The state of science reporting has always been pretty dismal, with a few major luminaries outshining the combined light of their lesser colleagues. I suspect that the science desk is the place where novices and interns get parked until they can finagle their way into “real” reporting, and they recognize that since their audience is for the most part even less informed about the subject they cover than they are, they are free to do a really bad job of it.

There are three or four major tricks that “science” writers and editors play to garner page views:

1) Silently swap a conventional definition with a generalized one: take a well-known concept, like electrical resistance or temperature, that has a numerical value that is strictly constrained to a particular domain, like positive values. Cover a researcher who is using a generalized version of that concept that permits values outside that domain: phase velocities greater than the speed of light, resistances or temperatures lower than zero. Report, preferably in the headline, “Researchers Achieve <<VALUE OUTSIDE DOMAIN>>” and (this is the important part) no where mention that you are strictly reporting on the generalized version of the concept. This will ensure your article conveys no useful information, that you will confuse your lay readership, and you will look like a complete pillock to those of us who understand what you’re doing.

In the case of a recent low in “science” reporting, a junior staff writer for Nature has produced an article of surpassing incompetence, which describes the achievement of negative (generalized) temperature in an ultra-cold gas. The trick is to replace the usual ratio-based definition of temperature (the average energy per degree of freedom) with the more general and more abstract differential-based definition used in statistical mechanics, which equates T with dE/dS, where S is the Boltzmann entropy. For systems where the available states are bounded from above, you can get into a situation where adding more energy squeezes the particles into fewer available states, which causes the generalized differentially defined temperature to be negative while still having a perfectly positive ordinary value of the average energy per degree of freedom.

By equivocating on which definition of temperature you are talking about, you too can be a complete failure as a science communicator, reporter, and human being.

2) Use an excessively general concept that covers vastly different things. We see this in reporting about dark matter, which is a vague catch-all concept that physicists and cosmologists use in a highly context-dependent way and “science” reporters use as if it always referred to exactly the same thing. There are two basic dark matter problems, and possibly more. The first is “galactic dark matter”, which has absolutely nothing to do with more cosmologically-relevant “non-baryonic dark matter”. “Galactic dark matter” is surmised to account for the rotation curves of spiral galaxies, which are far flatter than the amount of matter implied by their luminosity. That is, the amount of luminous matter drops of steeply as you move into the outer reaches of galactic arms, but the total amount of matter does not. Galactic dark matter is almost certainly made of perfectly ordinary protons and neutrons, just like us.

“Non-baryonic dark matter”, however, is stuff that is not like us. On very large scales, the universe does not have enough matter to account for the observed rate of expansion. We know the universe’s density in baryons from Big Bang models, and there simply aren’t enough protons and neutrons out there to account for how slowly the universe is expanding. There needs to be a lot more something to have slowed the expansion down, and we don’t know what that something is. Again, this is completely unrelated to galactic dark matter, and if you confuse the two you will confuse your readership, and look ignorant and sad to those of us who know what’s going on.

3) Adopt the most obfuscatory language possible. This, admittedly, is a sin the scientists have committed as well, although often at the behest of their editors, like Leon Lederman’s in “The God Particle”. The most egregious instance of this is “quantum teleportation”, which probably set back the public understanding of quantum information theory by a solid decade, as scientists simply lied and said nonsensical things like a “the universe is the same if we teleport the quantum state of an electron to the moon or if we send an electron with that state to the moon”. Since obviously the electron number of the moon differs by one in these two cases you’d think it would be fairly obvious that this is gibberish, but “science” reports slavishly repeated similar nonsense for over ten years. Fortunately this particular instance of egregiously misleading language has been corrected so many times now that it has ceased to do much new damage.

4) Talk about “breakthroughs”. There are almost no breakthroughs. Science and technology are almost always incremental enterprises, slowly accumulating better and better techniques and understandings. The “New Sensationalist” is full of examples of “science” reporting of this kind. Check any back issue from five years ago and calculate how many “breakthroughs” have resulted in significant change five years later.

5) Report on people and controversy rather than science and results. Scientists are people and sometimes we get into fights with each other. It can be amusing to watch, but if you’re reporting on it, you’re doing a human interest story, not a science story.

6) Wildly exaggerate the extent or importance of an effect, so that something that explains 7% of the variance is reported as “the cause”. This is done most commonly in sociology and medicine, leading to an endless stream of false and misleading stories that claim “Gene X Causes Cancer Y” (in 2% of the cases, but we won’t mention that) and “Situation P Results in Outcome Q” (although no causal analysis has been done and it only explains 10% of the variance at most.) The latter results in claims like “whiny kids grow up to be conservatives” and the like, because there is a few percent increase in the odds a person will have been uptight and rule-bound as a child if they are a conservative as a result. Of course, about half of such kids also grow up to be liberals, because about half of everyone grows up to be liberal.

This kind of story is simply innumerate pandering to people’s most simplistic prejudices, and attempts to create the illusion of mechanical causation where none exists.

That’s it off the top of my head. I hardly ever read “science” reporting these days, and when I do it’s purely so I can write a clearer explanation of the actual results here or in Facebook. I’m pretty sure “science” reporting has always been this way, and maybe all other kinds of reporting as well, but it’s a pity there are never more than one or two decent science reporters out there, given the impact that science has on all our lives.