Apsidal Shrines

Last year there was news of discovery of a 2000 year old Shiva temple complex in Uttar Pradesh, one of the oldest in India. Besides the age, what was interesting was the shape: the temple was apsidal. It was widely believed that the apisdal shape had Buddhist origins and was used by Hindus later. Historians like Romila Thapar have argued that if Hindu temples had such shape, they were converted Buddhist chaityas or shrines
This theory, in fact, cannot be credited to Marxist historians; they evolved out of a colonial myth. Colonial archaeologists, who found that the written record of India was imperfect, resorted to studying the history of art. This study, they hoped, would give a better historical record as well help understand the relation between various Indic traditions.
This Marxist/Colonial explanation — that Hindu traditions replaced Buddhist shrines — actually makes sense if you follow a linear chronology. There is no dispute over the fact that there was a resurgence of Hinduism  inspired by bhakti and hence it can be logically argued that this resurgent Hinduism or traditions usurped Buddhist chaityas. Also, Buddhist shrines have been around since the 4th century BCE while Hindu apsidal temples make their appearance few centuries later.
There are three reasons why the Colonials and Marxists are wrong.
First, archaeology has disproved many cases. For example, one site where an apsidal temple was found was Barsi in Maharshtra. According to the British, a Buddhist shrine was converted to a Trivikrama temple, but later archaeological excavations found an apsidal brick temple with a wooden mandapa. The mandapa was not a later addition, but an integral part of the temple. A similar theory was proposed by the British for the Kapotesvaraswamy temple in Guntur and the Durga temple at Aihole, but both were disproved.
Second, this theory ignores another possibility – co-existence. In Nagarjunakonda valley, which was settled from third millennium BCE to sixteenth century CE, there is evidence of both Buddhist establishments and Hindu temples, both using the same plans in different areas. Besides this, Naga traditions too  used the same style. Between second century BCE and seventh century CE, there is a rise in apsidal Buddhist shrines in peninsular India. Hindus also constructed new apsidal temples. For example in Kerala, after 800 CE, numerous sanctums with apsidal plans were constructed, especially the ones dedicated to Ayyappa. All these show that  various branches of Hindu traditions shared style and space with a number of domestic and regional traditions.
Finally, a point regarding the origin of this style. The earliest elliptical shrines are seen in Vidhisha (second century BCE), Nagari in Chittor (first century BCE) and Etah, Uttar Pradesh (200 – 100 BCE). An apsidal mud platform was also found in Ujjain (500 – 200 BCE). That’s not it. At Daimabad (1600 – 1400 BCE), a complex with a mud platform having fire altars and an apsidal temple with sacrificial activity were found. Similarly at Banawali, a Harappan site on the banks on the Sarasvati bed, there were fire altars on an apsidal structure. Thus the elliptical shape had religious significance from a much ancient time and there is only tradition which still builds fire altars the way people in Banawali did.
References

1. Himanshu Prabha Ray, The apsidal shrine in early Hinduism: origins, cultic affiliation, patronage, World Archaeology 36, no. 3 (2004): 343. (Thanks Ranjith)
2. Michel Danino, Lost River: On The Trail of the Sarasvati (Penguin Books India, 2010).
3. Upinder Singh, A History of Ancient and Early Medieval India: From the Stone Age to the 12th Century, 1st ed. (Prentice Hall, 2009).