Twenty-Five Years and Counting

Greg Wilson, who co-founded Software Carpentry with Brent Gorda in July 1998, reflects on how The Carpentries has contributed to the world of research as it is today, 25 years later.

This post was originally published with The Third Bit.

Next Monday (July 24, 2023) is the 25th anniversary of the first Software Carpentry workshop at Los Alamos National Laboratory. The Carpentries is hosting a roundtable to celebrate, and to prepare for that I’d like to talk about how the world looks today:

1. The Carpentries have taught data and computing skills to tens of thousands of researchers. It’s easily the most impactful thing I’ve ever been part of.

2. It’s still just a drop in the bucket. For example, about 35 wet lab biologists work with me here in Toronto. Most have graduate degrees, but only 3–4 had anything more than rudimentary programming skills before I arrived and only one had ever heard of the Carpentries.

Everywhere else I look I see the same drops and buckets. Open access is more common than it was, but most research is still locked behind paywalls. More researchers share open source software than ever before, but building that software still won’t get you tenure. And while most companies, universities, and labs have diversity initiatives, life is more precarious for junior researchers with each passing year, particularly those from marginalized backgrounds.

Those problems might seem too big for a small non-profit to tackle, but the Carpentries is already two-thirds of the way there:

1. “Would you like to do more research in less time and with less pain? Let us teach you some basic programming skills.”

2. “Would your life be easier if your colleagues knew how to use version control? Let us teach you some basic teaching skills.”

3. This is the new bit: “Are you tired of having to beg for space and funding for your workshops? Let us teach you the basics of community organizing.”

Here’s the trick: if you teach people how to get onto the committee that oversees the training budget, they can use those skills to get a practical data science course in the undergrad curriculum. Doing that will give them the experience and the connections they need to tackle promotion and tenure rules or to run for the board of the ABC, the AGU, the MPG, and organizations so venerable that they predate acronyms.

This is often how change happens: an organization created to address one problem becomes a training ground for solving others, while lessons learned in the small prevent mistakes in the large. I also think that aiming at a larger target would help address the fatigue that mature volunteer organizations inevitably experience. “Come for the coding, stay for the teaching, leave knowing how to change the world” might get people excited again.

Five years ago I spoke about what it was like to step away from Software Carpentry after putting so much into it for so many years. I said then that if we want to get through the next fifty years we’re going to need more science and more courage. I now believe that we also need to learn how to organize, just as we have learned how to code and how to teach. It’s a lot to ask of people who are already working a lot harder than they should have to, but the alternative is to spend the rest of our lives rolling the same rocks up the same hills. So, one more time:

Start where you are. Use what you have. Help who you can.

Further reading:

• Brown: Building Powerful Community Organizations
• Beckerman: The Quiet Before
• Manns and Rising: Fearless Change