|
||||||||
|
FORUM
Why Science Can't Afford to Be Sacred
Alex Carter So, what do you think when you lug home a couple of pounds of journal articles to read at the end of the day? That you're being studious, responsible, good? That's what I thought. I hoped to construct my experiments on the foundations the latest luminaries had laid in the most prestigious journals. How could I go wrong? You see, I thought I had learned graduate school's first lesson: that textbooks are not bibles, that half of what's in them is wrong and the other half outdated. So I turned away from the false god of textbooks and put my faith in the latest literature. That's why I've been surprised to hear more senior researchers question these papers, from the quality of the data to the validity of the conclusions. What? You mean I'm making the very same mistake as before, elevating the literature to the status of canon? Well, no more! I have finally seen the light and realize that from the beginning I should have placed my complete trust in our faculty and ... oops, there I go again. Maybe it's a natural though sometimes misleading impulse to want to believe in scientific explanations—narratives that make sense of the world—especially when they are told by those who have more experience. Why must scientists curb this impulse to buy into these stories too readily, even their own? Aren't we all trying our best to report the truth insofar as our experiments can capture it? The answers to these questions seem to fall into one of three categories: money, man, and machine. MoneyThis year a relatively generous 25 percent of neuroscience research proposals were funded. But it is still easy to imagine that fierce competition for research dollars might put a lot of pressure on some labs to devise the "sexiest" packaging for their project—maybe by emphasizing its clinical implications or any other claims that can be substantiated by the data, sometimes just barely.This has also been a period of economic prosperity in which many have jumped on the entrepreneurial bandwagon. At a recent discussion organized by the Harvard Biotechnology Club, one speaker described a new kind of hot-shot faculty around the country who will only consider positions that come with guaranteed opportunities to work as consultants, license the technology they develop, or even start their own companies. When a researcher at an academic institution is also invested in a company that depends on the success of his research, the potential for conflicts of interest is obvious. ManAs human beings, I believe we're all compulsive storytellers. Scientists maybe even more so. The typical narrative has a beginning, middle, and end. The scientific narrative is focused on describing process, establishing causality, and defining mechanism.Our very brains are designed for this task of recognizing patterns and maybe even identifying causal relationships. These aspects of human psychology may predispose us to see causality in data where only correlation exists. We seek to establish causality because it provides us with a sense of order and closure that we long for at the end of all stories. Even though we realize full well that in science no story is ever complete, we persist with our process-meets-molecule, process-loses-molecule, process-gets-molecule-back scenario. Combine our storytelling propensity with the pressures to compete for funds, and it is easy to see how biases might creep into experimental design or data interpretation. MachineToday's research instruments allow us to peer into previously inaccessible biological processes. That is, provided the machine is turned on and properly calibrated, and that you've set that knob over there to 13 and this one to -2.5, even if you don't know what they're for—and that you're recording from the right spot. I've been told by more experienced researchers that with some methods, it's possible to get any result you might want.Recently, I've been using a fancy microscope and software to look at synapses, the tiny specialized connections between nerve cells. But how can I be certain that what I'm seeing is real and not just the computer hallucinating because I fed it bad numbers? Sure, these are useful tools. However, their come-hither interfaces and willingness to give you what looks like the data you want are seductive. You think you've got the instrument wrapped around your finger when in fact you don't know the first thing about it. To entrust yourself to technology's charms is temptingly easy but also dangerous. Are We There Yet?Finally, we are living in a megahertz-driven world where faster is the order of the day. We're becoming accustomed to an acceleration in the pace of technological change.Sure, we can compile our data more rapidly now with computers. But we still have to design good experiments, perform all the appropriate controls, and show that our results are reproducible. These may well be the rate-limiting aspects of scientific research that prevent us from cutting corners without threatening the integrity of the enterprise. Will waiting two weeks for my mouse pups to be of the right age for an experiment become intolerable when I can download the entire contents of the Encyclopedia Britannica in two nanoseconds? Might I become inclined to do fewer repetitions of the experiment, which will save time but compromise the interpretation of my results? How do I remain hopeful that I can ever know about this world when modern science is riddled with all these caveats springing from our culture, our tools, and even our very makeup? These pitfalls only remind me of what we already know: that the process of science is imperfect. We run our observations through a filter of criticism and a prism of reason. But after the data has passed all these rigorous tests and we have dutifully reported our findings publicly, we get to ask ourselves in the privacy of our own thoughts if we believe our—or anyone's—conclusions. Accept this final, personal step as an act of faith, and science becomes just one way of knowing about the world. —Alex Carter, a sixth-year MD–PhD student in the HMS neuroscience program |
