Introduction

1.1. Theoretical Background

The concept of core vocabulary has been central to theories of language acquisition (Bloom, 2000), frequency-based learning (Ellis, 2002), and psycholinguistic modeling (Brysbaert & New, 2009). Core words are assumed to be the first acquired, most stable, and resistant to external change (Swadesh, 1955; Pagel et al., 2007). These words form the backbone of communication, appearing across diverse contexts and genres. However, core vocabulary is not synonymous with uniformity. Historical linguists such as Trask (1996) and Campbell (2004) note that even basic lexicon is subject to variation due to phonological erosion, semantic bleaching, and analogical restructuring. For instance, irregular verbs like go/went or be/was/were persist not because of systematicity but due to entrenched usage and frequency (Bybee, 2001).

Lexical divergence refers to the internal variation and transformation of native vocabulary over time, shaped by phonological shifts, semantic drift, morphological irregularities, and syntactic reclassification. While much linguistic research focuses on external lexical influences – such as borrowings from Latin, French, or global Englishes – comparatively little attention has been devoted to the divergence occurring within a language’s core vocabulary. These words, which are neither borrowed nor peripheral, but foundational to communication, often display unexpected irregularities that arise due to historical, social, and cognitive pressures (Bybee, 2001; Geeraerts, 1997; Traugott & Dasher, 2001).

The English language, a member of the Germanic family, retains a significant set of inherited lexemes from its Anglo-Saxon roots. This core includes high-frequency verbs (go, eat, see), concrete nouns (man, sun, night), and basic adjectives (good, old, hard) that form the backbone of everyday discourse (Nation, 2001; Swadesh, 1955). Despite their deep historical roots and relative resistance to borrowing, these words are not immune to change. Divergence occurs as they evolve in usage, pronunciation, meaning, or grammatical function across time and dialects. For example, the word girl, which once meant a child of either gender, has narrowed semantically to denote a female child or young woman. Similarly, silly has undergone semantic degradation, shifting from “happy” or “blessed” to “naive” or “foolish” (Blank, 1999).

Phonological divergence is particularly visible across regional dialects. Words such as route, either, and schedule are pronounced differently in American and British varieties of English, often reflecting deeper sociolinguistic identities (Wells, 1982; Labov, 1994). These phonetic variants, although seemingly minor, raise important considerations for standardization, second language instruction, and perceptions of linguistic correctness. Morphological divergence is evident in irregular forms that persist despite regular paradigms: go → went, foot → feet, and child → children. These examples demonstrate the tension between historical retention and contemporary system regularity (Plag, 2003; Lieber, 2004).

Syntactic divergence further underscores the flexibility of core vocabulary. Words such as text, run, and light have shifted grammatical categories over time through a process of zero derivation, now functioning as both nouns and verbs without affixation (Bauer & Valera, 2005). This fluidity contributes to the productivity of English but also complicates parsing, teaching, and lexical categorization.

This paper investigates phonological, morphological, semantic, and syntactic divergence among core English words using both empirical and theoretical lenses. We adopt a hybrid framework inspired by computational linguistic evaluation methods – such as the Analytical Hierarchy Process (AHP) and regression-based pattern recognition – to model and assess divergence systematically. While such frameworks are often reserved for second language acquisition or external borrowing studies, we apply them innovatively to native lexicon analysis.

The objectives of this study are:

• To identify and categorize divergent elements in core English vocabulary;
• To apply a ranked model based on multidimensional linguistic feature analysis;
• To present comparative data tables and examples illustrating divergence;
• To contribute a hybrid methodology for evaluating internal variation in native lexicon.

Additionally, the study explores the relationship between lexical frequency and semantic shift, and examines how syntactic reclassification affects teaching, lexicography, and cognitive processing. A comparative evaluation of traditional versus computationally supported models will highlight the utility of combining descriptive linguistics with structured analysis tools.

To guide the study, we formulated the following research questions:

1. What types of divergence are most frequent among core English words?
2. How can we categorize and quantify divergence using a structured linguistic model?
3. What are the implications of these divergences for teaching, lexicography, and language processing?

Methodology

2.1. Research Design

This study adopts a qualitative and descriptive linguistic research design that emphasizes interpretative depth over numerical generalization. The objective is to identify, examine, and interpret divergent elements within core English vocabulary – particularly those that have shifted phonologically, morphologically, semantically, or syntactically across time or regional varieties.

Qualitative approaches are widely recognized in linguistic research aiming to explain processes of variation and language change through detailed textual and contextual analysis (Creswell, 2014; Dörnyei, 2007). This study aligns with the work of Aitchison (2012), who explores the evolving mental lexicon, and Labov (1994), whose sociolinguistic framework highlights internal variability in language.

We integrate traditional descriptive linguistic methods with computationally inspired frameworks to enhance explanatory rigor. Specifically, we adapt the Analytical Hierarchy Process (AHP) and comparative feature-ranking models to evaluate the relative importance and interaction of divergent features. Though not implemented in full statistical form, AHP principles are applied to prioritize divergence types based on linguistic frequency and functional impact (Bosshard et al., 2021; Jatnika et al., 2019).

2.2 Data Collection

A stratified sample of 50 high-frequency native English words was compiled from the British National Corpus (BNC) and the Corpus of Contemporary American English (COCA). Selection excluded known borrowings, ensuring focus on native lexemes with historical continuity from Old and Middle English. The sample was balanced across lexical categories: 20 nouns, 15 verbs, 10 adjectives, and 5 function words. Particular attention was given to words showing divergence in pronunciation, inflection, semantic range, or syntactic role. This approach draws on guidelines from Crystal (2003), the Oxford English Dictionary, and Etymonline databases.

Divergences were coded using categorical markers:

• P – Phonological
• M – Morphological
• S – Semantic
• G – Grammatical/Syntactic

2.3 Reliability and Validity

To ensure the reliability of the analysis, all coded features were independently reviewed by two trained linguists specializing in phonological and syntactic analysis. Each researcher worked separately to classify divergence types using the defined coding system (P, M, S, G), with coding sheets kept blind to each other’s input during the first round. Inter-rater reliability was assessed using Cohen’s Kappa (κ = 0.86), indicating substantial agreement (Landis & Koch, 1977). For example, both reviewers consistently marked go and be as morphologically divergent due to their suppletive forms, while the word light raised initial discrepancies due to its overlapping semantic and syntactic categories. Disagreements were resolved through collaborative discussion and cross-referencing with authoritative linguistic resources such as the Oxford English Dictionary (OED), the Cambridge Grammar of the English Language, and historical corpus data.Triangulation was applied to enhance validity. Usage frequency and collocational data were retrieved from both BNC and COCA, while phonological and morphological properties were validated across multiple linguistic resources, including Wells’ (1982) dialectal phonology study, the Cambridge Pronouncing Dictionary, and Etymonline’s etymological records. For instance, phonetic variants of route (as /ruːt/ in the UK and /raʊt/ in parts of the US) were cross-checked using both IPA notation and audio samples from regional corpora. Similarly, the etymological development of silly from "blessed" to "naive" was traced through diachronic entries in the OED.

2.4. Corpus Overview

The British National Corpus and the Corpus of Contemporary American English together span over 900 million words, offering balanced datasets representing both British and American English across spoken and written genres. For this study, we focused on:

• BNC Spoken Subset: 10 million words reflecting informal, everyday speech in British English, including regional dialects and transcribed conversations.
• COCA Spoken and Magazine Subsets: 30 million words each, providing a combination of casual American spoken English and edited journalistic prose, ensuring lexical diversity across registers.

All selected words had a frequency threshold of at least 5,000 occurrences across both corpora, making them statistically representative of core vocabulary. Additional selection criteria required that each word appear across at least three different genres (e.g., academic, conversational, fiction), to ensure broad functional use. For instance, the word run appears commonly in sports commentary, medical texts (e.g., "run a test"), and everyday conversation, highlighting its polysemous and syntactically flexible nature. This high dispersion confirmed its core vocabulary status.

The corpora also allowed investigation of context-sensitive patterns. For example, text showed increased verbal use post-2000, particularly in spoken registers, reflecting technological influence on syntactic reclassification. Such patterns would not be detectable through dictionary-based approaches alone, underscoring the value of corpus triangulation.

2.5. Coding Summary Table

Table 1. Sample Divergence Feature Coding

 (1 = divergence present; 0 = no divergence)

The coding system enables transparent comparison of divergence types across core lexemes. Phonological divergences were generally region-specific, as seen in route and aunt, whereas morphological irregularities (e.g., child → children) often reflected historical retention. Syntactic flexibility, such as run functioning as both noun and verb, revealed dynamic category transitions, frequently correlated with semantic generalization.

Results

Phonological variation

Phonological variation refers to differences in pronunciation across dialects, regions, or social groups, often influenced by historical sound changes and standardization practices. In English, many core words – those in frequent use and of native origin – exhibit considerable variation in their phonological realization.

Table 1. presents examples of phonological divergence:

These phonological differences are well-documented in dialectology and sociolinguistics literature (Wells, 1982; Labov, 1994). The influence of regional identity, education, and prestige are crucial factors in maintaining variation. Additionally, global communication and digital media continue to blur some of these distinctions while reinforcing others.

Morphological Irregularities

Core English words often retain irregular morphological patterns, largely due to their deep historical roots. These patterns deviate from the modern productive morphological rules of regular verb conjugation and noun pluralization.

Table 2. illustrates several core words with morphological irregularities:

These forms resist regularization due to frequency and entrenchment in language use (Bybee, 2006). High-frequency usage preserves irregular forms over time, a phenomenon documented in cognitive linguistics and diachronic morphology (Lieber, 2009; Haspelmath & Sims, 2010).

Semantic Shift

Semantic divergence is one of the most dynamic types of lexical change. Core words often undergo shifts in meaning – either narrowing, broadening, amelioration (positive shift), or pejoration (negative shift).

Table 3. provides examples of semantic shifts in native English words:

Such changes highlight the relationship between word meaning and socio-cultural evolution (Traugott & Dasher, 2002). Semantic shifts are especially influenced by metaphor, euphemism, slang, and discourse context.

 Syntactic Flexibility

English is known for its syntactic economy and flexibility. Core words can shift between grammatical categories through zero derivation (conversion) without morphological change. This functional expansion reflects the productivity of English syntax.

Table 4. includes examples of syntactic category shifts:

These changes are widely documented in English corpora and grammar texts (Huddleston & Pullum, 2002). They illustrate how syntactic class can be flexibly assigned based on context and usage frequency – a characteristic feature of analytic languages.

Conclusion

The analysis of divergent elements within core English vocabulary has demonstrated that even the most fundamental, non-borrowed words are not linguistically static. Across phonological, morphological, semantic, and syntactic dimensions, divergence emerges as a natural outcome of linguistic evolution, functional need, and sociocultural influence. This study provided empirical evidence that core words exhibit variability that challenges assumptions of regularity and uniformity in native vocabulary.

Phonological divergence was shown to arise both inter-dialectally and intra-dialectally. Examples such as route (/ruːt/ vs. /raʊt/) and either (/ˈiːðər/ vs. /ˈaɪðər/) illustrate the role of regional, social, and register-based variation. These phonetic differences are not merely stylistic but reflect underlying sociolinguistic identities and shifting prestige norms.

Morphological irregularities – such as go → went and foot → feet – persist due to entrenchment and high token frequency. These forms resist analogical leveling and preserve historic inflectional patterns that would otherwise be lost in a purely rule-based system. Their survival exemplifies Bybee’s (2006) theory of frequency-driven resistance to change.

Semantic divergence, as seen in silly (from “happy” to “foolish”) and meat (from “food” to “animal flesh”), revealed how sociocultural forces and metaphorical extensions drive shifts in word meaning. These changes are deeply tied to discourse practices and shifts in conceptual framing across generations and domains.

Syntactic reclassification further highlighted the dynamism of core vocabulary. Through zero derivation, words such as text, Google, and run transition fluidly between grammatical categories. This functional elasticity showcases English’s analytic structure and raises considerations for parsing, teaching, and computational modeling.

The study introduced a hybrid approach that combined descriptive linguistic methods with a computational mindset. By applying a weighted divergence framework inspired by the Analytical Hierarchy Process, we quantified divergence patterns and identified their relative prominence. This approach, typically reserved for structured decision-making or applied linguistics, proved effective for analyzing internal variation in a qualitative domain.These findings underscore the importance of re-evaluating how core vocabulary is treated in linguistics, lexicography, and education. Far from being monolithic or rule-bound, core words are vibrant, evolving elements shaped by usage, context, and communicative need. Their divergence is not an exception but a defining feature of their centrality in the lexicon.Future research may expand this model to diachronic corpora, compare divergence across languages in the Germanic family, or use AI-powered tools to automate feature classification on a larger scale. Sociolinguistic dimensions – such as age, gender, and regional identity – could also be incorporated to better understand how divergence patterns vary across communities.Ultimately, this research contributes to a more nuanced understanding of language change, core lexicon dynamics, and the mechanisms that sustain linguistic complexity. Even the most common words carry within them the imprints of history, interaction, and adaptation – offering ongoing insight into how language lives and evolves.