One of the challenges we struggle with at Science Commons is how to communicate the work to two vastly different audiences: the scientific audience (who might benefit most from our work and actually use it!) and the cultural audience (who tend to dominate the user space of Creative Commons licenses). In this post I’m going to try to examine why it’s hard and how we might bridge the gap.

My short version is: science is harder than culture. Science takes a level of exactness and accuracy that other fields don’t (and in many ways harms cultural creativity). Science is institutional in a way that other fields aren’t. And last, the technical and social infrastructures for cultural creativity simply aren’t very well tuned to support science – we can’t just point biologists at web 2.0 or “get data online” and expect things to change.

more after the jump.

We are surrounded by pervasive infrastructure for cultural creativity: it’s cheap and easy to create, consume, transform, post, and share cultural content. The infrastructure for that is multilayered, starting with the obvious networks and web technologies, but extending to cheap digital cameras, businesses that provide photo and music sharing storage on the network, fast networks to homes and public locations, embedded software for manipulating, mashing, mixing, and republishing content, new systems for individuals to characterize and tag content like tagging, and the existence of an entire legal ecosystem for managing and expressing the legal rights associated with content. It takes all of this just to get to a world where there are enough photos online for flickr’s tagging system to analyze related tags – or a world where an open licensing system like Creative Commons makes sense for normal people to find and use.

These systems were designed for different purposes. And some of them – the networking protocols, the Web, and fast networks especially – work for science too. Those are the systems that were designed as abstract elements of infrastrucutre. But many of the others were designed far more narrowly for cultural use. Hardware and software are the obvious ones (digital cameras or music technologies). Obviously, this stuff isn’t going to support science all that well.

But let’s zoom in most closely on web 2.0 technologies and how they intersect with science. Web 2.0 is a lot things (besides being a tired business cliche). To me the real core of what most smart people mean when they talk about web 2.0 is that it’s about users putting stuff on the web and having that user-generate stuff snapping together a la wikipedia, digg, reddit, etc. – thus bringing us another tired line about the wisdom of the crowds. This is a good thing, whether cliche or not, as anyone who uses wikipedia knows. And the crowds agree, which is why wikipedia is constantly at the top of my google search results.

But is the infrastructure that builds this world strong enough to sustain a similar explosion in the sciences? Or even in narrower fields like life sciences, or climate science?

I’m not sure. One of the big selling points for web 2.0 is how easy it is – anyone can start a web 2.0 company, and you can even get free advice on mission statements , names , and buzzwords. I’ve got a favorite fake web 2.0 company called Thoughtsphere (we share data-driven folksonomies, don’tcha know) taped up on my wall over my monitor to remind me not to fall into hypespeak. The whole point is, it’s easy. Use AJAX, start facebook from your dorm room, let users generate the content, (insert a revenue miracle here), and change the world. Anyone can do it, and anyone can do it fast and cheap.

Viz, wikipedia, where the goal is to replicate an existing resource and then beat that existing resource. I’m going to Korea soon, so I check the entry on Seoul. The existing infrastructure is there: people know about Seoul, wiki software lets them add text and links, and open licensing ensures it’s all legal. As a result, the work on Seoul snaps together into a single article, and in turn, with other articles about Korea, and so on and so forth. Bam, as the cook says. An open, free, user-built encyclopedia, built on the same kind of software I use for drafting (sorry though, I use MoinMoin on my desktop).

But science isn’t easy. Science is hard. Life sciences and climate change are actually terrifyingly hard. It’s not like we are trying to replicate an existing resource in those sciences, or even to do engineering. We’re still trying to discover the rules of the game in life and climate sciences. For all we know about small pieces of puzzles, we are a long, long way from being able to meaningfully use the data we generate to inform future decisions.

On top of that, one of the other assumptions of the web 2.0 world is that wisdom of crowds bit. But in the sciences, the stuff the crowd believes tends to either be boring (because it’s already been proven) or wrong (because it’s based on flawed or incomplete data). That’s a real problem in the uptake of tools. But it’s a tiny problem compared to the social infrastructure of the sciences. Because scientists get rewards for withholding data until publication (you win for being first to publish in most fields) the average scientist does not want to be followed to the water hole.

I want that science 2.0 future as much as anyone. But I’ve come to believe that the sciences require a new set of infrastructure, and in some cases new institutions, to be built before we get to science 2.0. That infrastructure has to be robust and detailed and accurate, and user-generated foksonomies aren’t going to cut it, no matter how useful they may be in finding photographs or interesting web posts. That means a burden of using exact names, new technological services, ontologies, data integration, and more. All of the 1000+ molecular biology databases need to be as integrated as web pages are before they can really start to sing, and no wisdom of the crowds is going to integrate those databases. It’s going to take boring old technology work by dedicated people to get there. Let’s just hope their work is public domain. Web 2.0 is people in fast cars; Science 2.0 needs roads, bridges, and research into internal combustion engines.

On to data infrastructure. It isn’t enough to post data sets online – it takes pages of annotations to make data useful to other scientists, because although it’s very easy to simply mix two data sets together, it’s very hard to do that in a way that is meaningful. You have to know if the numbers in the data set come from a gene chip using flourescence or radiation – and that is the easy stuff. You have to know why the experiment was done. You have to know if you trust the experimenter to have run the protocol correctly. You have to hope the lab was at a constant temperature. And you have no idea if the data reflected most of the data in the course of the experiment or was an outlier used to prop up a much-needed publication to help a postdoc achieve escape velocity in his or her career. You need lots of annotation. Think of the metadata that comes off your camera on every flickr shot – we need to get to that level of infrastructure, but across a dizzyingly wide space of instruments and assays. More roads, more bridges.

Another missing piece of the infrastructure is institutional. eBay and Amazon are great for culture consumption. But those models don’t port to science, because most scientists work at institutions. They don’t practice science at home in the garage (we can and should have a discussion about how lame that fact is, but it’s the truth). Thus, their science is subject to an entire set of controls that don’t apply to culture. Until the infrastructure supports the reality of institutional involvement, we’re going to be frustrated at the uptake of science 2.0.

Yet another missing piece of the infrastructure is legal. That’s the day job , and I won’t go into it here.

I could go on and on, and I’ll come back to this theme again. We need the roads and the bridges for science, an infrastructure that is what anthropologists might call culturally relevant for science, if we are going to really create the same revolution in science that we’ve seen in culture.