There is currently much debate over whether nosology (classification / taxonomy) should be dimensional or categorical in psychiatry. the dimensional vs categorical debate seems to have a couple of different applications.

- Firstly, whether there is a categorical or dimensional difference between the people who are mentally disordered and the people who aren't. if there was a fact as to whether every individual either is or isn't mentally disordered then a categorical approach would be justified. if there was a matter of degree (with funny borderline cases where there is no further fact of the matter) then a dimensional system would be appropriate.

- Secondly, whether there is a categorical or dimensional difference between the people who have a certain kind of mental disorder and the people who don't. if there is a fact as to whether every individual has a particular kind of mental disorder or not then a categorical approach would be justified. if there was a matter of degree (with funny borderline cases where there is no further fact of the matter) then a dimensional system would be appropriate.

part of this debate involves mapping behavioural symptoms in state space. the notion is that we list the symptoms that people do exhibit and then over time we try and see whether there are correlations to be found between different symptoms. if we found that certain symptoms were found to be clustered together in a fair few people then we may have grounds for regarding these clustered symptoms to indicate a natural kind of disorder. this is fairly much the approach that Kraepelin took when he distinguished (roughly) schizophrenia from mood disorders. Bentall has some interesting things to say about the number of people with schizoaffective (who seem to suggest that there is a continuum of symptom clusters from the schizophrenic spectrum through the mood spectrum).

- Thirdly, whether there is a categorical or dimensional difference between the people who have a certain symptom (e.g., delusion) and the people who don't. if there is a fact as to whether every individual is deluisonal or not then a categorical approach would be justified. if there was a matter of degree (with funny borderline cases where there is no further fact of the matter such as belief in ghosts, god, aliens etc) then a dimensional system would be appropriate.

The DSM maintains (there is a little blurb in it somewhere) that while it offers a categorical system mental illness is more approprately regarded as dimensional (it doesn't say in precisely what respects it is regarded as dimensional). The DSM states that the main reason it persists with a categorical approach (on all three of the above issues) is because there is no workable dimensional system of classification at present. The DSM doesn't say this, but basically, while theorists have been pushing for a dimensional approach there is very little consensus from these theorists as to what a dimensional system should look like. There is also the very reasonable concern that the system must be workable in practice (ie teachable to new clinicians and it mustn't take too long to diagnose patients).

The psychodynamic diagnostic manual (PDM) has attempted to dimensionally define personality disorders. Not a bad attempt, I reckon, though the psychodynamic terminology is likely to result in the manual isolating itself from more biologically or cognitive behaviour therapy oriented clinicians (to say nothing of insurance companies). It attempts to get around the insurance company issue (I think) by using the same coding system as the DSM (though I could be wrong on this).

I think that the main reason why the DSM wants to stay categorical is that it makes it look more like medical nosologies / systems of classification. At first glance anyway there is a fact as to whether someone is invaded by a certain pathogen or not. There is a fact as to whether someone has a fractured leg or not. In the attempt to be a 'real science just like medicine' psychiatry presents itself as being more determinate than it really is.

This move is of course supported by the drug companies (anti-depressants rectify a certain kind of illness just like medication for diabetes!!). This move is also supported by consumers feel like being labelled validates their distress and (in some cases) gives them hope for improvement. This move is also supported by those who advocate that mental disorders should be reimbursed by health insurers comperably to the reimbursement received for physical conditions.

The reasons for remaining categorical are extra-scientific in the sense of being politically and socially driven.

And so... The science suffers :-(

As do the people who are told they are chronic and the like :-(