The Badger Left
Dog bones and bona fides
My dachshund digs holes in my couch. It doesn’t work. His nails are trimmed and I’m not sure he could scar the cushion even if they weren’t. Every so often he tries again, wanders off, then comes back for another go, each attempt as fresh to him as the first. Is he silly or am I?
He was bred for badgers. Long body for the tunnel, short seal-shovel legs, loose neck skin so a bite couldn’t reach anything that mattered. A bark engineered to carry up through dirt. Stubbornness, because a dog that quits halfway down dies halfway down. The badger asked a question and his body was the answer.
The badger left. The dog stayed. The spine herniates now. The bark annoys the neighbors. He is excellent at a job no one is hiring for.
A credential is a bred answer too and the question it was bred for is older than most of the people defending it remember. Before the credential, trust did not travel. A stranger in one city had few ways to trust a stranger in another. You needed personal acquaintance, or you needed some institution to say: this one has been seen. The guild letter, the teaching license, the bar admission, the medical board, the civil service exam. Not all paper does the same work. Some of it licenses, some of it trains, some of it merely blesses. But all of it promises to make a person legible at a distance.
This worked, for a long time, well enough. The MD meant something because medicine was dangerous, training was controlled, and medical schools were scarce. The PhD meant something because the apprenticeship was long and the gatekeepers were few. My certs were bred for a market that needed to sort, at a broad scale, the people who understood health systems from the people who didn’t. It put me in rooms I would not otherwise have entered. The badger was there.
Slowly, then quickly, the asymmetry the credential was bred to solve has been routed around. The internet lets reputation travel without institutional vouching. Open work lets strangers verify each other directly through artifacts rather than paper. Network density inside professional classes has grown to the point where the relevant question in many hiring decisions is no longer “does this person have the credential” but “does anyone I trust know this person.” The credential is still checked. It is noticed mostly when absent.
I have a stack of AI governance certificates. Each taught me something. Each is also the dog on the couch. The worst part is that I knew this while I was doing it. The dog probably knows too, in whatever intuitive way a dog knows things. The knowing does not stop the digging. I have looked at course pages with the half-ashamed hope that one more line on the page might make me easier to believe.
The harder thing to say is that credentials were never only a way to trust strangers. They were also a way to decide which strangers would count. They restricted entry into professions to protect the people already inside them. They distributed access to status. They taught institutions whose names to recognize and whose to pass over. The guilds did this openly. The professions do it through accreditation. The institutions I have spent my career inside do it through a different vocabulary that arrives at the same place. When the trust function decays, the rationing function does not steadfastly decay with it. It persists without its old costume, its old custom, and starts to look like what its purpose always partly was.
Some rooms are the credential’s natural habitat. They are places where competence must first be understood before it can become persuasive. In those rooms, the paper does something prior to judging competence. It decides who is allowed to enter the category. The people doing comparable work without it are not lower on the list. They are not on the list.
The dog was bred to find one specific thing in one specific kind of dark. He also cannot find much of anything that does not look like a badger in a tunnel. Even if he wants it, even when the dog bone is right in front of his snout.
This predates AI by decades. The humanities PhD has spent decades producing people for a market that no longer resembles the one that justified it. Law school survived one labor-market correction and keeps selling certainty. The MBA became, in some places, less a formation than a tollbooth. AI did not create the problem. It has only made the incessant barking harder to ignore.
The temptation now is to call for credential reform. Well-intentioned people propose this every twenty years and it does not happen, because the institutions that issue credentials are the same institutions that benefit from the rationing function. The dachshund cannot redesign itself. The kennel club cannot either. What will happen, what is happening, is that the credential as institution will decouple further, the rationing function will become more visible, the people inside will defend it more loudly as it becomes harder to defend, and the people outside will build parallel structures that route around it. None of this is good or bad in the abstract. It is what happens when a bred answer outlives its question.
I am inside a system that still rewards the bred answer and I can see enough of the room to know the couch is a couch and not fertile hole-digging turf. Some of these neat little papers are still teaching me things. Some of the rooms I want to be in still ask for it. And stopping feels less like a choice than a confession, a resignation.
He comes back to the couch. Digs. Wanders off. Comes back. The badger is not in the couch. The badger was never in any couch. He is doing what he was made to do, in a room I arranged for him. He is not the only one who keeps digging.



"He also cannot find much of anything that does not look like a badger in a tunnel. Even if he wants it, even when the dog bone is right in front of his snout."
I wonder if the dog might not actually engage with such a bone if you managed to hide one in a badger-scented tunnel? Maybe even an intriguing toy?
I notice some of the work in biology by Michael Levin which seems to point to deeper mechanisms - perhaps not mechanisms at all in our usual sense - than the merely genetic. Affordances provided by fields which we are not aware of.
I got a liberal arts degree a long time ago (not even a BA but the foo-foo latinized "AB") in an era when it was broadly considered a credential to an equally broadly-considered field of fields (note the doubling usage here). Employers needed someone who could think. Now, honestly, this did not work out so precisely well for me and I ended up inventing a field for myself - even to the extent of hybridizing as a stay-at-home dad - but I think the principle was real. Though short-lived as we entered the hard narrowing of utilitarian capitalism.
In the age of AI this may seem even less likely. But I do wonder at the possibility that we might humanely (odd usage) train AI in such a way that we can partner with them and broadly-considered "human reasoning" might be a facility of mutual benefit. Just sayin' Thanks.