File-naming conventions

File-naming is a crucial aspect of digital fieldwork. Ideally your data should have the same name from creation through to your rediscovery of it in 20 years and then into the archives for posterity. In this way all of the work you put into annotating the file will be resolvable to an archival version by you and by others in the future.

While there are lots of ways of making up filenames, a basic principle is that you should NOT change filenames after they are created as you will have a problem tracking your own data over time. If you use a tool that names files automatically, as most cameras and recorders do, you may want to rename them as soon as you copy them to your computer so you do not have multiple versions of the same file with different names.

It is also worth talking to an archive about what kind of issues there could be with non-standard characters (like non-alphanumeric charcters for example) or unicode characters. If you have a choice about the names then make them more robust by avoiding known problems (like excessive length or the arbitrary switching of upper and lower case for example).

• A discussion on file-naming protocols (labelling and metadata) can be read in the RNLD discussion list archive (May 2010).

• PARADISEC has published a set of principles for its file-naming here.

• Jane Simpson has a useful post on the Transient Languages blog about file-naming in a field methods class (March 2007).