Application of ASP in Second Language Acquisition: Formal Input Processing Theory

2.1. Shirin Saitunan Amsa (ASP)

ASP is a declarative programming paradigm based on the stable model (answer set) semantics of logic programming. It excels at representingdefault reasoning, incomplete information, and dynamic domains.—these are the core features that simulate human cognitive processes. Rules in ASP take the form:head :- body., where if the body is satisfied, the head is true. Defaults can be elegantly represented using negation as failure (not).

2.2. Ka'idar Sarrafa Shigarwa

The IP theory proposed by VanPatten posits that second language learners, especially beginners, due to limited processing resources (working memory) and incomplete grammatical knowledge, employ a set ofdefault heuristicsExtract meaning from input. A key principle isFirst Noun Principle: Learners tend to assign the agent/subject role to the first noun or pronoun they encounter in a sentence. This leads to systematic misinterpretations, such as misinterpreting the passive sentence "The cat was bitten by the dog" as "The cat bit the dog".

3.1. Ƙirƙirar Tsarin Tsoho

IP principles are encoded as ASP rules. For example, the First Noun Principle can be expressed as a default rule that applies when grammatical cues (such as passive voice markers) are not processed due to resource limitations:

% 默认：将施事角色分配给第一个名词
assign_agent(FirstNoun, Event) :-
    sentence_word(FirstNoun, Position1, Noun),
    sentence_word(Verb, Position2, VerbLex),
    Position1 < Position2,
    event(Event, VerbLex),
    not processed(grammatical_cue(passive, Verb)),
    not overridden_by_grammar(Event).

not processed(...) The condition captures resource limitations, making this rule non-monotonic.

3.2. Representing Learner Knowledge and Resources

The model integrates a dynamic representation of the learner's state:

• Lexical knowledge: Such as knows_word(learner, 'dog', noun, animal). The fact.
• Grammatical knowledge: Internalized rules (e.g., rules about the passive voice).
• Processing resources: Modeled as a constraint limiting the number of grammatical features that can be processed simultaneously in a given sentence.

The interaction between default strategies and acquired grammatical knowledge is modeled through rule prioritization or rule cancellation.

4.1. System Architecture

PIas (Processing Input as a System) is a prototype system that takes English sentences and a learner profile (approximate proficiency level, known vocabulary/grammar) as input. It uses a formal ASP model to generate one or more predicted interpretations (answer sets).

4.2. Experimental Prediction and Verification

Wannan takarda tana nuna fitarwar tsarin ga misalin gargajiya. Game da jumlar wuce gona da iri "The cat was bitten by the dog" da kuma fayil na mai farawa:

• Hasashen Fahimta 1 (na asali): Mai aikatawa = cat, aiki = bite, abin da aka yi masa = dog. (Kuskuren fahimtar aiki mai ƙarfi).
• Sharuɗɗan Fahimta daidai: Only when the learner's profile contains processed knowledge of passive morphology (processed(grammatical_cue(passive, 'bitten'))) will the model predict the correct passive interpretation, thereby overriding the default rule.

These computational predictions align with empirical observations in SLA research, validating the model's face validity. This formalization also reveals potential ambiguities in natural language theory and suggests improvements.

5.1. Core Logical Formalism

The core of the model can be abstracted using logical constraints. Let $L$ be the learner's knowledge state, $S$ be the input sentence, and $R$ be the available processing resources. An interpretation $I$ is a set of semantic roles and relations. The IP theory $T$ defines a mapping function $F_T$ constrained by default rules $D$:

$I = F_T(S, L, R) \quad \text{subject to} \quad \sum_{g \in G(S)} \text{cost}(g) \leq R$

Here, $G(S)$ is the set of grammatical features in $S$, and $\text{cost}(g)$ is the cognitive load for processing $g$. If $g \notin \text{processed}(L, R, S)$, the default rule $D$ applies.

5.2. Analytical Framework Example

Case Study: The First Noun Principle in Different Syntactic Structures.

Input: "John ya ba littafi ya ba Mary." (Ya ƙunshi madaidaicin fi'ili mai biyu a cikin jumlar wanda aka yi masa aiki).
Fayil ɗin Malami: Mai farawa; ya san kalmomin 'littafi', 'ba', 'Mary', 'John'; bai aiwatar da siffar jumlar wanda aka yi masa aiki ko tsarin baƙi ba.
Aiwar Tsarin ASP:
1. Binciken ƙamus: BOOK, GIVE, MARY, JOHN.
2. Kasa aiwatar da jumlar wanda aka yi masa aiki ('was given') da kuma maɗaurin kai ('to Mary').
3. Ka'idar suna ta farko ta tashi: BOOK an ba shi matsayin mai yin aiki.
4. Dabarar tsari ta layi: An fassara jerin a matsayin mai yin aiki-aiki-mai karɓa-? (Matsayin JOHN ba a sani ba).
Hasashen Fitowa: Multiple answer sets may arise, such as {agent(BOOK), action(GIVE), recipient(MARY), other_participant(JOHN)}, leading to confusing interpretations like "the book gave something to Mary (with John also involved)." This precisely identifies the specific areas where learners might become confused, allowing teachers to target their instruction accordingly.