Start with your NAAN and decide on a short, fixed extension that you will add to it. The extension should
- start with one or more lowercase letters,
- end with a digit (non-zero preferred), and
- contain no vowels or the letter ‘l’ (ell).
Many people make the initial mistake of adding a “/” between the end of the shoulder and the rest of the ARK, for example,
ark:/12345/x5/wf6789/c2/s4.pdf ^ WRONG!
It’s natural to want to visually mark the shoulder’s end, but it’s prohibited by ARK rules.
Why? The reason is that adding a “/” after “/x5” makes two false assertions:
- that ark:/12345/x5 also names an actual object, and
- that the original object (ark:/12345/x5/wf6789/c2/s4.pdf) is contained in it.
Adding a “/” might make the shoulder boundary obvious to in-house ARK administrators, but recall that they are trained specialists. The end user has no business knowing your internal operational details, and if they did you would risk their trying to hold you to account for their inferences (eg, about consistent support levels across objects sharing the apparent containing object). Less transparency about administrative structure hides messy details and can save you user-support time in the end.
In fact, in-house ARK administrators always know where the shoulder ends, provided it was chosen using the “first-digit convention”. A primordinal shoulder is a sequence of one or more betanumeric characters ending in a digit. This means that the shoulder is all letters (often just one) after the NAAN up to and including the first digit encountered after the NAAN. Another advantage of primordinal shoulders is that there is an infinite number of them possible under any NAAN.
There are different ways to implement a shoulder. Fundamentally, a shoulder is the deliberate practice of assigning ARKs that start with a particular extension to a NAAN. You could implement two shoulders simply by assigning ARKs beginning ark:/12345/x8 only to apples and ARKs beginning ark:/12345/x9 only to oranges.
If you use a service that stores ARKs in the N2T.net resolver, such as ezid.cdlib.org, then you can supplement that practice in two different ways. First, you could take advantage of N2T’s suffix passthrough feature by creating a short ARK, such as ark:/99152/p0, that looks and acts like a shoulder. To make it work, it suffices for that ARK to redirect to a server URL that can handle all the ARKs on that shoulder (eg, the Smithsonian does this), and you wouldn’t have to store or manage any other ARKs on that shoulder at N2T. Second, the EZID service (and perhaps others), associates a shoulder with a minter service and an API access point.
A completely different kind of shoulder “creation” step is needed to implement a shoulder under one of the few shared NAANs (described under namespaces).