When Wine and Apple Both Help the Production of Grapes: ERP Evidence for Post-lexical Semantic Facilitation in Picture Naming

Abstract

Background: Producing a word in referential naming requires to select the right word in our mental lexicon among co-activated semantically related words. The mechanisms underlying semantic context effects during speech planning are still controversial, particularly for semantic facilitation which investigation remains under-represented in contrast to the plethora of studies dealing with interference. Our aim is to study the time-course of semantic facilitation in picture naming, using a picture-word “interference” paradigm and event-related potentials (ERPs).Methods: We compared two different types of semantic relationships, associative and categorical, in a single word priming and a double word priming paradigm. The primes were presented visually with a long negative Stimulus Onset Asynchrony (SOA), which is expected to cause facilitation.Results: Shorter naming latencies were observed after both associative and categorical primes, as compared to unrelated primes, and even shorter latencies after two primes. Electrophysiological results showed relatively late modulations of waveform amplitudes for both types of primes (beginning ~330 ms post picture onset with a single prime and ~275 ms post picture onset with two primes), corresponding to a shift in latency of similar topographic maps across conditions.Conclusion: The present results are in favor of a post-lexical locus of semantic facilitation for associative and categorical priming in picture naming and confirm that semantic facilitation is as relevant as semantic interference to inform on word production. The post-lexical locus argued here might be related to self-monitoting or/and to modulations at the level of word-form planning, without excluding the participation of strategic processes.

A novel multi-word paradigm for investigating semantic context effects in language production

“…At longer negative SOAs (-1000 to -400ms), the effect transformed into semantic facilitation-using the same stimulus materials. Similarly Python et al [38] find facilitation from categorically and associatively related distractor words at an SOA of -400ms. These findings indicate that at longer SOAs, conceptual priming outweighs lexical competition.…”

“…As opposed to typical single-word PWI paradigms, in our novel paradigm a negative stimulus onset asynchrony (SOA) of 6 s was used, which is much longer than in the typical single-distractor-word paradigms. Indeed, previous studies have shown that SOAs of -1000 ms or -400 ms led to facilitation rather than semantic interference for categorical distractor words presented prior to the picture [36,38]. We used the long SOA to ensure that each word, especially from the categorically related distractor words, was fixated and processed.…”

“…Timing in the PWI paradigm has been shown to greatly affect naming speed: Prominently, the interval between a distractor word and target (the stimulus onset asynchrony, SOA) influences the polarity of the context effect [5,[36][37][38]. Manipulating the SOA systematically with different time intervals, Zhang et al [36] demonstrated that a semantic interference effect from categorically related word distractors only occurred at an SOA of -100ms before, or of 0ms, that is simultaneously to, target onset.…”

A critical review of the behavioural, neuroimaging, and electrophysiological studies of co-activation of representations during word production

“…The results can thus be read as a more negative N400 in the presence of a second co-activated representation. When comparing the nature of the activated representations, findings consistently point to a lower-amplitude N400-like waveform for related compared to unrelated distractors, both when the effect of semantic distractors has been interference (Blackford, Holcomb, Grainger, & Kuperberg, 2012;Dell'Acqua et al, 2010;Rose et al, 2019;Wong et al, 2017) and when it has been facilitation due to the long negative SOA (Python et al, 2018b).…”

“…These opposing effects appear to have different timelines: The facilitation induced by the non-categorical relationships is most robust in early stimulus onset asynchronies (SOAs), when the distractor is presented -450 to -100 ms before the target (or even earlier; see Python, Fargier, & Laganaro, 2018b). In comparison, the interference induced by the categorically related items is more robust when the distractor is presented around the same time as the target, i.e., SOA of -200 ms to 150 ms (Alario et al, 2000;La Heij, Dirkx, & Kramer, 1990;Lupker, 1979).…”