Lightening up the lecture

I write occasional commissioned pieces for Jobs.ac.uk, and today I’m mostly pointing you to my latest one over there.

The lecture gets a bad rap. It’s pretty easy to find essays deploring the whole idea of teaching via the lecture – compared to other techniques, you’ll read, it’s outdated, it’s ineffective, it’s boring. But the lecture isn’t going anywhere, for at least two reasons. First, in some situations it’s actually quite effective; and on top of that, lectures used in combination with other teaching approaches can make them more effective. (Yes, there’s literature establishing this.) And second, even if lectures weren’t effective, academics would lecture anyway. We’re conservative; and perhaps surprisingly, so are our students.

So if (at least some) lectures are here to stay, it becomes important that there’s a world of difference between a bad lecture and a good one.

What would you like to see in a 3rd edition of ‘The Scientist’s Guide to Writing’?

The second edition of The Scientist’s Guide to Writing isn’t quite three years old, but I’ve been talking with the publisher about a third edition. And I want your help with it.

I have to admit that thinking about a 3^rd edition makes me slightly uncomfortable. I hate edition greed: the relentless production of new textbook editions every couple of years, just to undercut the used-book market and make students pay more. You’ve all seen it: a “new” edition of a 1^st-year biology text that puts the chapters in a different order, swaps in a few new examples in place of older ones that worked just as well, and sports a couple of shiny new digital bells and whistles that don’t really add anything worthwhile. I don’t want to do that!

But there’s one major reason that The Scientist’s Guide needs an update: AI.

What our new book looks like: “Teaching and Mentoring Writers in the Sciences”

About a year ago, I posted some basic information about my new book, coauthored with the brilliant Bethann Garramon Merkle. It’s time for an update: as books grow up, they change and presumably improve. The book is now in production (with the University of Chicago Press, for release in Fall 2025). It has a new title, new organization, and some new content. So it’s time to tell you more about Teaching and Mentoring Writers in the Sciences: An Evidence-Based Approach.*

What does good writing advice look like?

“Writing Advice is a Lie”, opines Henry Oliver on Substack, with an argument that writers (of fiction mostly, I think?) should ignore all the writing advice that surrounds them on the grounds that “Almost all of it is wrong. Flat wrong. Plain wrong. Waste-of-time wrong”.

As you can imagine, this took me a little bit aback. After all, I give writing advice a lot; heck, I teach courses consisting largely of writing advice, and I’ve published a book full of it.*

If it were only that a random Substack post made me tear my hair out, you could argue that I ought to find something better to do than lose sleep over it. But “Writing Advice is a Lie” made me think a bit, because I will cheerfully admit that some writing advice is a lie. There really is a lot of bad writing advice out there. So what distinguishes good writing advice from bad writing advice?

Trolley problems in scientific writing

Last week I wrote about the impossibility of deciding on the single best way to write anything – because readers differ, and what works well for one reader will work poorly for another. (Don’t panic, this isn’t a completely nihilistic position. There’s definitely bad, good, and better writing; there just isn’t best.)

There’s an interesting consequence of this: it brings trolley problems into scientific writing. Not, I mean, literally: I suppose you could fold up a copy of your latest paper and jam it, just right, into the tracks to divert the trolley – but let’s leave that to the A-Team. What I mean, instead, is that it opens up the possibility of weighing, for a particular writing decision, readers advantaged (or recruited) vs. readers disadvantaged (or repelled).

There’s no best way to write something

What’s the best way to write your next paper – or, to ask a simpler question, the best way to construct your title, write the opening sentence of your Introduction, or plot the pivotal data set? The bad news is that there isn’t a best way. The good news is exactly the same.

When I suggest that there isn’t a best way to write a particular bit of text, I don’t mean that there aren’t bad ways and better ways; of course there are.

Buffon’s eulogy, Pliny vs. Newton, and is science a big pile of facts?

A couple of months ago, I reviewed Jason Roberts’s new book Every Living Thing: The Great and Deadly Race to Know All Life. It’s the story of two eighteen-century naturalists, Carl Linnaeus and Georges-Louis Leclerc, the Comte de Buffon, and their contrasting approaches to studying biology. In a nutshell, Linnaeus was looking for structure and order in Earth’s biodiversity, while Buffon argued quite explicitly that there wasn’t any and we shouldn’t look for it. (For example, he wrote: “The abstract does not exist; there are no simple principles. All is compound.”)

It’s a fascinating book, stuffed with bits of information that made me think. But here’s something that really brought me up short, and that may seem trivial, but I think is not trivial at all.

ChatGPT has become Nickelback

In our new book on teaching and mentoring scientific writing, we’ve included a chapter considering the use of so-called “AI” tools in scientific writing. (“AI” writing tools such as ChatGPT are better called LLMs, or “large language models”, and I’ll make that switch right now.) It’s an unavoidable topic these days; students are using LLMs to write, instructors are freaking out about it, and on social media, breathlessly incendiary takes are everywhere.

As I’ve mentioned a few times, I’m fascinated by the ways we, as scientists, think about what we do as scientists.

Not-so-wonderful Latin names: The “stupid treehopper”, Carynota stupida

Thanks to a tweet from PJ Liesch, I now know about the “stupid treehopper”, Carynota stupida. Some Latin names are wonderful. Some roll off the tongue. Some have fascinating or poignant stories behind them. Some teach us interesting and important lessons about the history and culture of science. And some are hilarious.

Then there’s Carynota stupida.

What, I wondered, did this particular treehopper do to deserve being labelled stupida?

Is scientific writing getting better or worse?

It’s tempting, and easy, to reminisce about the good old days. The world was better when we were young (except in all the ways it wasn’t), and it was even better in the Golden Age before we joined it (except, again, in all the ways it wasn’t). You can spend as many hours as you like debating this; some things were clearly better in the past (1850s atmospheric CO₂!), some were clearly worse (1850s infant mortality!), and some things are more matters of opinion (1850s mens’ fashions!). But what about scientific writing? Has it gotten better since some vaguely defined past, or worse?*

Before we can say anything useful about that, we need to think about what “better” would mean. I’ll claim that recognizing “better” for scientific writing is quite easy in principle.