When the meaningless make sense: Wordlikeness and affective norms for 4,800 pseudowords and 1,200 Spanish words

Abstract

Most research using pseudowords has focused on the contribution of sublexical properties to the study of word processing. However, recent evidence suggests that pseudowords can also provide insights into the semantic and emotional aspects of language processing. Research in this field would greatly benefit from datasets providing estimations for pseudoword stimuli in the various lexicosemantic variables that have been shown to affect word recognition. Such datasets are currently lacking. In the present study, we introduce normative data for 4,800 pseudowords and 1,200 Spanish words, which were rated by 1,210 participants on three dimensions: wordlikeness, valence, and arousal. The stimuli were derived from emotional and neutral base words, and the morphological structure of the pseudowords was manipulated to create four versions combining real roots and suffixes with non-roots and non-suffixes. Additionally, we computed the normalized Levenshtein distance, the number of orthographic neighbors, and the mean Levenshtein distance to the 20 closest orthographic neighbors (OLD20) to examine the influence of objective measures on perceived wordlikeness. The results showed that the morphological structure of the pseudowords had a gradual effect on wordlikeness, valence, and arousal scores, with those combining real roots and suffixes being rated the highest. Furthermore, affective variables were found to consistently predict perceived wordlikeness ratings, whereas objective measures only accounted for a small proportion of the variance. This empirically validated set of well-controlled pseudowords is a valuable resource for researchers interested in the effects of morphology and affect on word processing. The complete database can be downloaded from: https://osf.io/baues/.