If Poynter blogger Andrew Beaujon had been a science writer, he might have noticed something fishy about the graphs he posted in Why Journalists Drive Scientists Crazy, in Graphs. The graphs do not appear to be based on any data. There is not a data point to be found.
If data were involved in any way, there’s no mention of where this data came from or how it was obtained. A science writer would ask about the data, and the error bars, for that matter. Are these even really graphs, or just illustrations? Whatever they are, the purpose seems to be to express how one scientist, Sabine Hossenfelder, feels about science journalists. In a blog post, she expresses some frustration.
And Beaujon seems to agree:
The journos feel they have to elide detail so a general audience can read them. The scientists feel the resulting “knowledge transfer” to readers is pitifully low.
Do very many science journalists really feel this way? Is Beaujon assuming that Hossenfelder’s view can be extrapolated to other scientists? Over the years I’ve talked with hundreds of scientists and I see little evidence that we drive them crazy. The hands down biggest complaint I hear from scientists about science journalism is there isn’t very much of it in newspapers or other mainstream publication these days.
But apparently we drive Hossenfelder crazy because, she writes, we sacrifice accuracy and substance for accessibility. In one graph she claims to plot information content against popularity, though, again, where are the data?
The whole conceit feels like a false dichotomy. It’s like saying that food can either be tasty or nutritious but not both. In writing, if anything, readability and accuracy tend to go together. One hallmark of B.S. is that it’s hard to follow because it can’t be explained with logic.
If there’s anything journalists sacrifice accuracy for, it would be sexiness. A reporter may get more clicks on a story that promises a new drug will cure cancer than a story that simply explains how a new drug works and why scientists hope it will help. More people might click on a story that says GMOs will kill you than one that carefully lays out potential risks and benefits of a particular GMO product. But sexiness should not be conflated with readability. If anything, the less sexy but more accurate stories are likely to be clearer because the reporter can back up the headlines with evidence and logic.
Hossenfelder also laments that science journalists aren’t more like sports writers, since sports writers get to assume their readers know something about sports and don’t have to dumb things down.
Hossenfelder seems to be saying she’d rather read technical articles full of equations than journalistic stories aimed at the hoi polloi. Do other scientists feel this way? Do they really disdain readable stories that give context, history and some level of interpretation, even in fields outside their areas of expertise?
There is reason to doubt this.
Just look at what scientists like to read – and pay to read. Many read the journal Science. The back part of Science is made up of fairly technical scientific papers but the front section consists of interesting stories written by journalists. These stories are mostly devoid of equations and clear enough to be understood by anyone. The journal’s news section has been that way for a long time and the scientists seem to like it that way.
Some years ago I wrote about physics and astronomy for Science and then switched to a newspaper. You might think I had to dumb down my writing, but in fact, I found writing about physics for biologists, psychologists and geologists is not that different from writing about general science topics for cops, firemen, waitresses and bus drivers. There are extremely intelligent, interested people in all walks of life.
While Hossenfelder (and Poynter) concern their posts with the subtraction of information from stories, much of what we do in the interview process is to push scientist to fill in information that’s missing from their technical papers – motivations, caveats, and logical steps they’ve skipped between their data and their conclusions. We ask them if they can fill in holes that their colleagues and rivals poke in their arguments.
Several prominent science journalists defended the profession in excellent comments. Here’s a good one from Chris Joyce:
I have to disagree with this analysis. For one, the graph commits the very crime described: it looks great and lends legitimacy, but one needs the underlying data it's based on to verify whether it's really representative. I suspect it may oversimplify or "dumb down" the details. For another, go to specialized sports media sites; compared to newspapers or broadcast (I'm a science journalist in the latter business), the level of detail is orders of magnitude higher. General media indeed "dumb down" the level of detail in sports writing, because many readers are not sports fanatics. And last, scientists more often than not overestimate the value of "their" details in communicating their work. They write, and have been taught to write, for their peers–a small group with its own set of standards. Those standards are designed to help other scientists test and replicate their work, and also to prove to skeptical peers that they've designed their protocol correctly and properly interpreted that data. That goal is so very different from communicating science to non-scientists, who mostly will be bored/confused/disinterested in, for example, methodology, which is only rarely important to understanding the results. As for layering levels of sophistication, we already do: You can read Scientific American, listen to NPR, or read the New York Post–and everything in between.
Well said. George Musser wrote this comment under Hossenfelder’s original blog post:
Our present system DOES have levels, shading from Physics Today and Nature N&V to the daily papers. In universities, we have everything from graduate seminars to physics for poets. Do you not think this system is working? If not, what about it is failing?
I'm struck that you keep mentioning terminology as important. I think this is definitely a difference between science writers and researchers. Researchers tend to see education as professional preparation, in which case you do need jargon and technique. But journalists are catering to an audience of dabblers.
I find that it is often the SCIENTISTS, not the journalists, who push the knowledge curve toward the left. They preemptively dumb down the discussion, sometimes because they think the public can't handle it, and sometimes because they themselves haven't done the hard interpretive work of their own theories. They say things like, "It is as it is because the equations say so", rather than attempt to spell out the physical principles. Or they say, "This work has implications for X", and when I ask what the implications are, they can't tell me.
Again, well said. Scientists are not universally good at interpreting their work or putting it into context. Beyond that, scientists often radically disagree with one another on questions that have a crucial bearing on environment concerns or public health. Journalists can weight multiple perspectives and therefore give readers a more complete picture than any one scientist could do by simply describing his or her own research. Many scientists appreciate such stories, which is why they so often complain that there aren’t more of them.
And finally, a question to Poynter: Isn’t there usually some tension between journalists and those we cover? If there was never any tension, would we really be doing our jobs?
Leave a Reply
You must be logged in to post a comment.