Author = Nasri, Abbas
Number of Articles: 2
Distributed Morphology, Derivational Modularity and the Problem of Circularity

Distributed Morphology, Derivational Modularity and the Problem of Circularity

Volume 16, Issue 2, March 2026, Pages 267-289

https://doi.org/10.22059/jolr.2026.401945.666937

Abbas Nasri

Abstract This article aims to defend the Parallel Morphology Hypothesis as the theoretical foundation of the Optimal Distributed Morphology (ODM) framework. Unlike the standard rule-based model of Distributed Morphology (DM), which assumes a rigid, feed-forward derivational architecture comprising sequential morphosyntactic modules, the parallel hypothesis posits that all post-syntactic operations occur simultaneously within a constraint-based evaluative system. Building upon insights from Trommer (2001a, 2001b) and Foley (2017), the article argues that adherence to the Derivational Architecture Principle (DAP) in classical DM leads to a form of circular reasoning when analyzing irregular inflectional forms. Specifically, decisions regarding operations such as feature bundle fusion or fission often presuppose knowledge of which Vocabulary Items will be inserted. A knowledge that, in a strictly sequential model, should not yet be accessible. Using representative data from languages such as English and Basque, the article demonstrates that attempts to resolve this paradox within a strictly modular, rule-based DM architecture require extensive stipulations and ad hoc rules. These additions not only undermine theoretical elegance but also compromise the model’s psychological and computational plausibility. In contrast, adopting a constraint-based approach grounded in Optimality Theory (OT), as formalized in the ODM framework, allows for a more economical and predictive analysis. Here, the derivational output is determined through the competition of candidate structures evaluated by violable constraints, rather than deterministic rule application.A constraint-based analysis of irregular plural formation (e.g., mice vs. mouses) is provided, incorporating three key constraints: *UNDERSPEC, *STRUCT, and CRSP-D. These constraints regulate under-specification, structural economy, and faithfulness to input feature bundles, respectively. Through a formal ranking procedure (Recursive Constraint Demotion), the paper shows how ODM naturally accounts for both regular and irregular morphology without recourse to excessive derivational opacity or lexical idiosyncrasies. The article concludes that the Parallel Morphology Hypothesis not only offers a more descriptively adequate account of inflectional irregularities but also aligns more closely with broader cognitive and psycholinguistic principles of linguistic architecture

Electrophysiological Evidence of Neurological Representations of the Phonological and Phonetic Properties of Persian Vowels in the Auditory Cortex

Electrophysiological Evidence of Neurological Representations of the Phonological and Phonetic Properties of Persian Vowels in the Auditory Cortex

Volume 11, Issue 2, February 2021, Pages 223-245

https://doi.org/10.22059/jolr.2021.313724.666663

Abbas Nasri, Gholamhosain Karimi doustan

Abstract How the brain encodes the speech acoustic signal into phonological representations is a fundamental question for the neurobiology of language. The following paper is aimed to investigate the relationship between the phonological and phonetic properties of Persian simple vowels and neurophysiological events corresponding to them. To achieve such goal, we employed electroencephalography to map the Persian vowel system onto cortical sources using the N1 auditory evoked component. We found evidence that the N1 is characterized by asymmetrical indexes in the auditory areas of the cortex, structuring vowel representations. Properties of these ERPs were analyzed and modelled on one hand by the landmarks in the spectral window of their respective stimulus (such as F1, F2 and F2-F1) and on the other hand by the phonological distinctive features categorizing them (namely, height and place). The results revealed that the responses contain at least two distinguishable modulations of N1 components: a symmetric N1a which peaked between 113 to 149 milliseconds after the onset of the stimulus and a heavily left-leaning N1b which peaked between 149 to 170 milliseconds thereafter. Both N1a and N1b subcomponents showed strong correlations with a variety of parameters of both phonological and acoustic nature of the respective stimuli. However, N1a was significantly better modelled by acoustic factors while N1b displayed a better fit to a model based on phonetic factors. Based on such results, this paper argues that firstly the perceptual procedure of vowel categorization is a gradient process starting from demarcation of the stimulus signal according to acoustic landmarks which is done almost symmetrically then the processing load shifts significantly to the left hemisphere for the categorization of the input based on its perceived distinctive features. And secondly, that such information can be exploited to draft a ‘tonochronic’ map of such perceptual processes and define a perceptual field for every vowel and distinctive feature in the tonochronic space.