One of the things that investigating Lakshmibai's story has driven home is the sheer difficulty of knowing what to believe. There are some stories which are obvious nonsense, but others are more difficult to gauge. Take the example of her death. There are two principle and incompatible versions of it. They have two elements in common, that her companion, Mandar, was killed at the same time that Lakshmibai was mortally wounded with Lakshmibai surviving for a while afterwards.Otherwise the stories could be about two entirely different people.

The written evidence has to be treated with some degree of scepticism since the author will be presenting certain facts that support their case, and possibly suppressing others. They may be making assumptions of knowledge on the part of the reader to which we are not privy. Intelligence reports are decidedly murky, and have the added twist that if the person spied on knows about the spy, then they may present the spy with false, or misleading, information.

At least with written evidence we can be certain that the words we read today are the same words that were written then. This is not the case with verbal evidence. Mahasweta Devi mentions that from family history that Lakshmibai's stepmother claimed that after the peace treaty between Jhansi and Orchha that the two Rani's embraced like sisters, perhaps, but hostilities continued between the two 'sisters' right up until the British arrived. So what did she really say, and mean?

Then there is a problem with introduced facts. Also in Devi's book she rejects Mrs Mutlow's testimony because Mutlow claims to have hidden in a Hindu memorial in the Jokan Bagh and there are no such structures in the Bagh. However the photograph of the site of the massacre, ie of Jokan Bagh, taken shortly afterwards, shows what, to me, look like Hindu structures of some sort but perhaps I am wrong.

Nonetheless by looking at what the British wrote about her, we can be reasonably certain that if there is a bias it will not be in her favour and that any favourable conclusions we draw from it should be at minimally true. The actual truth could be even better.