|
||||||||
|
FORUMDiscovering and Inventing the Scientific Paper The experiments have been performed, the data tabulated, and the flag of statistical significance raised in victory. However, the scientist's job isn't done. All this effort will never see the late-night glow of a curious researcher's desk lamp unless it gets written up and published. Part historical record, part recipe, and part whodunit, the scientific paper is a strange chimera. As a result, this brand of writing involves much more than just taking one's laboratory notebook and dumping out its contents. Learning the art of writing a successful scientific paper has deepened my understanding of every facet of research, from improving the design of our experiments to assembling our experimental observations into a compelling story. Talking the TalkUnencumbered by experience, I was inclined to launch into sections of my paper with confidence: "Our study is the first to conclusively demonstrate that Process X...." The more experienced co-authors on the paper would redirect me with, "You can't say demonstrate, that's too strong." OK. "Our study suggests that Process X causes Effect Y and that...." "You can't say causes since we really haven't demonstrated causality." Oh. Then: "Our study suggests that Process X is associated with Effect Y and that...."I was fortunate in that after agreeing on a general direction for the paper my adviser would leave me to produce a draft we would then edit and I would rewrite. This iterative process has greatly enhanced my ability to bring experimental results into sharp focus and draw appropriate conclusions from those data. Some of the data you obtain is like finding the smoking gun with the suspect's fingerprints on the trigger, while other data are circumstantial. The importance of which experiments researchers choose to do and how to design them has also been driven home by the writing process. At no other time are clear, interpretable results so cherished. If They Only KnewThe general public believes that science lives in the ivory tower of objectivity. That's only half the story. The data may point to a suspect, but it's up to the author to decide which suspect and then secure a conviction. To do so, the author builds a case that guides the jury (the paper's reviewers) from Exhibit A to Exhibit B and so on such that the conclusion reached by the paper seems to all (or at least two out of three reviewers) the most reasonable one. Scientific papers therefore give the impression that research is done in a very linear manner. They do not reveal the multiple experiments being tried at once, all the false starts and dead ends. Such a chronology would be incomprehensible. Therefore, a scientific paper is a historical paradox. It is a record of things that did happen, just not of things as they happened.Now, I can already hear the shouts of indignation: "Science not a record of things as they happened? Blasphemy! Take away that man's notebook!" But wait. Consider filmmaking. I remember being a little boy and finding out how movies were really made. I was disappointed when I discovered that movies were not filmed in the sequence I ultimately saw, that you could shoot the end first and the middle last and then swap it all around in the editing room so it would make sense. I knew that what I was seeing on the screen wasn't real life, but I still felt duped. Only later did I come to understand that there was a truth behind the finished product that was deeper than the truth about which scene was shot on which day. So it is with the research article. Each set of experiments "took place." But the author has the freedom, the responsibility even, to explore and interpret the relationship among those experiments to create a meaningful whole. So when I swap Figure 5 with Figure 1 because I think that this arrangement will make the point of my paper more convincing, I tell myself, "I'm not trying to mislead anyone. It's my responsibility." But I wonder how the general public would react if they only knew how much of the value in scientific writing emerges from interpretation and therefore from invention, not just from discovery? Would they feel misled like I did as a child? Or would they come to see that the new ideas and new models that come out of the writing process constitute real advances? I hope they would see these as tools that take us further in our understanding. One Size Fits None?Ironically, although most papers present novel findings and are born out of personal interpretation, many papers also sound like they were written by the same sleep-inducing Überauthor, capable of weighing down the eyelids of the most caffeinated graduate student. The prevailing attitude has been that the data alone should be sufficient to captivate the reader's interest. If this were true, would we go to such lengths to create a coherent story? Authors strive to convince the reader of the logic of their conclusions but do nothing to pull the reader into the drama of the work and engage their imagination.I'm particularly fond of times when a paper reveals an especially unexpected finding, and the best we can manage is to start a paragraph with a lackluster "Surprisingly." I've been told that in the past, authors could be recognized by their writing style alone. Style is no substitute for content, but it is not at all clear how to develop a voice of one's own in today's literature. The profoundly influential Spanish neurobiologist Ramòn y Cajal was both an author and an auteur. Today, some call Cajal's prose flowery, but it is also inspiring and alive 70 years after his death. I wonder how many of today's "successful" papers will be able to speak in the same way. —Alex Carter is a seventh-year MD–PhD student in the neuroscience program at HMS |
