Like many of you I have been horrified to hear of the scandalous treatment of the Windrush generation in the UK, in particular the relatively recent destruction of important documents relating to their invited migration to the UK from the Caribbean. This truly highlights the power archives and the fact that they can at times quite literally make the difference between life and death.
Archives and records management are hugely political acts, but in thinking about this general topic my thoughts have turned to archives and records that have not survived due to disasters or warfare. I have recently read a book about the great Lisbon earthquake of 1755 (I'm wondering if anyone has made a study of similarities in earthquake recovery over time, or whether this topic will end up on my list of PhDs I'm never going to write?) and apparently it is next to impossible to research the important early modern economy of Portugal because almost all relevant documentation was destroyed.
If you want to do research into 6th Battalion, Lincolnshire Regiment at Gallipoli it's a bit tricky. Their war diary - the official day to day record of what the unit was doing - is probably at the bottom of the Aegean Sea. Indeed, in the paper finding aids at the UK National Archives there's a question mark where the item's record number should be. Similarly a lot of similar documentation was lost, along with an awful lot of equipment and those left to cover the retreat, at Dunkirk.
For much of my time as specialist military librarian I answered family history enquiries from members of the public. Many, many people wanted to know what their relatives had done during the First World War. Yet approximately two thirds of service records from that war were destroyed in an air raid during the Second World War. (More about the fire and what was lost can be found on the excellent Long Long Trail website.) What this means that it is not easy to find out exactly what your family member did during the war if they served in the British Army. There are work arounds but essentially a lot of my previous work was defined by archival absences and telling people that there's a good chance their never really going to find the information they're looking for.
I have no idea whether any of these losses could have been prevented or not and even with the best will in the world there will always be archival absences, so willful destruction and dodgy retention policies really don't help matters.