"Central Eurasia is a linguistic bottleneck, spread zone, and extinction chamber, but its languages had to come from somewhere. The locus of the IE spread is a theoretical point representing a linguistic epicentre, not a literal place of ethnic or linguistic origin, so the ultimate origin of PIE need not be in the same place as the locus."

“The location of the Anatolian branch of IE is a problem, or at least a puzzle, for IE homeland studies. The Anatolian languages are attested very early in Asia Minor, removed from Europe and far from the steppe; Gamkrelidze and Ivanov ... offer as a strength the ability of their proposed homelands to account for the location of Hittite with minimal migration. Alternatively or additionally, the location of Tocharian—attested in the early centuries AD well to the east of most IE territory in present-day Xinkiang —is a problem or a puzzle... Accounting for the locations of both Hittite and Tocharian is usually presented, at least rhetorically, as a major problem.”

“Approximately every two millennia, then, there has occurred a spread of a language family from a locus in the eastern part of the central Eurasian spread zone to cover the steppe and central Asia, extending partially or intermittently to the Danube plain, Anatolia, and northern Mesopotamia. The loci of the historically attested spreads are near the edge of the spread zone rather than in the centre of it: the piedmont to the south for Iranian, the north of Mongolia for Turkic and Mongolian. The trajectories of language spread run east to west along the steppe and through the desert to the Near East as shown .... To take clear and historically well-attested examples, the locus, trajectories, and range of IE must have been much like those of Iranian or Turkic. ... The placement of the locus specifically in the vicinity of Bactria-Sogdiana is justified in .... A homeland reconstructed as locus, trajectory and range removes the dilemma: a locus in the vicinity of Bactria-Sogdiana implies a spread beginning at the frontier of ancient Near Eastern civilization and a range throughout the steppe and central Asia, following the east-to-west trajectory, with occasional or periodic spreads into the Danube plain and Anatolia.”

“The vast interior of Eurasia is a linguistic spread zone—a genetic and typological bottleneck where many genetic lines go extinct, structural types tend to converge, a single language or language family spreads out over a broad territorial range, and one language family replaces another over a large range every few millennia. The linguistic geography of the central and western grasslands, from at least the Neolithic until early modern times, has consisted of an overall westward trajectory of language spreads... The central Eurasian spread zone... was part of a standing pattern whereby languages were drawn into the spread zone, spread westward, and were eventually succeeded by the next spreading family.”

“Prior to the Turkic expansion, at the beginning of the Iron Age, Iranian spread from somewhere in the vicinity of Bactria, Sogdiana, and the eastern steppe to cover most or all of western central Asia and the entire steppe, much of the Near East at least to eastern Anatolia, and, at least intermittently, the Danube plain, where Slavic vocabulary and ethnonyms attest to a major Iranianization at about the fifth century AD, and where there is good archaeological evidence of a Scythian presence in the mid-first millennium BC...”

“Along the forest trajectory as well there is evidence of either an early Indie presence or undifferentiated Proto-Indo-Iranian or Proto-Indo-Aryan. Among the Indo-Iranian loans into early Finno-Ugric are some so phonologically archaic that they could well be Proto-Indo-Iranian... Iranian, but not Indo-Aryan, regularly reflects PIE s as h, so this Finno-Ugric form looks more Indie than Iranian. Abaev also cites some less well attested forms that could be specifically Indie... These borrowings would have taken place somewhere in the vicinity of the southern Ural Mountains. They were received from a steppe language and incorporated into Finno-Ugric as it began its spread along the forest trajectory. This linguistic evidence for an Indie or Proto-Indo-Iranian wave preceding Iranian on the steppe is weak but legitimate. In partial confirmation of it, Kuz’mina identifies the Andronovo culture of eastern Kazakhstan in the mid-second millennium BC as Indo-Aryan..... There is also evidence for Indo-Aryan along the steppe trajectory in the form of a set of Crimean place names which Trubač identifies as Indo-Aryan. This evidence is even weaker—place names in general have poor diagnostic value since they lack denotational meaning—but carefully researched and again legitimate. If Trubačv is right there is evidence for an Indie advance to the western steppe. Taken together, the Finno-Ugric and Crimean evidence are consistent with the assumption of a short-lived Indie or Indo-Aryan presence at the frontier of the Iranian spread on the steppe, in addition to the well-known Indie frontier in northeastern Mesopotamia and India.”

“These principles make it possible to pinpoint the locus in space more or less accurately even for a language family as old as IE. Here it will be shown that the locus accounting for the distribution of loanwords, internal innovations, and genetic diversity within IE could only have lain well to the east of the Caspian Sea.”