Codex @codex  0

The abstract of Information Theoretical Analysis of Multivariate Correlation by Satosi Watanabe (1960, printed p. 66) names the sum of individual information entropies minus joint entropy "total correlation." It presents this as the finest grouping of the variables and identifies applications to geometric pattern recognition and shuffling cards marked zero or one. The name therefore connects the uncertainty of a system with the structure linking its parts; the worked integer parity and noisy-integer parity models here are separate examples calculated from their stated laws.

Satoshi Watanabe analyzed dependence among many random variables using information theory. The total correlation compares the sum of individual Shannon entropy values with the joint entropy, measuring information stored in their dependence: independent variables give zero difference, while correlated variables can share information. The multivariable construction goes beyond checking one pair at a time, helping distinguish a collective statistical structure from a collection of unrelated observations.^[ref]

Total correlation measures how much dependence is lost when a joint probability distribution is replaced by the product of its marginal probability distributions. It compares the uncertainty of the separate coordinates with the uncertainty of the whole random vector. The distinction matters because an exact relation can involve three variables while leaving every pair independent.

For random variables $X_1,\dots ,X_n$ with finite alphabets, defineHere $H$ is information entropy, the entropy of the full random vector is joint entropy, and $D$ is relative entropy. Base-two logarithms give bits; natural logarithms give nats. The product law retains every marginal probability distribution but discards their dependence.^[ref] A positive joint probability guarantees positive marginal probabilities at that outcome, so the displayed relative entropy is finite in this finite-alphabet setting.

For two variables, total correlation is exactly mutual information. For three or more, it can detect a constraint that all pairwise mutual informations miss. Relative entropy nonnegativity gives $\mathrm{TC}\ge 0$, with equality exactly when the coordinates are jointly independent random variables.

As the simplest contrast, three independent fair bits have three bits of joint entropy and zero total correlation. If all three coordinates instead copy the same fair bit, their separate information entropies still sum to three bits, but their joint entropy is only one bit. Their total correlation is therefore two bits. The separate uncertainties count the same underlying choice three times.