The only way you can estimate environmental influences is by measuring them.

There is almost no evidence that diagnoses such as 'schizophrenia' and 'bipolar disorder' correspond to discrete entities ('natural kinds' in the language of philosophy).

If we are to believe the evidence from clinical trials there are many effective pharmacological and psychological treatments for mental illness. Epidemiological data, on the other hand, says otherwise.

My interest in the psychological roots of psychosis has both personal (my brother Andrew committed suicide) and professional origins (I was trained in a behaviorist approach to psychology which - whatever its limitations - at least taught me to see human behavior in its social context).

Conventional psychiatry has emphasized the genetic roots of psychosis based on the claim that twin and other studies show that schizophrenia is 80% heritable, which means that 80% of the cause is genetic.

The only way you can estimate environmental influences is by measuring them.

There is also evidence from epidemiological studies that psychotic-like experiences are much more common than has hitherto been thought (with about 10% of the population affected) and that these experiences exist on continua with healthy or 'normal' functioning: instead of the world falling into two groups (the psychotic and the non-psychotic) people vary in their disposition to psychosis and only a minority of people who have these experiences require or seek help.