Informative use of “not” is N400‐blind

Abstract

While sentence processing is generally a highly incremental and predictive process, negation seems to present an exception to this generalization. Two-step models of negation processing claim that predicate negation is computed only after the meaning of the core proposition has been computed. Several ERP studies eliciting the N400 (an index of semantic integration or lexical expectation) have found a "negationblind" pattern of N400 results, suggesting that the negation has not been integrated into the overall sentence meaning by the time the critical word for the N400 is encountered. Recent research, however, showed that the N400 was sensitive to the negation-modulated truth value of the sentence when negation was pragmatically licensed. We investigate the possibility that negation-blind N400 is due to underinformativeness of stimuli in past experiments. We found that ERPs to simple classexclusion statements ("A hammer is not a bird") still exhibit negation blindness, even when negation is presented in a more meaningful context. The current findings provide new support for late/non-incremental interpretation of negation even when negation is pragmatically licensed.

Inhibitory Mechanisms in the Processing of Negations: A Neural Reuse Hypothesis

“…Most ERP studies found no evidence of the constraints imposed by negation on lexical processing, indicating seemingly a “clean” simulation step for the negated information. More specifically, negation seemed not to reverse (or reduce) the semantic relatedness effect that is typically obtained for the N400 component (e.g., Fischler et al, 1983 ; Kounios & Holcomb, 1992 ; Dudschig et al, 2019 ; Palaz et al, 2020 ). In fact, both affirmative and negative sentences showed the same type of relatedness effect, namely: larger N400 amplitudes for semantically unrelated than related words.…”

Multivariate pattern analysis of EEG reveals nuanced impact of negation on sentence processing in the N400 and later time windows

“…Although language processing theories to date almost unanimously argue that sentences are parsed online in an incremental fashion (Crocker, 1994;Kamide et al, 2003;Kempen and Vosse, 1989;Nieuwland and Kuperberg, 2008;Schlesewsky and Bornkessel, 2004;Stabler, 1994), such accounts are at odds when cases of negation processing are being examined. In the case of electrophysiology, event-related potentials (ERPs) literature from the past two decades has consistently reported that negation during the processing of sentences-under different contexts and in its various forms-does not have any impact on the N400 effect arising from semantic relatedness and congruency (Dudschig et al, 2018;Fischler et al, 1983;Palaz et al, 2020). These data may thus speak for the non-incremental two-step processing accounts of negation (Kaup, 2001;MacDonald and Just, 1989).…”

Negation Processing in Children with ADHD: The Generic Problem of Using Negation in Instructions

“…The specific increase of error rates in behavioral tasks following the use of negation has been of particular interest for studies investigating mental control -and thus gained influence beyond the basic language comprehension literature -and has been named the ironic effects of negation (Wegener, 1998). Second, if looking at electrophysiological data, there are a variety of studies showing that negation operators are not instantly integrated into the comprehension process (Dudschig et al, , 2019Fischler et al, 1983), or in other words that core markers of semantic processing (i.e., the N400) are negation-blind (Palaz et al 2020). Although some of the negation-related difficulties can be overcome if negation is used in a manner that sentence continuations are predictable (Nieuwland, 2016;Nieuwland, & Kuperberg, 2008), not all types of negation licensing result in full negation integration (see Palaz et al, 2020).…”

“…Second, if looking at electrophysiological data, there are a variety of studies showing that negation operators are not instantly integrated into the comprehension process (Dudschig et al, , 2019Fischler et al, 1983), or in other words that core markers of semantic processing (i.e., the N400) are negation-blind (Palaz et al 2020). Although some of the negation-related difficulties can be overcome if negation is used in a manner that sentence continuations are predictable (Nieuwland, 2016;Nieuwland, & Kuperberg, 2008), not all types of negation licensing result in full negation integration (see Palaz et al, 2020). Also, for the use of negation in imperatives -which are of particular interest for the present study -there has not been any evidence that licensing avoids the additional processing difficulties.…”