Glottochronology is a controversial linguistic method developed by Morris Swadesh in the 1950s, aimed at estimating the time when two related languages diverged from a common ancestor. It is based on the principle of lexicostatistics, which posits that the core vocabulary of a language changes at a relatively constant rate over time.

At the heart of glottochronology is the 'Swadesh list,' a standardized list of 100 or 200 basic vocabulary items (e.g., 'I,' 'you,' 'hand,' 'water,' 'sun') that are considered to be resistant to borrowing and cultural change. The method involves comparing the percentage of cognates (words derived from a common ancestral word) in the Swadesh lists of two related languages. A mathematical formula is then applied, which assumes a constant rate of retention for these core vocabulary items (typically estimated at around 86% per millennium for the 100-word list).

The formula used is t = (ln C) / (2 ln r), where 't' is the time depth in millennia, 'C' is the percentage of cognates, and 'r' is the constant retention rate. For example, if two languages share 70% of their core vocabulary, glottochronology attempts to calculate how many millennia it would take for that much divergence to occur.

Despite its initial appeal, glottochronology has faced significant criticism. Critics argue that the assumption of a constant rate of lexical change is flawed, as language change can be influenced by various factors like contact with other languages, social dynamics, and historical events. The 'basic' nature of vocabulary items can also be subjective, and borrowing can still occur even in core vocabulary. Furthermore, the method's accuracy tends to decrease significantly for deeper time depths (beyond 5,000-7,000 years). While it provided an early quantitative approach to historical linguistics, its reliability is generally considered lower than that of the comparative method, which focuses on systematic sound changes and grammatical reconstruction. Today, glottochronology is rarely used as a primary dating tool by historical linguists, though lexicostatistics can still be useful for establishing degrees of relatedness between languages.