Science of the Times: Newspapers and popular science
Are you a science journalist? (Quick, pre-post request): Please answer five, super-quick questions for Sylvia Tippmann (@SyTipp), a science journalism student based in London.
For more than two centuries, newspaper journalism has helped shape the world view of millions of people - offering us a window on global events that would otherwise have remained unknown. Distant wars. Economic upheaval. And, of course, the great scientific discoveries of the age.
To approach my OpenSciLogs project - in which I aim to investigate historical trends in popular science reporting - my first instinct was to start by looking at back-editions of magazines, such as The Popular Science Monthly (now just Popular Science) and the stalwart Scientific American. But while these kind of publications will certainly be useful, I've decided to begin from a different angle: I'm going to take the paper route.
At first glance, newspapers might not seem like the best jumping-off point for a popular science study. After all, of the couple of hundred articles you can expect to find in a typical daily paper, chances are that only a small percentage will be dedicated to science. Surely, then, concentrating on the science-only titles - the SciAm's and the PopSci's - would be more fruitful?
But the more I've thought about it, the more I'm convinced that newspapers are the way to go. The science stories in the dailies may occupy a smaller fraction of the total content, but spread that small fraction out over 365 editions per year, and multiply it by a huge readership, and it's clear that even the least scientifically-inclined tabloid will trump any magazine in terms of coverage.
This means that when it comes to popular science in its strictest sense - that is, the science that the majority of the population encounter - journalists have, historically, had the most influence.
A Century of Science in Seconds
When it comes to constructing a historical overview, though, the sheer scale of print journalism makes an "article by article" approach impossible. Say, for example, you want to build up a picture of the flavour of science reporting in a particular newspaper across the whole 20th century. You might try to look at one issue per year, and classify the scientific articles by subject, physics, psychology, evolutionary biology, etc. It would be a chore, but possible. You sit back, satisfied. A hundred years covered.
But in reality, what you'd have actually looked at is a tiny part of the picture - less than one third of one percent of the total number of articles. For all that effort, not a particularly strong scientific sample. Luckily though, there's an easy way to scan the whole hundred percent.
The bulging archives of the British Library contain every edition of various major newspapers - including The New York Times and The Guardian - every single word of which can be scanned in a matter of seconds. (The digital age is gradually amputating the first letter from research).
I started with the NYT.
Ever wondered how many New York Times articles contained the word "biology" in the 20th century. Well, exactly this many.
I've done my best to conceal it with these high school-style bar charts, but what I've just done is something quite amazing. Between the years 1900 and 2000, there were around 5 million individual articles published in the NYT. Average word count of each? Let's be ultra-conservative and say 100. That's half a billion words. (I shudder to think how long it took some shut-away team of researchers to compile all that.)
Powerful doesn't quite begin to describe this search engine. But before I declare the history of science journalism completely solved, there is an important drawback that needs addressing. What the search engine does, quite brilliantly, is tell you the number of articles which mention your search term. What it doesn't do, is tell you what kind of articles they are. Simply put, if an in-depth piece on this year's World Series, or Prince George's dimples, happened to mention, even a single time, the word "chemistry", then the search would count them along with all the worthwhile articles on titration.
This all-inclusive type of search obviously isn't ideal for my purposes; I want to focus on articles which are just about science. But I think, with a careful choice of search term, I can isolate them.
Finding the Words
The two searches I showed earlier - for biology and chemistry - used fairly broad terms. The word "biology" could conceivably be used in pretty much any kind of article. But what if, instead, I'd gone for "natural selection"? People don't really wheel out that phrase unless they're writing about evolution.
Hey presto. Different term, different pattern - and one which I can be fairly sure hasn't been overly diluted by non-scientific articles.
By using more exclusive search terms like this, I think I can home in on the truly scientific articles in these epic archives. That's what I'm going to do next. So I'll end this post with an open question: What else should I put in the box?