Semantic similarity promotes interference in the continuous naming paradigm: behavioural and electrophysiological evidence

ABSTRACT

We investigated within-category semantic distance effects in the continuous naming paradigm with reaction times (RTs) and event-related potentials (ERPs). Cumulative semantic interference and ERP effects were observed only for closely related members of basic level categories with high feature overlap (e.g. apes: orangutan, chimpanzee), indicating that shared broad semantic category membership (e.g. animals: orangutan, donkey) without considerable semantic feature overlap is insufficient to induce semantic interference. ERP modulations were characterised by an enhanced P1 at about 100–150 ms, that may reflect early co-activation of visual-conceptual feature information, and a relative posterior positivity starting at about 250 ms that was positively correlated with RTs, reflecting lexical selection. Furthermore, a posterior negativity between 450 and 600 ms was observed and associated with semantic-lexical calibration processes. These findings suggest early conceptual and lexical loci of semantic interference and underline the importance of converging activation spread triggered by shared semantic features during speech planning.

KEYWORDS:

• Language production
• semantic distance
• semantic interference
• continuous naming
• electrophysiology

Acknowledgments

We thank Meggie Danziger and Marie Borowikow for her assistance in data collection, and Guido Kiecker for technical support.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCiD

Rasha Abdel Rahman http://orcid.org/0000-0002-8438-1570

Notes

1. We also conducted an ANOVA with the factor repetition of list. Overall, naming latencies decreased significantly over repetition, F₁(2,46) = 70.3, p < .001, η_p² = .75; F₂(2,70) = 245.0, p < .001, η_p² = .87. However, as has been reported before (Costa et al., 2009; Navarrete et al., 2010; Rose & Abdel Rahman, 2016), this factor did not influence the ordinal position effect (Fs < 1). As the purpose for including repetitions was to increase the number of EEG segments, we will not discuss these data any further.

2. We also conducted an ANOVA over all 62 electrodes that supports the results of the ROI analysis, except of the time windows between 450 and 500 ms and between 550 and 600 ms that show a statistical trend in the overall analysis, p < .07.

3. Interestingly, a similar P1 modulation has been found by Costa et al. (2009, not reported) and interpreted as a repetition effect that should be strongest for the first vs. second presentation (personal communication).

Additional information

Funding

This research was supported by a grant [AB 277/ 4] from the German Research Council to Rasha Abdel Rahman.