In his keynote article, O'Grady provides the newest pieces in his research program, the overarching goal of which is to reinterpret what linguistic theory has traditionally called 'syntactic knowledge' as being a side-effect of constraints imposed by a domain-general processing system. Here O'Grady proposes that first language (L1) syntactic acquisition is explainable in terms of strengthening processing routines in response to linguistic input, whereas second language (L2) syntactic development involves transfer of the better-entrenched L1 routines, except when properties of the L2 favor a simpler processing strategy. O'Grady's work is part of a broader literature which eschews domain-specific rules and representations in favor of general processing strategies. An important and unique contribution of his research program is that it provides a coherent processing-based account of many complex linguistic phenomena that form the core evidence for many linguists' claims that recourse to abstract linguistic representations is conceptually necessary.At the heart of this theory is 'the processor, ' which takes over the duties of 'the grammar' from more traditional syntactic theories. O'Grady's theory, like those proposed by other emergentists, assumes that cross-linguistic universals and the seeming similarity of L1 speakers' knowledge and processing of language can be explained in terms of the processor's fundamental architecture. For O'Grady, this means a highly capacity-constrained working memory (WM) system. However, as with 'the grammar' in generative linguistic theories, the neurocognitive nature of 'the processor' in O'Grady's framework remains relatively unexplored. In my commentary, I would like to focus on an implicit assumption in this work, namely that the processor uses the same types of information in all individuals when processing linguistic dependencies. In doing so, I will point out some recent findings using neural measures of language processing, namely event-related brain potentials (ERPs), from both my own research and from others. These findings show that the language processor has multiple streams of information processing,