One Brahmanical practice, evident from c. 1000 BCE onwards, was to classify people (especially Brahmanas ) in terms of gotras . Each gotra was named after a Vedic seer , and all those who belonged to the same gotra were regarded as his descendants. Hence, Statement 1 is incorrect. Two rules about gotra were particularly important : women were expected to give up their father’s gotra and adopt that of their husband on marriage . Hence, Statement 2 is correct. members of the same gotra could not marry. Hence, Statement 3 is correct. One way to find out whether this was commonly followed is to consider the names of men and women , which were sometimes derived from gotra names. These names are available for powerful ruling lineages such as the Satavahanas who ruled over parts of western India and the Deccan (c. s econd-century BCE-second century CE ). Several of their inscriptions have been recovered , which allow historians to trace family ties , including marriages. Some of the Satavahana rulers were polygynous (that is, had more than one wife). An examination of the names of women who married Satavahana rulers indicates that many of them had names derived from gotras such as Gotama and Vasistha , their father’s gotras . They evidently retained these names instead of adopting names derived from their husband’s gotra name as they were required to do according to the Brahmanical rules. What is also apparent is that some of these women belonged to the same gotra . As is obvious, this ran counter to the ideal of exogamy recommended in the Brahmanical texts. In fact, it exemplified an alternative practice, that of endogamy or marriage within the kin group, which was (and is) prevalent amongst several communities in south India. Such marriages amongst kinfolk (such as cousins) ensured a close-knit community. It is likely that there were variations in other parts of the subcontinent as well, but as yet it has not been possible to reconstruct specific details.