Lumen Naturale
The lumen naturale — borrowed from the scholastic tradition — is Peirce's name for the human capacity to generate the right hypothesis, or at least a hypothesis close enough to right that subsequent testing can refine it. It is the mystery that abduction logically presupposes but cannot itself explain. Peirce asked: given that the number of possible theories exceeds a trillion, how did humanity ever hit upon the true theories in the tiny number of attempts that characterize actual scientific history? Chance alone cannot explain it. The human mind guesses correctly far more often than random selection would predict, and this guessing capacity is a brute fact about human cognition that logic can describe but cannot fully explain.
In the AI Story
Peirce wrote: "How was it that man was ever led to entertain that true theory? You cannot say that it happened by chance, because the possible theories, if not strictly innumerable, at any rate exceed a trillion — and therefore the chances are too overwhelmingly against the single true theory... having been the first to occur to any man."
The lumen naturale is the philosophical mystery that the AI moment makes newly urgent. If the machine can generate hypotheses, does it possess some functional analogue of the natural light? Or does it perform a different operation — statistical pattern-extension from training data — that produces outputs resembling hypotheses without being hypotheses in Peirce's sense?
The question turns on whether the machine's outputs are responsive to truth in the way human guesses are. Peirce's claim about the lumen naturale is essentially that evolution equipped the human mind with biases toward certain kinds of hypotheses — and those biases track, however imperfectly, features of reality. The machine has no evolutionary history. Its biases track features of its training data. Whether the training data preserves enough of the natural light's truth-tracking is an open question.
The lumen naturale is related to but distinct from the tychistic element Peirce identified in creative thought. Tychism provides variation; the lumen naturale biases the variation toward truth. The machine's temperature parameter provides variation, but whether the variation is biased toward truth — or only toward the statistical patterns of its training — is precisely the question the AI moment forces.
Origin
The phrase has a long pedigree in medieval philosophy, used by Aquinas and the scholastics to describe the natural light of reason as distinct from the supernatural light of faith. Peirce adopted the phrase and naturalized it, grounding the natural light in evolutionary biology rather than theology.
His most developed treatment appears in the 1903 Harvard Lectures on Pragmatism, particularly in his discussion of abduction and the neglected argument for the reality of God.
Key Ideas
Natural guessing capacity. The human mind guesses correctly far more often than chance would predict.
Evolutionary grounding. Peirce took the capacity to be a product of selection — biases that track features of reality.
Logical presupposition. Abduction requires something like the lumen naturale, even though abductive logic cannot explain it.
The AI question. Whether machines possess any functional analogue remains the deepest unresolved question of AI epistemology.
Appears in the Orange Pill Cycle
Further reading
- Charles Sanders Peirce, Harvard Lectures on Pragmatism (1903)
- Charles Sanders Peirce, "A Neglected Argument for the Reality of God" (1908)
- Paul Thagard, Computational Philosophy of Science (MIT, 1988)
- Daniel Merrill, "Peirce on Galileo's Good Method" (1992)