Syntactic Complexity as a Compensatory Heuristic: Towards a Model of Evaluative Judgment in EFL Writing Assessment

Abstract

Existing accounts of second language processing explain how EFL learners extract meaning from text, manage limited working-memory capacity, and register salient linguistic features. However, none specifies what happens when a reader must convert that processing into a single evaluative judgment of writing quality. This paper addresses that gap by proposing the Compensatory Model of Evaluative Judgment (CMEJ), which integrates Input Processing Theory, Cognitive Load Theory, and the Noticing Hypothesis into a single, testable mechanism. CMEJ predicts that evaluative judgment follows one of two pathways depending on a reader's prior topic knowledge. Under sufficient familiarity, judgment rests on content-based criteria such as coherence and evidential strength. Under limited familiarity, judgment defaults to structurally salient heuristics, principally syntactic complexity, as a proxy for quality. The relationship between topic knowledge and reliance on syntactic complexity is modelled as continuous and directional rather than threshold-based and is specified simultaneously as a statistical moderation and a cognitive mediation. The paper positions CMEJ against the three frameworks it builds on, examines the model's own opportunities and boundary conditions, and outlines theoretical and pedagogical implications for rater training, rubric design, and assessment validity in EFL writing assessment. CMEJ is offered here as a conceptual, falsifiable contribution; an empirical test of its central prediction is identified as the direct next step.