I've written about Peter Bellwood's section on Pama-Nyungan in First Migrants and now I think it's time to say something brief about his section on Indo-European.
Bellwood believes that Indo-European originated in Anatolia with the very first farmers. This is not the consensus among linguists and archaeologists, with the more widely accepted location being north of the Black Sea at the tail end of the Neolithic, when wheeled vehicles, wool, horse-riding, and large herds of cattle - all reconstructible to proto-Indo-European - became commonplace on the western Eurasian steppe. Bellwood takes this to such an extreme that he even believes Indic languages to have arrived in South Asia with the first farmers there, something not even considered reasonable by Colin Renfrew (who originated the idea of an Anatolian origin within archaeology). So his account focuses on farming and on distorting the linguistic record to demonstrate an Anatolian origin.
He accepts unquestioned a series of dates putting proto-Indo-European at 6700 BCE, dates which rest on debatable premises. He mentions that food production can be reconstructed to proto-Indo-European, and so asserts that its lexicon perfectly fits the early Neolithic despite the presence of words for wheeled vehicles and wool, which he neglects to mention are found only in archaeological contexts after 4000 BCE. He claims that Indo-European has Afroasiatic loanwords (dubious but plausible) but neglects the much better attested loans from and into Uralic languages (and also neglects to mention that Afroasiatic speakers are a plausible candidate for Neolithic migrations into eastern Europe, perhaps accounting for loanwords in Indo-European). He deliberately fails to mention that the Anatolian branch is often considered to have split from PIE first and that 'classical' PIE is a different language to the pre-proto-Indo-European of which Anatolian is considered to have been a part, all so as to justify the idea that PIE was spoken by an early Neolithic population instead of a much later Chalcolithic one.
Bellwood endorses the claims of Casule, who believes Burushaski to be an Indo-European language (either Phrygian or an even earlier IE breakaway than Anatolian). These claims are not widely accepted, or even accepted by anybody other than Casule, as far as I'm aware. Bellwood avoids mentioning the presence of Brahui, a Dravidian language, in the area once inhabited by the Indus (Harappan) civilization, presumably so he can get away with the bizarre claim that the Harappans spoke an Indo-European language. He seems to have no awareness of what Indo-European culture was actually like and uses no comparative cultural information in his reconstruction of Indo-European history. He mentions none of the genetic evidence showing a plausible Indo-European origin of the R1a1a NRY haplogroup with a present concentration north of the Black Sea, in Russia/Ukraine/Poland, roughly where David Anthony and others put the speakers of PIE.
When I say that Bellwood ignores reconstructed PIE culture, what I mean is that he doesn't bother with any of the evidence from historic-period Indo-European-speaking societies. The people who composed the Rigveda were clearly pastoralists who valued horses, quite unlike the people of Harappan civilization. Love of the horse and the concept of cattle as the moveable wealth par excellence are found throughout the Indo-European-speaking world, in, as far as I'm aware, every attested branch, from Italic and Germanic to Indic and Armenian. The world of PIE speakers was not one in which horses were merely wild animals of little economic importance, known only as quarry (as the wild horses of early Neolithic Anatolia would have been). It was a world in which horses were familiar companions and workmates.
Moreover, Indic and Iranic societies share obvious similarities as they enter the historical record, including many similar rituals, concepts, holy plants (soma in Indic, haoma in Avestan), and an emphasis on cattle and horses. Bellwood's IE diagram places the split between Indic and Iranic at 2600 BCE, but the implication of the text is that the split must have been much older, unless he's making the implausible claim that people in Iran and India maintained effectively the same language for four thousand years, agriculture having arrived in India and Iran around 7000 BCE.
There's also the problem - effectively unaddressed in Bellwood's book - presented by the total absence of Indic and Iranic speakers in the Near East or Persia until the very late second millennium BCE (when Indic shows up in Syria, amazingly, after which Iranic speakers showed up in Iran - the Medes, Persians, and so on). If the earliest farmers in Iran and India spoke IE languages and got to those places direct from Anatolia, then why do the earliest recorded people in Persia and the Near East - the path between Anatolia and India - speak a variety of languages (Elamite, Sumerian, Afroasiatic) that are definitively non-Indo-European?
There are other errors and omissions. It's probably the shoddiest piece of work Bellwood has written, and all the while he claims that linguists and archaeologists who reject his claims are 'intransigent'.
If you read First Migrants, and I certainly recommend it, do bear in mind that some of what Bellwood says is not widely accepted and that many of his arguments have fatal flaws, as especially so with his discussion of Indo-European.