Positional Encoding Mechanisms of English Morphemes

<p dir="ltr">This thesis investigates positional encoding mechanisms of English morphemes, a relatively unexplored aspect of visual word recognition with implications for models of morphological processing. Collecting behavioural and neuroimaging data from children and adult readers, the experiments herein employ the morpheme transposition paradigm to clarify the developmental trajectory, encoding nuances, and spatiotemporal dynamics of positional constraints on stems and affixes. Chapter 1 provides a general introduction to relevant reading science literature with an overview of orthographic/statistical/morphological/semantic influences on visual word recognition, the key experimental methods applied in subsequent chapters, and the reading models that provide a theoretical framework for our findings. Building on our previous work confirming the free recognition of stem morphemes and the relatively position-dependent recognition of affixes, Chapter 2 reports a series of experiments that explore the age at which these fundamental distinctions are established in the reading system. Two lexical decision experiments with children ages 7-12 reveal stem and affix transposition effects resemble those found in adult readers by the end of primary school, with significant morpheme transposition effects emerging and growing more pronounced as the age of participants increased. These findings indicate that accumulated exposure to morphological regularities in text, modulated by reading proficiency, allows distinct positional constraints on morphemes to solidify as reading expertise evolves. Chapter 3 explores the case of English words with multiple affixes (e.g., bookishness), which suggest there must be further nuance to the positional constraints imposed on affixes in the reading system to facilitate cases where they occur in atypical locations but still convey meaning. Two lexical decision experiments investigate the positional encoding of mid-embedded suffixes in nonwords (e.g., bestersell from bestseller). The resulting morpheme interference effects clarify the requirements for activation of English affixes, revealing that suffixes are not as strictly positionally coded as assumed under most models of reading and can instead be recognised without edge-alignment and while preceding stems in a letter string. Three EEG experiments in Chapter 4 explore the temporal dynamics of these positional encoding mechanisms, analysing key ERP components and the influence of semantics on morpheme transposition effects. They reveal nonwords derived from English compounds (e.g., dreamday from daydream) create interference effects relative to controls across N250/N400/P600 components that are modulated by semantic transparency, while items containing affixes that violate positional constraints do not. Finally, Chapter 5 synthesises these findings and their implications for models of reading that we hope these novel insights can contribute to, along with discussion of future research directions.</p>