A Search for Bilingual Advantages in Creativity: Evidence for Capacity- Limited, Not Resource-Based, Cognitive Architecture

A long-standing claim in the bilingualism literature is that bilingual experience confers broad cognitive advantages, including enhanced creativity, by expanding a pool of general cognitive resources. We evaluated this claim by testing competing architectural accounts of how creativity is supported by cognitive capacity, control, and experience. Using structural equation modeling, we compared three theoretically distinct models—Modular, Resource Pool, and Hybrid—that were specified over identical measurement structures and covariate sets and fit to the same data. All three models exhibited acceptable global fit, indicating that fit alone cannot adjudicate among competing cognitive architectures. Critical differences instead emerged in the structure and interpretability of parameter estimates. In the Modular model, creativity was predicted by attention control and orientational disposition, with general fluid intelligence contributing indirectly through control processes and no independent contribution of bilingualism. In contrast, the Resource Pool and Hybrid models constructed shared latent factors whose meaning depended on partially antagonistic contributions from attentional control and bilingual experience, yielding ambiguous interpretations of creativity as a resource-driven outcome. Taken together, these results do not support the existence of a shared cognitive resource underlying creativity. Instead, they favor a capacity-limited, control- based architecture in which creative performance reflects regulated exploration under cognitive constraints rather than the accumulation of general cognitive resources.