Work

Statistical Verification of Positional Structure in Undeciphered Scripts: A Monte Carlo Approach to the Harappan Corpus

Year: 2026
Type: article
Abstract: This study applies a Monte Carlo permutation test to verify whether the positional structure of the undeciphered Harappan (Indus) script reflects a statistically significant combinatorial system rather than random distribution. A corpus of 419 sign instances from published Harappan inscriptions is analyzed for positional co-occurrence, sign-repetition patterns, and terminal-sign behaviour. For each statistic, a null distribution is generated by random permutation of sign positions across the corpus (10,000 permutations); observed values are then evaluated against this null model. The results show that the positional structure of the corpus exceeds chance expectations at statistically significant levels, supporting the hypothesis that the Harappan sign system encodes an organized (likely linguistic) information structure. The method is language-independent by design: it makes no assumptions about the underlying language, script direction, or sign phonetic values, and therefore can be applied to any undeciphered corpus. The procedure is fully replicable: all statistics, permutation parameters, and significance thresholds are documented. This paper presents the methodological framework and results; implications for sign identification and corpus segmentation are discussed. The approach offers a quantitative foundation for future decipherment attempts, replacing manual analogy-based comparisons with a falsifiable statistical standard.
Language: English
Open Access status: green
OpenAlex ID: W7232836348
FWCI:
Cites:
Cited by:
Related to: 10