Yesterday I had a meeting with a fellow researcher, Clare Martin. Some of the research that we work on concerns multirelations (a model for expressing two kinds of non-determinism), and we were pondering how best to write up our latest results on multirelations, so that whoever reads it can see why it’s useful.
Why do we want to motivate the research? A more immediate answer is that if we don’t show the reader why the research is useful, then a reviewer of the paper won’t be inclined to recommend it for publication, but it’s also the case that if we want to write good well-written papers (yes! we do!) then part of writing a good paper is making the reader care about the research presented there.
Sometimes research can be easy to provide motivation for: for example if I have a new algorithm to solve a specific problem then when I write up the research in a paper, it’s pretty easy to show the reader of the paper what the purpose of the work is and thus get the reader to care that that research exists. If the problem is interesting and grabs the reader’s attention, then so much the better. Even if the problem isn’t interesting enough that it could be used for nerd sniping, it still provides a rationale for the paper’s existence.
However, multirelations live in the dry theoretical regions of computer science, and for the more theoretical results, it’s not easy to show the relevance of the mathematics of multirelations to solving actual computational problems because the connection is fairly indirect. So it’s difficult to show why the results are useful and motivate the reader of a paper about multirelations to care about the results.
Coincidentally I’ve just been reading a very useful book “Made to Stick” by Chip and Dan Heath which is about making information more memorable and there’s a lot of lessons in there that can be applied to writing papers.
Some of the information in the book addresses presenting ideas and making people care about them. One example they used concerned a duo piano team who were asked to write a mission statement, and they came up with something like “… Duo Piano exists to protect and preserve the music of duo piano.”. Whilst this was true, it didn’t make the other teams see why they should care about duo piano. But when the other teams started asking the duo piano team questions about why duo piano was important (the questions they asked eventually teased out the information that it combined the breadth of orchestral music with the intimacy of chamber music) then the others started to care.
So the way we’re going to approach writing the paper is something similar. I’ve just had a teaching-heavy semester and haven’t been able to work much on multirelations, so my mind is a bit fresher on the topic. I’m going to play devil’s advocate and ask ourselves lots of questions like
- “Why do we need this bit about lifting functors in the paper?”
- “How does this result about set-valued functions tie in to the later section?”
- “What is this extension an extension of?”
- “Why do boolean algebras matter in this setting?”
Hopefully, providing the answers to these questions will lead us to be able to articulate what it is that these theoretical results about multirelations contribute, and if we can articulate that ourselves, then we have a fighting chance of being able to convey that to our readers.
0 Responses to “Presenting Research”