Centum and satem languages

Languages of the Indo-European family are classified as either centum languages or satem languages according to how the dorsal consonants (sounds of "K", "G" and "Y" type) of the reconstructed Proto-Indo-European language (PIE) developed. An example of the different developments is provided by the words for "hundred" found in the early attested Indo-European languages (which is where the two branches get their names). In centum languages, they typically began with a /k/ sound (Latin centum was pronounced with initial /k/), but in satem languages, they often began with /s/ (the example satem comes from the Avestan language of Zoroastrian scripture).

Quotes

 * The Indo-European languages were initially divided on the basis of the velar-palatal distribution of cognate terms into a Western branch, called the centum group (Latin centum < PIE *kmtom 'hundred'), which had preserved the Proto-Indo-European velar phoneme -k and a satem (Avestan satem/Sanskrit satam < PIE *kmtom 'hundred') branch, which had developed a palatal phoneme -s for the same term. The centum group included Celtic, Greek, Italic, Germanic, and Anatolian, and the satem group, Balto-Slavic, Indo-Iranian, Albanian and Armenian. This neat east-west division, however, was short- lived. A centum Indo-European language called Tocharian was found as far east as Chinese Turkestan (Xinjiang). Moreover, Melchert (1987) has argued that the Anatolian language Luvian is neither satam. nor centum, thus questioning the heuristic value of the entire divide. In any event, Elst's proposal that earlier tribes could have emigrated from India bearing the centum characteristics and, after the velars had evolved into palatals in the Indian Urheimat, later tribes could have followed them bearing the new satem forms (while the Indo-Aryans remained in the homeland), cannot actually be discounted as a possibility on these particular grounds. This scenario would receive some support if the recent report by Zoller of a centum language spoken in the North of India proves well-founded. This language, called proto-Bangani, although largely assimilated by the surrounding dialects, has supposedly preserved some centum vocabulary, especially, and significantly, in the context of old stories.
 * Bryant, E. F. (2001). The Quest for the Origins of Vedic Culture : the Indo-Aryan migration debate. Oxford University Press.


 * Koenraad Elst (1996) tries his hand in actually offering a falsifiable model of how migrations from a South Asian homeland might have occurred from within the parameters of the relevant dialectal relationships: We propose that there is no necessary link between the fact that Sanskrit is not the oldest form of IE and the hypothesis that India is not the oldest habitat of IE. It is perfectly possible that a Kentum language which we now label as PIE was spoken in India, that some of its speakers emigrated and developed Kentum languages like Germanic and Tokharic, and that subsequently the PIE language in its Indian homeland developed and satemized into Sanskrit.
 * Bryant, E. F. (2001). The Quest for the Origins of Vedic Culture : the Indo-Aryan migration debate. Oxford University Press.