On the early left-anterior negativity (ELAN) in syntax studies

Within the framework of Friederici’s (2002) neurocognitive model of sentence processing, the early left anterior negativity (ELAN) in event-related potentials (ERPs) has been claimed to be a brain marker of syntactic first-pass parsing. As ELAN components seem to be exclusively elicited by word category violations (phrase structure violations), they have been taken as strong empirical support for syntax-first models of sentence processing and have gained considerable impact on psycholinguistic theory in a variety of domains. The present article reviews relevant ELAN studies and raises a number of serious issues concerning the reliability and validity of the findings. We also discuss how baseline problems and contextual factors can contribute to early ERP effects in studies examining word category violations. We conclude that – despite the apparent wealth of ELAN data – the functional significance of these findings remains largely unclear. The present paper does not claim to have falsified the existence of ELANs or syntax-related early frontal negativities. However, by separating facts from myths, the paper attempts to make a constructive contribution to how future ERP research in the area of syntax processing may better advance our understanding of online sentence comprehension.

► The view that ELANs are specific reflections of automatic syntax-first processing (Friederici, 2002) is critically evaluated.
► We show that core claims about ELANs are not supported by ERP data and that some ‘ELAN’ findings must be artifacts.
► Paradigms used to elicit ELANs are methodologically highly problematic and prone to context-driven ERP artifacts.
► Local ELANs (100–300 ms) in auditory studies can be understood as sustained negativities that are superimposed by P600s.
► Our new account for apparent ‘semantic blocking effects’ further supports alternative views of how parsing unfolds on-line.