Presentation

Dave Kenneth Cayado¹, Samantha Wray^2,3, Linnaea Stockall¹; ¹Queen Mary University of London, ²Dartmouth College, ³New York University Abu Dhabi

Previous studies from various languages have shown that morphologically complex words are subject to early, form-based decomposition mechanisms, taking place within the first 200ms of word recognition (M170) and localized in the left fusiform gyrus, (the Visual Word Form Area – VWFA). Gwilliams and Marantz (2016) developed a functional localizer for this response for English, building on original work in Finnish, which contrasts words and symbol strings embedded in varying levels of Gaussian noise. The current study has two objectives: (1) validate a functional localizer specific to Tagalog, in order to allow (2) investigations of VWFA-based morphological decomposition as modulated by phonological alternations and morphophonological variability. METHODS: 18 right-handed native Tagalog speakers (aged 19-52, X̄ = 35.46) have so far participated in two tasks, with concurrent MEG recording using a 1000 Hz sample rate on a 208-channel axial gradiometer system: the localizer, lasting 10 min, a passive looking task with 4 stimulus types (i.e., a 4-letter Tagalog word, a single letter, a symbol, or a string of symbols, either unmasked, or masked by Gaussian noise level 24), and a lexical decision task, lasting 20 min, with three conditions, all involving the prefixes paN- or maN-: (a) No-Change; words that do not exhibit phonological alternations (e.g., paN + halo ‘mixer’ = paNGhalo); (b) Nasal-Assimilated: the nasal N is pronounced as /m/ , /n/ or /ŋ/ to match the place of articulation of the stem initial obstruent (e.g., paN + pook ‘district’ = pampook); (c) Nasal-Substituted: the prefix-final nasal and the stem-initial obstruent are substituted by a single phoneme that is homorganic to the original obstruent (e.g., maN + palo ‘slap’ = mamalo). Nasal substitution makes the boundary between the prefix and stem opaque, and whether a stem triggers Assimilation or Substitution is not entirely predictable, but Substitution is more likely to occur for voiceless-initial than voiced-initial stems. We predict that the effects of morpheme-boundary opacity and the predictability of the prefix allomorph should be detected as variance in the M170. LOCALIZER RESULTS: we conducted a two-stage regression analysis in which regressions were fit at each time point and source point per-subject for factors of String Type (symbols, letters) and Stimulus Type (1-character string, 4-character string). Spatio-temporal cluster-based permutation tests were conducted in the bilateral occipitotemporal region over a 130–180ms time window. We found a type II response for letter/word vs. symbol in the medial surface of the lingual gyrus that is approaching significance (p = 0.059), with higher activation for letter/word than symbol, thereby replicating the canonical activation direction previously found in English, Greek, and Finnish. However, this response is in the right hemisphere, contrary to expectation. Given that the right-hemisphere VWFA analogue has been implicated in morphological processing (Zweig and Pylkkanen 2009), the region identified in the current study will serve as an fROI for the M170 analysis of the lexical decision task, aiming to shed light into how phonological rules influence early, form-based, morphological decomposition.

Topic Areas: Morphology, Reading