Central Eurasia is a linguistic bottleneck, spread zone, and extinction chamber, but its languages had to come from somewhere. The locus of the IE sp… - Johanna Nichols

The Iranian family, which was next to sweep across the steppe and deserts, finds its region of greatest diversity in the central Asian mountains, and its ancestral Indo-Iranian family finds its own greatest diversity in the mountain region from central Asia to northern India (i.e. Bactria- Sogdiana and parts just south).

No migrations are required to derive the attested IE distribution from a reconstructed homeland consisting of a locus in western central Asia and a range over the steppe and desert. Sometime in the fourth to early third millennium, PIE spread along the steppe and southern trajectories to occupy the entire reconstructed range: the steppe, the desert of wester central Asia, part of the adjacent mountains, and perhaps some of south- west Asia. At this time its distribution was continuous, and that distribution had been achieved not by migration but by expansion.

The archaeological sources on which this summary is based... most often describe the westward cultures as derived from, or extensions from, the eastern ones. Mallory (1989) and Anthony (1991, 1995) interpret the directionality of cultural derivation as west to east. It is the east-to-west directionality of cultural derivation that would be consistent with the east-to-west linguistic trajectory, since spread of a whole culture is likely to involve language spread (and vice versa). A predominantly east-to-west directionality of cultural derivation and descent for the Eneolithic steppe would be a strong indication that the spread zone with its westward trajectories had taken shape. This is a purely archaeological question, but it is important to dating the rise of the linguistic spread zone.