It's not surprising, therefore, that Mary reintroduces a notion of psychosis, however much she may think that the concept of schizophrenia is unscientific. But, neither is psychosis an absolute concept. Even the BPS attempt to explain psychosis and schizophrenia in everyday language still uses the terms (see previous post). The usefulness or validity of terms like 'psychosis' and 'schizophrenia' may depend on the ability to identify certain patterns between different patient presentations (see another previous post).
For example, the experience of hearing voices can be a dissociative symptom. This situation may have been used as a rather superficial argument for the abandonment of the diagnosis of schizophrenia but it does create a category of 'dissociative voice hearing'. As Green says, if the new catgory provides "more clinical information than the DSM, there is a chance that [it] will be adopted and applied instead, or even incorporated into that manual". Dissociation, in fact, does feature as a category already in DSM, including dissociative identity disorder (see changes made in DSM-5). It's a weakness of the BPS report, mentioned above, that it makes no attempt to discuss the difference between psychosis/schizophrenia and dissociation. The point I'm making is that it is a meaningful discussion to have and we do need to have words to communicate about it.
As Green concludes, "no psychiatric language is able to 'do justice' to the particulars of any given case". However, it serves a function in giving rise to a general form of pragmatic knowledge. DSM-5 has failed in its attempt to move from symptom-based diagnoses to aetiologically-based diagnoses using the latest advances from neurosciences and genetics (see previous post). We need to make the most of this failure (which should have been predicted anyway) to have a better diagnostic understanding of terms like psychosis and schizophrenia, if they have any meaning at all (see another previous post).