Flexible and fast: Linguistic shortcut affects both shallow and deep conceptual processing

Abstract

Previous research has shown that people use linguistic distributional information during conceptual processing, and that it is especially useful for shallow tasks and rapid responding. Using two conceptual combination tasks, we showed that this linguistic shortcut extends to the processing of novel stimuli, is used in both successful and unsuccessful conceptual processing, and is evident in both shallow and deep conceptual tasks. Specifically, as predicted by the ECCo theory of conceptual combination, people use the linguistic shortcut as a "quick-and-dirty" guide to whether the concepts are likely to combine into a coherent conceptual representation, in both shallow sensibility judgment and deep interpretation generation tasks. Linguistic distributional frequency predicts both the likelihood and the time course of rejecting a novel word compound as nonsensical or uninterpretable. However, it predicts the time course of successful processing only in shallow sensibility judgment, because the deeper conceptual process of interpretation generation does not allow the linguistic shortcut to suffice. Furthermore, the effects of linguistic distributional frequency are independent of any effects of conventional word frequency. We discuss the utility of the linguistic shortcut as a cognitive triage mechanism that can optimize processing in a limited-resource conceptual system.

What have labels ever done for us? The linguistic shortcut in conceptual processing

“…Liu et al (2018) explored the mechanism in more detail by introducing a competing measure of ease of simulation (i.e., a normed variable that reflects how easily a phrase could result in a coherent simulation of meaning) and comparing its effect to that of distributional cooccurrence frequency in metaphor processing. Following Connell and Lynott (2013), Liu et al found that people were more likely to reject a sentence as nonsensical when the constituent words seldom co-occurred across language (e.g., illness can be bright) compared to when they co-occurred more often (e.g., supply can be bright), and additionally showed that this effect occurred independently of ease of simulation. The speed of rejection, however, was more LABELS IN CONCEPTUAL PROCESSING 21 complex.…”

“…Humans do not have infinite executive and memory resources to devote to conceptual processing, and so it would greatly enhance cognitive efficiency if there were some form of triage mechanism available that could flag up at an early stage when a particular conceptual process is likely to prove excessively difficult and is not worth further effort. Connell and Lynott (2013; propose that the linguistic shortcut can offer such a cognitive triage mechanism. Since concepts that are often discussed in the same context are likely to be easily integrated in a shared situation , linguistic distributional information from concept labels offers a guide to the likelihood of successful simulation.…”

“…As previously discussed, Connell and Lynott (2013) asked people to either judge the sensibility or provide an interpretation of novel noun-noun compounds (e.g., elephant complaint): a conceptual combination task that required creating a new representation. However, Connell and Lynott did not only examine "yes" responses that represented successful conceptual processing (i.e., responses where the compounds were judged as sensible or an interpretation was provided), but also "no" responses that would traditionally be discarded as representing unsuccessful processing of the task (i.e., responses where compounds were deemed to make no sense or have no meaningful interpretation).…”

“…Only when difficult associated false fillers were included (e.g., banana→monkey) did people begin to process the concepts more deeply, and linguistic association could no longer predict responses. Connell and Lynott (2013) took the depth of processing proposal one step further by examining the same stimuli in two related tasks: shallow sensibility judgement of novel nounnoun phrases (e.g., yes or no: does elephant complaint make sense?") and deep interpretation generation (e.g., yes or no: can you think of a meaning for elephant complaint?).…”

Classification systems offer a microcosm of issues in conceptual processing: a commentary on Kemmerer (2016)

“…A likely possibility, consistent with how other concepts are represented, is that multiple representations work together, including grounding, compressed abstractions, and distributed linguistic representations (e.g., Andrews, Frank, & Vigliocco, 2014;Barsalou, Santos, Simmons, & Wilson, 2008;Connell & Lynott, 2013;Louwerse, 2011;Paivio, 1986).…”