Grammatical morpheme density: typology, ontogeny, phylogeny
Course author
Professor, Max Planck Institute for the Science of Human History
Course annotation
What is an isolating language, and how do such languages differ from other language types? The conventional answer is that in isolating languages, the number of morphemes per word is small, approaching one. However, this answer is problematic because it presupposes that we can unambiguously identify whether a given form is a word, as opposed to being part of a larger word, or alternatively consisting of two or more smaller words. In actual fact, determining where one word ends and another one begins is a complex task even in well-studied languages, and all the more so in languages about which we know relatively little.
We propose an alternative typology appealing to the notion of Grammatical-Morpheme Density. Typically, isolating languages are of low grammatical-morpheme density: syntagmatically, they exhibit a low ratio of grammatical as opposed to contentive morphemes in texts, while paradigmatically, they display predominance of grammatical markers that are optional as opposed to obligatory. Languages with low grammatical-morpheme density are characteristic of the Mekong-Mamberamo linguistic area extending from Mainland Southeast Asia through the Indonesian archipelago to western New Guinea. Low grammatical-morpheme density is also associated with early child language and can be argued to represent an archaic stage in the evolution of language.