Hide/Show Apps

The Effects of cross-morphemic letter transpositions on morphological processing in turkish: a psycholinguistic investigation

Çağlar, Ozan Can
This study investigates whether Turkish native speakers have access to semantic information in the course of morphological decomposition at the early stages of visual word recognition. Two masked priming experiments were conducted to test the effects of semantic transparency on the recognition of target words. The main prime conditions of the study were the following: (a) semantically transparent (e.g., çizim-ÇİZ, Eng. drawing-DRAW), (b) semantically opaque (e.g., tuzak-TUZ; Eng. trap-SALT), and (c) form overlap (e.g., kasap-KAS; Eng. butcher-MUSCLE). Transparent pairs were both semantically and morphologically related whereas opaque pairs shared no semantic but a pseudo-morphological relation. Form overlap pairs displayed overlapping orthographic features only. The letter order/identity of the primes were also manipulated at the morpheme boundary for each condition (e.g., transposed-letter primes: çiizm-ÇİZ, replaced-letter primes: çiurm-ÇİZ) to see how cross-morphemic transpositions would inform the debates on the role of semantic information in the early processing. The results showed significant priming effects of both semantically transparent and opaque forms. However, it turned out that opaque forms revealed no priming effect when they included letter transpositions at the morpheme boundary. The significant priming effects obtained from transparent forms, on the other hand, were not decreased by cross-morphemic transpositions as dramatically as that obtained from opaque forms. The findings contest the form-first account which supports the view that the early processing of morphologically complex forms is blind to semantic information. The observed priming effect patterns were consistent with the predictions of dual-route models of morphological processing, which assume parallel activation of morpho-orthographic and morpho-semantic information.