Tom Morris

On Sunday, I wrote about the new Research Excellence Framework and it's pretty phillistine approach to academic research and 'impact'. It's not just philosophers who are pissed off about it - scientists are too. And they've started a petition on the Number 10 website with some fightin' talk: We request the reversal of a policy now being applied by the UK Research Councils. This policy directs funds to projects whose outcomes are specified in advance. Science has never worked in this way, and never could. The real world is blind to our hopes, fears, and aspirations. Scientific research seeks to describe this world, replacing ignorance and error with knowledge and understanding. Where a specific outcome can be predicted with confidence, then there is no research. and We call upon the Research Councils to return to their mission of advancing the frontiers of human understanding. Public support for science must renew its investment in discovery if it is to create prosperity and well-being.

The same applies, in a slightly modified form, to history and philosophy and a few other humanities and arts subjects too: imagine I've set myself the task of reading a particular philosopher, like Quine. It's going to take me three months, say, to read about Quine - to read his work, to read the secondary literature, read his critics, read responses to those critics etc. Three months may be a bit optimistic - Quine did write a lot. Let's say I apply to the Arts and Humanities Research Council and say "I really want to learn about Quine. I think there is something interesting he might have to say about, oh, the problem of universals."

And they say "well, what impact is that going to have on society?"

To which I might reply, "well, I don't know. Since the likelihood of me walking into a random pub anywhere in Britain, asking who has heard of Willard Van Orman Quine and getting back a response in the affirmative is closely approximating zero, and Quine had pretty much nothing to say on the topic of religious fundamentalism in the early twenty-first century, improving health outcomes or getting more people to wander around the National Gallery, probably pretty much fuck all, but, you never know, I might discover some interesting new facet to Quine's work which I might be able to publish in a journal which other people interested in Quine could read and learn from. And wouldn't that be nice?"

Their answer would be "well then, go fuck off. If you don't know that this is going to make the world a better place before you have even started, we have no time for you. I hear Burger King need people."

So, yes, I'll be signing the anti-REF petition. It'd be very good if scientists were to really vigourously oppose REF. The problem as I see it is that those in the social sciences will be supporting it as they'll get a bigger slice of cash. Economics and sociology will definitely be getting more. Practical sciences like engineering (including computer science in as much as it works on producing practical stuff) and medical research are going to benefit from REF if they are producing stuff that is going to benefit industry. And the people who are going to lose out are humanists (in more abtruse bits of literature, in philosophy, some parts of history, classical studies, religious studies, art history) and basic research in science. What's bloody irritating about it is that basic research is still vitally important and still teaching us really great and interesting things - things like the Large Hadron Collider is a perfect example of a basic research project that would fail pretty completely if you were to measure the impact (in the goofy research council sense of the word - since it is literally a machine to smash atoms together and measure the effects of the impact!).

That said, in the spirit of charity, I propose a compromise: the government could hold off on implementing the impact section of REF for ten years. In that ten years, a marked improvement in the public understanding of science and scholarship would have to take place, and a similar improvement in the intellectual content of newspapers, TV, radio and schooling. If at the end of the ten years, we've got a nation which can broadly appreciate the work of humanists and basic research scientists and can tell the difference between useful research, PR fluff and bullshit that's designed solely to get 'impact' money, then we can implement this system. Until then, we should have academics deciding whether or not academic research is up to snuff.

It's been kind of depressing to watch people react to Google Wave. I wrote my review back in August and I still agree with it. Google Wave is utterly pointless. It's a bit of marketing flash, combined with a brainfuck of a product that mashes together Campfire, Gmail, Etherpad/Google Docs, Picasa/Flickr, YouTube and Google Talk in an utterly unappealing headfuck of a product. But still people are heavily into the hype cycle. It took me five minutes to see what was wrong with it: huge amounts of needless clutter in the UI. The whole thing is what you'd get if you retrofitted a VCR to also serve the functions of a toaster.

Wave is going to be attractive to the sort of people who do e-mail wrong, as I described in the previous post.

But I do have to take issue with one critic of Wave - Steve Rubel. He describes Wave as Google Wave 1.0 = RSS, the Sequel. In Other Words, DoA... for Now.

It's a commonly understood law of logic that a conclusion can be right even if the premises do not support that conclusion. Let me illustrate:

1. Strictly Come Dancing is the best thing to ever happen to television.

2. Therefore, the bottle of wine on my kitchen table is red.

The truth of (2) is not determined by (1). It's utterly irrelevant. You can imagine a world in which (1) is true but (2) is false.

And so it is with Mr. Rubel. Firstly, he says that RSS is "dead". Leaving aside the fact that technical standards aren't really a property which exhibit life in the first place, and so finding out that they aren't alive should come as no real surprise, and understanding it in the metaphorical sense of "no longer relevant", let us consider this claim: is RSS dead? Of course not. It works perfectly well for what it does, and lots of people use it. Not many people ride motorcycles, but we don't all of a sudden say that motorcycles are dead. RSS doesn't need mass market adoption. Just like e-mail doesn't. It doesn't make my experience of reading RSS feeds better or worse to know that my neighbour is doing it. This whole thing is so chronically dumb and pointless that it beggars belief that people even spend a moment of their short lives considering it.

So, the argument is something like this:

1. RSS is 'dead' - or TechMeme-dead (it's like dead, but you know, you don't have to call a coroner).

2. Google Wave has some properties that make it like RSS.

3. Therefore, Google Wave is dead, or TechMeme-dead or whatever, too.

Okay, but that doesn't really get to the rather key differences: Google Wave is like e-mail - you need other people to use it effectively. RSS is not. My use of RSS is unaffected if my friends or family don't use it. My use of a messaging and collaboration platform is not. But that doesn't necessarily mean that the service won't be useful - I mean, I don't think it will, but that's my personal judgment. A messaging and collaboration platform may be useful but not get mass-market adoption. Look at something like bug or issue trackers. Bug trackers are very useful - lots of people I know spend a lot of time using bug trackers. And they are pretty useful - they help programming projects keep track of issues. Might it be that Trac or Bugzilla are also TechMeme-dead? It seems so. The way in which Google Wave or RSS is being alleged to be dead seems to be pretty closely analagous to how one might declare Trac or Bugzilla to be dead.

This tells us a lot about the TechMeme-dead predicate: namely, that it's completely fucking useless! The only way for something to not be dead according to this crtieria is for it to get mass-market adoption. Anything less than total market dominance equals 'dead'. To use a human analogy, the only way to be truly alive is to be a movie star. A TechMeme-dead product may be financially viable, be making money, have an active user community, solve real business needs and so on. But it's still TechMeme-dead. Look at something like Basecamp from 37signals. It's financially viable, makes money, has an active user community, solves real business needs, is continually updated, is supported and used. But as it doesn't have universal appeal and reach, it is TechMeme-dead.

If you suddenly start declaring that a lot of things have failed or are "dead", your expectations and your criteria of success have been needlessly inflated.

As I said, the premise can be wrong, and the conclusion can still be right: Wave is utterly pointless. It's been designed to fix a social problem - namely, that of people sucking at e-mail. My preferred solution to that problem is to point people to RFC 1855 on a daily basis, grumble about it on Twitter, write little scripts that automatically publish to my blog any e-mails that contain 'confidentiality notices', and to ruthlessly filter out idiots who haven't figured out how to send e-mail. That includes top-posters, who really need to get shipped off to re-education camps. And, no, working for large companies doesn't excuse your failing at Internet. I said it before, and I"ll say it again: if you do e-mail properly, Wave is pointless.