h1

Filtering

December 2, 2009

Columbia University Professor Peter Kelemen has what I consider an even-handed and reasonable article in Popular Mechanics on the East Anglia e-mails. Among points that he argues: science should be transparent, journals should be independent, climate science is not a house of cards, scientific skepticism is par for the course, and that there are a few things we know for certain about the climate.

I’d like to pick on him, however, for his sloppy reference to post-modern philosophy. He says the following:

I have noticed a little bit of post-modern philosophy creeping into some scientific discourse that has the potential to subvert this process. The reasoning goes like this: “It is impossible to perceive anything without filtering raw sense impressions, ergo … I filter my data!”

Presumably, this relates to claims like mine from yesterday, asserting that, effectively, all data is filtered.

It’s easy to pick on the post-modernists. I can do it too, since I’m not a post-modernist. But the claim that our data is always filtered is not uniquely post-modern. Indeed, it also comes from the hardest-line positivists in the philosophy of science.

The point is really a conceptual one — we can’t avoid the filtering, that’s why we have acceptable methods of filtering unacceptable data. I’m not saying that we don’t want to get as close to the instrumentation as possible.

Generally speaking, the more information we have, the better. If I have a raw NEF file or a lossy JPG file, it’s probably better for me to have the raw NEF file. Provided I know how to operate the Photoshop dashboard, I can do more with the raw file than the lossy one, and I can also deal with jpeg interpretations that might otherwise screw up my image. At the same time, it takes quite a bit more space to store the raw files, it is harder to process them, and because I can do more with the raw file, there’s more latitude for me to screw up the image myself.

Yet even with this latitude, I can’t alter what the instrument has recorded — it’ll always be a flat, two-dimensional representation filtered through a given lens, distorted by the sensing ability of that lens, subject to ISO, f-stop, filters, ambient visual noise, framing, and the capacity of my sensor. Indeed, ramping up the ISO or dialing down the f-stop effectively “enhances” the raw image.

The parallels aren’t exact, but there is an important lesson: our instrumentation is owned and operated by people, all of whom make judgments about what to keep and what not to keep. When I make a judgment about what data I’m going to record by shooting with a wider aperture, I’m ensuring that I don’t have a sharp focus on my background. That judgment affects the raw data that I have; and such judgments are always, always, lurking in the background of all of our data. If I clean up the data by transcribing it from a notepad into my computer, I’m filtering and enhancing it. My action doesn’t necessarily erode the integrity of the data. It just introduces the human element more palpably.

Some have pointed out that the data exists in other places as well, like GISS, so perhaps this is a reason to discard the raw data. Why store it in East Anglia if it or similar data is also stored somewhere else? Don’t know the truth of that claim. I’m more interested in the general observation that this data is what we have, there is no reason to disregard it because it has been “enhanced,” and that simply because it has been discarded doesn’t necessarily cast a dark shadow on CRU.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: