Paradigm complexity: entropy and implicative structure over cells #
Two views of the Paradigm Cell Filling Problem over a ParadigmSystem.
The quantitative layer measures cell distributions: Shannon entropy
of a cell, conditional entropy between cells (the integrand of
[AM13a]'s i-complexity), and the derived
implicative-structure predicates. The qualitative layer is the
zero-entropy (categorical) case: a set of cells Predicts another when
the forms filling it determine the form filling the target across every
inflection class — the implicative relation [BB16]
quantify with conditional entropy, here as a plain relation over the
system's rows. Entropy quantification stays prose-level per repo
discipline; Predicts is its zero-entropy extreme.
Main declarations #
ParadigmSystem.cellEntropy,ParadigmSystem.conditionalCellEntropyParadigmSystem.isImplicative,ParadigmSystem.isTransparentParadigmSystem.Predicts,ParadigmSystem.IsPrincipalPartSet— the categorical implicative relation and principal-part sets
iComplexity (the paper-specific average conditional entropy) and
LCECHolds (the LCEC threshold predicate) live in
Studies/AckermanMalouf2013.lean because they are particular
aggregations the paper defines, not substrate primitives.
Shannon entropy of the empirical distribution at cell c (in nats).
Equations
- ps.cellEntropy c = match Morphology.distOfList✝ (ps.cellDistribution c) with | (support, prob) => Morphology.entropy✝ support prob
Instances For
Conditional entropy H(C_i | C_j) of cell ci given cell cj (in nats).
The integrand of [AM13a]'s i-complexity.
Equations
- One or more equations did not get rendered due to their size.
Instances For
A cell pair (ci, cj) is implicative iff knowing the form at cj
perfectly determines the form at ci (zero conditional entropy).
Equations
- ps.isImplicative ci cj = (ps.conditionalCellEntropy ci cj = 0)
Instances For
A paradigm is transparent iff every off-diagonal cell pair is implicative — every cell perfectly predicts every other.
Equations
- ps.isTransparent = ∀ (ci cj : Fin n), ci ≠ cj → ps.isImplicative ci cj
Instances For
Qualitative implicative structure #
The categorical (zero-entropy) face of the Paradigm Cell Filling Problem [BB16]: which sets of known cells determine an unknown cell across the inflection classes.
A set of cells S predicts cell j when the forms filling S
determine the form filling j across the system: any two inflection
classes agreeing on every cell of S also agree at j. The
zero-conditional-entropy case of the implicative relation
[BB16] measure.
Equations
Instances For
A set of cells is a principal-part set when it predicts every cell of the system: knowing those forms determines the whole paradigm.
Equations
- ps.IsPrincipalPartSet S = ∀ (j : Fin n), ps.Predicts S j
Instances For
Prediction is monotone in the known cells: enlarging the predictor set never destroys a prediction.