Another spoke broke on my commute this morning. At the start of commuting season a spoke failed in exactly the same way, breaking at the threads in the nipple, not at the elbow in the hub flange.
As a wheel builder I take spoke breakage personally. I can't do anything about the fatigue failure of heat treated rims or the eventual fatigue failure of really elderly spokes, but spoke quality has improved so much in recent years that reputable builders are actually reusing them when replacing rims. Such a thing was never done a couple of decades ago.
The type of failure made me wonder if the spokes were a tad too short, creating a stress riser in the threaded portion. The rim depth of the Sun CR 18 does not allow for more than a millimeter more length. Then I remembered that stainless steel is prone to anaerobic corrosion in places where moisture can get in and air is excluded. It's a peculiar problem with stainless that can have a serious impact in marine applications.
In areas where roads are salted, the water around spoke nipples can be a tad on the briny side as well, adding electrolytic components to the possible mix. But anaerobic corrosion does not require saltiness.
I don't know for sure whether the spoke failures I suffered stemmed from anaerobic corrosion. Symptomatically it seems quite possible. I'm completely respoking the wheel today so I don't get caught doing the bent-rim samba down the road again.