TRADITIONAL psychiatry uses the approaches of medicine to try to understand mental health problems and guide treatment.
This means relying on diagnosis - identifying what are believed to be mental illnesses from their signs and symptoms, in the same way that doctors in other branches of medicine diagnose physical health problems.
There are two widely used systems in psychiatry: the World Health Organization's International Standard Classification of Diseases, Injuries and Causes of Death - or ICD - and the American Psychiatric Association's Diagnostic and Statistical Manual - or DSM.
The latest edition of the DSM was published this weekend.
Both ICD and DSM were first published immediately after the World War II and have been revised extensively over the years.
But there hasn't been a new edition of DSM since DSM-IV was published in 1994 - almost 20 years ago.
ICD is technically the international standard classification system and forms the basis for NHS procedures in the UK.
But the American DSM is also extremely influential and is widely used in research and academic fields and for planners, for example keeping hospital records.
That's why the publication of its fifth edition is important.
The nature of psychiatric diagnosis
The diagnosis of mental health problems is extraordinarily complex - and controversial.
The basic aim of diagnostic manuals is to explain the underlying nature and structure of mental health problems.
They attempt to describe patterns observed in nature, for example how a patient behaves, without (the authors claim) making assumptions about why.
However, the complexity of mental health problems can lead to difficult decisions.
It also means the manuals themselves are also complex: what criteria are included; the rules about which disorders are included and which aren't; and the relationships between different families in the manuals, for example between obsessive compulsive disorder and impulse control disorder.
This also leads to significant differences in opinion.
Families of illnesses
The manuals are designed to group similar types of diagnoses together.
For instance, diagnoses that are all concerned with anxiety of various kinds are listed together.
And they are generally seen as separate from problems such as learning disabilities.
Including problems such as children's learning disabilities, relationship and personality difficulties, emotional problems and problems of later life such as dementia, can be problematic. DSM-5 has come under critcism for changes in some of these areas.
One example discussed widely is that idea that it might be possible to receive a diagnosis of "major depressive episode" when one is still grieving for the death of a loved-one.
Bereavement was specifically excluded from previous versions.
The ICD and DSM are different, and to an extent are rival systems, but there is huge overlap.
This allows researchers and clinicians to translate diagnoses from one system to another - a bit like cross-referencing between two dictionaries.
DSM uses what is called a "multi-axial" scheme to classify diagnoses. Psychiatrists use multiple axes to diagnose and treat patients.
Primary diagnoses form a first tier called Axis I and includes depression and schizophrenia.
So-called developmental and personality disorders lie in Axis II and includes autism.
Related issues such as the degree of disruption caused to a person's life are assessed on remaining axes.
In practice, Axes I and II diagnoses tend to be used in a similar way.
Psychiatric diagnosis echoes and resembles conventional medical diagnosis, but there are no useful biological markers or tests for illnesses like you might get if you were treating someone with diabetes - which makes many people sceptical of biological explanations per se.
Diagnosis of a person's problems is inevitably based on their descriptions of their feelings, thoughts and behaviour and on the observations of the person trying to make the diagnosis.
It also means that decisions about the criteria for each diagnosis - the structure and content of DSM and ICD - are essentially made by committee. In the case of DSM, a taskforce.
Different clinicians - and particularly psychiatrists and psychologists - differ as to what particular problems should be included or what the criteria should be.
Some also question the reliability of psychiatric diagnoses, whether we should think of problems as illnesses to be treated or that a broadening of psychiatric diagnoses means a wider variety of personal problems could attract a diagnosis.
One widely discussed example is that it might now be possible to receive a diagnosis of "major depressive episode" when one is grieving for the death of a loved-one.
Others fear the opposite: that diagnoses, and therefore psychiatric support, will be taken away.
This has particularly been the case with changes to the definition of autism and the exclusion of Asperger's from DSM-5.
While the publication of DSM-5 has catalysed criticism, it is also pushing new approaches into the spotlight.
The director of the US National Institute of Mental Health, the largest funder of mental health research in the world, said this month that it was moving away from a DSM-style approach to focus on biology, genetics and neuroscience, allowing disorders to be defined by causes, not symptoms.
New research will continue to develop our understanding of the causes and treatment of mental illness.
But public debate and controversy over the way we should approach it won't be very far away.
Update your news preferences and get the latest news delivered to your inbox.