Doctoral Dissertation Research: The constraints of dual morphological systems on visual word processing in Maltese

University Of Arizona, Tucson AZ

Investigators

Abstract

All languages have strategies for combining independently meaningful linguistic units, called morphemes, to build words. Most languages, like English, primarily combine these units by linearly appending them to one another (e.g. re- + write -> rewrite); such languages are said to have linear morphology. Semitic languages, like Arabic, are famous for having nonlinear morphology: discontinuous units which express meaning and grammatical information are combined non-linearly (e.g. Arabic k-t-b 'writing' + aa-i 'doer, NOUN' -> kaatib 'author'). Maltese is particularly unusual in its morphological characteristics: as a Semitic language, Maltese has primarily used non-linear morphology historically. However, for the past millennium it has been in close contact with a series of languages that use linear morphology (Sicilian, Italian, French, and English), and speakers have heavily borrowed words and adopted linear morphological strategies from these languages. Today, Maltese comprises a "mixed" language: at least half of all words are borrowed from non-Semitic languages (and so do not exhibit non-linear morphology), and speakers use both linear and non-linear word-building strategies extensively. This project investigates how the presence of distinct morphological systems in Maltese affects word processing, exploring differences in how Maltese readers recognize Semitic words vs. non-Semitic borrowings. It explores the impact of historical language mixing on the processes underlying daily language use, and so has implications for the study of languages in similar contact situations. The results of a series of experiments will comprise the data analyzed in the Co-PI's dissertation. The project will also produce a database of Maltese word recognition data, the first such database for a language with non-linear morphology, which other researchers may use to study language processing. The Co-PI, a doctoral student at the University of Arizona, will conduct a series of psycholinguistic experiments investigating how the existence of distinct lexicons exhibiting unique morphological characteristics (native Semitic words vs. non-Semitic borrowings) affects visual word recognition in Maltese. Previous research has found (1) that Maltese readers are faster to recognize Semitic words than non-Semitic borrowings and (2) that, when primed by letter strings consisting of a subset of a word?s consonants (e.g. frx primes FIREX 'to spread'), they are faster to recognize Semitic words (for which such strings comprise a morpheme) but not non-Semitic ones (for which such strings are non-morphemic). The latter finding conflicts with the "subset priming" effect found in non-Semitic languages, namely that exposure to such letter strings typically facilitates word recognition (e.g. 'csn' primes CASINO), owing to the role of consonant letters in constraining lexical candidates, and suggests that language-specific morphological properties such as Semitic nonconcatenative morphology may further shape the role of consonant letters in lexical processing. The proposed research includes (1) a Maltese visual lexical decision megastudy which explores processing differences between Semitic and non-Semitic words across a wide range of items when potential confounding factors like orthographic neighborhood density and morphological structure are controlled for (and which will result in the publication of a database of Maltese lexical decision responses) (Experiment 1); and (2) a series of visual masked priming studies which test for subset priming effects in non-Semitic Maltese words when confounding factors, namely the morphemic status of the prime, are controlled for (Experiments 2-3). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.