In my own little corner, I've been learning a lot about R while coordinating a series of statistics workshops in R for our department. I hope to post more about this soon, but I've been thinking a lot recently about my workflow in R, and how to organize all my files (aka Project Management).
While reading other blogs about R and workflow, I found a link to Drew Conway's list of ten reasons why grad students should blog. Hey! We're grad students (& Post-Docs)! We Blog! Ok, we share a blog, but this has advantages over trying to forge an individual identity on the web:
- We can provide more content as a group than individually
- We don't have to try to compete for readers' attention: because there really aren't enough blogs already out there (*sarcasm*).
- We can collaborate on content, or provide direct feedback, even having discussions all in one place!
Why do members of our lab blog? Many of the reasons mentioned in Drew Conway's list certainly apply, but I would add that it also serves as a record of progress along our scientific journey in various areas: snapshots of how we were thinking at the time. Like an academic paper, but less rigorous and not peer-reviewed (and therefore faster but less reliable). It reminds me of a McGill yearbook I recently found from 1978 in Thomson House, including a summary of News from that year, and major events on campus: I was struck my how little some parts of lower campus have changed even since then (McGill is OLD), and how much the fashions and administrative issues have changed. There was an article from one professor predicting that the paper publishing business was going to get too expensive, compared to microfiche cards that cost pennies and could easily be used by students with a microfiche reader for a mere $100! Ok, maybe he hadn't heard of the internet yet, but that just goes to show that even expert predictions often fail. I prefer to stick to explanation rather than prognostication, but Your Mileage May Vary.
Scientific knowledge is characterized by historical facticity, and today's truth may become tomorrow's urban legend: do we act on what we know, or keep waiting to resolve uncertainty? Blogging provides a record of the thought, reasoning and discussion process that eventually turns into peer-reviewed publications, media reports, etc. and becomes part of the public record. So, blogging provides more fodder for historians and philosophers of science and provides people a peek into the developmental process of "Scientific knowledge".
This also suggests another reason to blog: to spark philosophical discussions that are difficult to publish in a forum read by scientific peers. It sometimes feels like scientists take the philosophical aspects of their work for granted all too often, despite a few of us jumping up and down about how your philosophical approach can affect the way you interpret results, or even the questions that you ask in the first place. Science - at least academic science - is just as social and cultural as it is academic. I may not have anything novel to say about it, but it is something that occasionally bears repeating to provoke thought in others.
So, blog away my fellow Biodiversity Scientists, and join the party. I'm waiting for you to blow my mind.
Posts
I think it's really interesting that you privilege the explanation aspect of science over prediction. I have said before that I think science is a way for people to form 'narratives of understanding' about the world, with the aim to managing their interactions in it. As you acknowledge that the facticity of explanation is temporally dependent, why not simply treat explanatory theories as a heuristic device for generating predictions?
ReplyDeleteI agree that scientific explanations are essentially a heuristic for generating predictions. Predictions are also used to test a theory: a theory with no *testable* predictions is arguably useless, and "unscientific" in the Popperian sense.
ReplyDeleteMy concern is that some see prediction as the goal of science, and that's where I disagree. I like your narrative approach, and I think that's largely true. My aversion to prediction as a goal also stems from the fact that it is sometimes just plain impractical. Science may be able to explain the trajectory of a projectile on the surface of the earth, and the idea is that if you have enough data on mass, wind direction and velocity, gravity, force exerted, etc. then you can predict the path a projectile will take. This approach was used to great effect in the second world war, and is still used for sending things to space. But when you just want to throw a rock at something, being able to do calculus is much less important than practice and experience.
I think part of my struggle with the goal of science is dealing with uncertainty. Uncertainty is much less of a concern when explaining something that trying to make a prediction: Will the planet warm 1 or 3 degrees in the next 50 years? We worry a lot less about the same degree of error when reconstructing historical temperatures, at least on a personal level. Ya know why? It's already happened! What we actually care about is how we understand and relate to it. But the same difference in the future can affect how we prepare for it.
Certainly, learning from the past makes it easier to modify our world the way we want. But, we have to accept that no prediction is bullet-proof, so I worry less about that and more about explanations, which are more practical, and arguably more to the point. Explaining my own data is hard enough!