Gradations of interpretability in spoken complex word recognition

Schuster, Swetlana
Lahiri, Aditi

University of Oxford, UK

^*Corresponding author. Faculty of Linguistics, Philology and Phonetics, Clarendon Institute Oxford, University of Oxford, Oxford, OX1 2HG, UK.

E-mail address:[email protected]; [email protected]

ARTICLE INFO

Received December 2, 2018; Received in revised form April 11, 2019; Accepted April 13, 2019.

Available online 17 April 2019

Neuropsychologia 133:p 107065, October 2019. | DOI: 10.1016/j.neuropsychologia.2019.04.006

ABSTRACT

In an ERP experiment with cross-modal priming, we found evidence for graded interpretability of complex pseudowords, expressed as differences in the degree of priming between a complex pseudoword and its embedded base word. Only pseudowords that were constructed according to the morphological constraints of the language (for instance that nouns ending in {-ung} in German must be derived from a verbal base) led to significant attenuation of the N400 component relative to an unrelated control. Novel formations that consisted of the concatenation of a stem and an incompatible affix (e.g. wirr A.,‘confused’ and {-lein}, which can only add to nouns) did not modulate the N400 component. Moreover, gradations in degree of priming were observed with novel pseudowords that were the result of several morphological operations as a function of the composition of the full sequence of morphological operations. Any positions within the sequence that were filled by possible, but non-existing formations (thus compare hübsch (A., ‘pretty’) > *hübschen (V., ‘beautify’) > *Hübschung (N., ‘beautification’) versus spitz (A., ‘sharp’) > spitzen (V., ‘sharpen’) > *Spitzung (N., ‘sharpening’), whereby in the former the intermediate position is expressed through a possible, but non-existing verb form) led to reduced activation of the base word. These findings corroborate previous behavioural and neuroimaging results in visual word recognition that found support for step-wise morphological decomposition in multi-affixed pseudowords and reveal that the composition of the derivational path can lead to gradations of interpretability in the recognition of spoken complex pseudowords.

Highlights

Cross-modal electrophysiological study with different sets of complex novel forms.
Priming only obtained in novel forms that obey the structural rules of the language.
Superficial parsing of novel forms into suffixes does not suffice for priming.
The degree of priming depends on the internal composition of the complex novel form.