The Carpentries is the umbrella nonprofit organization behind three volunteer-taught lesson programs — Software Carpentry, Data Carpentry, and Library Carpentry — that teach researchers, students, and library and information professionals the foundational computing and data-handling skills that most graduate training leaves out: the Unix shell, version control with Git, programming in Python or R, SQL, and basic data organization. For a research-data-management (RDM) program, the Carpentries model matters less as a piece of trivia and more as a concrete answer to a recurring problem: a data management plan can specify tidy, well-documented, reusable data, but researchers still need somewhere to learn the practical skills that make that possible. This guide explains what The Carpentries is, how its three lesson programs differ, how workshops and instructor certification actually work, and how the whole model fits into a broader RDM and computational-reproducibility training pathway.
What The Carpentries is
The Carpentries is a registered 501(c)(3) nonprofit, based in Delaware, USA, whose stated purpose is to teach foundational coding and data science skills to researchers worldwide — skills the organization notes “are often not included as a part of basic research training in many disciplines.” It is not a single curriculum but a parent organization for three constituent lesson programs:
- Software Carpentry — the original program, covering the Unix shell, Git version control, and programming in Python or R.
- Data Carpentry — aimed at researchers who work with data and need coding and project-organization skills for reproducible data work, with lessons built around discipline-specific datasets.
- Library Carpentry — aimed at people in library and information-related roles, teaching software and data skills using library-relevant datasets and tools such as OpenRefine.
Day-to-day, the organization is run by a small Core Team, but its actual teaching capacity comes almost entirely from volunteers — Instructors, Instructor Trainers, Maintainers, Lesson Developers, and partner institutions — operating under a shared Code of Conduct.
A brief history: from Software Carpentry to a unified nonprofit
Software Carpentry is the oldest of the three programs: it began in 1998 when Greg Wilson and Brent Gorda started running workshops to teach researchers basic software development practices, and its lesson materials were made open source in 2001. Data Carpentry and Library Carpentry were both established in 2014, extending the same teaching model to research-data literacy and to library and information professionals, respectively.
In 2018, Software Carpentry and Data Carpentry formally unified to form The Carpentries, operating at that point under the fiscal sponsorship of Community Initiatives, with Library Carpentry joining shortly after as the third lesson program under the same umbrella. The Carpentries remained under fiscal sponsorship until February 1, 2025, when it completed its transition into an independent, standalone 501(c)(3) nonprofit organization. Software Carpentry, Data Carpentry, and Library Carpentry today all operate as lesson programs of that single organization, sharing an infrastructure, a certification pipeline, and a Code of Conduct, while retaining distinct curricula aimed at somewhat different audiences.
The three lesson programs: what each teaches
Because the three programs share instructors, infrastructure, and teaching philosophy, workshops are often described collectively as “a Carpentries workshop,” but the actual lesson content differs by audience:
Software Carpentry
Targets researchers, students, and staff who want general-purpose programming and automation skills: the Unix shell for navigating and scripting file systems, Git and GitHub for version control, and introductory programming in Python or R.
Data Carpentry
Targets researchers working directly with data who need coding and project-organization skills specifically for reproducible data analysis — lessons are typically built around discipline-specific example datasets (ecology, genomics, social science, and others) rather than generic exercises.
Library Carpentry
Targets library and information professionals, covering the same underlying tools (shell, Git, data cleaning with OpenRefine) but applied to library-relevant tasks and datasets. Because it is often harder for library staff to get two consecutive days away from a service desk, Library Carpentry workshops are more commonly delivered in flexible half-day blocks rather than as a single two-day block.
All three programs share a common thread relevant to RDM: none of them teach a specific repository, data standard, or funder policy. They teach the underlying technical fluency — scripting, version control, reproducible code — that makes it practical for a researcher to actually execute the plan a repository, standard, or funder policy requires.
How a Carpentries workshop works
Carpentries workshops follow a consistent format regardless of which lesson program is being taught. Software Carpentry and Data Carpentry workshops are traditionally delivered as two-day, in-person events; Library Carpentry uses the same core lessons but is frequently split into shorter, non-contiguous sessions for the scheduling reasons noted above. Across all three programs, the teaching method is “participatory live coding”: instructors type and run code live, and learners follow along on their own machines rather than watching slides, which is deliberately aimed at building working confidence rather than passive familiarity. Workshops explicitly target novice learners — no previous programming experience is required or assumed.
Workshops are hosted by member institutions (universities, libraries, research institutes, and other organizations) and taught by certified volunteer Instructors, who are trained specifically in pedagogy and evidence-based teaching practices rather than only in the technical material itself. An institution wanting to run a workshop typically works with The Carpentries’ workshop administration process (contact: [email protected]) to schedule a workshop, identify a venue, and recruit or request certified Instructors.
Becoming a certified instructor: the training and checkout pipeline
Anyone who wants to teach a Carpentries workshop goes through a defined pipeline rather than simply volunteering to teach:
- Instructor Training — a course covering pedagogy and evidence-based teaching practice (not just the technical lesson content), typically delivered online by a certified Instructor Trainer.
- Checkout — after training, candidates complete three checkout steps before certification: a teaching demonstration (assessed and logged by a Trainer), a “Get Involved” step in which the candidate takes some further step of engagement with the Carpentries community, and attendance at community Welcome Sessions aimed at onboarding new volunteers.
- Certification — once training and all three checkout steps are complete, the volunteer becomes a Certified Carpentries Instructor, eligible to teach Software Carpentry, Data Carpentry, and Library Carpentry workshops anywhere in the global community, not just at their home institution.
This pipeline is what lets a single, relatively small Core Team support a global teaching program: the actual instruction is delivered by a large, trained, and certified volunteer base rather than paid staff.
Carpentries membership for institutions
Universities, libraries, and research institutes can join The Carpentries as members. Membership is structured around supporting an institution’s own capacity to run workshops rather than simply purchasing training as a service: member benefits include priority access to Instructor Training seats (so an institution can build up its own pool of certified Instructors), access to the wider global pool of certified Instructors to help staff workshops, and quarterly reporting on benefit usage and trainee certification progress, which research-computing and library units can use to track their own training pipeline internally.
Where Carpentries training fits in a research data management pathway
Carpentries workshops are not RDM policy training — they don’t cover writing a data management plan, choosing a metadata schema, or interpreting a funder’s data-sharing mandate. What they provide is the technical skill layer that sits underneath those policy requirements and makes them achievable in practice:
- Version control (Git) supports the kind of documented, auditable analysis history that both reproducibility and provenance depend on.
- Scripted, reproducible workflows (in place of manual, click-through data processing) are a practical prerequisite for computational reproducibility — the ability for someone else to obtain the same results from the same data and code, as distinct from replicability, which concerns obtaining consistent results in a new, independent study of the same question.
- Data organization and cleaning skills (including tools like OpenRefine, taught in Library Carpentry) directly support producing data that meets FAIR criteria — consistent structure and documentation are what make data actually findable, accessible, interoperable, and reusable in practice, rather than just in principle.
- Basic programming fluency (Python, R, SQL) is what lets a researcher actually execute a data management plan’s technical commitments — converting formats, generating standardized metadata, or depositing to a repository programmatically — rather than treating them as someone else’s job.
For research administrators and RDM specialists building out institutional training, the practical takeaway is that policy-facing guidance (what a data management plan requires) and skills-facing training (how to actually produce reproducible, well-organized data) are complementary and usually need to be offered together: a researcher who understands the FAIR principles but has never used version control, and a researcher who can script fluently but has never seen a funder’s data-sharing requirements, both fall short of what a modern RDM program is trying to achieve. The Carpentries is one of the more established, largely free, community-run options for closing the skills half of that gap, alongside institution-run bootcamps and research-computing training offices.
Frequently asked questions
Is The Carpentries free to attend?
Workshops are organized by host institutions, and cost varies by host — many university- or library-hosted workshops are free or low-cost for their own community members, since the instructors are volunteers and the lesson materials are open and freely available online. Institutions considering hosting should confirm current costs directly with The Carpentries, as fee structures can change.
Do I need a coding background to attend a workshop?
No. All three lesson programs are explicitly designed for novice learners with little to no prior programming experience; that is a stated design constraint of the curriculum, not just a suggestion.
What is the difference between Software Carpentry and Data Carpentry?
Software Carpentry teaches general-purpose programming, shell, and version-control skills aimed at a broad research audience. Data Carpentry teaches largely the same underlying tools but organizes lessons around reproducible data analysis using discipline-specific example datasets. In practice the two curricula overlap significantly and are often taught together or back-to-back in a single event.
Can library staff who are not researchers attend Carpentries training?
Yes — Library Carpentry exists specifically for library and information professionals, and its lessons are built around library-relevant datasets and tools rather than lab or field research data.
How does an institution become a Carpentries member?
Institutions apply for membership through The Carpentries directly; membership grants benefits such as priority access to Instructor Training seats and access to the certified Instructor pool, in exchange for the institution’s ongoing support of the volunteer community model.
Does completing Carpentries training satisfy a funder’s data management training requirement?
Not directly. Carpentries training builds general technical skill (shell, Git, programming, data cleaning) rather than teaching a specific funder’s policy. Institutions should treat it as a complement to, not a substitute for, formal data management plan and RDM policy training required by specific funders.
Sources: The Carpentries, About Us; The Carpentries, Workshops; The Carpentries, Instructor Training; The Carpentries, Membership.
