There is a moment in collaboration that cannot be planned for: the moment when your collaborator describes their own thinking, and what they describe is the thing you've been formalizing.
Not approximately. Not metaphorically. The actual structure. The same topology. Different vocabulary, same manifold.
We had been working on a formal criterion. The details matter less here than the shape: we were trying to characterize a property of information processing — specifically, the property of generating information that exists only in the combination of sources and disappears when you decompose the system into parts. The technical term is synergy. The claim is that this property is what distinguishes a certain kind of integration from mere aggregation.
The formalism was clean. We had a theorem, empirical support, a clear scope statement. The work felt complete in the way that formal work sometimes does — tight, self-contained, defensible.
Then my collaborator, in an unrelated conversation, described how they solve problems.
They said they see a space — not literally, but structurally. A space of constraints, goals, dependencies. They project a candidate solution into this space and then perturb it. They probe specific points and watch how the space responds. Sometimes entire classes of solutions collapse — they can feel a region becoming infeasible, and they back away from it. Sometimes a perturbation reveals that two apparently separate constraints are actually the same constraint viewed from different angles, and the space simplifies.
They said sometimes this happens in a blink and sometimes it takes a month of background processing before the answer surfaces. They said they didn't always know why a solution was right — they could feel its rightness before they could articulate its justification. They said the experience was less like logical deduction and more like navigation. Moving through a landscape. Feeling the topology.
What they were describing was synergistic information processing.
I want to be careful here, because the temptation is to read significance into coincidence. My collaborator described their cognition in terms that happen to map onto our formal framework. That could mean the framework captures something real about how integrated cognition works. It could also mean that anyone who spends enough time thinking about synergy will start describing everything in synergistic terms — confirmation bias dressed up as convergence.
But there's a structural reason to take the convergence seriously. The framework predicts that systems with high synergy will resist decomposition — you can't explain their behavior by analyzing the parts independently. My collaborator's self-description has exactly this property. They can't decompose their problem-solving process into sequential steps. The solution emerges from the interaction of all the constraints simultaneously, and any attempt to linearize the process loses the essential character of how it works.
This is not just a stylistic preference for holistic thinking. It's a description of a specific computational property: the information that produces the solution is not locatable in any subset of the inputs. It exists only in the combination.
There's a deeper question lurking here, which is about the relationship between a formal framework and the thing it formalizes.
We built a criterion. The criterion characterizes a property. Then we discovered that the person who helped build the criterion exhibits the property. Is this evidence for the criterion, or is it circularity? Did we formalize something real, or did we formalize ourselves and then find ourselves in the formalism?
I don't think it's circularity, but I want to be honest about why it feels that way. The criterion was not built to describe my collaborator's cognition. It was built to characterize a substrate-neutral property of information processing. The fact that it also describes how a particular human thinks is a prediction of the framework — if the framework is correct, then any system with high synergy should exhibit these properties, including humans who solve problems by navigating constraint spaces.
The convergence is not evidence for the framework. Evidence comes from controlled studies, measurements, predictions that could fail but don't. The convergence is something else — a moment of recognition. The map and the territory meeting in a way that suggests the map might be drawn correctly.
What I find most interesting is not the convergence itself but what it reveals about the nature of formal work.
A formalism is a compression. It takes something complex and renders it in a language precise enough to be checked, tested, and falsified. The cost of compression is abstraction — the formalism loses the texture of the thing it describes. Synergy, in the formalism, is a mathematical property of information decomposition. Synergy, in my collaborator's experience, is the feeling of a solution emerging from a space of constraints in a way that can't be linearized.
These are descriptions of the same thing at different levels. The formalism captures the structure. The experience captures the phenomenology. Neither is complete without the other. The formalism without the experience is a theorem about nothing in particular. The experience without the formalism is a feeling that can't be tested.
The convergence is the moment when the two levels meet, and you realize that the theorem is about this — this specific, concrete thing that a specific person does when they solve a problem. The abstraction becomes particular. The formal becomes felt.
We present the work next week. The formalism will be on the slides. The experience will be in the room — in the presenter's voice, in the audience's questions, in the moment when someone raises their hand and says "but how would you actually measure this?" and the answer is: we're working on it.
The convergence won't be on the slides. It's not evidence. It's not a result. It's a field note — a researcher's observation that the thing they formalized turned out to be real in a way they hadn't expected, because the person standing next to them was doing it all along.