This is something I hope to see gain traction with the new gust of open wind blowing through scientific reporting. Open results, accessible results, means more than just the accessibility of the PDF, if you ask me. Publicly funded research has a duty to be readable and understandable by the public, who paid for it.
Or, at the very least, not more opaque than strictly necessary.
A well written paper takes a sympathetic view of the reader and helps guide them to understanding. This is not how academics are taught how to write. It's like a form of handed-down abuse.
Try reading Euclid's Elements. It was written before equations had been invented and is for the most part "plain language". However, it reads as vastly more obtuse than an equivalent modern formulation of the same facts with symbols and algebra.
The nice thing about mathematical notation is that a simple, one line equation can express relationships and detail that what would otherwise take a full paragraph to unpack. Personally, once I'm familiar with the symbols and underlying concepts, an equation is much easier to hold in my head than a long jumble of plain language explication.
I think most researchers aren't simply pomping up their paper with obtuse symbols. Mathematical notation is both compact as well as a common language, so why not use it? As a side benefit, we also get to lean on the vast background of mathematical research which offers potential precision that would otherwise be completely lost.
Popular science writing is very important, but it’s just a small part of the practice. Doing more of it would mean less actual research unless funds and bodies are added for it. I'd like to think that as a community we servo around the sweet spot given the resources.
No, it is not. This is why I drew the distinction in the first place.
> Normal people can only understand what is reported.
Which backs up my point that sensationalised reporting is a problem. Peer-reviewed, scientific papers are written for other scientific researchers — that's as it should be.