I'm just going to go over a few bits of M. L. West's Indo-European Poetry and Myth to show how cultural traits have to be made explicable by language origin hypotheses. It is true that this evidence is comparatively weak, but that's not the point. The point is that the nature of Indo-European culture and society points to a steppe origin, however weakly. It seems to me that the cultural evidence means that the null hypothesis should be that Proto-Indo-European was spoken on the steppe by horse-riding pastoralists, given that there is a slight controversy over Indo-European origins.
Indo-European-speaking groups in historic times loved horses. Whether they lived in France or Persia, they waxed lyrical about equids. Horses were the subjects of ritual: in several documented cases, stallions were sexually assaulted by kings and then slaughtered in great ritual events. Events of this nature were documented in Ireland and India, both at extremes of the Indo-European distribution - Ireland and India are clearly not related through any means other than IE ancestry. In India, the ritual was referred to as aśvamedhá, 'horse sacrifice'.
West describes the ritual (p.417-418):
Its preparation took a year. The central act was the sacrifice of a prize white or grey stallion which had had no contact with mares during that time and which had drawn the king's chariot. After it was killed and before it was dismembered, the principal queen lay down and passed the night with it under covers, and verses were chanted encouraging it to impregnate her. So here too the king's consort was required to enact a kind of supernatural hieros gamos [as in similar Greek traditions].He continues:
Scholars have noted an analogy between the aśvamedhá and a disgusting ritual recorded from twelfth-century Donegal. When a new king was consecrated, the people assembled and he, declaring himself to be a horse, copulated with a white mare in front of them. The animal was at once sacrificed and dismembered and the meat was boiled. A large barrel was filled with the broth. The naked king climed into it and sat there, dipping his mouth in to sup, while the gobbets of boiled meat were distributed among the spectators.Furthermore:
...it is related that the assumption of the Swedish kingship by the pagan Svein was ratified by the sacrifice, dismemberment, and consumption of a horse.So here we have evidence of a similar ritual activity in three parts of the Indo-European world: India, Ireland, and Sweden. Or, Sanskrit (Indo-Aryan), Irish (Celtic), and Norse (Germanic). Similar rituals were also present in Rome and, according to Roman sources, Gaul. Humans copulating with horses also appear on Swedish rock art of the Nordic Bronze Age. These groups are separated by vast distances and were not related by any means except their shared Indo-European ancestry. They all sacrificed and ate horses, sometimes copulating with them - and horses were assuredly not part of the Neolithic expansion into Europe that some people would associate with the spread of Indo-European languages. The distribution of this tradition doesn't make sense given a Neolithic agricultural expansion. It does make sense given a steppe origin in the Eurasian Bronze Age.
Again, this evidence is comparatively weak. But it should still be taken into account.
Horses also appear in several standard phrases. The phrase 'horses and men' appears in the poetic language of many Indo-European traditions, in that order - almost always 'horses' first and 'men' after, often in poetry describing combat. This is found in Persian, Greek, Latin, and several Celtic traditions, not the mention the Aryan tradition of the Vedas. And, of course, the word for 'horse' can be reconstructed to Proto-Indo-European.
Perhaps the strongest cultural and literary evidence for Indo-European steppe pastoralism is the presence of cattle raiding in almost every IE tradition. In fact, West, who is professor emeritus at All Soul's College (and, I think, not someone to take lightly), examines the literary evidence wholly through the lens of a steppe origin (p.451):
A form of aggression often celebrated in the Indo-European literatures is the cattle raid. The domestication of the horse allowed the early pastoralists of the Eurasian steppe to herd much larger numbers of animals than before, roaming over a vaster area. It also provided a means of driving off other people's flocks and herds. This was the easiest way to acquire wealth, which was commonly measured in cattle. But it was liable to provoke fighting.West then goes through evidence from everywhere. The mythological Theban War was fought, according to Hesiod, 'on account of Oedipus' flocks'. In the Rigveda (RV 10.38), there is a prayer for success in a cattle raid:
In this glorious battle, Indra,There is also evidence from the Mahabharata (4. 29-61); from the yasnas of Zarathushtra (Y. 12. 2) ('I abjure thievery and cattle-raiding, despoiling and devastating the great Mazdakayasnian clans'); from Ireland (táin, meaning 'cattle raid'; see especially the Táin Bó Cúailnge, the famous 'cattle raid of Cooley'); and from pre-Germanic Britain ('Eithinyn the son of Boddwadaf "attacked in force for the herd(s) of the East" [Y Gododdin 434]'). Speakers of Indo-European languages appear in all of these traditions as pastoralists for whom cattle were wealth. They were belligerent, tribal, and pastoral, and they used horses or chariots to conduct raids on one another's flocks. This appears clearly in literary traditions from every corner of the Indo-European world.
this energetic tumult, urge us on to win,
in the cattle raid where among the bold beringed ones [warriors]
the arrows fly in all directions for men's defeat.
Yet again, I appreciate that this evidence is relatively weak. Cultural traditions can change readily, and something as visceral as a horse sacrifice associated with kingship could, I suppose, have spread through a different method than by association with the Indo-European language family. But far from being primarily agriculturalists for whom the horse was a later innovation (as the Anatolian hypothesis claims), Indo-European speakers appear in their traditions as pastoralists for whom horses and the machines they pulled were important. Equine traditions with consistent themes and obviously shared ancestry appear throughout the Indo-European world, including the easterly and westerly extremes, and they also appear in the archaeological evidence from the southern Russian steppe.
The evidence for Indo-European origins should come from as many corners as possible. I'm concerned when I see studies that rely on lists of a hundred cognates or that only look at one segment of evidence. Even 'weak' evidence, like that from literary sources and cultural traditions, should be taken into account. Where phylogeographic models and statistical assessments are strongest - as with, say, Arawak - the models also accord with the cultural and literary data from the societies in question. The recent phylogeographic models of the Indo-European languages published in Science do not do this, and they neglect almost every sort of data beyond that represented by the narrow statistical methodology employed. They also neglect languages of known ancestry but for which we lack a list of cognate terms, including the Iranic languages of the Pontic-Caspian steppe.
Just because a method appears to be more scientific than another doesn't mean that it actually is. Science should use all of the available evidence, not merely the quantifiable kind.
I want to make something very clear. I have seen this repeated in several places, so I'll tackle it here. The Indo-European expansion did not happen because of chariots. Wheeled vehicles probably were involved - but they were wagons, not chariots. Chariots require the rider to stand up. They can carry little in the way of luggage or necessary items. They are fast-moving platforms for riders to use when racing or fighting, not the primary method of Indo-European transportation. They would have made Indo-European migration very uncomfortable indeed. It is a strawman to suggest that migrations occurred because of chariots according to the steppe hypothesis.
Furthermore, I have seen some people suggest that the Anatolian hypothesis is innately superior because wheeled migration into prehistoric Europe is a ridiculous proposition. It was a heavily-forested area, unsuitable for chariots and wagons, according to these people. Well, Europe was heavily-forested - but this was true in historic times as well as prehistoric. Britain, for instance, was covered in thick forest at the time of the Roman empire. And yet it is very clear that the Britons fought on chariots and moved in wagons, as did the Romans, Germanic tribespeople, the Greeks, the Balts and Slavs, and the continental Celts. The forest was no impediment in historic times, so why infer that it was in prehistoric times?