Saturday, October 18, 2014

How to write/present science: BABY-WEREWOLF-SILVER BULLET

As an editor, reviewer, supervisor, committee member, and colleague, I have read countless papers and proposals and have seen similarly countless presentations. Some work well and some don’t. Beyond the picky details of slides that are too wordy, speaking that is too fast, sentences that are poorly constructed, and so on – the most critical problem is making clear why the work is interesting and important. Why should we read further rather than moving to the next paper on the pile? Why should we give you money as opposed to your competitor? Why should we listen to your talk instead of tweeting about the party last night? This simple and yet pervasive inability to engage the reader and have them buy into your work is likely the single greatest flaw in the writing of every student (and many postdocs and faculty members). In this post, I will explain a simple metaphor that can help you to solve this problem in each and every one of your papers/proposals/presentations.

The metaphor emerged from a comment by McGill’s Dean of Science, Martin Grant, about what makes a good proposal. He suggested that you need a werewolf and a silver bullet. With a werewolf, a funding agency and their reviewers can see the problem that needs solving. With a silver bullet, the agency and reviewers can see that you have a realistic chance of solving the problem. When translating this logic to my own students, I have modified it somewhat to better fit the ideal outline of a paper/proposal/talk. The basic idea is that the structure of your paper/proposal/talk should follow this sequence:
  • CUTE BABY: First explain to the reader/listener the general umbrella under which the work falls – some umbrella that will make the reader/listener sit up out of their sleep-deprived torpor and say to themselves “Oh, OK, this talk is about something that is interesting and important. I better pay attention to what new insight they might bring.” Cute babies can be things that are important, such as ecosystem health and human well-being. A common cute baby here is biodiversity and its contribution to ecosystem services. Cute babies can also be things that are interesting, such as theories, with examples from ecology and evolution being the equilibrium theory of island biogeography or the ecological theory of adaptive radiation. In developing this cute baby, it is critically important to not overtly state that the baby is cute. “One of the most important topics in ecology is the maintainance of biodiversity” – this is a “motherhood” statement (http://en.wiktionary.org/wiki/motherhood_statement) that just annoys the reader by making them feel manipulated. Instead the reader should make up their own mind when reading about the baby that it is indeed cute. Stated another way, if you have to say that your study area/work is important, then people will think you are trying too hard. Instead, the reader should think “oh, that is important” without you having to say it.
Cute babies.*
  • SCARY WEREWOLF: Next explain how that cute baby is somehow threatened, so the reader/listener shifts forward on their seat and begins to empathetically furrow their brow in shared concern, thinking “Yes, that’s true, that really is an unsolved problem that could hurt the baby.” Scary werewolves for biodiversity and ecosystem services might be climate change and habitat loss and, well, pretty much anything. Scary werewolves for theories are things like potentially inappropriate assumptions, or the lack of empirical tests, or the failure to include an important idea, or low explanatory power. Here (as opposed to the above) it is more often OK to state that the werewolf is scary but it is still more effective to avoid motherhood statements and let the fear emerge within the reader’s mind. (How often do stories say “the werewolf was scary”? Instead they say that “a hulking beast with dark, tangled mats of hair emerged from the darkness dripping blood from its fangs with its eyes glinting in the moonlight.”)
Scary werewolf.*
  • SILVER BULLET: Finally, explain how the work that you did/will conduct has the potential to slay – or at least severely wound – the werewolf. If you do this effectively, the reader/listener will begin to unfurrow their brow and nod: “Ah, yes, that would be a great way to solve the problem.” Silver bullets can be applied solutions to problems, such as a new design for corridors that reduces the negative effects of habitat loss. Or they can be new experiments that address outstanding gaps in knowledge, or particular study systems that are ideally suited to show how some theory needs to be modified. Once again, you ideally don’t say “I have the silver bullet”; instead, the reader has this emergent thought while reading about the study system. And, of course, NEVER say your system is “ideal,” which means “couldn’t possibly be better,” as everyone who works on a different system will immediately think “no it isn’t.” Instead your system is “excellent.”
Silver bullet.*
  • DEAD WEREWOLF: For work that has already been conducted, one would ideally show that the silver bullet (new method/theory/experiment/observation) has killed the werewolf and thereby saved the cute baby. In reality, however, it is just as likely – and effective – to show how you have wounded or exposed the werewolf or how you have shown that the werewolf is scarier than originally thought or how you have found a new werewolf. These alternative end points nicely establish the need for further work. For work that has not yet been conducted, such as in a proposal, the dead/injured/new werewolf is not actually shown, but the reader has to see what it might look like. That is, they can visualize the werewolf lying dead on the ground while the cries of the baby fade into giggles while the baby bounces up and down on the werewolf’s belly. (Or alternatively, the werewolf is just a hairy uncle bouncing the baby on its knee.)
Dead werewolf.*
In implementing this schema, I suggest working from the goal that your Introduction will follow the above structure. The first paragraph (or section) describes the cute baby (the general area of research, subtly making clear why it is important), the next section describes the scary werewolf (the problem/gap/limitation of current knowledge/work), the next section suggests a silver bullet (the study system/experiment/new theory), and the final section postulates what the werewolf might end up looking like (the predictions/questions/hypotheses). In the context of a presentation, the entire talk should follow this structure, with the methods falling into the silver bullet part and the results/discussion taking the form of the dead/injured werewolf. Note also that studies sometimes examine multiple questions, in which case the baby-werewolf-bullet approach can take on a fractal appearance: the whole study, the individual components, and sometimes even within individual sections.

At this point, I am sure you have some thoughts or criticisms of the above plan. Although I presumably can’t predict all of them, here are some likely ones.
  • But my work just isn’t that important – it won’t solve world hunger, it won’t halt the loss of biodiversity, and it won’t overthrow the ecological theory of adaptive radiation. Surely I shouldn’t try to pretend it will. Indeed you shouldn’t, but don’t throw the cute baby out with the scary werewolf. Instead, you simply need to scale your baby/werewolf/silver bullet accordingly. If you have a relatively small problem, give us a clearly defined but only modestly scary werewolf. If your silver bullet is unlikely to slay the werewolf, then give us some silver pepper spray. The key point is that the above logic and outline applies regardless of the size of the problem or the actual outcome of the study – yet, it is true that you can’t oversell your baby-werewolf-bullet.
  • But I don’t want to oversell my work – reviewers will see through my attempt to make it seem more important than it is. This concern is related to the one immediately above and, again, it is correct that you shouldn’t promise something you can’t deliver or outline a werewolf you can’t kill. However, you can outline nested werewolves – like Russian doll werewolves where you can slay some small ones thus getting closer to the big one. Conveniently, the solution is the same as above – scale the baby/werewolf/bullet to the scope of your study and what you can deliver.
  • But I didn’t actually kill the werewolf. No problem. Explain how your silver bullet was tarnished (polishing will fix it up) or was made of aluminum (I need a new experiment) or how it missed the werewolf altogether (the werewolf was in our imagination or was really just a hairy uncle seen in low light). This outcome is just as satisfying in most instances.
Maybe it wasn't that scary after all.*
Or maybe it is the baby that is scary.*
Well, there it is – a suggested plan for writing every paper/proposal/presentation for the rest of your career. I hope it helps. I have certainly found it immensely helpful for improving the logical flow and engaging narrative of my students’ work, as well as my own. Of course it doesn’t always work and of course it doesn’t guarantee acceptance (rejection can occur for many other reasons), but I think it solves the problem of how to structure the presentation of ideas and how to make clear the importance of your study. And, even within this framework, many other improvements can be made to the grammer/writing/presentation. Here are my own suggestions: http://redpath-staff.mcgill.ca/hendry/writingtips.pdf and I list some additional ones below.
The Science of Scientific Writing

On whimsy, jokes, and beauty: can scientific writing be enjoyed?

Five ways to improve your science writing

-----------------------------------

* I did not take these photos but found them on google - the original source (and copyright holder) is not clear.

6 comments:

  1. Andrew,

    You also have to be careful to not throw the baby out with the bath water. I KNOW that common usage allows for sloppy subect/verb and pronoun agreement. However, some of us old timers get sidetracked by that sloppy usage. All we see when we read a carefully prepared article/posting is lack of agreement. The baby section of your posting is FULL of subject/verb and pronoun disagreement. Check out https://owl.english.purdue.edu

    Your fussy/particular Mom

    ReplyDelete
    Replies
    1. I will try harder next time Mom - but then I thought it was your job to correct my English!

      Delete
  2. The ware-wolf-baby-bullet analogy has already been immensely helpful for me. But imagine if I actually wrote a grant on ware-wolf predation and biological control with silver bullets. That would be some heavy metal. Thanks!

    ReplyDelete
  3. This is great advice! However, I would also suggest that there are additional ways to structure an introduction to a subject in manuscripts and grants that goes from the big idea to the specific question/problem being addressed. See: https://www.northwestern.edu/climb/resources/written-communication/index.html
    Combining the two approaches may be optimal. Dough

    ReplyDelete

A 25-year quest for the Holy Grail of evolutionary biology

When I started my postdoc in 1998, I think it is safe to say that the Holy Grail (or maybe Rosetta Stone) for many evolutionary biologists w...