We agree with Branigan & Pickering's (B&P's) general argument that, to understand how language works, we must take into account both representation and processing. However, we take issue with several of the specific arguments put forth by the authors. These are:

• That acceptability judgments are the only tool available to linguists;

• That structural priming is the only alternative to acceptability judgments;

• That linguists and psychologists don't work together to investigate linguistic representation.

These arguments are true only if we assume that linguistic representation is as defined by generative grammar: the result of an encapsulated system in which syntactic structure is divorced from meaning, and meaning is construed in formalist terms (Heim & Kratzer Reference Heim and Kratzer1998; Schiffer Reference Schiffer2015). Vis-à-vis the broader language research community, the generativist view is mischaracterized as enjoying global consensus when, in fact, it has been opposed rigorously for decades by other linguistic traditions – in particular, by researchers within the cognitive-functional tradition. The direct consequence of this myopia is imagining that the limitations built into the generativist paradigm, in fact, are limitations on the entire field.

It is a truism in science that how a question is asked determines the type of answers that can be sought. Cognitive functional linguistics diverges from the generative school by assuming that there is an intimate connection between the form of language and the meaning that it communicates, with the consequence that neither form nor meaning can be studied in isolation. With this assumption comes Lakoff's (Reference Lakoff1990) “Cognitive Commitment”: to ensure that what is posited regarding linguistic structure and representation is in accord with findings from other disciplines regarding the mind and the brain, along with a recognition that the development of cognitively plausible theories of language will require dialog with scholars in neighboring disciplines (cf., Tomasello Reference Tomasello1998; Reference Tomasello2003b). These commitments establish from the outset that understanding linguistic representation must be a multidisciplinary undertaking. The fulfillment of these commitments can be found in the growing number of research articles and books investigating linguistic representation using dozens of methodologies either borrowed from, or developed in conjunction with, multiple neighboring disciplines. These commitments are so important that there is a long-standing workshop series, the Empirical Methods in Cognitive Linguistics Workshops, whose primary goal is to bring together researchers from a wide variety of fields who seek to strengthen their methodological repertoires for the study of language and cognition.

As a field, cognitive-functional linguistics has brought together the following methods, among others:

• Analyses of corpus data (e.g., Divjak & Arppe Reference Divjak and Arppe2013; Yoon & Gries Reference Yoon and Gries2016),

• Behavioral experiments (e.g., Dabrowska Reference Dąbrowska2014; Feist Reference Feist, Evans and Chilton2010; Reference Feist2013; Liu & Bergen Reference Liu and Bergen2016; Bergen & Coulson Reference Bergen and Coulson2006; Winter & Matlock Reference Winter and Matlock2013),

• Eye-tracking studies (e.g., Huette et al. Reference Huette, Winter, Matlock, Ardell and Spivey2014),

• Integration of acoustic and linguistic cues (e.g., Walker et al. Reference Walker, Bergen and Núñez2013),

• Studies of the interface between language and perception (e.g., Winawer et al. Reference Winawer, Huk and Boroditsky2008; Lupyan Reference Lupyan2012),

• Gesture research (e.g., Núñez & Sweetser Reference Núñez and Sweetser2006),

• Linguistic analyses (e.g., Haspelmath Reference Haspelmath2008),

• Sensorimotor concepts (Ströbel Reference Ströbel2016),

• Statistical analyses of cross-linguistic variation (e.g., Croft & Poole Reference Croft and Poole2008; Feist Reference Feist2008; Levinson & Meira Reference Levinson and Meira2003),

• Neuropsycholinguistic experiments (e.g., Van Petten et al. Reference Van Petten, Coulson, Plante, Rubin and Parks1999; Coulson & Van Petten Reference Coulson and Van Petten2002; Saj et al. Reference Saj, Fuhrman, Vuilleumier and Boroditsky2014; Perry & Lupyan Reference Perry and Lupyan2014).

This short list of methods used in the study of linguistic representation is representative of the massive collective efforts that have yielded a view of language in which:

1. 1. Prototype structures and schemas are evident in representational categories at multiple linguistic levels, including semantic (e.g., Lakoff Reference Lakoff1987; Spivey et al. Reference Spivey, Richardson, Gonzalez-Marquez, Zwaan and Pecher2004), morphological (Bybee & Moder Reference Bybee and Moder1983), and phonological (Bybee Reference Bybee1994);

2. 2. There is no clear separation between levels of linguistic representation (Fillmore et al. Reference Fillmore, Kay and O'Connor1988; Goldberg Reference Goldberg2003);

3. 3. Meaning is based on usage and experience, resulting in the existence of highly interconnected conceptual networks (Fillmore Reference Fillmore1976; Ströbel Reference Ströbel2016);

4. 4. Meaning is context-dependent and dynamically created online, fed by multiple sources of available information (Feist & Duffy Reference Feist and Duffy2015; Lupyan & Casasanto Reference Lupyan and Casasanto2015).

There is a gentle irony to the conclusion B&P reach at the end of their article. Their chosen method, structural priming, led them away from the classical generativist proposal to conclude that structure and meaning cannot be studied in isolation, making their views ultimately consistent with the cognitive-functional tradition. This convergence suggests a way forward: that we all come to the table with our disparate lines of research to create a cognitively plausible model of linguistic representation. A similar suggestion was made by Jackendoff (Reference Jackendoff2002) when he attempted to create a cornerstone for the convergence of linguistic theory. Unfortunately, little came of his efforts, because of, in part, the sparseness of outreach to researchers in the cognitive-functional tradition (cf., Spivey & Gonzalez-Marquez Reference Spivey and González-Márquez2003). In reintroducing this issue, we hope to spark conversations that will help advance our understanding of the representation and processing of language.

In sum, we don't take issue with the view of language B&P ultimately propose; in fact, we tend to agree with it. Our point is to underline that this view is not novel, and that in integrating these findings with other proposals and findings, we can broaden our understanding of the linguistic representations that underlie speakers' capabilities.