A case of “order insensitivity”? Natural and artificial language processing in a man with primary progressive aphasia.
This is the second (and certainly final) paper reporting data from WR, a man with primary progressive aphasia who showed a very atypical sentence comprehension profile. This paper unites one strand of my research, artificial grammar learning (a paradigm that looks at processing of sequences), with more classic clinical work (showing patients pictures and sentences). Because WR's language pattern goes against common clinical expectations and artificial grammar learning is a tricky paradigm, we waited until we had published a first case report and some related artificial grammar learning work before submitting this manuscript.
It appears that WR was to some degree insensitive to the order in which elements appeared both in language and in experiments with non-linguistic stimuli. Depending on the theoretical framework, this observation explains his language pattern. If we are right about this "order insensitivity", English is a very bad language if you have it because English phrase order is so important for understanding. Our work supports general cognitive, instead of language specific, explanations for aphasia, though a lot of questions remain.
One minor question is how to call the system we describe in the paper. "Sequence processing" is too general and can refer to many things we are not addressing, "configuration" is linguistic but, depending on the theory, means something very different.
Because the new data reported here are "only" from one patient and one experiment (though we present them within a bigger context), we decided to try Cortex' "Notes" format, which limits manuscripts to 3,000 words. However, reviewers always want more, never less, information, so we ultimately had to move some details to an online supplementary.
I am going to publish a longer blog post about WR soon.
If you use this link before August 8, 2015, you can get the paper for free regardless of where you work.