True scientific news doesn’t make the news

Journalism and science have very different time scales. A week-old newspaper is barely worth reading. A week-old scientific paper is still warm from the photocopier. Journalism neglects this discrepancy and pressures science to hurry up. Numerous headlines with the eye-roll-inducing opening “A new study shows” imply that only the newest (or weirdest) science is worthy of attention. Google declares about 2000 times as many results for that phrase compared to “an old study shows”.

To make sense of the differences between science and its representation in the media, it first helps to figure out what does and doesn’t make news, and what does and doesn’t get widely shared or discussed. With that established, we’re better positioned to overcome the challenges of communicating all scientific research, not just the sexy aspects that make good headlines.

The first step for a news writer to get their piece past the editor is to have a hook. This advice for pitching an op-ed, for example, repeatedly notes the importance of framing it in terms of a current event. Matt Shipman in his Science Communication Breakdown identifies this framing as the primary obstacle for science writers wanting to write about everyday science questions:

There is one clear challenge to this approach: the lack of a timely news hook. Why, an editor may ask, are we writing about this now?

Scientific writers concentrating on newly published research automatically have a hook: their article is only timely if it comes out when the study is published. Yet this timing is arbitrary for most scientific fields. I’d guess that the general public, and even a well educated subset, would be hard-pressed to estimate the dates to within ±5 years that randomly chosen scientific findings were made. Here, try it. Name the year when Paris Sabeti and others determined that tolerance for lactose in humans is actually a rather recent (last 5000–10 000 years) evolutionary adaptation. (Before this, adult humans were unable to digest cow’s milk.)

The lactose finding is at least 15 years old judging by this article in the American Journal of Human Genetics, with the finding building on existing evidence from papers dating back to the 1970. Chances are that a scientific writer could dress up this finding and present it to the audience as brand new, published just yesterday even. With the exception of some doctors and nutritionists, the general public would be none the wiser.

There would actually be benefits to a science writer discussing a 15-year-old article, with its age facilitating much more discussion. What other findings did the article incite? Have their results influenced the scientists’ subsequent research. Did the result get confirmed, refuted, or forgotten? (Remember, peer review doesn’t confirm a manuscript is correct.) Such contextual questions cannot be answered when discussing a newly published study. A headline from a Science News article posted in December 2018 makes this abundantly clear: These 2018 findings could be big news — if they turn out to be true.

Regardless of whether the lactose adaptation was discovered in 2004, back in the 70s, or just yesterday, it’s an interesting result that deserves mention in the news. Of course, the date certainly matters if we’re discussing how and whether this news spreads. Back in 1988 in his A history of news, Mitchell Stephens wrote that the fastest medium with the largest potential audience will disseminate the bulk of a community’s breaking news. He was discussing how radio and television were overtaking newspapers, but his argument is just as applicable today when shared articles on social media determines much of the news we read.

Sharing in any form is key to the spread of news. The Wikipedia page on News notes that humans exhibit a nearly universal desire to learn and share news, which they satisfy by talking to each other and sharing information. Not just any news gets widely shared though. There are two key ingredients. The news story should signal something about the person sharing, and it helps if it is controversial. Steve Omohundro describes this in the framework of costly signalling. As he puts it, nobody is tweeting 2 + 2 = 4 because everyone agrees with it, and it says nothing about the tweeter themselves. Yet tweeting about a polarising issue actually requires more effort as it generates antagonism and leads to time-consuming arguments. Nevertheless, from a signalling standpoint, there’s a benefit to making noise.

Trilling et al. (2016) list several article characteristics that enhance “shareworthiness”. Not that it lasts. They point to other studies showing that a news article’s proliferation on twitter usually saturates after one day. Assuming this short time frame holds true for sharing of a new scientific finding, this implies a short collective attention span for any given study.

Don’t get me wrong, many scientific studies deserve their 15 minutes of fame. And social media provides a platform that enables wide dissemination of these studies, which has to be a good thing. But we can’t forget the stories that news and social media are not conducive to highlighting. As physicist Max Tegmark puts it:

If we define “interesting” in terms of clicks and Nielsen ratings, then top candidates must involve sudden change of some sort, whether it be a discovery or a disaster. If we instead define “interesting” in terms of importance for the future of humanity, then our top list should include even developments too slow to meet journalist’s definition of “news”.

A lack of disaster has actually made NASA too successful argues Nicholas Christakis. By making manned and unmanned space travel safe and routine, it has become boring to the public, thereby discouraging Americans to think the country should increase NASA’s budget. In a similar vein, physicist Lisa Randall notes that the true sign that dark matter searches have succeeded will be that the discovery will be taken for granted and will afterward cease to be news.

Weather prediction is a particularly good example of a scientific field that is gradually, but persistently, improving. We make use of weather predictions daily, but seldom appreciate the continual improvements in skill of these forecasts. Samuel Arbesman, in his aptly titled piece, weather prediction has quietly gotten a lot better, notes that the accuracy in predictions is increasing by about an extra day per decade. The cause? A steady accumulation of scientific advances and technological improvements. Even the most complex systems become more orderly as different pieces of knowledge fall into place.

A day per decade equates to about 30 seconds per day. No wonder that improvements in forecasting seldom make the news. Of course, this is the wrong way to look at it. The story of weather prediction, particularly numerical weather prediction (as opposed to old wives’ tales or rules of thumb for sailors) goes back more than one hundred years. In 1916, the British mathematician Lewis Fry Richardson tried to use mathematics to make a weather forecast for 7am, May 20, 1910. Yes, that’s right, a forecast for six years earlier. Although the particular forecast was a failure, the method he invoked is essentially the one we use today. I’m guessing his failed forecast didn’t prove newsworthy at the time. But in hindsight, it represents a historic moment well beyond only meteorology. By the way, it wasn’t until 1953 that a computer finally outpaced the weather and resulted in a prediction 90 minutes into the future. (Credit to Andrew Revkin and Lisa Mechaley for these weather history morsels.)

The forecasting improvements make clear how important time scale is in discussing scientific advances. Another example would be the slow and steady increase in life expectancy over the last 100 years as noted by peace researcher John Galtung. Senior scientists who have spent their careers in a particular field may be able to appreciate the past gradual progress. But for grad student and postdocs, it’s easy to forget the long game and instead highlight the importance of the newest findings and how they fit into the scientific landscape right now.

I recently read a fascinating piece explaining New Zealand’s involvement in sea ice research over the last 40 years. It’s the story of the plainly titled project K131, an ongoing endeavour that has included research on the mechanical strength of glaciers and sea ice, the propagation of ocean waves through ice-covered waters, the inner structure of the ice as determined by nuclear magnetic resonance, the oceanography of a cavity beneath a 200-m thick ice shelf (you have to melt a very long hole for this), and the occasional tangent like the development of earthquake isolators for New Zealand’s parliament building and its largest museum. The threads that sustain this project involve a few key scientists and some simple (but also clever and thriving) methodologies like converting shipping containers to lab space for on-ice field work. Big science in serial-meandering mode is how the authors described this long-running project, adding that such an approach is essential in order to work over a long period of time on sometimes unfashionable topics (keeping in mind that climate science was once unfashionable) with changing funding.

I would, of course, find K131’s story fascinating as I was one of a couple dozen students to be part of it. The story made clear how my one year of master’s research fit into a much longer saga. At the time, I could see obvious links to previous studies in the literature, but it was never completely clear how my particular project came about or how it tied into the overarching and long-term goals of New Zealand’s sea ice program.

When discussing your science, it is beneficial to the general public, the scientific community as a whole, or just your specific field, to communicate your work on long time scales. For example, did an instrument developed decades ago make your current work possible? Have recent technological improvements now enabled you to test a theory put forward many years back? What scientific theories are gradually developing consensus? These facets are just as important as the flashy and exciting future problems that your science may address.

If you do want your latest scientific finding to gain traction in the news or social media right now, it evidently helps to find a hook and tie your work to current events, or just have a provocative tag line. If, instead, you remember that science and journalism are different, you can be content with realising that your finding may actually be more important 10 years from now. It may never go viral, but that doesn’t imply it is unimportant. As scientists, we’re used to delayed gratification. We need to maintain that even when we think our latest findings should be shared with the world.

Author: Ken Hughes

Post-doctoral research scientist in physical oceanography

%d bloggers like this: