"WR" had primary progressive aphasia, a type of dementia which starts out as a language disorder. His grammatical profile showed a pattern that, according to many theories of grammar and aphasia, should not exist: WR was good at understanding passive sentences, such as The lion is killed by the man, but very poor at actives such as The man kills the lion, which are considered easier and more resistant to brain damage. More importantly, many theories predict that if someone has difficulties with actives, that person should find passives at least as difficult.
We found that only one other publication reported a similar case. WR came at the right time as I was starting to look at theories which are not that mainstream in clinical linguistics, in particular usage-based construction grammar. The theory has wonderful explanations for what we thought and drove a quite thorough experimental design. I am also glad that this work got me into a working relationship with the great Ewa Dąbrowska. Unfortunately reviewers did not like the way we elaborated on the theory, but there will be other opportunities for us to do that.
So how rare is WR's profile among people with aphasia? We don't know, and I share a suspicion raised by Alfonso Caramazza and colleagues that many cases do not get published, or are difficult to get published, because they do not fit the common predictions about aphasia.