planet.donarmstrong.com

I've often been disparaging of the community efforts of big academic collaborations, because it seems like they rarely communicate with the outside world well - this is particularly true of interim (not-yet-publishable) results and software. Over the years I've evolved a theory that big consortia are so busy communicating within that they have no energy for communicating without. This robs the larger scientific community of insight and scientific results in a way that I feel like smaller collaborations do not - you could probably come up with "communication per $$", or something, as a metric, and I bet large consortia would show poorer numbers.

I particularly admire open source communities here, because the communication is often so good (compared, at least, with consortia, or really academics of any kind) and rather fine grained. Since many open source communities are both distributed and asynchronous, they really seem to excel at information sharing in useful ways.

(See Max Ogden's excellent doc about how to run an async team if you're interested in some of the lowdown here.)

I am hoping to use my coordination position within the #CommonsPilot to facilitate better communication, and we've even hired some people to do that. So imagine my frustration to be in exactly that "silent" situation with the #CommonsPilot! I can now partly confirm my initial theory, and elaborate upon it, with the benefit of about 6 months of experience.

Without further ado, here are:

The top N reasons why I think big consortia are unusually silent.

1. We're too busy talking to each other by e-mail!

By the time I finish reading and responding to e-mails (and, ahem, sending new ones) from the #CommonsPilot each day, I'm out of time and energy.

1. We're too busy talking to each other on teleconferences!

Most information is passed via in-person teleconferences, from which very little information actually escapes. This is exacerbated by people's interest in ONLY communicating this way, because it leads to more focused and thoughtful engagement by busy academics. It's high bandwidth, sure, but it's also isolating - only the people who have the time and energy to show up for all the teleconferences are in the know.

One takeaway that I got from this excellent blog post, aturon.log: listening and trust, part 1, about the Rust community, is that "all major project decisions must go through the RFC process" - which must involve written communication that clearly recapitulates anything discussed on a phone call. We were already instituting this in the Data Commons before this blog post, but now I have extra reasons to do so :)

1. Consortium wide decisions require multiple rounds of discussion before consensus is reached and can be communicated externally.

I really don't want to post things that people disagree with, but it takes a lot of time to figure out what that is (and isn't).

1. Rules for communicating with the outside world aren't clear.

Funding bodies and senior PIs are often risk averse, and figuring out what is and isn't a risk is tough. We've finally gotten some blogging and Twitter guidelines approved and we'll post them when we can.

1. Hierarchies interfere.

Typically the people most familiar with social media are junior in collaborations, and (for better or for worse) are worried about irritating those senior to them by speaking out of turn.

Here I have an edge, since I'm both a PI and a coordinator on this project, and my proposal focused on outreach (and this proposal was accepted by the NIH). So I have a mandate.

1. Communicating externally takes time, energy, and willpower.

Usually, there's no one whose job it is to communicate with the community. To which I say...

...welcome, Dr. Rayna Harris :).

Wait, why should we be communicating anyway?

I started with the implicit assumption that consortia should be communicating with the outside world. Why??

I think there are many reasons. It's not just about communicating science more effectively, although that's part of it; it's also about:

• communicating about what big, expensive consortia are doing that's worthwhile; think "accountability to taxpayers and stakeholders".

• gaining buy-in for consortium decisions from the wider community. This is particularly important for efforts like the #CommonsPilot, where we are hoping to identify and implement good standards and build a community of practice.

• getting feedback (negative and positive) on consortium decisions. If we're picking tech that is out of date or old or bad, we should know - and we don't always!

Perhaps the best reason, though, is that external communication can help people internal to the Consortium understand what's going on. I've often found that there are relatively few people truly "in the loop" in any given situation, and a commitment to external communication of internal decisions can actually help communicate those same internal decisions internally.

Or, to put it another way, if you're not communicating in one venue, you're probably not communicating well in any venue, and this is probably harming your consortium and limiting the contributions of people - especially junior people.

So, what's the status, anyway?

Conveniently, Rayna Harris just wrote a nice blog post about some of our coordination efforts and the current status of the Data Commons!

More soon, I hope :)

--titus

p.s. Thanks to VM Brasseur for her comments and suggestions on this post!