contingency-respondent-dsl

🇯🇵 日本語版 README

Tier B Pavlovian (respondent) procedure extension for contingency-dsl. This package defines 26 classical conditioning procedures that extend the Core respondent layer without modifying its grammar, by registering production rules through the ExtensionRespondentPrimitive extension point.

Relationship to contingency-dsl

The Core package keeps the respondent layer deliberately minimal — only the foundational two-term contingency primitives (R1–R14: Pair.*, Extinction, CSOnly, USOnly, Contingency, TrulyRandom, ExplicitlyUnpaired, Compound, Serial, ITI, Differential). The Core respondent grammar exposes a single extension hook:

ExtensionRespondentPrimitive ::= Identifier "(" ArgList? ")"

Any third-party respondent vocabulary that needs deeper Pavlovian coverage registers its identifiers through this point. contingency-respondent-dsl is the first and canonical consumer.

Scope — 26 Tier B primitives

Grouped by conceptual family (see spec/en/tier-b-primitives/_index.md):

• Acquisition & higher-order — higher-order conditioning (Pavlov, 1927; Rizley & Rescorla, 1972), sensory preconditioning (Brogden, 1939), counterconditioning (Dickinson & Pearce, 1977).
• Cue competition — blocking (Kamin, 1969), overshadowing (Mackintosh, 1976), overexpectation (Rescorla, 1970), super-conditioning (Rescorla, 1971), retrospective revaluation (Van Hamme & Wasserman, 1994).
• Inhibitory learning — conditioned inhibition (Rescorla, 1969), occasion setting (Holland, 1983), inhibition of delay (Pavlov, 1927), summation / retardation test (Rescorla, 1969).
• Preexposure — latent inhibition (Lubow & Moore, 1959), US preexposure (Randich & LoLordo, 1979).
• Extinction & recovery — renewal (Bouton & Bolles, 1979), reinstatement (Rescorla & Heth, 1975), spontaneous recovery (Rescorla, 2004), latent extinction, Pavlovian partial-reinforcement extinction effect, contextual extinction (Bouton, 2004), reconsolidation interference (Monfils et al., 2009).
• Specialized procedures — conditioned taste aversion (Garcia, Ervin, & Koelling, 1966), stimulus generalization (Hovland, 1937), contextual fear conditioning (Fanselow, 1990), peak procedure (Roberts, 1981), mediated conditioning (Holland, 1981).

Repository layout

contingency-respondent-dsl/
├── spec/en/                      formal specification (English canon)
│   ├── architecture.md           how this package plugs into Core
│   ├── design-philosophy.md      additive-only extension discipline
│   ├── integration-with-core.md  extension-point contract
│   ├── grammar.md                EBNF for all 26 Tier B primitives
│   ├── theory.md                 Pavlovian learning framework (pointer-level)
│   └── tier-b-primitives/        one file per primitive + _index.md
├── spec/ja/                      Japanese mirror of spec/en/
├── schema/
│   ├── grammar.ebnf              the 26 Tier B productions
│   ├── ast.schema.json           JSON Schema 2020-12 for AST nodes
│   └── extension-registry.md     how a program loads this registry
├── conformance/
│   ├── README.md
│   └── tier-b/                   one .json fixture file per primitive
└── docs/
    ├── en/README.md
    └── ja/README.md

Status

This package currently provides:

• The EBNF productions for 26 Pavlovian procedures that extend Core's ExtensionRespondentPrimitive (see spec/en/grammar.md).
• A JSON Schema for the resulting AST nodes (see schema/ast.schema.json).
• Per-primitive operational definitions with primary APA citations.
• A Python parser skeleton (src/contingency_respondent_dsl/) with:
  • Frozen dataclasses for all 26 primitives (ast.py).
  • A registry that dispatches identifier + keyword-args bags to AST constructors (registry.py).
  • A dict serializer compatible with the AST schema (serializer.py).
  • A thin parse_primitive() entrypoint (parser.py).
• Conformance fixtures for 5 demonstration primitives (blocking, renewal, latent inhibition, overshadowing, reinstatement) with concrete expected_ast dicts. The remaining 21 fixtures have populated input strings but retain expected_ast: null until their argument schemas are exercised end-to-end through the parser.

Installation (for local development)

This package uses mise + venv per the upstream OperantKit policy. uv is not used.

# From inside contingency-respondent-dsl/
mise exec -- python -m venv .venv
.venv/bin/python -m pip install -e .
.venv/bin/python -m pip install pytest
.venv/bin/pytest tests/ -q

The Core parser (contingency-dsl-py) is a separate sibling package and is not required to exercise this package's own unit tests: the Tier B parser operates on already-evaluated keyword arguments at the extension boundary, and nested Tier A expressions are treated as opaque dicts on that boundary.

Usage

from contingency_respondent_dsl import (
    parse_primitive,
    to_dict,
    from_dict,
)

node = parse_primitive(
    "Blocking",
    positional_args=[],
    keyword_args={
        "phase1": {"type": "Pair.ForwardDelay", "cs": "a", "us": "shock"},
        "phase2": {
            "type": "Pair.ForwardDelay",
            "cs": {"type": "Compound", "elements": ["a", "x"]},
            "us": "shock",
        },
        "test":   {"type": "Extinction", "cs": "x"},
        "phase1_trials": 40,
        "phase2_trials": 20,
        "test_trials":   10,
    },
)
# node is a frozen contingency_respondent_dsl.ast.Blocking instance

serialized = to_dict(node)  # -> schema-shaped dict
restored   = from_dict(serialized)
assert restored == node

To attach the Tier B extension to a registry-like object:

from contingency_respondent_dsl import register

bag: dict = {}
ext = register(bag)   # dict: stores ext under "respondent-tier-b"
ext.parse_primitive("Renewal", None, { ... })

The register() helper accepts a dict, any object exposing a register_tier_b(ext) method, or a plain namespace.

Coverage status

All 26 primitive identifiers are registered (unknown identifiers raise UnknownPrimitiveError). End-to-end parse paths are validated for 5 demonstration primitives; AST classes for the remaining 21 are defined and registered but their round-trip paths are exercised only through from_dict / to_dict in the conformance runner as their fixtures are populated.

Navigation

• Formal specification (English): spec/en/
• Formal specification (Japanese): spec/ja/
• EBNF grammar: schema/grammar.ebnf
• AST JSON Schema: schema/ast.schema.json
• Per-primitive index: spec/en/tier-b-primitives/_index.md
• Conformance fixtures: conformance/tier-b/

References (selected)

• Bouton, M. E. (2004). Context and behavioral processes in extinction. Learning & Memory, 11(5), 485–494. https://doi.org/10.1101/lm.78804
• Kamin, L. J. (1969). Predictability, surprise, attention, and conditioning. In B. A. Campbell & R. M. Church (Eds.), Punishment and aversive behavior (pp. 279–296). Appleton-Century-Crofts.
• Mackintosh, N. J. (1975). A theory of attention: Variations in the associability of stimuli with reinforcement. Psychological Review, 82(4), 276–298. https://doi.org/10.1037/h0076778
• Pavlov, I. P. (1927). Conditioned reflexes. Oxford University Press.
• Pearce, J. M., & Hall, G. (1980). A model for Pavlovian learning: Variations in the effectiveness of conditioned but not of unconditioned stimuli. Psychological Review, 87(6), 532–552. https://doi.org/10.1037/0033-295X.87.6.532
• Rescorla, R. A. (1967). Pavlovian conditioning and its proper control procedures. Psychological Review, 74(1), 71–80. https://doi.org/10.1037/h0024109
• Rescorla, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning. In A. H. Black & W. F. Prokasy (Eds.), Classical conditioning II (pp. 64–99). Appleton-Century-Crofts.