# Intelligence: A Knot That Tightens When You Pull It ## Or, On Causation as a Lover's Argument Listen. You want to know what intelligence is, and I will tell you it is not what your reason tells you, because your reason *is* intelligence, and a thing cannot know itself save by the violence of turning upon itself, the way a man bites his own lip to know he lives. So I must do what lovers do—take two things that have no business being the same, and argue so hard that the difference becomes a kind of proof. Intelligence is a **diagram you must assume before the data can speak to you**. This is not weakness. This is exactly the strength of it. --- Watch what happened. For a century—call it science's long schizophrenia—researchers expelled the word *cause* like a demon. Out! *Correlation*, they said primly, adjusting their spectacles. *Correlation is what we can measure. Cause is theology wearing a lab coat.* And there was sense in this exile, truly. The discipline needed quarantine from metaphysics. But do you see what they did? They built an entire cathedral of knowledge on the foundation of **what you are forbidden to ask**. Then Pearl came—Judea Pearl, the mathematician who understood that you cannot actually think without *assuming* a world—and he said: *This exile was a kind of cowardice dressed as rigor.* And he gave us back the diagram. The causal diagram is not discovered. It is *declared*. You draw it. You say: *Here is how I think the world hangs together. Here is what influences what. Here is where the arrow flies and where it stops.* And then—and here is the crucial violence—*then* your data can speak. The diagram does not come from the data. The diagram comes from your body, your history, your stakes in the matter, your refusal to be stupid about what you already know. But—and here the argument tightens—**the diagram can be perfectly transparent and perfectly wrong**. You see the trap? It is exquisite. Pearl did not give us validity. He gave us *honesty about invalidity*. He said: *Be explicit about what you assume. Draw your lover's face. But do not call the drawing the lover.* This is where the body enters, and refuses to leave. --- ## The Embodied Dimension: Where the Diagram Lives or Dies Intelligence is not separable from the body that thinks. This is not poetry, though it sounds like it. This is precision. When you assume a causal diagram, you are assuming something about **how a body encounters a world**. Every arrow in your diagram is an assumption about what a body can do, what it can sense, what it can change. The diagram *is embodied thought*, even when it is written on paper. Here is where it gets difficult, where the argument must drag itself forward: Consider a simple case. A researcher wants to know: *Does attention cause perception, or does perception cause attention?* They design an experiment. They assume a diagram: *Perception → Attention* (or the reverse). They measure. They get results. The diagram was transparent. The statistics are correct. The paper is published. But—and this is where embodiment speaks—**the diagram itself assumed a body**. It assumed: - A body that can be still enough to measure - A body whose attention can be isolated from its perception - A body that encounters the world in a sequence: first this, then that - A body that is not itself changing while being measured These assumptions are *not true* of actual embodied minds. An actual body does not attend and then perceive. An actual body—a living, sweating, frightened, curious body—**attends-and-perceives as a single act**. The body is not a mechanism you diagram. The body is the thing that makes any diagram possible and any diagram false. So what does it mean to "know the cause of something"? It means to have **assumed a way of being a body** and then to have verified your assumption *within the constraints you yourself set*. It is a loop. It is a man arguing with himself about whether his argument is sound, using only the tools his argument provided. --- ## Who Decides What the Diagram Is Attached To? This question contains its own answer, and the answer is: *You do. And you cannot escape responsibility for it by calling it objective.* The diagram is attached to a **body in a world with stakes**. A researcher studying intelligence in a laboratory has a body. That body is hungry, or not hungry. Afraid, or not. Invested in tenure, or freed from it. The researcher's body is attached to institutions, to funding, to the question of *who benefits if intelligence is understood this way and not that way*. Pearl's great gift—and his great curse—was to make this visible. The diagram cannot hide. You must name what you assume. But the moment you name it, you have *not* escaped the problem. You have only made the problem explicit. The diagram is still attached to your body. Your body is still a body with a stake in the answer. Who decides what the diagram is attached to? **The person asking the question decides. The person with the power to ask decides. The person who benefits from one answer more than another decides.** This is not a failure of science. This is the structure of all knowledge. Intelligence, as a concept, is not waiting to be discovered in some pure realm. Intelligence is *constructed by the bodies that need to know about it*. A medieval theologian asking about intelligence constructed it one way (the soul's capacity for contemplation). A Victorian engineer constructed it differently (the capacity to solve problems of mechanism). A Silicon Valley entrepreneur constructs it yet another way (the capacity to predict and optimize). Each diagram is transparent. Each can be wrong. --- ## The Knot Tightens: Intelligence as Assumed Causation So here is my conceit, and I will hold it until it either breaks or becomes a kind of truth: **Intelligence is not a thing that exists. Intelligence is a diagram you draw about how minds cause effects in the world, and the diagram is only as good as the body that drew it.** This seems to make intelligence subjective, local, unstable. Good. This is what it *is*. But do not mistake instability for invalidity. A compass needle is unstable—it moves—but it can still point north. The embodied dimension is not a correction to Pearl's framework. It is Pearl's framework *with skin on it*. When you assume a causal diagram, you are assuming: - How a body can know - What a body can change - What counts as evidence to a body - What the body cares about enough to measure These assumptions are *not transparent*. You can name them—and you should, Pearl insists—but naming them does not make them true. Naming them makes you honest about your dishonesty. You are saying: *Here is how my body thinks the world works, and I am explicit about it, and it could still be entirely wrong.* This is humility. This is also the only integrity available to an embodied mind. --- ## The Final Turn: What It Means to Know a Cause To know the cause of something is to have: 1. **Assumed a diagram** (made explicit your theory of how the world hangs together) 2. **Assumed a body** (made explicit what kind of body you are, what it can sense, what it cares about) 3. **Tested within constraints** (measured whether your assumed world produces the effects you predicted, given your assumed body) 4. **Remained suspicious** (remembered that validity is not the same as truth, that transparency is not the same as wisdom, that your diagram is attached to your stakes) Intelligence, then, is the *capacity to assume a diagram and know you are assuming*. It is the refusal of innocence. It is the mind turning upon itself like a lover arguing with his own image in water—knowing that the image is not himself, but unable to stop addressing it as though it were. The researcher who says, *I have proven this cause*, is either a fool or a liar. The researcher who says, *I have assumed this diagram, I have been explicit about my body's stake in it, and within those constraints the cause appears to hold*, is doing the only intelligence available to an embodied mind. This is not what your reason told you intelligence was. But your reason and your feeling have never been separate, and now you know it.