Another blog about scientific writing appeared on Friday by Adam Ruben.
It contains all of the standard complaints, albeit in a humorous way, about why scientific writing is basically dull, turgid and opaque.
Adam concludes that:
But there’s a reason scientific journal articles tend to be dry, and it’s because we’re writing them that way. We hope that the data constitutes an interesting story all by itself, but we all know it usually doesn’t. It needs us, the people who understand its depth and charm, to frame it and explain it in interesting ways.
Are scientific journals dry and dull? Beauty is, often, in the eye of the beholder.
I have been running some NMR experiments this week. Nuclear Magnetic Resonance Spectroscopy (NMR) is a great technique. By using atomic nuclear spin, you can use NMR to measure the environment atoms find themselves in.
This is a simple example of a hydrogen 1-D NMR spectra for ethanol.
The single peak on the far left is the OH hydrogen, the peaks in the middle are the CH2 hydrogens and the peaks on the right are the CH3 hydrogens. If you make a new molecule which has hydrogens, you can dissolve it in ethanol and then see where this new molecule’s hydrogens show up. From the new peaks which appear in the NMR spectrum you can make a deduction about the structure of your new molecule.
Did I spend the week running NMR spectra for new lovely molecules I made? No. I spent the week collecing NMR spectra for all of my solvents. Eventually, I will dissolve my new lovely molecules in these solvents, eventually. NMR on regular old solvents is honestly a bit dull, but it does tell you if your solvents are pure which is kind of important. You want to make sure the apperance of a new peak is really about the molecule you WANT to see not just an impurity.
Unfortunately I had some random peaks in my NMR spectra for some of my solvents. How did I know what they were? I read a paper – which is one of the scientific world’s less interesting papers. Its called:
NMR Chemical Shifts of Common Laboratory Solvents as Trace Impurities
(J. Org. Chem. 1997, 62, 7512-7515)
Is this paper entertaining to read? Not really. Is this paper highly important and useful? Absolutely. Its been cited 851 times and I would think used much more than indicated by its citations.
Does this paper need to be written in a more accessible way? Not really. I, like most people who use this paper, don’t read the text so much as look at the tables, so I can identify that annoying little tiny impurity peak in my NMR solvent spectra.
If you look at the publication lists of many established scientists, these lists usually show publications in a variety of scientific journals – from the more generalist high impact journals (such as Nature and Science) to the Journal of Very Technical Details. Not all science belongs in the same journals and the writing styles are different for these different journal types. In a specialist journal, I personally don’t want to be entertained with flowery words and prose, I just want them to get to the point. I don’t care if its written in the 3rd, 4th or 8th person as long as it is clear. I also find these ‘dry’ journals pretty damn interesting because I am interested in the subject.
While I do take the point that science can be and maybe is often written in a dull way, not all science is going to be interesting to all people. The way science is written in some journals is just fine, just like any technical report on any subject.
Everything should be written in a way that achieves its particular goals. A phone directory written as poetry isn’t a very good phone directory. The ee cummins telephone tome would not only be unneccessary, it would actually be unhelpful. A scientific article in a technical journal is not written for a general non-specialist audience. A Nature paper should be written for a scientifically literate audience and an article about science in The Sun should be written for everyone.
Here, here!
I have to admit not reading the blog you cited complaining about scientific writing, but as you noted, it is technical writing and when properly done I do enjoy the clear, crisp and cautious way in which papers are written.
Several years ago I would have been bored looking at NMR spectra, but since entering into a collaboration that has used NMR to solve the solution structure of a protein domain we are interested in, I can’t get enough NMR! Our initial collaborative paper was a 2 dimensional assignment of the 133 amino acid domain and appeared in the Journal of NMR. As you say, not fancy stuff, but it led to some really exciting findings which were later published in more lucrative journals.
But my point is that science writing is highly specialized, and with the increasing knowledge coming in, what’s fascinating to one scientist will be dull to another, and I don’t think the writing has anything to do with that.
I am going to have to bend my brain a bit to remember NMR from undergrad chemistry lab days… the CH2 hydrogens each show up as a doublet? That sounds vaguely familiar…
Still, running solvents through does have the side benefit that the results ought to be nice and predictable, and easy to understand (impurities notwithstanding), which I think might be a good thing for, say, a Friday afternoon. 🙂
Hmm… takes me back.
As a chemistry undergrad in the early 80s I learned a fair bit about 1H NMR spectra in structure determination, and I later did a whole PhD in 31P NMR – though as that was mostly bits of living tissue (see e.g. here) there was none of that high-resolution nonsense with its fancy spin-spin coupling.
Anyway, to get back to the main point, it would definitely be hard to make a compendium of chemical shifts read super-interesting. Of course, that applies to a lot of necessary but not desperately exciting experiments.
PS My favourite piece of obscure NMR terminology was the ‘Nuclear Overhauser Effect’ – entirely for the name, which always sounded like some kind of special modification to a Star Trek Warp drive.
My least favourite thing about NMR was having to shim the magnet manually, for any with memories of NMR in the old days, which used to feel like a true exercise in extended futility. I imagine they have computers to do it nowadays.
PS Actually, the other thing about the Nuclear Overhauser Effect, which Sylvia and maybe Richard W might appreciate, was that it reminded me of the name Offenhauser. So maybe it wasn’t a Star Trek motor – more something from an early 1970s Indy 500…
That’s a bit of North American auto racing history I knew nothing about – thanks Austin.
I was thinking Offenbach, which is a whole different thing entirely.
Am I off-topic yet?
You still have to shim… Or should I say one still must shim the magnet
Good grief – still?
I’m genuinely astonished it hasn’t been computerised, as I’d have thought it was a classic minimisation/maximisation problem. Can anyone explain why not?
I think because shimming just optimises the magnetic field, and each sample is different so has a different effect. It doesn’t take that long so maybe it’s just faster if you do it yourself… I don’t know, honestly
There’s writing clearly and succinctly, and there’s turgid. Prose doesn’t have to be flowery to be readable. There is no excuse for bad prose, even (especially!) in technical articles.
That is true, I agree, but that is a different topic. Writing badly I think is usually the problem, not active vs. passive voice or any of the other accusations people have about science writing. There is no excuse for bad prose in technical articles, I agree, because technical writing should be the opposite of prose…
No, absolutely not. Look up ‘prose’.
Technically it is prose, but when I read ‘prose’ I think literature – which admittedly may be just me. The blog this guy wrote was prefaced by his supervisor saying ‘you don’t write like a scientist’ but maybe his writing was just poor in the context? Who knows. Writing is a very plastic thing, and sometimes ‘bad’ vs. ‘good’ writing can be a matter of opinion and sometimes ‘dull’ vs ‘interesting’ can be too.
Of course nobody is expecting technical articles to be great literature. But if you can’t get your point across, then you are writing badly. Well-written scientific articles, even if they’re just a list of chemical shifts, are a joy and should be encouraged. I’d even say a necessity: well-written articles make finding the information you require so much easier (and there’s more chance of getting the right information…)
There’s no excuse for anything else.
Dude, I agree with you, I am not defending bad writing…
that’s all right then… dudette. 😉
I would have said mate, but that doesn’t sound right coming out of my mouth. I also think learning to write well in a variety of different contexts is not easy, it takes years to hone written skills.
That is why we usually treasure things like Perspectives and Reviews on the Nature back half treadmill, the style is much more free than for the hard-core research articles and make for a nice change. They usually do not hang around for long (neither do big-dino papers)!
When editing primary research though our overriding aim is clarity, not fancity, if I may coin a word. So we’ll break rules, the style guide and convention if it makes the paper easier to understand. Style is OK but we never let it get in the way of the science.
For the really arty stuff we have Futures…
I don’t think any journal should be aiming for fancity. Even the guff in Nature (by which I mean the stuff everybody reads, not the research papers. In case that wasn’t clear…) shouldn’t be flowery. It can be expressive and paint lovely mind pictures–but you can achieve that without being fancy shmancy.
And as you say, you can put the really crap stuff into Futures where it belongs.