While I was away I noticed on, I think, Twitter, which I was scrolling through while waiting at a bus stop/train station/whatever, somebody getting into a froth over somebody deleting their tweets upon
mature reflection, and how this was The Death of History.
To which my own reactions were:
a) Archivists have been thinking about the problems posed by the fragility of the digital record for a good couple of decades plus, this is not something no-one has noticed before. (Wasn't the Library of Congress archiving Twitter, and presumably there are some measures against tampering, if so? - hah, I see that there have been problems of processing and it's not actually accessible, or wasn't as at last year.)
b) Quite apart from the dangers of fire, flood and insect or animal depredation to which records in the more traditional forms have been exposed, there has been a fair amount of deliberate curating of the record over the centuries, by deliberate destruction or just careful concealment (whether it's the Foreign Office secret archive or the concealment of Turner's erotic drawings under a misleading file title).
c) While you can delete or destroy a particular record, you cannot always get rid of the information that it did exist - presumably it was other people commenting on the now-deleted tweets or retweeting them that led to the decision to delete them, but that doesn't eradicate the fact of their existence. This may even draw attention to the deleted record: this is why when I was still being an archivist we used to persuade donors not to ask for closures apart from those mandated by Data Protection, because the idea that something is *CLOSED* causes some people's ears to prick up in a supposition that there will be *HIDDEN SECRETS* (this was very, very, seldom the case).
I might also invoke the case that came up in Prince of Tricksters, where Netley Lucas under one of his identities was communicating with different officials and departments, possibly, it is suggested, as a means to confuse his trail: but, due to the growth of bureaucracy, as well as the social networks they belonged to, could also communicate among one another to discover that this was all the same guy.
There is also the phenomenon that I have mentioned to researchers, that yes [organisations of a certain ideological bent] have been very coy about placing their archives anywhere where people might do research in them; BUT the organisations and people they were against kept tabs on their activities, collected their literature, etc.
Also that if person/organisation's own papers do not survive, you can find out a good deal from the surviving records of those they interacted with.