Comparing the valence, emotionality and subjective familiarity of words in a first and a second language

ABSTRACT

Extant research has compared the processing of affectively laden words between L1 and L2. However, most studies used verbal stimuli that were validated for a single language or for both languages but using independent samples. We systematically compared ratings of valence, emotional intensity and subjective familiarity of negative, neutral, positive and taboo words, presented in L1 (European-Portuguese) and their equivalent in L2 (English), produced by the same individual. All participants (n = 230) were native European Portuguese speakers that self-reported being fluent in English (M_{Age of acquisition}= 8.62 years old, SD = 2.94). Most participants (55.1%) reported having learned English in a formal context (i.e. school). As expected, words in L1 (vs. L2) were rated as more familiar and extreme in valence. Surprisingly, higher emotional intensity ratings in L1 (vs. L2) were only observed for taboo words. These findings contribute for the bilingualism research by emphasizing that differences in the affective processing of different languages may be more noticeable in specific evaluative dimensions (e.g. valence) or specific word types (e.g. taboo words). Subjective norms for the full set of 640 words (evaluated by a sample ranging from 26 to 32 participants) are available at https://osf.io/va2tj/

KEYWORDS:

• L1
• L2
• negative words
• neutral words
• positive words
• taboo words
• valence
• emotional intensity
• subjective familiarity
• normative ratings

Acknowledgements

The authors would like to thank Ana Domingos for her help in data collection.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 The initial sample included 248 participants. However, 14 participants were excluded because they reported a different nationality or dominant language and four participants were excluded due to the high percentage of missing cases.

2 English proficiency was assessed by asking the participants (European Portuguese native speakers) to report how fluent they were in speaking, understanding, reading and writing in English (from 1 = Not fluent at all to 5 = Very fluent). The four items were combined in a single proficiency index (Cronbach’s alpha = .92).

3 Percentage of translations = [(number of participants that evaluated a given word * 100)/number of translations provided]; Percentage of correct responses = [(number of translations provided – errors) * 100/number of translations provided].

4 Mauchly’s test for the main effect of word type, χ²(5) = 129.12, p < .001 (ε = .78), and for the interaction between language and word type, χ²(5) = 48.01, p < .001(ε = .88 for the interaction).

5 Mauchly’s test for the main effect of word type, χ²(5) = 119.20, p < .001 (ε = .74), and for the interaction between language and word type, χ²(5) = 36.31, p < .001 (ε = .90).

6 Mauchly’s test for the main effect of word type, χ²(5) = 42.49, p < .001 (ε = .90), and for the interaction between language and word type, χ²(5) = 127.68, p < .001 (ε = .71).

Additional information

Funding

Part of this research was funded by Fundação para a Ciência e Tecnologia (PTDC/MHC-PCN/5217/2014) and by a Marie Curie fellowship (FP7-PEOPLE-2013-CIG/631673) awarded to the first author.