|Anatolian||Hittite | Luvian | Palaic | Lycian | Lydian | Carian|
|Greek||Attic Greek | Greek Dialects|
|Indo-Iranian||Sanskrit | Middle Indic | Old Persian | Avestan | Pahlavi|
|Italic||Latin | Old Latin | Oscan | Umbrian | Venetic|
|Germanic||Gothic | Ancient Nordic|
Indo-European is the most recent common ancestor to the largest group of modern languages. From India to Britain and the Americas, the descendants of this mother tongue are still spoken and evolving. In ancient times, the Indo-European languages were similarly diverse, being composed of several branches that spanned over Asia and Europe. These were the Anatolian, Indo-Iranian, Greek, Italic, Celtic, Germanic, and Armenian branches. The Tocharian, Balto-Slavic, and Albanian branches are only attested too recently to be considered ancient language groups.
These various branches all broke off from a "proto-tongue" around the middle of the fifth millennium BC. This Proto-Indo-European language (PIE), although completely unattested, has been reconstructed by linguists using the Comparative Method. Upon breaking away from PIE, the language groups went through complex changes, generally involving the simplification of the complex phonological and morphological qualities of the mother tongue. What distinguishes members of post-PIE language groups is their shared innovations, that is, the modifications to PIE that they share. However, all of the groups are known to have evolved from PIE due to their shared retentions, or those qualities of the mother tongue that they retained. These specific innovations are described in the respective language group pages.
Two important phonological innovations that span over several groups categorize two "sections" of the PIE daughter languages. One is the Centum (Latin for 100) group, which includes the Anatolian, Greek, Italic, and Celtic languages, and the other is Satem (Avestan for 100) group, which includes the Indo-Iranian and Armenian groups. In PIE, there was a set of dorsal consonants including pure velars, palatovelars, and labiovelars (see Glossary for terminology). In the Satem languages, the labiovelars merged into the pure velars, and the palatalized velars became fricatives and affricatives produced at the front of the mouth. Thus, PIE *ḱm̥tóm became Avestan satem, with the palatovelar "k" becoming an alveolar voiceless fricative. The Centum groups, on the other hand, merged the palatovelars into the pure velars and kept the labiovelars (e.g. "qu" in English). The result is the pure velar "k" in centum.