Exploring Tamil–Korean linguistic parallels: A computational and historical analysis of possible Pre-Hangul contact

This study investigates potential linguistic parallels between the Dravidian language Tamil and the Koreanic language Korean, evaluating whether observed similarities may reflect typological convergence, lexical coincidence, or indirect historical contact through maritime trade networks before the creation of Hangul in 1443 CE. The analysis compares phonological systems, consonant–vowel organization, morphological features, orthographic (letter stroke) patterns, and lexical correspondences, while applying computational string-similarity metrics to a dataset of 100 Korean–Tamil vocabulary pairs. Phonological inventories and historical script forms were compiled from established linguistic sources. Lexical similarity was measured using six computational models: Damerau–Levenshtein distance, Jaro similarity, Longest Common Subsequence (LCS), Cosine similarity, Jaccard similarity, and Ratcliff–Obershelp similarity. Hierarchical clustering (UPGMA) was employed to classify similarity levels among lexical pairs. The analysis produced the following scores: 0.5736 (Damerau–Levenshtein), 0.2255 (Jaro), 0.6001 (LCS), 0.4683 (Cosine), 0.3441 (Jaccard), and 0.5716 (Ratcliff–Obershelp), yielding an overall average similarity of approximately 56% across the dataset. Clustering results further identified groups of high, moderate, and low similarity. Both languages exhibit typological features commonly associated with agglutinative systems, including suffix-based morphology, Subject–Object–Verb (SOV) word order, and consonant–vowel syllable organization, as well as limited resemblances in orthographic stroke patterns.