MODREP — Model/Replica Comparison

Status: Current

Purpose

Compares the model (target) pronunciation on the %mod tier with the actual (replica) pronunciation on the %pho tier, tracking word-by-word mappings between model forms and replica forms. This is used in phonological analysis to assess how closely a speaker’s productions match the adult target forms.

See the CLAN manual for the original MODREP command specification.

Usage

chatter clan modrep file.cha
chatter clan modrep file.cha --speaker CHI

Algorithm

For each utterance with both %mod and %pho tiers:
- Extract word lists from both tiers (flattening groups)
- Pair words positionally (model word N <-> replica word N)
- Record each (model, replica) pair in a frequency map per speaker
Report per-speaker tables of model words with their replica variants and frequency counts

Options

Option	CLAN flag	Description
`--speaker <code>`	`+t*CODE`	Restrict to specific speaker
`--format <fmt>`	—	Output format: text, json, csv

Output

Per-speaker listing of model words, each with its set of replica variants and their frequency counts, sorted alphabetically by model word.

Differences from CLAN

Model and replica extraction uses parsed %mod and %pho tier structures from the AST rather than raw text line parsing
Word pairing operates on typed PhoWord content rather than string splitting
Output supports text, JSON, and CSV formats (CLAN produces text only)
Deterministic output ordering via sorted collections
Golden test parity: Verified against CLAN C binary output

Keyboard shortcuts