Subliminal Emotional Words Impact Syntactic Processing: Evidence from Performance and Event-Related Brain Potentials

Abstract

Recent studies demonstrate that syntactic processing can be affected by emotional information and that subliminal emotional information can also affect cognitive processes. In this study, we explore whether unconscious emotional information may also impact syntactic processing. In an Event-Related brain Potential (ERP) study, positive, neutral and negative subliminal adjectives were inserted within neutral sentences, just before the presentation of the supraliminal adjective. They could either be correct (50%) or contain a morphosyntactic violation (number or gender disagreements). Larger error rates were observed for incorrect sentences than for correct ones, in contrast to most studies using supraliminal information. Strikingly, emotional adjectives affected the conscious syntactic processing of sentences containing morphosyntactic anomalies. The neutral condition elicited left anterior negativity (LAN) followed by a P600 component. However, a lack of anterior negativity and an early P600 onset for the negative condition were found, probably as a result of the negative subliminal correct adjective capturing early syntactic resources. Positive masked adjectives in turn prompted an N400 component in response to morphosyntactic violations, probably reflecting the induction of a heuristic processing mode involving access to lexico-semantic information to solve agreement anomalies. Our results add to recent evidence on the impact of emotional information on syntactic processing, while showing that this can occur even when the reader is unaware of the emotional stimuli.

Isolating the Effects of Word’s Emotional Valence on Subsequent Morphosyntactic Processing: An Event-Related Brain Potentials Study

“…This is true even regardless of the manner of processing the word, i.e., whether it was normally (RA task) or hardly read, that is, with content access inhibition and conflict, among other factors (ES task). Overall, the finding would reveal that emotional information is powerful enough as to affect syntactic processing even when processed in impoverished circumstances, in line with previous studies with subliminal presentations of emotional material within the sentence ( Jiménez-Ortega et al, 2017 ).…”

“…Thus, evidence supports that emotional information influences subsequent non-emotional tasks, including language processing (e.g., Gupta and Raymond, 2012 ) and, particularly, sentence-morphosyntactic processing (e.g., Jiménez-Ortega et al, 2012 , 2017 ; Verhees et al, 2015 ). This has been a highly relevant finding, given that syntax has been traditionally described as automatic, modular and encapsulated, blind to other processes ( Hauser et al, 2002 ; Friederici, 2004 ).…”

“…Some studies have explored the impact of emotional information on morphosyntactic processing when emotion is enclosed within the sentence to be syntactically analyzed (short-term effects) ( Martín-Loeches et al, 2012 ; Jiménez-Ortega et al, 2017 ). Overall, these studies have reported differential effects as a function of valence, and the results have been related to divergent cognitive strategies induced by emotions during morphosyntactic processing.…”

Neural capacity limits during unconscious semantic processing

“…It is important to note that the observed N400 effect in visual masking relies on different neural mechanisms from those observed in earlier studies using the attentional blink (Batterink & Neville, ) and inter‐ocular suppression (Axelrod et al ., ), as the extent of neurocognitive processing elicited by unconscious stimuli is known to vary depending on the nature of experimental procedures used to render stimuli invisible (Raymond et al ., ; Fogelson et al ., ). Specifically, activation reduction in early visual cortex has been shown to play a key role in the disruption of subjective awareness during attentional blink (Williams et al ., ; Hein et al ., ) and CFS (Jimenez‐Ortega et al ., , ). By contrast, briefly flashed words under visual masking can produce strong feedforward activation of the occipitotemporal cortex as well as weak and transient activation of the frontoparietal cortex (Dehaene et al ., ; Kouider & Dehaene, ).…”

Positive emotional reactions to loved names

“…Emotional words have the same properties of relevant names in terms of visual format and structure and lack of unique intrinsic properties. This affective written stimulus can modify syntactic processing even when the subject is unaware of its content (Jiménez‐Ortega, Espuny, de Tejada, Vargas‐Rivero, & Martín‐Loeches, ). Their emotional value comes from the element to which it is referring.…”