The Discipline for Reading an Incomplete Record
Share
The Discipline for Reading an Incomplete Record
Before we get started, this newsletter grows almost exclusively through trusted networks. If you find the conversations in the CX Pod or the newsletter useful, please consider sharing it on.
▎ "We keep treating this as a binary — the old human skills, and the new technical skills — as if they're two different things. We haven't figured out how to put that together." —Kathleen deLaski, Founder, Education Design Lab; author, Who Needs College Anymore
Most institutions teaching toward employment still sort skills into two bins. One holds the durable things: writing, critical thinking, collaboration. The other holds the moving target: whatever the current technical wave requires. The assumption underneath the sorting is that these are separate curricula, taught separately, sequenced one after the other. Kathleen deLaski, who runs Education Design Lab's work inside colleges building alternatives to the four-year degree, has watched enough classrooms to know the binary doesn't survive contact with actual work. A student who prototypes an app with an AI coding tool in an afternoon isn't applying a technical skill and then a human one. They're doing one thing, and the curriculum has no box for it.
The obvious response is to replace the binary with a model built from evidence of what integrated work actually requires. This is where the second problem shows up, and it isn't separate from the first — it's the reason the first one persists. The four-year degree has decades of tracking behind it: registrars, employer surveys, wage data collected at scale, all the infrastructure needed to say with some confidence what a credential produces. The alternative pathways don't have that infrastructure yet. Almost nobody tracks a graduate long enough after they leave a program to know whether it paid off. "We just don't have the data," as Delaski puts it; not because the outcomes aren't there, but because nobody built the pipe to see them.
That produces a genuine bind, not a solvable inconvenience. The category you can measure is the one you increasingly suspect is wrong. The category you believe is closer to true, you can't yet prove — and building the tracking infrastructure to prove it will take years the labor market has already shown it won't wait for. An institution asked to replan its skills model on evidence is being asked to use a dataset that doesn't exist yet, on a timeline that doesn't allow for waiting until it does.
There's a discipline built for exactly this condition, and it isn't data science. It's history. A historian never gets a complete record. Ledgers survive by accident, letters are missing their replies, a census undercounts the people it was least designed to see. The discipline's entire method is built around that gap — triangulating fragments, reading what's absent as evidence in itself, constructing a working account that is understood from the start to be provisional and gets revised as new fragments surface. Historians don't wait for the complete archive before they'll say anything. They've never had one.
That method doesn't transfer cleanly. A historian revises a narrative on their own schedule. A dean deciding what to fund next year does not have that luxury — someone has to commit to a working answer this budget cycle, on data that is thin and will stay thin for a while. The harder work isn't borrowing the historian's comfort with incomplete evidence. It's building an institution willing to act on a provisional account and revise it in public, rather than waiting for a complete dataset that will arrive, if it arrives, after the workforce it was meant to describe has already changed shape again.
The record will never be complete enough to justify the wait. The discipline built for acting on it anyway has been sitting in the same institutions the whole time.
—--------
Universities don’t hire Education Design Lab for another pilot. They hired the lab specifically for the design thinking process behind deciding what to keep and what to kill: "We needed a process that could set up success metrics — green-light and sunset. For everything you build, you eventually have to sunset something." That process is documented, not proprietary to one lab. The HCD Guide Series walks a practitioner through building it — the measurement step and the sunset step both — without commissioning it from scratch.
"We don't have the data that says these pathways are a sustainable path to financial success. We just don't have it." That's not just a higher-ed problem. All organizations share this issue.
Data will never be complete—even in a world with too much of it. Get comfortable with incomplete data, even in a world that’s drowning in it.
[Book an introductory conversation →]
What we’re into
Aaron
I was so excited to hear Kathleen say the words “design thinking”!!
Scott
I know! I have to be honest, I haven’t been sure if design thinking had survived contact with 2026. After all, even though we can’t trust most AI studies, design thinking has been, well, just awful at justifying its existence.
Aaron
It got productized and flattened, as Maura Shea said on the Pod a few weeks ago.
Ana
An episode that had me wondering if the HCD Guide series was just a massive waste of time…
Scott
No, I think that’s the wrong perspective to take. What Kathleen is talking about is the actual material of design thinking: the generative part. That’s what she means by “weaving”. It’s making newness out of having so much data and so many choices.
Aaron
I would agree with that, but as a data guy, I’m also like “I want real, concrete answers.”
Ana
But concrete doesn’t really exist, Aaron.
Aaron
And that’s why we work together, Ana. Because it does. You know it does.
Ana
And you know it doesn’t.
Scott
Philosophy! This is what Kathleen was talking about.
Ana
Touché, Scott, Touché.
Credits
Guest: Kathleen deLaski / Host: Sheev Dave / Line Producer: Aaron Meyers / Producer: Ana Monroe / Words: A. Adams / Artwork: Study, James McNeill Whistler, via the National Gallery / Why this artwork?: The metaphor of study, its subject matter, and the sketchy execution of this image reflects the current moment of unease, experimentation, and hope in the future of education.