“Caste-ing my vote” (June 7) makes an interesting point about why the census should collect data on castes. Several examples are provided of how, when viewed differently, data can yield new meanings. So, what looks like rank discrimination at first sight looks different when another prism is applied to it. The problem with this, however, is that the author thinks events follow a rational course. Once data on caste is collected, no one is going to apply the filters the column talks of. Only the raw numbers will be bandied about. And that is why caste data must not be collected.
Sujata Kumar, Noida
