So I have a question. I will detail the following research. For ‘free will’ to be true, it has to explain the following. Or more accurately, the following has to be fully explicable within the free will hypothesis. How does it do that?
The second question would be “What does free will predict?” Because free will, to be a robust theory, must have predictive value. The following research, for example, is fully predicted by determinism. Expected. On free will, however, at best one must ad hoc rationalise which decreases the probability of the thesis being true and valid.
Let’s start with an obvious case. Huntington’s disease is a monogenetic disorder. This means that it is expressed as a result of a single gene, which is more unusual since many disorders are polygenetic. What this means is that at some point in the sufferer’s life, their personality will utterly change. One gene. (See things like herpes simplex encephalitis and CJD for big changes with little tweaks). As Robert Sapolsky states: “Alter one gene among tens of thousands and, approximately halfway through one’s life, there occurs a dramatic transformation of personality.”
Schizophrenia has genetic drivers, but is much more complex. It seems sufferers only get this psychological issue in certain scenarios, most notably when they are immigrants who feel disconnected from the society around them (eg, say, a Korean immigrant to the US who feels different to the society in which they live). This mix of genes and environment trigger a psychological state of affairs. As I mentioned elsewhere, with regard to tumours, why see causal determinism only in cases of psychological ‘abnormality’ and not in neurotypical cases? That is special pleading.
But on to the main point. Consider the work of Stephen Suomi and his team with monkeys. He raises monkeys in their natural settings, observing every detail of their lives. He saw that monkeys began to exhibit different personalities to each other from an early age. Virtually every social behaviour was developed, practiced and perfected by 4-6 months of age. Suomi was able to combine these observations with regular blood testing of hormones and metabolites and genetic analysis.
He found among the baby monkeys that:
- 20% displayed social anxiety (reacting to novel, mildly stressful social situations with fearful and anxious behaviour and long-lasting elevations of stress hormones).
- 5% were overly aggressive (impulsively aggressive and inappropriately belligerent behaviour, low levels of blood metabolite related to breakdown of the neurotransmitter serotonin).
- There were 2 different alleles (gene variants) for a protein involved in transporting serotonin (serotonin transporter 5-HTT gene) – long and short forms. The monkeys with short form showed more violence, the long form normal.
- The outcome depended entirely on the mix of these genetic variations and environment.
- There were two ways the monkeys could be reared: mothers (good) of peers (insecure).
The results:
| Raised with peers | Raised with mothers | |
| Short allele | Aggressive | Fine |
| Long allele | Fine | Fine |
This tells us 2 things: good mothering gives monkeys resilience to mitigate genes; good genes give monkeys social resilience no matter what.
After the success of these studies, people naturally started looking at humans. We have pretty much the same biology after all. In 2001, for example, Caspi and team looked at genes for depression. Are there? Sort of. There are genes which predispose one to depression, bu tit depends on life environment. Looking carefully at life events and genetics (specifically serotonin regulation). They isolated one gene which both parents carry. This meant people have three combinations: short/short, short/long and long/long. Short/short predisposed people to clinical depression, BUT ONLY in combination with bad life events. Being lucky enough to live a ‘good life’ for want of a better term mitigated the genetic predisposition for depression.
Caspi looked at whether abused children themselves go on to abuse, commonly claimed (something I heard in the staff room and corrected only yesterday). This is sort of true. Some do, some don’t. Are there causal drivers? It turns out a particular predicted gene variant differentiated the children who went on to abuse from those who didn’t. It turns out if you had the gene variant and were abused, you were much more likely to abuse, rather like the table above.
The same Caspi looked at adult psychosis and smoking marijuana. Again, the same gene-environment interaction was found.
Gao et al found that poor fear conditioning at age ¾ scarily predicted criminality 20 years later.
Walter Mischel, in the famous marshmallow/cookie experiments showed that delayed fgratification ability at age 5 or 6 led to better life outcomes as adolescents and adults. Being able to put off taking one cookie now, whilst the interviewer left the room with the promise that if the one remained, the child would get 2 on their return, at the age of 5 or 6, predicted a range of outcomes for the children later in their lives. This is why quit smoking product adverts have to include the idea of having willpower. But we know that willpower is biologically conditioned, and we know this empirically. These experiments have been built upon to show the internal-external interplay more robustly analysed.
So it is not genes, necessarily, or environment, necessarily. But more often than not, a combination of the two.
In sum, as the neuroethicist Martha Farah puts it, if an antidepressant pill “can help us take everyday problems in stride, and if a stimulant can help us meet our deadlines and keep our commitments at work, then must not unflabbable temperaments and conscientious characters also be features of people’s bodies? And if so, is there anything about people that is not a feature of their bodies?”
There are tens of thousands of studies I could cite. Every single one needs to be explained on the free will hypothesis. All are explained by the determinism hypothesis. Better still, they are predicted.
Notes
Caspi et al – Role of genotype in the cycle of violence in maltreated children
Caspi et al – Influence of life stress on depression: Moderation by a polymorphism in the 5-HTT gene
Caspi et al – Moderation of the effect of adolescent-onset cannabis use on adult psychosis by a functional polymorphism in the COMT gene: Longitudinal evidence of gene environment interaction
Farah – Neuroethics: A Guide for the Perplexed
Gao et al – Association of Poor Childhood Fear Conditioning and Adult Crime
Mischel – the Marshmallow experiments
Sapolsky – The frontal cortex and the criminal justice system
Suomi – Risk, Resilience and Gene-Environment Interplay in Primates
