Nobody can date the Rig Veda by excavation, because it names no site that has been dug and matched. The dating that holds academic consensus is linguistic. Rig Vedic Sanskrit is markedly more archaic than the Sanskrit of the Brahmanas, which is in turn older than classical Sanskrit as codified by Panini around the fifth or fourth century BCE. It is also strikingly close to the language of the older Avesta, the Iranian scripture, close enough that whole lines can be converted between the two by regular sound rules. That relationship, plus the rate of change visible across the Vedic corpus itself, puts composition around 1500-1200 BCE, with the family books earlier and books one and ten later.
The material world the hymns describe is consistent with that. They know the horse and the spoked-wheel chariot, cattle wealth, and rivers of the Punjab and the Sarasvati. They do not know iron, do not describe cities of the Indus type, and show no interest in the coastal south. What they call ayas appears to be copper or bronze.
Against this stand traditional and revisionist chronologies. Bal Gangadhar Tilak argued in 1893 from astronomical references in the hymns for a date near 4500 BCE; other writers, in what is usually called the Out of India position, place the Vedic corpus in the third millennium BCE and identify its composers with the Indus Valley civilisation, in order to deny any Indo-Aryan migration into the subcontinent. The astronomical arguments depend on reading precise observations out of allusive poetry, and the linguistic evidence, the shared Indo-Iranian inheritance in particular, runs the other way. The disagreement is real, and in India it is political as well as scholarly, but the weight of the evidence sits with the linguists.