neutral response I gave. “Yeah.”

  He slipped into a parody of Geddy Lee’s wheedling voice. “No his mind is not for rent. To any god or government.” I saw the tease for what it was.

  It wasn’t long after that before he escorted me away from the open house that was still in full swing and down to his lab. In the middle of the lab was the machine, a capsule about twenty feet long and twelve feet tall, shaped like a vitamin pill for a giant, painted an unfortunate chartreuse. I already knew something of the technical specs. Six layers of insulation, each at a lower temperature than the one before, until you reached the center, where one of the coldest things in the universe lay, trying to hear that long-sought, elusive message of gravity that emanated from the galactic center. Frost-coated pipes snaked from the machine’s top. Thick cables dangled from its side, tangled together, and led to consoles. I counted four colors of duct tape holding the thing together. It was amazing.

  “You see that?” DeWitt said, pointing at it, his other hand on my shoulder. “Science is going on in that thing, right here, right now. What’s going on out there in that conference room, the glad-handing and the butt-sniffing, that isn’t science. This is. I’m taking you away from that nonsense for a moment because I want to remind you of what really, truly matters. Because as you go through your career a lot of people who will look like they have your best interests at heart will try to make you forget. People come and go in this business. And it absolutely is a business. But Science will always be there for you. Remember that. No
matter what happens from here on out. Whether you end up with me as your adviser, or someone else.”

  He talked a good game. And the device before me looked like something I couldn’t wait to get my hands on. I was ready to coax it into yielding up new and spectacular truths. I was sold.

  —

  Years later, at a conference in Irvine, I ended up in a hallway conversation between sessions with an astrophysicist who told me of a saying that she in turn had heard among the community of science-fiction readers, who call it “Smullin’s Principle.”

  It is: Science fiction is a fantasy in which science always works.

  Only two of my cohort went to the DeWitt lab that year. I am not sure whether the department as a whole considered us the dregs of the barrel, but DeWitt, at least to me, gave me the impression that I had become his advisee because I was destined to be one of the elite. I can’t say the same for the other grad student who signed up for his lab.

  Her name was Claudia Pierson. She was a geek, in the days before geekdom became a badge of honor and then a marketing category. She was very much into Star Trek: The Next Generation, even though, if I remember correctly, the series had finished its run by then except for the occasional feature film. She had strong opinions about the superiority of Tom Baker over all other actors who’d played the lead role in Doctor Who.

  I think she cut her own hair, because I remember her uneven bangs hanging in front of her face, and the way she was always pushing them away. She had a different idea of the boundaries of personal space than everyone else, too, so that when she spoke to you she’d come so close that the two of you were nearly nose to nose. I had to fight an urge to step backward, because if I did, she’d just step forward again with no consciousness of what was wrong, and so she’d end up slowly pursuing me, chattering all the while, until I was backed against a wall.

  In our second year in the DeWitt lab he put us on this project that involved using the cryogenic bar detector in an attempt to pick up a continuous wave source. I was more hands-on with the detector, while Claudia’s job was writing the code that used algorithms to look for patterns in the data the detector generated, attempting to see signals in the noise.

  The decision to shift to a search for continuous wave sources instead of one-time events represented a potentially interesting and fruitful change in thinking. Historically, gravitational wave detectors had been employed in the hope of observing catastrophic events like exploding stars. But because the signal from such an event could only be received by a bar once, the signal had to be detected simultaneously by multiple bars in multiple laboratories in order to be considered valid. Multiple bars in multiple labs meant multiple physicists who had to agree that something interesting had been observed before they could publish their results.

  Consider the faint nature of the phenomena being sought. Consider that not all detectors were “on air” at the same time. Consider that each detector was fashioned somewhat differently, and that each lab had its own in-house algorithms for data processing. The hurdles to be overcome before you could publish results that could stand up to the scrutiny of your peers were nearly insurmountable. At least that was DeWitt’s opinion.

  Continuous wave sources were different from one-off events in a number of important ways, though. Think of a pulsar, a spinning neutron star. The gravitational waves it emits will be far weaker than the usual events that operators of detectors usually search for, because pulsars aren’t converting dozens of solar masses into energy. But those waves will be periodic, in a narrow frequency band, and highly predictable.

  So since you’re now attempting to prove that an event occurs every few seconds, rather than trying to get people to take your word that an event happened once that isn’t going to happen again, a single detector ought to be able to produce credible results without the need to confirm those results with another lab. Observing continuous wave sources would eliminate some of the need to collaborate, and DeWitt preferred to minimize collaboration whenever possible. Collaboration meant there was a danger of things that were “not science” entering into the scientific process, the things that were a consequence of being human.

  —

  DeWitt believed that every hand that touched the information tainted it, making the conclusions yielded from that information a little less believable. The idea he had was that the computer would examine the data without bias. Since it had no ego and no desire for publication or a Nobel Prize, it would not yearn to see patterns where there might be none, and its judgments of the evidence were bound to be fair. One might argue that because the computer was fashioned by humans and given its instructions by other humans, then its output would be the result of human will, even if the hands of those humans had been concealed through indirection. But it can be hard to let go of ideas like self-determining computers and pools of data that interpret themselves once they reach a certain size. If the goal is the appearance of absolute rationality, whatever that might mean, such ideas are most appealing.

  DeWitt was always looking over Claudia’s shoulder, often literally, as she coded. And DeWitt didn’t really comprehend code that well, not nearly as well as someone like Claudia, who’d entered college knowing she was going into physics, and understanding that because of the direction the field was taking, it would necessarily require a strong background in computer science. DeWitt had a general knowledge of programming that would let him muddle through, and when he looked at Claudia’s code he was something like a native English speaker reading a French translation of Macbeth with the help of a dictionary, getting by not because he understood French but because he knew the beats of Macbeth’s story. And as far as actually providing criticism, he was like a rich patron of the arts who took pride in not understanding art, but knowing what he liked. And what he didn’t like was the absence of positive results that Claudia’s algorithms produced. They looked at the data produced by the gravity wave detector and saw only random numbers. But the evidence for the waves had to be there. It had to be. Claudia just needed to look harder.

  After a month of this it became clear that DeWitt thought the problem lay with Claudia’s programming, and so if she tweaked her code enough the pattern that was surely lying within the data, waiting to be discovered, would manifest itself. So there would be days when she’d come into the lab and sit down at her desk, and he’d pull up a chair next to her and watch her code. For hours. She can’t have liked that. It didn’t help that he was unaware that he didn’t quite get what Claudia was doing. It didn’t help that when he saw something he thought he didn’t like, he’d comment derisively as soon as she typed the line, not getting that sometimes you need to get the line out of your head and onto the screen to see what’s wrong with it.

  He wasn’t nearly as hands-on with me. He showed his tacit approval of my work by leaving me alone. In fact, the work I did on the gravity wave detector took very little of my days in the lab. I drew a decent salary for turning some knobs and making sure that the detector had enough duct tape on it, more or less, and the rest of the time I worked on my own project, the preliminary research on the relationship between gravity and time that would lead to my development of the causality violation device. It was equally as likely that my clumsiness with managing the interface between the software and the detector was to blame for our failure to pick up a signal as Claudia’s coding was. Why did he ride Claudia so much, then?

  To this day I think it’s too strong a statement to say that DeWitt cooked his data. You might say that he wanted to win an argument against the majority of the scientific community more than he wanted to be right, but even that implies an unjustified suggestion of deliberate duplicity. Perhaps you could say that thirty years in the field had taught him to trust his gut, and so he thought it far more likely that Claudia, the wet-behind-the-ears grad student, was wrong, rather than that he himself could be wrong. Trusting his gut had gotten him very far.

  —

  Arrogance. It comes in two flavors. The fi
rst is the arrogance of youth. A person new to the field is full of ideas and high on his own promise of disruption, and thinks he will change the world with new thinking and sheer exuberance. But the second is the arrogance of old age, and this is far more dangerous.

  The first form of arrogance can be easily routed around. A young person is easily dismissed or told that he is wrong. His arrogance, if he chooses to hold on to it, will most likely endanger only himself. But the arrogance of old age can cloak itself in the authority of past accomplishments, which can serve to confirm the belief that one’s arrogance is justly held. It can shield a man from the realization that his beliefs have calcified, that he can no longer assess a situation accurately at first glance, that the world has changed around him and left him behind. Guarded from this knowledge, he remains content.

  Thus it does not ruin the life of the old person whose career is established, and whose past laurels will not be revoked no matter what future missteps he makes. But it can, and will, ruin the lives of those younger people who come near him, and fall into his orbit.

  —

  Because Claudia had to spend so much time writing and rewriting code under DeWitt’s eyes, her personal research began to fail. And sitting next to her, DeWitt would ride her about that, too, telling her that she was “a physicist now. Not a poet, not a philosopher, not a biologist. A physicist.” He’d tell her that she needed to be more focused, and there were a lot of things she didn’t have time for anymore, like “macramé, or flower arrangements, or knitting a scarf for Doctor Who to wear: you are being socialized into the profession, and the profession is what matters.” DeWitt imagined that whatever Claudia was doing when she wasn’t in the lab was “not science,” and therefore unimportant, unworthy, uninteresting: “Have you noticed that even though I’ve been doing science for three decades I basically live with this detector? I do that so I can understand it backward and forward and know its little moods. And I’m sure that there’s a lot of potential left to be squeezed out of this thing if we just try hard enough—what is that you just typed? That’s a bunch of nonsense! Is that the best you can do?”