/ Criminals should feel terror

This topic has been archived, and won't accept reply postings.
Pete Pozman 03 Aug 2019

I don't know about criminals but Priti Patel certainly instils terror in me. 

4
Shani 03 Aug 2019
In reply to Pete Pozman:

It's a nice idea but a rather empty platitude from Patel. She is looking at crime from a comfortable, middle class position where she has a lot to lose - and so thinks everyone thinks and behaves as she does. She needs to look at the complexity of motivation. 

Are you really going to scare a suicide bomber with a death threat?

When people are drunk, are they in a position to rationalise their behaviour? Friday & Saturday night in the city suggests not.

When poor people beg, steal, deal in smuggled or stolen goods, they're often trying to provide for themselves and make their means go further.

At the other end of the scale, wealthy bankers will continue to commit crime because of greed. But they'll have little to be scared about.

Perhaps the worst thing is that Patel's statement pre-empts the criminal act.

Post edited at 10:38
7
Lusk 03 Aug 2019
In reply to Pete Pozman:

I thought she was referring to the cricket.

Pete Pozman 03 Aug 2019
In reply to Pete Pozman:

I think that for criminals to feel terror we all need to feel it. One of the things that makes me proud of this country is that ordinary citizens don't duck down a side street when they see a policeman approaching. This surprises many visitors: the fact that we are not afraid of the police.

1
wercat 03 Aug 2019
In reply to Pete Pozman:

she might be thinking her proposed treatment of the criminal peaceful protestors, previously good citizens, when Brexit delivers the real goods.

2
Timmd 03 Aug 2019
In reply to Shani:

With three out of the four people I've met who have been to jail, having gone to jail through theft related activities to fund drug habits which were a consequence of child abuse creating a need for them to escape their own heads, it beats me where extra terror would have helped things.

I find it kinda disquieting when the people in charge come out with things like she has...

Post edited at 14:23
DancingOnRock 03 Aug 2019
In reply to Timmd:

Yes. It would be good if maybe she took some advice from criminologists. Beats me why politicians think they should be tinkering in areas of which they have no expertise.

Think they’ll need more than 20,000 extra police. 

Stichtplate 03 Aug 2019
In reply to Timmd:

> With three out of the four people I've met who have been to jail, having gone to jail through theft related activities to fund drug habits which were a consequence of child abuse creating a need for them to escape their own heads, it beats me where extra terror would have helped things.

It’s laudable to try and see the best in others, but only to a point. I’ve met some very bad men in prison, not everyone deserves to be excused their actions.

> I find it kinda disquieting when the people in charge come out with things like she has...

I’d agree with you there.

1
Baron Weasel 03 Aug 2019
In reply to Pete Pozman:

I find it ironic coming from a woman who was fired for basically committing high treason.

Ciro 03 Aug 2019
In reply to Stichtplate:

> It’s laudable to try and see the best in others, but only to a point. I’ve met some very bad men in prison, not everyone deserves to be excused their actions.

"Excusing" (or not) is moralising, and that  should be irrelevant when it comes to policy making around preventing/reducing the effects of crime.

As Shani says above, you can't terrorise people who feel they have nothing to lose - you can lock them up and throw away the key, or rehabilitate them and give them a feeling they have something to lose (I know which I think will be more successful in creating a happy society), but you can't scare people into behaving. 

I assume there's nobody left who thinks it's a successful parenting policy for young children, why do we think it would work better to modify behaviour later?

3
Darron 03 Aug 2019
In reply to Pete Pozman:

She works for Boris. Empty rhetoric is now government policy.

2
Shani 03 Aug 2019
In reply to Stichtplate:

> It’s laudable to try and see the best in others, but only to a point. I’ve met some very bad men in prison, not everyone deserves to be excused their actions.

There's a difference between 'excuse' and 'explain'. Let's leave aside the various flavours of 'free will' and whether we actually have it (i have no choice), and look at the fact that brain injury (http://www.pnas.org/cgi/doi/10.1073/pnas.1706587115) can lead to criminal behaviour. Not only that but normal physiological development - ever wonder why young men in particular are involved in violence and risk taking? From hormones to a chronic lack of sleep, lots of factors out of our immediate control can drive our behaviour. 

If you were in dire poverty or with a loved one in dire danger, your 'rational' course of action may well stray in to illegality.

If you'd been born in Denmark in the 10th Century you'd be happily axing people to death. If you'd been a Briton in 1850s India you'd have been dutifully shooting locals. Even with your current birthdate, if you'd have been born in the foothills of Pakistan, you'd be stoning gays and adulterers to death.

So dont think that if you'd been born in abject poverty on a sink estate in Middlesborough surrounded by drugs, alcohol, spousal abuse, violence, burglary etc... you'd seek to ahem....'fit in' to some extent.

There's a great prison program going on at the moment where prisoners are being taught to cook. It is substantively reducing recidivism. Yes there needs to be a punitive element but what criminals really need is an opportunity. Fear doesn't really come in to it because many people COME from a world of fear.

BBC News - Prison restaurant serves up cut in reoffending https://www.bbc.co.uk/news/education-49150997

summo 03 Aug 2019
In reply to Shani:

> So dont think that if you'd been born in abject poverty on a sink estate in Middlesborough surrounded by drugs, alcohol, spousal abuse, violence, burglary etc... you'd seek to ahem....'fit in' to some extent.

I don't think there would be a huge conscious level of fitting in, you'd just be getting on doing what you have always known without thinking any different. 

What's the worst that can happen, they'd wreck the education they were already on course to fail, they wouldn't get the job that didn't exist for them, they'd go to prison where some of their friends already were. It's a tough cycle to break. Reform through education in prison is probably the only way, as they are nearly guaranteed to go there, compared to offering them the same course in a college for free where they'll just follow their peer group around the streets instead. There is an argument for a form of national service somewhere amongst all this too. 

6
baron 03 Aug 2019
In reply to Shani:

You’ve got to be extremely unlucky to actually get caught if you commit a crime with only 1 in 10 crimes resulting in anyone being charged never mind actually being convicted and sent to prison.

4
Donny M 04 Aug 2019
In reply to Pete Pozman:

When was the last time you saw a ‘police man approaching’ ? 

I live in the centre of Bristol and don’t think I’ve seen a copper on foot for years.  

MonkeyPuzzle 04 Aug 2019
In reply to Donny M:

You don't always see them approaching: My mate Ben was in a house party where, unbeknownst to those inside, a couple were having a huge barny out front. A neighbour was concerned enough to call the police. Meanwhile inside, as Ben had just finished chopping out a bunch of lines on a mirror on his lap he looked up to see a policewoman stood in the living room doorway staring straight at him. Ben nodded at her and said "I'll just get my coat if that's okay".

Post edited at 02:19
Timmd 04 Aug 2019
In reply to Stichtplate:

> It’s laudable to try and see the best in others, but only to a point. I’ve met some very bad men in prison, not everyone deserves to be excused their actions.

I wasn't aware of implying that everybody did, but that could be down to you inferring that from my post. I don't doubt there are some who need to be locked up for the protection of the rest of us, but it strikes me the high rates of illiteracy, and histories of child abuse, and childhoods spent in the care system among prisoners has to be more than a coincidence, too.

Post edited at 12:40
L Wayne45` 04 Aug 2019

I've done that Clink Restraunt programme and it was the best thing that I could have done in prison. It meant that when I got out I was able to get back working and have a focus in life and something to do. I just think that people don't realise how tough it is when you get out of prison and want to do the right thing. You have so many obstacles and people against you that people do need help with life after prison if you want them to stay out.

I had been in Young Offenders before but when I got out I couldn't get a job that I could do, had fallen out with my family and ended u[ back inside. Now I was from a normal family and I was the first one to be in prison so it was hard on them and I suppose getting a longer sentence was a good thing as it gave me time to think and sort things out and start to think about what I wanted in life. I think prison can actually be a good thing for peopler like me is there is actually a point to it and you are learning/doing something that can help you when you get out. 

earlsdonwhu 04 Aug 2019
In reply to Pete Pozman:

Her comments were only meaningful to her party faithful. Its like Blair and his "tough on crime and tough on the causes of crime" slogan. 

Tobes 04 Aug 2019
In reply to Pete Pozman:

> I think that for criminals to feel terror we all need to feel it. One of the things that makes me proud of this country is that ordinary citizens don't duck down a side street when they see a policeman approaching. This surprises many visitors: the fact that we are not afraid of the police.

I guess you might not fit their specific ‘profile’? Ask someone from a different background how they feel about the police and the answer could be slightly different I guess.  

1
Coel Hellier 04 Aug 2019
In reply to Timmd:

> ... but it strikes me the high rates of illiteracy, and histories of child abuse, and childhoods spent in the care system among prisoners has to be more than a coincidence, too.

You're right, it's not a coincidence.  But nor is the reason the one many people think it is.  The biggest reason is not that people who have crap childhoods are more likely to be in jail as a consequence of the crap childhood.

The biggest factor is what genes people are born with.  That's easily the biggest thing that affects how likely someone is to be sent to jail as an adult.  Yes, this then runs in families, but the primary reason for that is not bad parenting, it's that ne'er-do-well parents pass on ne'er-do-well genes to their kids.

The science of this has been pretty much established for a couple of decades now, but for whatever reason the truth of the matter has not yet seeped into public awareness at large.

19
Shani 04 Aug 2019
In reply to Coel Hellier:

> You're right, it's not a coincidence.  But nor is the reason the one many people think it is.  The biggest reason is not that people who have crap childhoods are more likely to be in jail as a consequence of the crap childhood.

> The biggest factor is what genes people are born with.  That's easily the biggest thing that affects how likely someone is to be sent to jail as an adult.  Yes, this then runs in families, but the primary reason for that is not bad parenting, it's that ne'er-do-well parents pass on ne'er-do-well genes to their kids.

> The science of this has been pretty much established for a couple of decades now, but for whatever reason the truth of the matter has not yet seeped into public awareness at large.

I'd like to see evidence of this - it sounds dangerously close go phrenology!

I'm very sceptical of any criminal gene/s. Phenotypic expression would arguably be way more important and even that would be far behind physiological and social factors.

Post edited at 19:30
Offwidth 04 Aug 2019
In reply to Coel Hellier:

More massive oversimplificatiom from our own bargain basement Peterson. There are all sorts of correlates with crime (yes including some with genes and violence) yet few seem to be regarded in the scientific papers as absolutely causal, even for those particular genes that correlate with violence.

https://en.m.wikipedia.org/wiki/Statistical_correlations_of_criminal_behaviour

The fact you think that, although jailed criminal families tend to live in deprived areas and this or they might influence their kids is irrelevant, is telling about your politics being more importamt than your regards for the science.

https://aic.gov.au/publications/tandi/tandi263

Post edited at 19:40
Coel Hellier 04 Aug 2019
In reply to Shani:

> I'd like to see evidence of this

A good recent summary is by Robert Plomin: https://www.amazon.co.uk/dp/B07CV4NPJC/

These days the evidence is not really up for dispute, several decades of studies have settled it.  

> I'm very sceptical of any criminal gene/s.

Complex traits are hardly ever the result of single or a few genes, they are the combined effects of thousands of genes.

15
Shani 04 Aug 2019
In reply to Coel Hellier:

> A good recent summary is by Robert Plomin: https://www.amazon.co.uk/dp/B07CV4NPJC/

Cheers, I'm on it.

wintertree 04 Aug 2019
In reply to Coel Hellier:

> The science of this has been pretty much established for a couple of decades now

Yeah, I read a Nature paper that said there’s something in the DNA that makes people rotten.  They don’t know what it is but the effect is clear to see.  They called it “Dark DNA”.  They needed to give one part in a million a charge of 0.000001 e to make their models of it match observations however...

Coel Hellier 04 Aug 2019
In reply to wintertree:

If you don't agree, can you explain why you think twin studies are getting this wrong?  

John2 04 Aug 2019
In reply to Coel Hellier:

Nature or nurture? You are born into a family of criminals and all of your parents' friends are criminals. You are brought up in an environment where criminal behavior is quite normal. That habituation to criminal behavior is what determines whether or not people are likely to end up in jail in adulthood.

Coel Hellier 04 Aug 2019
In reply to John2:

> That habituation to criminal behavior is what determines whether or not people are likely to end up in jail in adulthood.

No it isn't. Yes, many people think that is the case, they even regard it as "obvious", but it's not true.

The way to find out is twin studies.  Classic twin studies involve adopted children.   These can be identical twins (same genes), separated at birth and adopted into different familes, compared with fraternal twins (share genes to the degree that brothers and sister do), and compared to unrelated adopted children.  

By comparing the outcomes for the different types of twin, one can deduce the relevant importance of genes versus "shared environment" (the similarities in environment that siblings would experience) and "non-shared environment" (the differences that even siblings would experience. 

Twin studies show that "habituation to criminal behavior" is NOT the biggest factor in determining whether or not people are likely to end up in jail in adulthood, instead, genes are. 

Indeed "shared environment" (the thing you point to) is less important than genes *and* less important than non-shared environment. 

9
Timmd 04 Aug 2019
In reply to Coel Hellier:

> You're right, it's not a coincidence.  But nor is the reason the one many people think it is.  The biggest reason is not that people who have crap childhoods are more likely to be in jail as a consequence of the crap childhood.

> The biggest factor is what genes people are born with.  That's easily the biggest thing that affects how likely someone is to be sent to jail as an adult.  Yes, this then runs in families, but the primary reason for that is not bad parenting, it's that ne'er-do-well parents pass on ne'er-do-well genes to their kids.

> The science of this has been pretty much established for a couple of decades now, but for whatever reason the truth of the matter has not yet seeped into public awareness at large.

Do you have any links which thoroughly explore what you describe, and say to what degree genes play a role? A working class painter (I guess you'd call him working class in living on a council estate) I was talking to one time and knew through my family, who was (hopefully still is) a nice guy, once mused to me that 'idiots breed idiots' in relation to the Kelvin flats which used to be in Sheffield, before becoming run down and grim and eventually being demolished, he said they were nice before the idiots moved in. Anyway...enough of my ramblings due to tiredness, some links would be most welcome and interesting. 

Post edited at 21:19
elsewhere 04 Aug 2019
In reply to Coel Hellier:

> If you don't agree, can you explain why you think twin studies are getting this wrong?  

Do they indicate heritability greater than 0.5? If not then I think genetics smaller than environmental factors.

wintertree 04 Aug 2019
In reply to Coel Hellier:

> If you don't agree, can you explain why you think twin studies are getting this wrong?  

I think we know far to little - including Rumsfeld’s famous unknown unknowns - and risk over interpretation of these studies.  We’re only just cracking open the can of worms that is intra-generational epigenetics and that could cut a swathe right through your line of thought.

Edit: Any twin study that relies on separated twins being both more likely to end up criminal seems fundamentally flawed.  Twins are rarely separated without some fundamental problems in the family environment which isn’t something that can be controlled for in a remotely ethical way.

Post edited at 21:17
elsewhere 04 Aug 2019
In reply to Pete Pozman:

> I don't know about criminals but Priti Patel certainly instils terror in me. 

Terrorising criminals? Well I suppose we can follow the examples of ISIS barbarism and paramilitary kneecappings if we want to terrorise those we do not like.

However I think the UK government should aim higher than emulating ISIS or terrorists.

Post edited at 21:31
Timmd 04 Aug 2019
In reply to elsewhere: I sometimes smile at the idea of 'Shoplifters will be persecuted' signs in shops tho, see if anybody notices the difference and says anything, or if less shoplifting happens. 

Post edited at 21:36
elsewhere 04 Aug 2019
In reply to Timmd:

Trespassers will be prostituted - people who can't read a map have it grim.

Wilderbeest 04 Aug 2019
In reply to Wayne45`:

Thank you for posting...

Coel Hellier 04 Aug 2019
In reply to wintertree:

> Twins are rarely separated without some fundamental problems in the family environment

Nowadays, yes, but in the past no.  Before the ready availability of abortion it was common for young, unmarried mothers to give up babies for adoption and it was routine to separate twins.

wintertree 04 Aug 2019
In reply to Coel Hellier:

> Nowadays, yes, but in the past no.  Before the ready availability of abortion it was common for young, unmarried mothers to give up babies for adoption and it was routine to separate twins.

Sure, but when you look at the standards of safeguarding and the use of institutional care in the recent past - let alone the more distant past - it’s hard to imagine any such studies having any real credibility.

TobyA 04 Aug 2019

In reply to Coel Hellier:

I thought it was the Islamic ideology that made terrorists terrorists? Now are you saying it's their genes?

2
Shani 04 Aug 2019
In reply to Coel Hellier:

> If you don't agree, can you explain why you think twin studies are getting this wrong?  

Twin studies have been used by others. Tim Spector (Professor of Genetic Epidemiology at King's College London) has a book caled 'Identically Different' which goes in to heritability and phenotypic expression.

This is the last book i recall reading on genetics (back in 2012), but it was based on twin studies and did seem to pour scorn on the "a gene for X" type thinking. Methylation and histone midification seemed to be much more important. 

Coel Hellier 04 Aug 2019
In reply to Timmd:

> Do you have any links which thoroughly explore what you describe, and say to what degree genes play a role?

There is quite a lot one can read by googling things like Minnesota Twin Family Study and people like Thomas Bouchard, for example this one:

https://www.livescience.com/47288-twin-study-importance-of-genetics.html

But, for an up-to-date and authoritative account I do recommend the recent Robert Plomin book cited above, since developments are happening rapidly in this field, particularly as the ability to perform larger and larger "genome wide association studies" is moving on rapidly. 

Coel Hellier 04 Aug 2019
In reply to Shani:

> ... did seem to pour scorn on the "a gene for X" type thinking.

The "one gene, one trait" idea is certainly way over-simplified.   In reality, it's a vast network of genes mapping to complex traits. 

> ... Methylation and histone midification seemed to be much more important. 

But such things (genes controlling gene expression and genes influencing other genes) is just part of how genetics plays out. 

1
Coel Hellier 04 Aug 2019
In reply to TobyA:

> I thought it was the Islamic ideology that made terrorists terrorists? Now are you saying it's their genes?

Nothing I've said says that genes are the only factor.  Of course environmental factors matter also.   But, note -- again, the twin studies tell us -- that shared environment (the environmental similarities that siblings would share) matters a lot less than non-shared environment (environmental factors that differ between siblings).   That is counter-intuitive to most people, but, again, if you want to argue against it then you have a lot of hard science to rebut. 

And note that the vast majority of those exposed to Islamic ideology do *not* turn to terrorism. So of course lots of other factors are relevant, and twin studies tell us that genetics sets our personality, and thus how we act, to a much greater extent than most people realise. 

Coel Hellier 04 Aug 2019
In reply to wintertree:

> Sure, but when you look at the standards of safeguarding and the use of institutional care in the recent past - let alone the more distant past - it’s hard to imagine any such studies having any real credibility.

Then why the difference between identical twins, separated at birth and adopted into different families, versus fraternal twins separated at birth and adopted into different families? 

The science here is solid, nothing you've said even begins to be a rebuttal. 

wintertree 04 Aug 2019
In reply to Coel Hellier:

> Then why the difference between identical twins, separated at birth and adopted into different families, versus fraternal twins separated at birth and adopted into different families? 

I’ve already mentioned epiginetics as one example that you’ve ignored.  Then the woeful state of historic safeguarding to me makes the data garbage.

Shani 04 Aug 2019
In reply to wintertree:

> epiginetics

That is methylation (on/off) and histone modification (like a volume control). 

wintertree 04 Aug 2019
In reply to Shani:

> That is methylation (on/off) and histone modification (like a volume control). 

Yes, but Coel reduced it to “just a part of how genetics plays out” which ignores the growing number of examples where consequences of environmental conditions can persist for multiple generations beyond exposure to them through epigenetic pathways.  

Recently, I’ve been working with a model system that has epigentic memory of at least 3 generations.  Excluding biome, the model system is over 30x less bass pairs than humans.  We’re only just learning how parental environmental exposure changes the life of progeny for simple organisms so to go claiming criminality is definitively in the DNA of humans is a few steps ahead of where we are.  Family is more than just upbringing, it’s how your DNA is run-time configured to express itself.

Shani 05 Aug 2019
In reply to wintertree:

Interesting. I've come across it a few times in literature in famine. Particularly the Dutch Winter Famine. Like you say, generational changes can occur.

Shani 05 Aug 2019
In reply to Shani:

Knew I'd discussed some of this on UKC before (2013):

https://www.ukclimbing.com/forums/walls+training/how_easy_to_lose_weight_really-532894?v=1#x7173486

Post edited at 07:29
JoshOvki 05 Aug 2019
In reply to Wayne45`:

I ate at the one in Cardiff about a year ago and the food was amazing. I didn't realise until a few days ago that there were a few dotted across the country.

For those of you who don't know Clink Restraunt programme, it is basically fine dining.

https://www.bbc.co.uk/news/education-49150997

cb294 05 Aug 2019
In reply to elsewhere:

Most people intuitively get "environmental factors" wrong. We are largely not talking about parenting or school experience, that comes later and is for many traits a minor factor. Much more important is what influences development during pregnancy and early after birth, but is not coded for by the child's DNA. This includes e.g. disease or nutritional state of the mother, hormonal environment or drug exposure (alcohol or nicotine) in the womb etc. . I would therefore not be bothered about the exact level of heritability for any given trait.

Still, Coel's point stands (despite certain mistakes): Many aspects about our social traits are determined by genetics to a larger extent than people appreciate. This includes a tendency to disregard social norms or the well being of others, resulting in behaviour that a society may or may not classify as criminal criminal (as pointed out by someone else above, what is normal and what criminal is highly context dependent!). Heritability is even higher for anything associated with sexual behaviour, e.g. the tendency to promiscuity/infidelity/serial monogamy.

1
cb294 05 Aug 2019
In reply to Coel Hellier:

Please see my reply to elsewhere above: Twin studies systematically overestimate heritability, as even twins separated at birth share key bits of their "environment". This is actually a big advantage of GWA studies, which are, however, much harder to interpret in other respects.

CB

John2 05 Aug 2019
In reply to Coel Hellier:

'they discovered they both suffered from tension headaches, were prone to nail biting, smoked Salem cigarettes, drove the same type of car and even vacationed at the same beach in Florida'

'Genes can help explain why someone is gay or straight, religious or not, brainy or not, and even whether they're likely to develop gum disease'

'Another study of four pairs of twins found that genetics had a stronger influence on sexual orientation in male twins than in female twins'

'The Minnesota researchers found that about 70 percent of IQ variation across the twin population was due to genetic differences among people, and 30 percent was due to environmental differences'

'genetics account for 50 percent of the religiosity among the population'

'Other studies found a strong genetic influence on dental or gum health'

Nothing about criminal behavior in the article you link to.

Coel Hellier 05 Aug 2019
In reply to wintertree:

> Yes, but Coel reduced it to “just a part of how genetics plays out” which ignores the growing number of examples where consequences of environmental conditions can persist for multiple generations beyond exposure to them through epigenetic pathways.  

First, that's pretty irrelevant to twin studies, which are not multi-generational, but are comparing identical twins separated at birth with fraternal twins separated at birth.

Second, much about these claims of multi-generational epigenetic effects is exaggerated and based on small effects in marginal data. 

>  Then the woeful state of historic safeguarding to me makes the data garbage

That reply doesn't begin to be adequate. Why would such things have systematically different effects on identical twins separated at birth compared with fraternal twins separated at birth?

You're dismissing a whole field of established, solid science for no good reason, other than dislike of what it is telling us. 

Coel Hellier 05 Aug 2019
In reply to cb294:

> Twin studies systematically overestimate heritability, as even twins separated at birth share key bits of their "environment".

Yes, they share a womb and environment up to their separation.

However, that is not relevant to the claim that habituation to a criminal environment in childhood, and/or an abusing childhood, are the biggest factors in who ends up in jail.

2
Coel Hellier 05 Aug 2019
In reply to John2:

> Nothing about criminal behavior in the article you link to.

You're right, it's about genetic versus environmental differences in general. 

cb294 05 Aug 2019
In reply to John2:

It is, nevertheless, true that the risk of criminal behaviour has a genetic contribution. Actually, I don't believe that there is a conspiracy to hide this "non-PC fact". Rather, it seems blatantly obvious and therefore too boring to spend much scientific effort on the narrow question of heritability of criminal behaviour. Still, there is a steady stream of psychological/genetics papers on the topic. Here is one 2015 example from the journal "Psychological Medicine" picked from the top of a brief PubMed query using "criminality heritability" as search strings:

A Swedish national twin study of criminal behavior and its violent, white-collar and property subtypes

K. S. Kendler (a1) (a2) (a3), H. H. Maes (a1) (a3), S. L. Lönn (a4), N. A. Morris (a5) ... 

DOI: https://doi.org/10.1017/S0033291714002098

Abstract

Background

We sought to clarify the etiological contribution of genetic and environmental factors to total criminal behavior (CB) measured as criminal convictions in men and women, and to violent (VCB), white-collar (WCCB) and property criminal behavior (PCB) in men only.

Method

In 21 603 twin pairs from the Swedish Twin Registry, we obtained information on all criminal convictions from 1973 to 2011 from the Swedish Crime Register. Twin modeling was performed using the OpenMx package.

Results

For all criminal convictions, heritability was estimated at around 45% in both sexes, with the shared environment accounting for 18% of the variance in liability in females and 27% in males. The correlation of these risk factors across sexes was estimated at +0.63. In men, the magnitudes of genetic and environmental influence were similar in the three criminal conviction subtypes. However, for violent and white-collar convictions, nearly half and one-third of the genetic effects were respectively unique to that criminal subtype. About half of the familial environmental effects were unique to property convictions.

Conclusions

The familial aggregation of officially recorded CB is substantial and results from both genetic and familial environmental factors. These factors are moderately correlated across the sexes suggesting that some genetic and environmental influences on criminal convictions are unique to men and to women. Violent criminal behavior and property crime are substantially influenced respectively by genetic and shared environmental risk factors unique to that criminal subtype.

2
cb294 05 Aug 2019
In reply to Coel Hellier:

But it is very much relevant to a claim that some trait is predominantly caused by genetics. Twin studies will only give an upper bound, GWAS approaches are much better at identifying the actual genetic contribution.

Also, just for clarification, heritability in this context means the fraction to which some chosen subset of genetic differences (or, in the case of twin studies, the sum of all differences) explains the observed variation in a trait across the population.

We are thus not discussing the absolute likelihood whether "someone goes to jail", but the differences in this likelihood between some group and the chosen control group.

"Going to jail" is also a particular useless measurement for criminal behaviour if you want to disentangle educational, socioeconomic, and genetic causes: We are all aware of how conviction rates and harshness of sentences can vary for the same crime, depending of who stands before the jury.

I agree, though, that the general public appears to underestimate to which extent our behavioural and psychological traits are shaped by our genes, including - but certainly not only with respect to -criminality. Whether it is, for the latter, the "biggest" contribution of many appears IMO pretty much irrelevant.

CB

wintertree 05 Aug 2019
In reply to Coel Hellier:

> First, that's pretty irrelevant to twin studies, which are not multi-generational, but are comparing identical twins separated at birth with fraternal twins separated at birth.

I don’t think you understand.  Effects are passed intra-generationally through epi genetic effects.  This means there are shared, common environmental traits affecting twins separated at birth.  

> Second, much about these claims of multi-generational epigenetic effects is exaggerated and based on small effects in marginal data. 

Well, let’s revisit that once it’s not a fast paced, developing field.

> That reply doesn't begin to be adequate. Why would such things have systematically different effects on identical twins separated at birth compared with fraternal twins separated at birth.

Looking significance in tainted data.  It’s just not good science.  The more we find out about institutional rehoming of children the less I trust studies on children separated in past decades.  To many unknown variables.

> You're dismissing a whole field of established, solid science for no good reason, other than dislike of what it is telling us. 

“Solid” is a big stretch.  Twin studies are rooted in a rather old methodology that predates an understanding of epigenetics.   Me - I think it’s a bad time to be drawing strong conclusions from twin studies.

Post edited at 09:40
1
Coel Hellier 05 Aug 2019
In reply to wintertree:

> I don’t think you understand.  Effects are passed intra-generationally through epi genetic effects.  This means there are shared, common environmental traits affecting twins separated at birth.  

And, despite me asking the question three times, you've still given to reason why this is different for identical twins (separated at birth) versus fraternal twins (separated at birth), which is the crucial part of many twin studies. 

> Well, let’s revisit that once it’s not a fast paced, developing field.

There seems to be a lot of over-claiming about epigenetics.  For example, for a skeptical assessment, see:

http://www.wiringthebrain.com/2018/05/grandmas-trauma-critical-appraisal-of.html

https://whyevolutionistrue.wordpress.com/2018/05/29/the-flimsy-evidence-for-epigenetic-changes-in-dna-to-be-transmitted-between-generations-of-humans/

> Looking significance in tainted data.  It’s just not good science.  The more we find out about institutional rehoming of children the less I trust studies on children separated in past decades.  To many unknown variables.

Sorry, but that's just empty waffle.   You need a  substantive reason why this affects identical twins (separated at birth) differently from  fraternal twins (separated at birth), otherwise the studies are sound.

Coel Hellier 05 Aug 2019
In reply to Coel Hellier:

> There seems to be a lot of over-claiming about epigenetics.  For example, for a skeptical assessment, see:

I'll give a quote from this:

"The way this notion [multi-generational epigenetic effects] is referred to – both in popular pieces and in the scientific literature – you’d be forgiven for thinking it is an established fact in humans, based on mountains of consistent, compelling evidence. In fact, the opposite is true – it is based on the flimsiest of evidence from a very small number of studies with very small sample sizes and serious methodological flaws."

http://www.wiringthebrain.com/2018/05/grandmas-trauma-critical-appraisal-of.html

The piece then goes on to substantiate that assessment.

wintertree 05 Aug 2019
In reply to Coel Hellier:

> And, despite me asking the question three times

Sorry; I got derailed by you not recognising that it’s an intra-generational effect.

> you've still given to reason why this is different for identical twins (separated at birth) versus fraternal twins (separated at birth), which is the crucial part of many twin studies. 

As I've said before, the field has a lot of unknown unknowns.  Depending on which factors trigger epigenetic changes, and when they program the germ cells or the embryo, they could affect different sorts of twins in the same way or in different ways.

> There seems to be a lot of over-claiming about epigenetics.  For example, for a skeptical assessment, see:

Funny how you list popular science references here.

> Sorry, but that's just empty waffle.   You need a  substantive reason why this affects identical twins (separated at birth) differently from  fraternal twins (separated at birth), otherwise the studies are sound.

I’m parked at the “reasonable doubt” stage.  The methodology of such studies is nowhere near as clear cut as previously assumed for several independent reasons.  Not enough is known to prove them valid or invalid to a substantiate capacity.  There’s also the “replication crisis” to consider - more and more evidence of all sorts of studies being hard to replicate.  

1
Coel Hellier 05 Aug 2019
In reply to wintertree:

> Funny how you list popular science references here.

The piece I linked to is written by an expert in genetics at Trinity College Dublin.  This guy:  https://www.tcd.ie/research/profiles/?profile=KEMITCHE   

>  There’s also the “replication crisis” to consider - more and more evidence of all sorts of studies being hard to replicate.  

Which is VASTLY more relevant to studies claiming epigenetic effects of the sort that would confound twin studies.   Can you point to even one such study that is solid and replicated?

In contrast, twin studies have been replicated many times. The Minnesotta studies started back in 1983, and in the three decades since then many other studies have replicated the results and found much the same thing.   If you want to dispute what is now a well-established and solid field you need to do better than mere talk of "unknown unknowns". 

cb294 05 Aug 2019
In reply to wintertree:

Twin studies clearly show that all aspects of human behaviour, including a propensity for criminal or antisocial behaviour (where the former much more than the latter depends on arbitrary social definitions!) are shaped by genetics. The actual level of heritability is hard to assess using twin studies, especially if the trait to be quantified is ill defined and confounded (e.g. conviction rates are easy to measure, but correlation with the underlying behavioural / psychological trait influenced by genetics will differ between social groups).

Epigenetics is another hard one. I would argue that the intra-generational effect is obvious but boring , i.e., this is pretty much what is meant by "environment": Whatever happens to a developing fetus to influence, say, brain development, will be mediated by genes being turned on or off by epigenetics, either by chance or as a consequence of life history.

Evidence for inter-generational transmission of epigenetic states is weak for humans (with the exception of the few loci where imprinting mistakes cause major problems), and will for fundamental reasons be hard to separate from "environment", in particular nutrition during development (overall chromatin modification levels critically depend on the levels of certain intermediate products of the energy metabolism).

CB

Coel Hellier 05 Aug 2019
In reply to wintertree:

> Sorry; I got derailed by you not recognising that it’s an intra-generational effect.

Did you mean "intra-generational" there, and not "inter-generational"?

And yes, I do recognise that there are both intra-generational epigenetic effects and, possibly (though not on a secure evidence basis) inter-generational effects.       Again, there's no good reason why either would confound twin studies (other than handwaving at "unknown unknowns"). 

Intra-generational effects are clearly "environment" in twin studies (as cb294 says). As for inter-generational effects, unless they are systematically different for identical versus fraternal twins (any evidence of that at all?) then it also does not confound twin studies. 

Coel Hellier 05 Aug 2019
In reply to Coel Hellier:

Here's a later assessment by the Trinity College Dublin professor of Neurobiology and Genetics:

http://www.wiringthebrain.com/2018/07/calibrating-scientific-skepticism-wider.html

AllanMac 05 Aug 2019
In reply to Coel Hellier:

Is it not true that phenotypic expression would confer only a propensity towards criminal traits, while upbringing/habituation might allow its free expression - or not? In other words, if a genetic propensity towards criminality is genetically present, yet habituation and upbringing disallows its free expression, would that person remain observably law-abiding, despite the subliminal urges within genes?

cb294 05 Aug 2019
In reply to Coel Hellier:

That blog post looks, from a quick read, like a fair summary.

The one criticism I have is the box summarizing TGEI as leading from experience to phenotype again in a few specific neurons in the next generation. This is an unnecessary exaggeration. 

Overall conditions globally altering gene expression fixed by epigenetics , including but nor exclusively in the maternal brain (e.g. nutrition or stress hormone levels during fear conditioning in mice), leading to rather broad brush phenotypes (e.g. synapse density or whatever, I am not a neuro development guy), which then re-occur during development of the offspring is entirely more believable.

CB

Coel Hellier 05 Aug 2019
In reply to AllanMac:

> Is it not true that phenotypic expression would confer only a propensity towards criminal traits, while upbringing/habituation might allow its free expression - or not? In other words, if a genetic propensity towards criminality is genetically present, yet habituation and upbringing disallows its free expression, would that person remain observably law-abiding, despite the subliminal urges within genes?

Yes, true.  In all cases both the genetic recipe and the environmental influences and outplaying of that genetic recipe are important. 

Offwidth 05 Aug 2019

In reply to Coel Hellier:

You're simply an embarrassment to science with your willful exaggeration. You are not an idiot, so saying such things has to be political. If you want to play pots and kettles show us the evidence.

As a reminder, you said "The biggest factor is what genes people are born with. That's easily the biggest thing that affects how likely someone is to be sent to jail as an adult." Thats way beyond current science where at most some traits like a tendancy to excessive violence have been significantly linked to genetics, but often where drug and/or other treatments are said to reduce risk, clearly indicating other factors are at play. If there was evidence for anything close to such a massively grandiose claim across what you say are a majority of genetically pre-programmed in the incarcerated don't you think we would all know about it, or is this a tin hat time cover-up as well?

7
Offwidth 05 Aug 2019
In reply to AllanMac:

What might be, is just a hypothesis in science.  Coel is saying it's true in the face of a highly complex area of scientific interacting factors... genetics, foetal effects, family and environment; and he claims, godlike, that genes are the largest determining factor for someone ending up in jail. This is bullshit science, even if, when the evidence is much better at some future time, it rather surprisingly turns out to be true.

In the paper he links from Kevin  are some wise words:

"When I see a paper making an extraordinary claim I think it is appropriate to judge it not just on the statistical evidence presented within it, which refer only to the parameters of that particular experiment, but on the more general relative likelihood set up here: Which is more likely? That the researchers have happened on some truly novel and extraordinary biology or that something funny happened somewhere?"

Coel doesn't even have the statistics, other than everyone in jail having genes.

I guess from the dislikes above some think I'm holding him to too high a standard but I think that is beholdant on Science Profs when dealing with the scientific method... if a Social Scientist Prof was spouting similar unevidenced  grandiose claims in astrophysics based on exaggerating something that might be evidence in that direction  I would be calling them out  in the same way.

Post edited at 13:44
4
Coel Hellier 05 Aug 2019
In reply to Offwidth:

> In the paper he links from Kevin  are some wise words:

OK, so let's quote it again:

"When I see a paper making an extraordinary claim I think it is appropriate to judge it not just on the statistical evidence presented within it, which refer only to the parameters of that particular experiment, but on the more general relative likelihood set up here: Which is more likely? That the researchers have happened on some truly novel and extraordinary biology or that something funny happened somewhere?"

The idea that genes affect our personalities is not "novel and extraordinary biology", it's exactly what we expect! Do you think that genes don't do anything?  That they are just there for show? 

On the face of it, the idea that genes affect how our brains function (and thus our personalities) is no more a novel and extraordinary idea than the idea that genes affect how our livers function or how our cardio-vascular system works. 

1
Offwidth 05 Aug 2019
In reply to Coel Hellier:

I'm fully aware of the evidence of the twin studies. None of it shows its most likely that its genes that get you locked up, it just indicates in a few studies that its slightly more probable with certain genetic backgrounds in certain types of crime. Your claim is not even close to being justified in anything I've seen and would be massive scientific news if true.

https://law.jrank.org/pages/784/Crime-Causation-Biological-Theories-Genetic-epidemiological-studies.html

cb294 05 Aug 2019
In reply to Offwidth:

I don't normally agree with Coel, but he has a point in that the heritability of criminal behaviour is indeed much higher than normally appreciated, especially amongst people actively trying to change the environmental effects (which I find entirely laudable): That a parental influences, peer effects, education, poverty etc. don't help is obvious, but ignoring or underestimating the genetic component is actually counterproductive. Same goes for teaching and educational attainment, but that is a different issue (or maybe not, but in any case better left to a different thread...).

In contrast, for me as a biologist it is not at all surprising that the behavioural propensities eventually causing criminality are to a significant extent heritable and hence determined by genetics: This is true for so many other aspects of our behaviour, so it would be odd if it were different for those traits influencing our social interactions, empathy, altruism/egoism etc. that will in turn increase or decrease our likelihood of engaging in criminal behaviour. In addition, whether the resulting actions are considered criminal or not is a matter of societal conventions, which adds another layer of complexity.

In my posts above I have tried to address two issues that I think contribute to muddling the water: These are, first, the difference between heritability of a trait and its overall prevalence which sees a fair amount of abuse, and second, the difference between societal / family influences and what heritability studies sum up as "environment".

Whether you end up calling "nature" the single biggest contributing factor in explaining the variance in criminal behaviour, or the second or tenth biggest one, is IMO a bit meaningless anyway, as this ranking depends largely on into how many different factors you split the "nurture" side.

Also, given the unfairness of our judicial systems with respect to socioeconomic status, being in jail is a shit proxy for criminality! This does not change the fact that if you are jailed there is an increased conditional likelihood that you previously had a shit hand dealt in the genetic lottery, quite likely on top of a disadvantageous environment.

Actually, in most western criminal justice systems, the most common psychological / behavioural trait that is both clearly influenced by genetics and also definitely influences your chance of conviction is bound to be susceptibilty to substance addiction. I have no data on that, but would bet that this is much more common than psychopathy or a propensity towards violence.

Don't even get me started on various mental illnesses (which again show higher and higher heritability the better our genetic methods get) and your chances of ending up in jail or getting treatment, depending on your socioeconomic status.

CB

edit: Just had a look at your link. Funnily, the summarizing paragraph starts as follows:

"These data, obtained from three different countries and in different laboratories, lend support to the notion that criminal behavior appears to have a strong genetic component." 

Post edited at 15:25
Shani 05 Aug 2019
In reply to cb294:

The problem is, it's not a case of "gene/s = criminal behaviour".

That people with certain genes are more disposed to violence may get them a medal on a battlefield but land them in jail on civvy street. Genes are not deterministic. They're not destiny.

Offwidth 05 Aug 2019
In reply to cb294:

I'd pretty much agree with your entire post except your apparent assumptions about my views. Yes I fully agree some criminality seems to have a strong genetic component.. I believe in any proper confirmed evidence based on good research methodology. The article is one of many that fairly discusses that with other factors. However Coel claimed that "The biggest factor is what genes people are born with. That's easily the biggest thing that affects how likely someone is to be sent to jail as an adult."  This is shameful exaggeration from a Science Prof and looks very much politically motivated. No maybes, no broader context that he thinks its the largest of many signifcant factors, just something no one really knows, stated as fact.

1
cb294 05 Aug 2019
In reply to Shani:

Where have I claimed that? In fact I agree with you, but that does not change the fact that criminal behaviour is heritable to a larger extent than you seem to be prepared to accept. 

CB

Coel Hellier 05 Aug 2019
In reply to Offwidth:

>  This is shameful exaggeration from a Science Prof and looks very much politically motivated.

Feel free to continue your science denialism Offwidth, accompanied by personal attacks.  You're as bad as those who deny the mainstream consensus on climate change by huffing with irrelevant whataboutery. 

To everyone else, since several of my posts replying to Offwidth have been deleted (presumably Offwidth complained?, even though there was less personal attack in them than in his posts to me), I won't bother trying to discuss the science with him. 

2
cb294 05 Aug 2019
In reply to Offwidth:

Yes Coel's claim is quite likely deliberately exaggerated, but may possibly be true*, unlike the "nurture only" hypotheses I often hear from the teachers and social workers among my friends. Again, these topics come up more often with respect to education, but I think the issues are related, in particular with respect to "environmental" risk factors.

Since the genetic contribution to criminality clearly is not 100% we should definitely not give up in improving the lot of children suffering from the other risk factors such as poverty.  In fact, acknowledging the genetic component means that trying to influence the "environmental" circumstances becomes even more important and urgent (as there is a smaller fraction of the overall variation to play with).

CB

* As I said, whether genetics is the biggest or tenth biggest factor is irrelevant. However, for the benefit of society at large we should acknowledge the existence of a larger than normally appreciated genetic contribution to criminality, especially when designing policy aimed at reducing crime. For example, I am 100% sure that decriminalising drug abuse would immediately lower the genetic contribution to criminality, as susceptibilty to addiction definitely has a strong genetic component!

Offwidth 05 Aug 2019
In reply to cb294:

Given the history of genetic determinism and eugenics I'm surprised that anyone in science still thinks its a good idea to risk exaggerating genetic effects; some on the political far right still dishonestly push the opposite position almost as a theology. In such a political context it's hard work getting agreement to do the research let alone face real evidence, so I think being careful and exact is especially important on such a subject

Rob Exile Ward 05 Aug 2019
In reply to Coel Hellier:

It has seemed obvious to me for a very long time that if genetics could affect some organs e.g. skin, then they must equally have an affect on the brain. As you say, to think that genetics only applies from the neck down is nonsense.

However, I'm less clear where that leads or leaves us. Surely it is sensible to focus on those factors we can do something about - nurture, parenting, environment, education, opportunity - than over-emphasise those we can't - our genes. Otherwise these arguments will be adopted by the Right - they have been already - and that way a Brave New World lies.

L Wayne45` 05 Aug 2019
In reply to JoshOvki:

Ha glad you enjoyed it. I was in the one at Brixton though. People always said that they were surprised that the food was so good. Good to have a chat with "normal" people too. Really made me feel I could be good at something and gave me a focus

In reply to Coel Hellier:

> To everyone else, since several of my posts replying to Offwidth have been deleted

I've asked you once not to use Offwidth's real name in replies. It's not that hard is it?

cb294 05 Aug 2019
In reply to Offwidth:

I think you have this entirely the wrong way round. If anything, the "political mainstream" is still underplaying the influence of genetics and our evolutionary history, leading to misguided policy development on a wide range of topics. IMO the problem is too many social science / humanities graduates ignoring results from STEM subjects either because they don't fit the desired narrative (something of which the political right is clearly also guilty), or, as I suspect, because it is simply too complicated.

CB

edit: I am a life long social democrat / green, so I am arguing against my own tribe: Education (and crime) is where they consistently get it wrong for entirely predictable and obvious reasons. It drives me to despair!

Post edited at 17:28
1
Coel Hellier 05 Aug 2019
In reply to Rob Exile Ward:

> Surely it is sensible to focus on those factors we can do something about - nurture, parenting, environment, education, opportunity - than over-emphasise those we can't - our genes.

Absolutely, you should focus on such interventions. But to get them right you need to properly understand what the causes of things are!  If you understand what relates to genetics, what to shared environment, and what to non-shared environment, then you can properly implement and target your interventions. 

If, however, you ignore the science and go with intuition, perhaps just saying (quoting from up-thread): "You are born into a family of criminals ...  That habituation to criminal behavior is what determines whether or not people are likely to end up in jail in adulthood" -- which is largely wrong -- then you will get your interventions wrong. 

Post edited at 17:34
1
Coel Hellier 05 Aug 2019
In reply to Paul Phillips - UKC and UKH:

> I've asked you once not to use Offwidth's real name in replies. It's not that hard is it?

So Offwidth is happy making public personal attacks on someone using their real name, such as: "You're simply an embarrassment to science ...", but is unwilling to publicly stand behind his comment?

Is it within the rules to regard this as rather cowardly? 

3
John2 05 Aug 2019
In reply to Coel Hellier:

You have quoted my words as an example of why  people with whom you disagree are wrong. But as I pointed out earlier, the article to which you linked makes no mention of criminal behavior.

One of the UK's best climbers was the child of two keen but very middle of the road climbers. They used to take him climbing regularly from an early age, and by the age of 12 he was far surpassing their climbing achievements. Nature or nurture? He had the same genes as they did.

1
Coel Hellier 05 Aug 2019
In reply to cb294:

> Yes Coel's claim is quite likely deliberately exaggerated, but may possibly be true ...

I'd go with "baldly stated" rather than "deliberately exaggerated". 

Here are some quotes from Robert Plomin (perhaps the world's leading expert on this topic).

"For most of the 20th century, environmental factors were called nurture because the family was thought to be crucial in determining environmentally who we become. Genetic research has shown that this is not the case. We would essentially be the same person if we had been adopted at birth and raised in a different family."

"DNA isn’t all that matters, but it matters more than everything else put together in terms of the stable psychological traits that make us who we are."

https://blogs.scientificamerican.com/observations/in-the-nature-nurture-war-nature-wins/

"And yet the first essential proposition of Plomin’s book, that “we resemble our parents and our siblings because we are similar to them genetically, not because we have grown up in the same environment and experience the same opportunities or traumas”, is, says the author, purely consensual among geneticists. As is the implication from most research that, “We would be just as similar to our parents and siblings even if we had been adopted apart at birth and reared in different families.” Plomin tells me that so widespread is agreement on this nowadays, some of the abuse that used to go with espousing the science has disappeared."

https://www.thetimes.co.uk/article/robert-plomin-interview-why-genetic-testing-is-the-future-m2gcskpkv

Somewhat baldly stated, yes, and perhaps a bit exaggerated.  But the point is to overcome the refusal of swathes of society to properly address these issues, as exemplified by this thread, and for that a bit of stark language is in order. 

cb294 05 Aug 2019
In reply to John2:

> Nature or nurture? He had the same genes as they did.

Both, of course, and no he did not: With the same likelihood his hypothetical younger brother could be a less gifted climber than his parents. Think about it!

CB

cb294 05 Aug 2019
In reply to Coel Hellier:

> I'd go with "baldly stated" rather than "deliberately exaggerated". 

Fair enough! As you can see from my posts, I pretty much agree with the science, and am familiar with Plomin's work.

CB

Coel Hellier 05 Aug 2019
In reply to John2:

> But as I pointed out earlier, the article to which you linked makes no mention of criminal behavior.

And plenty of other articles do deal with criminal behaviour, including the one up-thread that cba cited and quoted. 

Again, the truths about the relative influences of genetic, shared environment and non-shared environment, are far broader than just crime, applying to all psychological traits. 

> They used to take him climbing regularly from an early age, and by the age of 12 he was far surpassing their climbing achievements. Nature or nurture? He had the same genes as they did.

A mixture of both -- as everything is.  And yes, aptitude for climbing will indeed have a strong genetic component (and yes, of course that will not be expressed if the environmental exposure to climbing does not occur).  And no, the kid does not have the same genes as the parents, he has half of one parent's genes and half of the other parent's, and thus has a novel mix that is similar to but different from either of them. 

Shani 05 Aug 2019
In reply to cb294:

> * As I said, whether genetics is the biggest or tenth biggest factor is irrelevant. However, for the benefit of society at large we should acknowledge the existence of a larger than normally appreciated genetic contribution to criminality, especially when designing policy aimed at reducing crime. For example, I am 100% sure that decriminalising drug abuse would immediately lower the genetic contribution to criminality, as susceptibilty to addiction definitely has a strong genetic component!

This is the crux for me. AFAIK epigenetics affect homosexual orientation. Whether that makes you a criminal depends on where you were born and in what year.

As an interesting aside to this debate, speed is genetic, but to be the fastest takes more; if you look at the progression of the 100m world record & add in the birth order of the record holder, something pretty interesting happens:

1. Usain Bolt (second of three)
2. Safa Powell (sixth of six)
3. Justin Gatlin (fourth of four)
4. Maurice Greene (fourth of four)
5. Donovan Bailey (third of three)
6. Leroy Burrell (fourth of five)
7. Carl Lewis (third of four)
8. Leroy Burrell (fourth of five)
9. Carl Lewis (third of four)
10. Calvin Smith (sixth of eight)

The distribution seems heavily skewed towards those born further down the birth order - and the same pattern apparently exists in the all time top 10 NFL running backs in rushing yardage.

wintertree 05 Aug 2019
In reply to cb294:

> Evidence for inter-generational transmission of epigenetic states [...]  will for fundamental reasons be hard to separate from "environment", in particular nutrition during development 

Totally agree.  In humans it’s going to be really hard to separate these factors.  

cb294 05 Aug 2019
In reply to Shani:

Yes homosexuality can be environmental (e.g. hormone exposure in the womb) as well as genetic. Both triggers will be translated first into differential gene activity during development and then a psychological phenotype. Both modes will, amongst other things, involve epigenetics in establishing the specific gene expression patterns. I would therefore not call homosexuality an epigenetic condition. 

The correlation of speed and birth order is quite obviously environmental*: Younger sibs are generally a PITA and thus have to run away from their older sibs a lot to escape their rightful punishment, giving them an unfair early training advantage.

CB, 1st of 3

* Seriously, this is interesting. Never seen this before. Will have to check whether there are scientific  papers on the potential mechanism!

Shani 05 Aug 2019
In reply to cb294:

One cool environmental factor involves French 100m runner Christophe Lamaitre (first white guy under 10s for 100m).

The TGV went past his childhood home and he used to race it between two fixed points on his street!

wintertree 05 Aug 2019
In reply to Offwidth:

> This is shameful exaggeration from a Science Prof and looks very much politically motivated.

It does rather when you book end it with this comment “But the point is to overcome the refusal of swathes of society to properly address these issues”

How do we properly address the issue of genetic contributions to criminality?  Eugenics?  Sterilisation?  Retro-virally reprogramming embryos?  Reduced sentencing for those at the mercy of their genes?  Offer chemical decriminalisation as an alternative to prison?  Genetically targeted intervention with young people?

The only “proper” way I can see of addressing the genetic component of the link is to prevent genetic discrimination as with all other aspects of life.  

As with other problems with a genetic component, the “proper” way to deal with it in my view is to reduce the societal and environmental exposure to the factors that trigger the genetic potential.  Genetics plays a role in risk of developing asbestosis but the correct response was to do away with asbestos.  

Post edited at 18:39
2
cb294 05 Aug 2019
In reply to wintertree:

Your list of options is pointless and needlessly inflammatory: It will not even be necessary to know the genetic makeup of an individual for society to benefit from knowing which aspects of criminal behaviour is to which extent heritable.

It is the same issue as with, say, women's engagement in STEM subjects, or improving educational attainment in problem demographics: If you do not understand in detail the cause of a problem it is hard to rationally devise policies to achieve the desired outcome. The current approach is that policies are routinely based on ideology (e.g., all children have identical potential and differ only in nurture), resulting in shit policies and predictable, continuous failure. Rubbish in, rubbish out.

In education, it is an easy problem to solve, though: Allow for some continuous lowering of standards, and voila, degrees all around. That this makes it in the long run even more unfair for gifted children from non-academic families (e.g. because prospective employers now have to look at other aspects than grades, e.g. unpaid internships that you have to first be able to afford), who cares?

Thus, on which factor statistically associated with an increased likelihood to lead to criminal behaviour would you spend the limited resources available? I have no idea, and am not sure anyone else does as yet, but I am sure that correctly identifying and attributing the genetic and environmental aspects, and then focusing on those where a realistic invention could yield the largest effect. e.g. because it contributes most to the variance on the "environmental" side, would be a reasonable first step.

By all means keep the risks you allude to in mind, but sticking the fingers in the ears and going lala just because knowledge of the genetic side may also used to construe unpalatable solutions is unhelpful. Dickens already knew that the same is true for the environmental side!

CB

edit: I absolutely agree with the second half of your post on reducing the factors likely to trigger genetic predispositions!

Post edited at 19:13
wintertree 05 Aug 2019
In reply to cb294:

I agree with tour post and I was being deliberately extreme - but that is to a degree what I read into the quote I made from Coel’s post.  

The problem is not that society doesn’t understand the details linking genetics and criminality.  There is much about environment and upbringing that is well understood and that is not acted on by society yet Coel appears to suggest our failing is society’s refusal to acknowledge genetics.  There are other societies with the same sort of understanding as the UK but far less criminality.  At best the genetic aspect is a red herring from problems we already know how to solve far better than we do; at worst it’s the start of a dark path.

I agree on basing policy on understanding not ideology but the gulf between those in the UK is already vast - I’m sure the details on genetic contribution will continue to firm up.  I’m not so sure they’ll make a jot of difference to UK policy, and I’m absolutely sure - to paraphrase Coel - the problem is not the refusal of vast swathes of society to recognise the genetic influence on criminality.

Edit: my experience of STEM inclusivity in the UK is at odds with what you describe - understanding how being shaped by different personal and societal backgrounds interacts with unintentionally loaded systems to benefit some more then others, and then asking how to make new systems with as little bias baked in as possible.  None of this involved understanding genetic links so much as understanding how different people think, act and behave.  

Post edited at 19:25
1
Coel Hellier 05 Aug 2019
In reply to wintertree:

> There is much about environment and upbringing that is well understood and that is not acted on by society ...

Can you give examples, to make clear what sort of thing you're referring to?

> At best the genetic aspect is a red herring from problems we already know how to solve far better than we do ...

I doubt that. Where genetics is a large part of the picture (most psychological traits), appreciation of that is necessary to optimise policies. 

> I’m absolutely sure - to paraphrase Coel - the problem is not the refusal of vast swathes of society to recognise the genetic influence on criminality.

That comment of mine was not specifically about criminality, it was about understanding humans and human nature in general. The prevalance of blank slate ideology and the lack of acceptance of genetic factors affects lots of other things besides criminality. 

> my experience of STEM inclusivity in the UK is at odds with what you describe - understanding how being shaped by different personal and societal backgrounds interacts with unintentionally loaded systems to benefit some more then others, and then asking how to make new systems with as little bias baked in as possible.  None of this involved understanding genetic links so much as understanding how different people think, act and behave.

But this is an example where it does matter.  For example, suppose we have a field where the sex ratio is 30:70.  Is that fact evidence of bias, evidence that the field systematically mistreats one of the sexes, or not?    If we don't even know the answer to that question, then how can we produce proper evidence-based policies? 

3
Rob Exile Ward 05 Aug 2019
In reply to Coel Hellier:

What might such an evidence based policy look like?

wintertree 05 Aug 2019
In reply to Coel Hellier:

> Can you give examples, to make clear what sort of thing you're referring to?

I could but I don’t want to be drawn down another tangent.  There’s no shortage of evidence about how many different things affect life chances.  

> That comment of mine was not specifically about criminality, it was about understanding humans and human nature in general.

That was not clear to me.  I’m not sure my view changes much mind you. 

> But this is an example where it does matter.  For example, suppose we have a field where the sex ratio is 30:70.  Is that fact evidence of bias, evidence that the field systematically mistreats one of the sexes, or not?   

A line I’ve seen trotted out by our “Athena Swan Resistance”.   For a given subject there’s often evidence from elsewhere in the world that different societies and cultures have very different gender ratios.  There’s also plenty of evidence that making changes to remove unintentional bias results in new cohorts being less biased.   This is all evidence - and it’s real, tangible, actionable and here and now.  It’s also ethically achievable with very large numbers and detailed statistics gathering nationwide and internationally.  But perhaps we should stop all that and wait for the verdict on if current minority groups (within STEM) are genetically less interested in or less capable in STEM subjects.  

> If we don't even know the answer to that question, then how can we produce proper evidence-based policies? 

Well, I suspect we’re not going to have good evidence on how genetics bias undergraduate STEM intake in my professional lifetime, so should I just give up?  What is the problem with a policy of identifying and removing unintended bias from systems?  In my experience that generally makes them better for all concerned.

Post edited at 19:54
1
Coel Hellier 05 Aug 2019
In reply to Rob Exile Ward:

> What might such an evidence based policy look like?

On what topic, criminality, or sex ratios in STEM or something else?

In STEM, a blank-slate ideology would maintain that a 30:70 sex ratio is necessarily the result of bias, and that one would therefore need to implement more and more draconian policies to root out bias and enforce a 50:50 ratio.

An acceptance of the role of genetics, however, would say that since genetics is systematically different between the two sexes, the attractiveness of STEM could well be systematically different (and likely is), and thus that a 30:70 sex ratio could be entirely appropriate and fine.  Given that, you would look for actual bias and actual discrimination -- if there actually still is any these days -- and correct those, but not worry about attaining any particular ratio as an outcome. 

5
Coel Hellier 05 Aug 2019
In reply to wintertree:

> What is the problem with a policy of identifying and removing unintended bias from systems?

Nothing at all, so long as one indeed is identifying and removing actual bias, and not just declaring: "the ratio is 30:70 therefore there is bias".

1
Offwidth 05 Aug 2019
In reply to Coel Hellier:

Thats just rhetoric. At no point do you show any actual evidence to back up yoir point on genetics being the main factor that will get you incarcerated. You then call people names when they call you out. Calling such claims 'baldly stating' is what politicians do, not scientists.

4
Coel Hellier 05 Aug 2019
In reply to Offwidth:

> Thats just rhetoric. At no point do you show any actual evidence to back up yoir point on genetics being the main factor that will get you incarcerated. You then call people names when they call you out. Calling such claims 'baldly stating' is what politicians do, not scientists.

The only "name" I called you was your actual real name!  (So, I guess your comment has a sort of literal truth to it.)  And you objected to that because you don't want to have to stand behind your online behaviour. 

Coel Hellier 05 Aug 2019
In reply to the thread:

Here's a real example of why identifying causes does matter: Autism.  

In the bad old days, autism was blamed on bad parenting.  If the parents didn't love their child then it turned out autistic.  After all, in a blank-slate ideology you have to find something in the parenting or the child's home or early environment (vaccines?) that is causing autism. So blame the parents.

But studies now show that the genetic contribution to the causes of autism is about 80% (huge!) and that the shared-environment contribution (parenting; child's home environment) is pretty small at about 3% to 5% (or maybe zero).  

That alone tells you that it's not the fault of the parenting or home environment, so no amount of looking there for the prime culprit is going to succeed.   (Which is not to say that intervention cannot ameliorate the effects; it can.)

Shani 05 Aug 2019
In reply to Coel Hellier:

When you talk of criminality,  what specific behaviours do you have in mind? The problem with law is that it varies from country to country, political whim and time, so would be quite different to the behavioural spectrum that encompasses autism, no?

Coel Hellier 05 Aug 2019
In reply to Shani:

> When you talk of criminality,  what specific behaviours do you have in mind? 

Well, to be honest, I'm not really talking about criminality! I'm talking about psychological traits in general, and criminality is just an example.  So I don't really have particular behaviours in mind, but by "criminality" I'd be talking about the sort of thing regarded as a crime in most societies, including violent assault, robbery, murder, etc.

> The problem with law is that it varies from country to country, political whim and time, ...

True, and indeed whenever we talk of "heritability" and the relative effects of genes and environment, we have to be talking about a particular range of environments, since obviously things change if the environment changes.

FactorXXX 05 Aug 2019
In reply to Coel Hellier:

> The only "name" I called you was your actual real name!  (So, I guess your comment has a sort of literal truth to it.)  And you objected to that because you don't want to have to stand behind your online behaviour. 

What makes that even stranger is that there is a link to his Personal Web Page on his UKC Profile.

Pefa 05 Aug 2019
In reply to Coel Hellier:

> A good recent summary is by Robert Plomin: https://www.amazon.co.uk/dp/B07CV4NPJC/

Read the very first comment of the review of his book in your own link and see that his book gets completely slaughtered. 

> These days the evidence is not really up for dispute, several decades of studies have settled it.  

It would be interesting to see one. 

Post edited at 21:58
TobyA 05 Aug 2019
In reply to Shani:

Yep, it was Coel's discussion of nerdowells (or however it's spelt) and going to jail that seemed ridiculous. My three children have exactly the same (slightly ridiculous) little and kinked little fingers as I do, so it's hardly like anyone is disputing genetics isn't importance in making us who we are, but from a social science approach it's so utterly random as to how your genetic heritage does or doesn't fit into society, saying that's what make someone a criminal, or even a ner' do well (?) such a stretch.

Coel Hellier 05 Aug 2019
In reply to Pefa:

> Read the very first comment of the review of his book in your own link and see that his book gets completely slaughtered. 

By an anonymous commenter of unknown standing; whereas Plomin has numerous peer-reviewed publications which have been cited oodles of times.

TobyA 05 Aug 2019
In reply to cb294:

As a teacher, (and my wife is a social worker! Full house!), I think the reason that many people involved in education might look to environmental factors in producing educational outcomes is firstly that those are the factors that they might (and I stress only might) be able to have some effect on. But far more importantly is that the education system is still based on the idea of a true meritocracy, despite the consistent and huge amount of evidence that suggest that we do not live in a meritocracy, and plenty of people fail within the education system because of the environments they come from rather than factors that are individual to them such as 'intelligence' however you want to define that.

I also suspect your teacher and social worker friends are not old and jaded! I hope they keep fighting the good fight even if the science shows they've always had a hand tied behind their backs

Coel Hellier 05 Aug 2019
In reply to TobyA:

> ... but from a social science approach it's so utterly random as to how your genetic heritage does or doesn't fit into society, ...

Then the "social science" approach is wrong.  You need a real science approach. 

2
Coel Hellier 05 Aug 2019
In reply to TobyA:

> ... and plenty of people fail within the education system because of the environments they come from rather than factors that are individual to them such as 'intelligence' however you want to define that.

You might think that, but that doesn't make it so. 

2
sammy5000 05 Aug 2019
In reply to Coel Hellier:

Did you get a retard gene!

FactorXXX 05 Aug 2019
In reply to sammy5000:

> Did you get a retard gene!

Careful, Offwidth might report you for using non PC language on UKC...

1
Shani 05 Aug 2019
TobyA 05 Aug 2019
In reply to Coel Hellier:

Might there be a genetic predisposition to becoming reliant on cannabis? Some people more likely to become "addicted" to weed than others? Or at the very least, some get more pleasure from being stoned than others? In which case whether you live in Colorado or Alabama makes a difference as to how your genetic predisposition fits you into society. Or whether you live in Canada now or Canada a decade ago.

If you're gay surely it matters whether you live Saudi Arabia or Norway as to whether your genes are going to lead to criminality?

TobyA 05 Aug 2019
In reply to Coel Hellier:

And you might think it's not so, but that doesn't make it not so. Surely you can accept that say GCSE grades are not a pure result of solely children's genetically inherited intelligence? You keep saying above it's a mix of genes and environment.

Ridge 05 Aug 2019
In reply to the thread:

I got lost at phenotypic expression and intra-generational epigenetics.

Can't we have a bit of 'hangings too good for them' ranting and bizarre and impractical ideas for corporal punishment for a change?

rogerwebb 05 Aug 2019
In reply to TobyA:

Crime covers the whole spread of human activity. After a quarter century defending criminals the only genetic link I can confidently see from empirical observation is they are more likely to be male than not. 

Post edited at 23:10
Pefa 06 Aug 2019
In reply to Coel Hellier:

> > Read the very first comment of the review of his book in your own link and see that his book gets completely slaughtered. 

> By an anonymous commenter of unknown standing; whereas Plomin has numerous peer-reviewed publications which have been cited oodles of times.

That is ignoring everything this person states and the guy certainly is an expert of that there is no doubt. Doesnt that illicit any doubt in you at all about the veracity of Plomin's book?

It seems to me that there are people on the extremes who want to state that it is all in the genes for political reasons so that they can excuse poor social conditions for creating everything from drug addiction to crime. 

Coel Hellier 06 Aug 2019
In reply to TobyA:

> Might there be a genetic predisposition to becoming reliant on cannabis? Some people more likely to become "addicted" to weed than others?

Obviously, yes.  As a rule of thumb, all such psychological traits have a strong genetic component.

> In which case whether you live in Colorado or Alabama makes a difference as to how your genetic predisposition fits you into society.

Sure, yes it does.  In the same way, people might have genes that give them a tendency to get fat, but that's only going to occur in a land of supermarkets and fast-food outlets, and not in a famine in a third-world nation 100 years ago.

> If you're gay surely it matters whether you live Saudi Arabia or Norway as to whether your genes are going to lead to criminality?

Though when people talk of "criminality", they're not asking whether something happens to be against the local laws, they're discussing particular behavioural traits. 

Coel Hellier 06 Aug 2019
In reply to TobyA:

> Surely you can accept that say GCSE grades are not a pure result of solely children's genetically inherited intelligence?

Obviously not!  A child 200 years ago would not have GCSE grades at all because there weren't GCSEs! 

> You keep saying above it's a mix of genes and environment.

Exactly. No-one is arguing that everything is 100% genetic.  

Here's a good primer: "Top 10 Replicated Findings from Behavioral Genetics"   https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4739500/

Coel Hellier 06 Aug 2019
In reply to Pefa:

> That is ignoring everything this person states and the guy certainly is an expert of that there is no doubt.

Again, it's an anonymous commenter of unknown credentials. And, again, loads of people have peer-reviewed Plomin's stuff.

> It seems to me that there are people on the extremes who want to state that it is all in the genes for political reasons ...

None of the scientists in this field are saying that it is "all in the genes", since the studies show clear influence from environment also (as is obvious).   

JoshOvki 06 Aug 2019
In reply to Wayne45`:

Nice not to be not treated like a criminal for a change! Out of curiosity are you still in the catering business or was it still hard to find work once you left? (I know this is quite an intrusive question so don't feel obliged to answer)

cb294 06 Aug 2019
In reply to TobyA:

I agree completely!

CB

rogerwebb 06 Aug 2019
In reply to rogerwebb:

> Crime covers the whole spread of human activity. After a quarter century defending criminals the only genetic link I can confidently see from empirical observation is they are more likely to be male than not. 

In reply to myself. 

I don't like classifying people as criminals and shouldn't have lazily used the term. 

Post edited at 08:24
cb294 06 Aug 2019
In reply to TobyA:

You are right in that the same genetic predisposition can lead to behaviour that will be judged criminal or legal depending on societal circumstances. Importantly, I believe that science can help devise better policies in these cases. For example, we now know that homosexuality is an entirely normal trait that is present in all species where biologists have looked, probably as a balanced polymorphism. This scientific insight should help overcome religious or ideological bigotry that historically led to the criminalisation of homosexuality.

Similarly, the degree of susceptibility to addiction is clearly genetic, while just as clearly the environment plays a role whether an individual becomes actually addicted. Acknowledging this biological fact could again help with devising a more reasonable approach to addiction. Ignoring these facts has, conversely, led to the ideological war on drugs that has completely failed, with disastrous consequences. Never mind the situation here, bad as it is for those people criminalized due to their addiction, just think of what happened in countries like Colombia, Guatemala, or Mexico!

CB

rogerwebb 06 Aug 2019
In reply to Coel Hellier:

> Obviously, yes.  As a rule of thumb, all such psychological traits have a strong genetic component.

> Though when people talk of "criminality", they're not asking whether something happens to be against the local laws, they're discussing particular behavioural traits. 

Which particular behavioural traits? 

Coel Hellier 06 Aug 2019
In reply to rogerwebb:

> Which particular behavioural traits? 

Tendencies to aggression, violence, impulsiveness, et cetera.

TobyA 06 Aug 2019
In reply to Coel Hellier:

We don't have criminal beatles or sea horses or even chimpanzees. Criminality is a social label for certain behaviours, but the label is a social construct.

cb294 06 Aug 2019
In reply to rogerwebb:

Intelligence, susceptibility to addiction, propensity towards violence, lowered empathy, dissocial personality structure, psychopathy and narcissism, schizophrenia and related psychopathologies (don't tell me that a significant fraction of the prison population does not actually suffer from mental/psychological problems and would be better of with treatment), stress resistance, you name it.

All of these traits have been shown to have, to varying extents, a heritable component.

CB

John2 06 Aug 2019
In reply to cb294:

But plenty of people possess these traits without lapsing into criminality. As I said earlier, those brought up in a consistently criminal environment will be far more likely to engage in criminality themselves. 

Coel Hellier 06 Aug 2019
In reply to TobyA:

> Criminality is a social label for certain behaviours, but the label is a social construct.

But the behaviours are real, biological and heavily influenced by genes. 

5
Coel Hellier 06 Aug 2019
In reply to John2:

> As I said earlier, those brought up in a consistently criminal environment will be far more likely to engage in criminality themselves. 

Do you have any evidence for that, or are you just reporting your gut feeling?

The evidence is that this "shared environment" is much less influential than many people suppose.  E.g. (added italics):

"Twin studies can be used to investigate the contributions of genetic factors, the common or shared environment, and the unique or non-shared environment to individual differences in a measurable characteristic. [...] The subjects for [this] study were 3226 male twin pairs ..."

"Genetic factors, but not the common environment, significantly influenced whether subjects were ever arrested after age 15, whether subjects were arrested more than once after age 15, and later criminal behaviour."

"The environment shared by the twins has an important influence on criminality while the twins are in that environment, but the shared environmental influence does not persist after the individual has left that environment."

https://www.ncbi.nlm.nih.gov/pubmed/8862870

2
cb294 06 Aug 2019
In reply to John2:

There is a subtle fallacy in your argument: No one argues that the heritability of these traits is 100%, and clearly the expression of the traits can be influenced by the environment. However, choice and modification of environment (e.g. for oneself and ones children!) has itself a significant genetic contribution. 

Thus, in the specific context, you cannot simply tally a "consistently criminal environment" on the non-genetic side. Instead, you need carefully controlled studies to distinguish genetic and true, non-heritable, environmental influences.

This actually holds for most behavioural or psychological traits that have been studied, in particular general cognitive abilities and intelligence.

If you are interested in this I would recommend reading the review paper Coel linked to above. Three of the ten findings discussed in there (7-9) specifically address the links between genetic and environmental influences on behavioural traits, and the difficulty of separating them.

CB

edit: I was referring to the post from 7.14

Post edited at 10:09
John2 06 Aug 2019
In reply to Coel Hellier:

I cannot point to any academic papers, but that is not to say that they do not exist. I remember someone I used to know who spent his gap year doing social work in the East End of London. He spent all of this period working with people such as I described earlier who had been brought up in an atmosphere of criminality and really did not seem to know that any other way of life existed.

rogerwebb 06 Aug 2019
In reply to Coel Hellier:

> Tendencies to aggression, violence, impulsiveness, et cetera.

Young men? 

Offwidth 06 Aug 2019
In reply to Coel Hellier:

John's claim is pretty mainstream:  you are the one making an exceptional claim without any evidence as yet to back it up. As ever in your links you miss out the bits that give the correlations and likely fractional influences. There are plenty of crimnal behaviours that link to genetics but no expert scientists that I'm aware of make your claim that its the most likely reason that people are incarcerated.

"Then the "social science" approach is wrong.  You need a real science approach. "

Can't help yourself can you (no irony given that Plomin is a Social Scientist).

More from Adam Rutherford on the dangers of such simplistic thinking

https://www.newstatesman.com/2016/09/why-we-can-t-blame-warrior-genes-violent-crime

Finally I'm greatful UKC supports the rights to a privacy level we all want. You chose to use your name, I don't (even if it's still easy to work it out from my profile). There is nothing mysterious about the benefits of having a user name different from a real name;  its been discussed at length many times here. You dishonestly claim here for maybe the 4th time now that I've had you censored by running to the mods when you know full well that you actually self censored by ignoring to respect the warnings of the moderators  (which given UKC clear moderation policies impacting sometimes on freedoms of speech I guess you despise). Most of us take deletions on the chin when we know the mods have a point .. I don't believe in public censorship within UK legal constaints (and think the law is too strict in some respects) but I understand organisations rights to have rules and  protect themselves (UKC isn't a public soapbox on a street corner).

rogerwebb 06 Aug 2019
In reply to cb294:

> Intelligence, susceptibility to addiction, propensity towards violence, lowered empathy, dissocial personality structure, psychopathy and narcissism. 

Who are you studying here? I recognise these traits but I don't think they apply to the bulk of people appearing in court rather than the extreme minority (who can be pretty extreme) 

>(don't tell me that a significant fraction of the prison population does not actually suffer from mental/psychological problems and would be better of with treatment),

I wouldn't dream of it, but neither would I necessarily define those people as criminals.

The courts are full of the sad and the mad. The bad are a surprisingly small proportion. 

> All of these traits have been shown to have, to varying extents, a heritable component.

The issue being how big is that component. 

Coel Hellier 06 Aug 2019
In reply to John2:

> I used to know who spent his gap year doing social work in the East End of London. He spent all of this period working with people such as I described earlier ...

Anecdotes don't really cut it compared to properly conducted studies. 

Coel Hellier 06 Aug 2019
In reply to rogerwebb:

> I recognise these traits but I don't think they apply to the bulk of people appearing in court rather than the extreme minority (who can be pretty extreme) 

We all have tendencies towards all of these traits.  It's a question of degree, and how they balance against our other traits.   And the degree to which we have such traits is strongly influenced by genetics.

John2 06 Aug 2019
In reply to Coel Hellier:

If the study to which you linked made reference to criminal behavior it would be more germane to this discussion. No one is denying that traits such as aggression and selfishness can be inherited through genes, what you seem determined to deny is that a quality such as aggression can either lead to a high-achieving sporting career or a life of criminality.

Coel Hellier 06 Aug 2019
In reply to Offwidth:

> John's claim is pretty mainstream: 

You're right, it is. Vast swathes of society presume a blank-slate account.   However, the evidence has long established that it is wrong.

> .... you are the one making an exceptional claim without any evidence as yet to back it up.

Nope, there is nothing exceptional about the idea that genes affect our brains and how they function just as much as they affect our livers and how they function and all other parts of our bodies.   This is the default presumption.   The blank-slate ideology is the "extreme" position.  You think that genes are just there for show?

And it is notable that you claim I am "without any evidence as yet to back it up" when the very post you are referring to cited a study.  You're getting as science-denialist as those who claim there is no science to back up climate change. 

The twin tactics of denying the science and personally attacking anyone who disagrees with ideology are sadly way to common among academics such as yourself.

> Finally I'm greatful UKC supports the rights to a privacy level we all want.

Sure, but I consider it cowardly of you to personally attack someone using their real name, when not being willing to do it under your own real name. 

> You dishonestly claim here for maybe the 4th time now that I've had you censored by running to the mods ...

Are you claiming that you didn't go "running to the mods"?

> ... when you know full well that you actually self censored by ignoring to respect the warnings of the moderators 

Wrong.  I did not ignore any warnings.  An email asking me to stop using your name came at the same time that posts were deleted.

Post edited at 11:02
Coel Hellier 06 Aug 2019
In reply to John2:

> If the study to which you linked made reference to criminal behavior it would be more germane to this discussion.

The study I cited in the post just above, 09:52, does refer to criminal behaviour.   So does the one cb294 cited above. So do lots of others. 

Here's another one: "In this paper, we describe a quantitative summary of 12 twin (n=3795 twin pairs and 3 adoption studies=338 adoptees) ... Medium to large effect sizes were found for genetic influences across studies, with approximately 50% of the variance in measures of antisocial behavior attributable to genetic effects."

https://link.springer.com/article/10.1007/BF02239409

> ... what you seem determined to deny is that a quality such as aggression can either lead to a high-achieving sporting career or a life of criminality.

Where did I deny that?  Are you just making things up?

cb294 06 Aug 2019
In reply to rogerwebb:

I think we largely agree. People end up in court and will be convicted of a crime (and will thus be labelled criminals) often because they are, as you put it, sad or mad.

Like the traits I listed above that potentially make people "bad", clinically diagnosed depression and psychosis are simply extremes of a continuous distribution in these phenotypes in the general population. For both, heritability is rather high, for depression around 40% in women and 30% in men (based on twin and adoption studies in Sweden), and between 40% and 80% for various psychotic conditions.

However, discussion in the general public seems to ignore these findings, with IMO catastrophic consequences for policy development, with respect to, say, substance abuse. Similarly, if you find that environment modification and choice have themselves a large heritability, it should have implications for issues ranging from education to housing policy (i.e., people are even less free in their choices than one would assume simply from socioeconomics alone, making early and generous support even more urgent). 

CB

rogerwebb 06 Aug 2019
In reply to cb294:

> I think we largely agree. 

I think we do. 

So far this week and by general defence and prosecution agreement, 5% bad 95% sad or mad. 

The system creaks. It runs on the principle of save £5 now on 'our' budget who cares if we then spend £5000 later on someone else's. 

TobyA 06 Aug 2019
In reply to Coel Hellier:

Yes, but they are not criminal until a socially constructed legal system labels them so. 

TobyA 06 Aug 2019
In reply to Coel Hellier:

You started all this by using the term ner do wells! And claiming being a ner' do well is genetic!

Post edited at 12:02
Coel Hellier 06 Aug 2019
In reply to TobyA:

> You started all this by using the term ner do wells! And being a ner' do well is genetic!

Yes and yes.  So?

John2 06 Aug 2019
In reply to Coel Hellier:

Your 9.52 link refers to sets of twins who both served in Vietnam - a significantly traumatizing event. I quote, 'The environment shared by the twins has an important influence on criminality while the twins are in that environment, but the shared environmental influence does not persist after the individual has left that environment'

Seems to me to support my argument.

TobyA 06 Aug 2019
In reply to Coel Hellier:

Being a ner'do well is a moral classification, not a genetic trait. You accepted all that above.

2
cb294 06 Aug 2019
In reply to John2:

You have that the wrong way round: There is still significant concordance in criminality between the twins. Roughly, what is left is due to the genetic propensity.

Essentially this is promising, because it means you are not screwed up for life even if you grew up in a "consistently criminal environment" , as you put this above. 

CB

Coel Hellier 06 Aug 2019
In reply to John2:

> Your 9.52 link refers to sets of twins who both served in Vietnam - a significantly traumatizing event.

No, it was only traumatizing for a minority; anyhow, it's only one study, there are lots of others.  

1
Coel Hellier 06 Aug 2019
In reply to TobyA:

> Being a ner'do well is a moral classification, not a genetic trait. You accepted all that above.

"Ne'er do well" is indeed an evaluation about someone's capabilities, behaviour and lifestyle, but all of those are behavioural and psychiatric traits and all of those are under strong genetic influence.  

John2 06 Aug 2019
In reply to cb294:

Exactly, I'm not claiming that all criminals are irredeemable undesirables who can never reform. I'm saying that many people grow up in an environment from which it is difficult to escape.

There is a completely different argument about whether prison is the best method of dealing with these people, since for a significant number time spent in prison is in effect a masterclass in advanced criminality.

Offwidth 06 Aug 2019
In reply to Coel Hellier:

If the mainstream is so wrong (the key Danish studies have been mainstrean for years now but like all I'm aware of only show a significant rather than overwhelming increased liklihood in very specific criminal traits) and you are right, why not stop waffling on and link the evidence that shows the main reason people are jailed IS due to their genes. I think you can't do that because the evidence doesn't exist. Its like your arguments on Islam you go on and on and on about evidence and freedom of speech and despite having evidenced fair concerns about specifics you can never deliver actual evidence to back your concluding overall extreme views.

Sure I ask the mods to delete posts that use my real name in threads.. thats not 'running' anywhere just asking for the level of anonymity I want. Its certainly nothing to do with censoring ideas or arguments. You carefully ignore that this isn't the first time you used my name in a thread where I complained and were I'm guessing asked by the mods not to do that. Implying censorship in this sense is more politicised bs.

1
Pefa 06 Aug 2019
In reply to Coel Hellier:

> Again, it's an anonymous commenter of unknown credentials. And, again, loads of people have peer-reviewed Plomin's stuff.

Yes obviously, but the fact the guy slaughters him and is an expert tells me all is not right about this Plomin's work and there are books in similar situations created by people making big claims and published in big peer reviewed magazines then touted around the press that are also then exposed as not being all they say they are. 

> None of the scientists in this field are saying that it is "all in the genes", since the studies show clear influence from environment also (as is obvious).   

Good, as its more than obvious that the environmental conditions people are exposed to especially at very young ages do affect their later life with respect to addiction, crime, mental health etc. We all know psychopaths are born that way so genetic have their part to play to. 

One question though if I may: in these studies of twins and criminality ; what of twins who did indulge in criminality but didn't get caught? Therefor no arrest and not added to the figures. Is that one factor taken into account and if not would it not skew the results? 

Offwidth 06 Aug 2019
In reply to Pefa:

Of course it would, as would trauma in utero. The scientists in the studies don't make the idiotic grand claims Coel does, some even warn against it and are clear that free will and intervention can mitigate genetic influnece.

Coel Hellier 06 Aug 2019
In reply to Offwidth:

> If the mainstream is so wrong ...

Which it is. Most non-scientist intellectuals *way* under-estimate the extent to which genetics affects our personalities and behavioural traits. 

> and you are right, why not stop waffling on and link the evidence that shows the main reason people are jailed IS due to their genes.

Lots of studies have been linked to.  Again, here's a primer: "Top 10 Replicated Findings from Behavioral Genetics"

Number 1 is: "All psychological traits show significant and substantial genetic influence".  And "[a] consistent finding is that heritabilities are substantial, often accounting for half of the variance of psychological traits." And: "Significant and substantial genetic influence on individual differences in psychological traits is so widespread that we are unable to name an exception."

> You carefully ignore that this isn't the first time you used my name in a thread where I complained and were I'm guessing asked by the mods not to do that.

Oh right, so you merely "guess" that I was previously asked by mods not to use your name? Wrong, I was not.  And on the basis of that wrong guess you claim I "carefully ignore" what did not even happen -- thus attributing deceitfulness to me?  

And on the basis of that wrong guess you earlier said that I: "... know full well that you actually self censored by ignoring to respect the warnings of the moderators ... " -- wrong, there was no previous request so I was not ignoring anything, and thus I did not "know full well" about it.

And indeed based on these false guesses of yours you concluded that I "dishonestly claim" in what I said earlier?? 

Sheesh, Offwidth, this sort of attack-dog posting style, personally attacking posters just because they disagree with you, is why I've totally lost any patience or respect for you.

3
Offwidth 06 Aug 2019
In reply to Coel Hellier:

Hardly evidence on your assertion that genes are the most likely reason for someone to end up in jail.

I apologise if I was wrong about you being told before.. looking forward to the same from you about your false statement that I wish to censor you on this discussion. 

Remind me of that philosophical work that you said you admired that recommends an attacking debating style again?

cb294 06 Aug 2019
In reply to John2:

I agree. However, the link between environment choice and modification and genetics is more complicated than generally appreciated (again, see the primer Coel linked to). Any policies aimed at helping people trapped in these circumstances (or to help society by reducing crime levels) need to take these findings from behavioural genetics into account. Treating children from such areas as blank slates because "this is how it morally should be" is bound to result in policies that are at best not working. The misguided "war on drugs" and its disastrous consequences should serve as a warning example.

CB

Coel Hellier 06 Aug 2019
In reply to Offwidth:

> Hardly evidence on your assertion that genes are the most likely reason for someone to end up in jail.

Combined with everything else (including much cited up-thread), it is. 

The biggest reasons for people ending up in jail are aggressive, violent and impulsive personalities, low IQ, low educational attainment, consequences of substance abuse, and similar things.  And all of those traits have high heritability and are under strong genetic influence!   In comparison, the contribution of factors such as "brought up in a household where criminality is the norm" (referred to as "shared environment" in the studies) is much lower. 

> I apologise if I was wrong about you being told before.. looking forward to the same from you about your false statement that I wish to censor you on this discussion. 

I've just gone through the thread with the "find" utility of my browser, and the only person who has used the word "censor" in this thread is you.  I've not used it, so what are you on about?

> Remind me of that philosophical work that you said you admired that recommends an attacking debating style again?

What are you on about? Which philosophical work?

Post edited at 13:33
2
Coel Hellier 06 Aug 2019
In reply to Offwidth:

> The scientists in the studies don't make the idiotic grand claims Coel does, ...

Oh right, so Robert Plomin did not say either of:

"For most of the 20th century, environmental factors were called nurture because the family was thought to be crucial in determining environmentally who we become. Genetic research has shown that this is not the case. We would essentially be the same person if we had been adopted at birth and raised in a different family."

"DNA isn’t all that matters, but it matters more than everything else put together in terms of the stable psychological traits that make us who we are."

> ...  and are clear that free will and intervention can mitigate genetic influnece.

I am also clear that intervention can affect outcomes, since no psychological trait is 100% genetic. 

But what do you mean by "free will" here?  Are you referring to some dualistic, non-scientific nonsense?

Because, if you're referring to any sensible and scientific account of our "will", then that will also be under strong genetic influence.  That's what it means to say that all psychological traits are under strong genetic influence.

Coel Hellier 06 Aug 2019
In reply to Pefa:

> One question though if I may: in these studies of twins and criminality ; what of twins who did indulge in criminality but didn't get caught? Therefor no arrest and not added to the figures. Is that one factor taken into account and if not would it not skew the results? 

The crucial factor for estimating the genetic component is to compare identical twins (monozygotic) to fraternal twins (dizygotic). (And, preferably, to study twins separated at birth and raised apart.)

Now, if the rate of not-getting-caught were the same for both   identical twins and fraternal twins (when averaged over hundreds o them), then it will have no affect on the outcome and not skew the estimate of genetic effect. 

In order for that to skew the outcome, there would have to be a reason why someone is less/more likely to be caught if they have an identical twin (separated at birth and raised apart) than if they have a fraternal twin (separated at birth and raised apart).

That doesn't sound very plausible to me.  The twin studies (especially the ones with separation at birth) really are a very clean and reliable method of study. 

Offwidth 06 Aug 2019
In reply to Coel Hellier:

Maybe it was in the deleted post or in the private emails I deleted (like you calling me something like a 'vile little shit'). You publicly stand up for free speech but do seem to get very angry when facing robust debate. You do rile me more than most as academics who become polemicists based on ideology over evidence really bug me.

Free will is the basis of most criminal justice systems: its assumed people choose to do bad things and are sentenced based on that alongside evidence of guilt.  I'm aware that will is less free than 'society' thinks, including genetic influences, but that doesn't mean its non existant: we are are far from being completely pre-programmed.

2
Coel Hellier 06 Aug 2019
In reply to Offwidth:

> Maybe it was in the deleted post or in ...

Or maybe you just imagined it (and no, it wasn't in the deleted posts, because nothing had been deleted when those were written).

> You publicly stand up for free speech but do seem to get very angry when facing robust debate.

It's not robust debate that I deplore, it's personal attacks as a substitute for actually debating the issues.   You have recently said on another thread: "Many people tell me they don't like the increased negativity of the site, compared to earlier days, ...", and yet you are one of the posters most likely to start personal attacks, as this thread shows.  And not as a last resort, often as a first resort. 

> You do rile me more than most as academics who become polemicists based on ideology over evidence really bug me.

Pot! Kettle! Black!  Come on, the evidence really is on my side here.

> Free will is the basis of most criminal justice systems: its assumed people choose to do bad things and are sentenced based on that alongside evidence of guilt. 

But any "will" we have derives from genes plus environmental influences.  (Where else?)   So when we talk about our "will" we are talking about psychological traits that are exactly what is under strong genetic influence. So talking about "will" overcoming genetic influence is incoherent.  The whole way that such genes work is by affecting our brains and thus our "will".

3
Pefa 06 Aug 2019
In reply to Coel Hellier:

Thanks for your reply I'm aware of studies on identical and non identical twins from other fields so I know how reliable and important a test this can be. My point however was to show that you cannot predict whether a person who commits a criminal offence will get caught. The reason being that a good deal of being caught is down to chance, could they be ID'd, was there any witnesses, is the victim going to report it etc I mean they say only 1 in 10 crimes committed is reported for example so does that not skew these results somewhat? Sorry if I didn't make that clearer earlier. I mean I'm sure we all know plenty of people who got away with crimes and some like drug dealers who do so for decades and look at all the abusers who got away with crimes for decades. And users of illegal drugs or people who would just not get caught or reported for violent acts.

So if many criminals get away with crimes and obviously won't admit that then a study which compares data that says you are only a criminal if you get caught cannot be accurate. No? 

Post edited at 15:09
Offwidth 06 Aug 2019
In reply to Coel Hellier:

Pots and Kettles..... "The only "name" I called you was your actual real name!"  vs something like 'vile little shit'. Hardly a fair let alone level headed response from someone genuinely dedicated to debate. Your spin would make Alistair Campbell blush.

2
Coel Hellier 06 Aug 2019
In reply to Pefa:

> My point however was to show that you cannot predict whether a person who commits a criminal offence will get caught. The reason being that a good deal of being caught is down to chance, ...

Sure, which is why such studies add up results for hundreds to thousands of twins, thus averaging over chance.  That's entirely standard scientific practice. 

Harry Jarvis 06 Aug 2019
Coel Hellier 06 Aug 2019
In reply to Offwidth:

> "The only "name" I called you was your actual real name!" 

Which was true, about the posts on the thread. 

> ... vs something like 'vile little shit'. Hardly a fair let alone level headed response from someone genuinely dedicated to debate.

So what would you consider an appropriate appellation for someone who merely *guesses* about the facts of the situation, and on the basis of a wrong guess, without bothering to establish the actual facts, accuses someone else, in public, of being deceitful and dishonest -- and yet is unwilling to stand behind the accusation with their real name, and indeed runs to the mods to ensure they can hide behind a pseudonym.  And then moans on another thread that the forums have got unfriendly!  

I consider my remark a pretty fair summary; wouldn't you?   I'd be severely embarrassed to have behaved like that. 

But, again, this attack-dog style seems to be your first resort these days, and you seem to regard it as acceptable towards anyone whose views you consider to be outside the approved ideology.

1
Offwidth 06 Aug 2019
In reply to Harry Jarvis:

Good point 

1
Offwidth 06 Aug 2019
Pefa 06 Aug 2019
In reply to Coel Hellier:

Isn't that a bit random? I mean if say just 1 in 10 criminals get caught then that means their results could be out by that same amount and where does it actually state that they take that into consideration and what number do they use if they do?

> The biggest reasons for people ending up in jail are aggressive, violent and impulsive personalities, low IQ, low educational attainment, consequences of substance abuse, and similar things.  And all of those traits have high heritability and are under strong genetic influence!   In comparison, the contribution of factors such as "brought up in a household where criminality is the norm" (referred to as "shared environment" in the studies) is much lower.

What of trauma? Where criminality is not the norm but certain circumstances or certain criminal acts lead to criminal behaviour, say drug addiction or prostitution? For example child abuse or other traumas have a massive effect on adults becoming addicts (criminals when the addiction is illegal drugs) or becoming prostitutes, just one example. There is a famous now retired GP who has worked with drug addicts all his life and stated he has never met one who was not traumatised as a child. 

Post edited at 15:45
TobyA 06 Aug 2019
In reply to Coel Hellier:

> "Ne'er do well" is indeed an evaluation

Yes, it's an evaluation, made by others in a social context. Let's go back to being gay. 50 years ago that definitely made you a ne'er do well. It doesn't now, but I presume it's not the changing genetics of the straight population that have changed that.

TobyA 06 Aug 2019
In reply to Pefa:

Plus there is the whole sociology of which crimes get prioritised, inequality within the criminal justice system, etc etc.

TobyA 06 Aug 2019
In reply to Coel Hellier:

> The biggest reasons for people ending up in jail are aggressive, violent and impulsive personalities, low IQ, low educational attainment, consequences of substance abuse, and similar things.  

Where? When? Are Americans a lot more prone to all of those things as they imprison people at higher rates than I think ANY other country? Do they account for imprisonment rates in apartheid era South Africa? Or in Xinjiang Province now?

What evidence could possibly prove that? 

1
L Wayne45` 06 Aug 2019
In reply to JoshOvki:

Yeah it is. Sometimes when I working there it I sort of frogot I was still in prison! Did you talk much to the lads when you were at the one you were at?

Still in the catering/hospitality business 3 years later and am hoping to get on a trainee manager course in the next few month. Prob not the best job in the world but beats construction for me although it pays less where most of the lads I know went to work.

And yeah it is hard to get work with a record like min. Even tried the Army which I wanted to join as a kid but got turned down for that so happy enough with what I have

John2 06 Aug 2019
In reply to Coel Hellier:

It’s pretty obvious, Coel, that you are just Googling a succession of academic articles that you imagine (not always correctly) support your thesis. I thought that excessive use of Google was a trait that academics were attempting to suppress in their students.

3
Shani 06 Aug 2019
In reply to Coel Hellier:

> Tendencies to aggression, violence, impulsiveness, et cetera.

At least two of those are strong characteristics of successful saleasmen and leaders, quick to seize opportunity and hard bargaining in tough negotiations.

Edit: Arguably 2/3 are exhibited by Trump, Putin, Johnson et al...and not the same 2/3.

Post edited at 19:09
Coel Hellier 06 Aug 2019
In reply to Pefa:

> Isn't that a bit random? I mean if say just 1 in 10 criminals get caught then that means their results could be out by that same amount ...

This does not matter. As above, it would only matter if there were a reason why someone is less/more likely to be caught if they have an identical twin (separated at birth and raised apart) than if they have a fraternal twin (separated at birth and raised apart).  That doesn't sound very plausible to me. 

Coel Hellier 06 Aug 2019
In reply to TobyA:

> Yes, it's an evaluation, made by others in a social context. Let's go back to being gay. 50 years ago that definitely made you a ne'er do well. It doesn't now, but I presume it's not the changing genetics of the straight population that have changed that.

But this is irrelevant.  Since all psychological and behavioural  traits are strongly genetically influenced, whatever the definition of "ne're do well", qualification for that category will be largely a matter of genetics. 

So, if in one society, people who are aggressive/impulsive are regarded as "ne're-do-wells", then -- since aggressive/impulsive traits are strongly genetically influenced  -- who ends up being a "ne're-do-well" in that society will be mostly about genes. 

But, if in some other society, people who are meek/mild/cautious are regarded as ne're-do-wells, then -- since meek/mild/cautious traits are strongly genetically influenced  -- who ends up being a "ne're-do-well" in that society will be mostly about genes. 

Coel Hellier 06 Aug 2019
In reply to John2:

> It’s pretty obvious, Coel, that you are just Googling a succession of academic articles that you imagine (not always correctly) support your thesis. I thought that excessive use of Google was a trait that academics were attempting to suppress in their students.

You thought wrongly.  And not for the first time.  And it's notable that you're not actually presenting evidence for your claims, which are merely based on gut feeling. 

Coel Hellier 06 Aug 2019
In reply to TobyA:

> Are Americans a lot more prone to all of those things as they imprison people at higher rates than I think ANY other country?

When we're talking about criminality in these studies we're taking about what happens within one nation. 

As I've stated, if one is instead considering a wider range of environments, as would be the case when including lots of different countries, then heritability estimates will be different.

Pefa 06 Aug 2019
In reply to Coel Hellier:

Correct me if I'm wrong but you mean if one identical twin gets away with a crime then so will the other as they will have some gene that will ensure that and likewise if they both get caught committing crime? 

Coel Hellier 06 Aug 2019
In reply to Pefa:

> Correct me if I'm wrong but you mean if one identical twin gets away with a crime then so will the other as they will have some gene that will ensure that and likewise if they both get caught committing crime? 

No, that's not what I mean. 

The chances of any particular person getting away with a crime depends on all sorts of factors including chance. 

But for the twin studies to be affected by this, it would have to be the case that:

The chances of John getting away with a crime are systematically higher if he happens to have an identical twin (who was separated at birth and adopted by some other family and raised in some other town), compared to the chances of getting away with the same crime if he happens to have a fraternal twin (who was separated at birth and adopted by some other family and raised in some other town).  

As I said, that does not seem to me at all probable.  Why on earth would that difference in a twin living in some other city affect his chances of getting away with a crime? 

Pefa 06 Aug 2019
In reply to Coel Hellier:

Most crimes are not even reported and most people have probably committed many crimes yet never been caught so how could this not skew their results?

And what of my point on the effects of childhood trauma with respect to peoples life, like addictions and crime? 

Post edited at 22:39
Coel Hellier 07 Aug 2019
In reply to Pefa:

> Most crimes are not even reported and most people have probably committed many crimes yet never been caught so how could this not skew their results?

It does not skew the results because it affects identical twins and fraternal twins equally, and something affecting both equally does not skew the comparison between identical and fraternal twins.

> And what of my point on the effects of childhood trauma with respect to peoples life, like addictions and crime? 

Again, the whole point of twin studies is that they can separate such environmental affects from genetic ones, since environmental effects will -- on average, averaging over hundreds or thousands of cases -- be the same for identical twins (separated at birth and reared apart) as for fraternal twins (separated at birth and reared apart). The only thing not the same is the genetic similarly, and that gives a clean experiment enabling us to separate genetic and environmental factors. 

Pefa 07 Aug 2019
In reply to Coel Hellier:

Thanks for your reply which clears up my question about crimes committed yet unrecorded. 

Pefa 07 Aug 2019
In reply to Offwidth:

> Some more discussions around the actual issues

That one puts it all into perspective. 

Offwidth 07 Aug 2019
In reply to Coel Hellier:

Twins can be similarly affected in utero by all sorts of things and this can't be easily separated from any genetic traits. Twins, are more likely to have in utero problems as well, more so still for twins from deprived areas (all of which can correlate to subsequent criminality, as per my earlier link, repeated below). Alcohol exposure in the womb which would affect twins similarly has also been linked to criminality (example link below).

https://en.m.wikipedia.org/wiki/Statistical_correlations_of_criminal_behaviour

"Associated factors include maternal smoking during pregnancy, Low birth weight, perinatal trauma/birth complications,"

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3571436/

". Social factors like domestic and family violence were frequent in the risk group, this also being associated to maternal drinking during pregnancy. "

The numbers in the well known twin studies are smallish, especially for identical twins seperated at birth (the significant results even from thousands still have statistical error bars at the percentile level so its hard to know the exact level of effect), the increases in likelihood are usually stated in the tens of percentage range and they only relate to specific links ito criminality (eg a tendancy to excessive violence). As far as I'm aware no one who did such studies claims what you do (in fact some say the genetic effects can be mitigated significantly by treatment). If Plomin or any other serious expert researcher claimed what you do, you could have quoted them. You seem to have no experience in the research area. You are certainly exaggerating in stating something as a fact that no one can know as yet (and if it were true, it would have massive implications for all criminal justice systems), and as such you are an embarrassment to science (and deliberately so... since you seem incapable of moderation or apology.. so it must be political).

It's true that genetic effects in such areas have been understated for years but these days the knowledge that there is a significant effect in some cases is pretty mainstream.

Edit: some high profile criticism on Plomin:

https://www.nature.com/articles/d41586-018-06784-5

https://genotopia.scienceblog.com/506/lies-damned-lies-and-gwas/

https://blogs.bath.ac.uk/eric/2018/09/28/blueprint/

Post edited at 14:21
2
TobyA 07 Aug 2019
In reply to Coel Hellier:

You didn't claim that before. So your claim -about why most people are in prison - where are you talking about? If it's the US presumably you then have to agree with people who say African American are more genetically prone to being criminals? What else could account for their levels of incarceration? It obviously couldn't be the legacy of slavery and institutionalized racism in the form of Jim crow.

cb294 07 Aug 2019
In reply to Offwidth:

> It's true that genetic effects in such areas have been understated for years but these days the knowledge that there is a significant effect in some cases is pretty mainstream.

It is within biomedical sciences, but not amongst policy makers or voters (otherwise we would see more evidence driven policy, e.g. with respect to custodial sentences, criminalization of drug abuse.

CB

Offwidth 07 Aug 2019
In reply to TobyA:

Plomin does apparently think blacks have a lower IQ (see my third critic's link above, including support for The Bell Curve). Yet even with his heavily criticised views, he doesn't seem to claim anything like as outrageous as Coel does: genes being the most likley reason you will end up in jail.

You forgot the money gene and the knowing a good lawyer gene in the case of incarcerated blacks in the US.

Post edited at 14:32
1
Offwidth 07 Aug 2019
In reply to cb294:

Sorry I meant to clarify that...I always meant it is in Science, and so it should be in lower level government policy input as well. I'm not sure if anything can be said to be mainstream in politics or population understanding these days.

Coel Hellier 07 Aug 2019
In reply to Offwidth:

> Twins can be similarly affected in utero by all sorts of things and this can't be easily separated from any genetic traits.  ...

But, again, about all such factors:

In order to confound the twin studies you'd need it to be the case that identical twins were affected differently from fraternal twins, such that identical twins were made more similar to each other but fraternal twins were made less similar to each other.  That's not impossible, but it's then getting to be second-order or third-order effects.

Simply saying that alcohol abuse affects the child in the womb is not a confounder unless it systematically affects one twin differently from the other one.

> The numbers in the well known twin studies are smallish, ...

They used to be, but no longer.  Typical studies can have tens of thousands now.  This meta-analysis of many studies totals to 14 million pairs of twins: https://www.nature.com/articles/ng.3285

Note from that abstract: "The data are inconsistent with substantial influences from shared environment ...". 

That's the main point that I came to the thread to make, "shared environment" (defined as environment that siblings would share) is not a large factor in the psychological traits we have as adults.  Yes I know that that's counter-intuitive, but the studies are consistently saying it is true.  

> As far as I'm aware no one who did such studies claims what you do (in fact some say the genetic effects can be mitigated significantly by treatment).

Yes, of course genetic effects can be mitigated significantly by interventions!  That's entirely in line with what I'm saying, indeed it's what I have said, since no trait is 100% genetic!   Read the Primer that I've pointed to several times!  Here it is again:

"Top 10 Replicated Findings from Behavioral Genetics"
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4739500/

> If Plomin or any other serious expert researcher claimed what you do, you could have quoted them.

I've already quoted this from Plomin twice on the thread: "DNA isn’t all that matters, but it matters more than everything else put together in terms of the stable psychological traits that make us who we are."

> You seem to have no experience in the research area.

True, but then nor do you, in fact you're not even a scientist, and from the stuff you've said on the thread you're making multiple conceptual errors in evaluating it.

> You are certainly exaggerating in stating something as a fact that no one can know as yet

No, I'm not.  We *can* know these things from twin studies, and the statements that I'm making are in line with those of Plomin and similar leading experts in the field. 

> (and if it were true, it would have massive implications for all criminal justice systems), ...

Well yes! And that's partly why Plomin wrote his recent book, go and read the later chapters.  https://www.amazon.co.uk/Blueprint-How-DNA-Makes-Who/dp/0241282071

> ... and as such you are an embarrassment to science

Says someone too embarrased to put his name behind his posting antics. Says someone who is not even a scientist.  As I've said, you're doing what too many do today: if you don't like what someone has said and don't want or can't argue on the facts, then just attack them personally.

3
Coel Hellier 07 Aug 2019
In reply to Offwidth:

> ... doesn't seem to claim anything like as outrageous as Coel does: genes are the most likley reason you will end up in jail.

So exactly what is it that you regard as "outrageous" about the claim?

I amplified it above, saying:

"The biggest reasons for people ending up in jail are aggressive, violent and impulsive personalities, low IQ, low educational attainment, consequences of substance abuse, and similar things."

Do you disagree with that?   If so, which factors would you delete from the list, what factors would you add to it?   By the way, the claim is intended to be about incarceration within a society such as the UK, so let's consider the UK. 

I then claimed:

"And all of those traits have high heritability and are under strong genetic influence!"

Do you disagree with that?    

Offwidth 07 Aug 2019
In reply to Coel Hellier:

Your arguments on in utero effects make no sense to me. If both babies are affected it distorts the data. You often can't adjust for its effect as a lot of the data is old and won't have recorded things like the mother's smoking and alcohol use.

If Plomin says it, you would have quoted him already and I've linked alternative critical scientific views on his work from experts in the area (including a scathing review in Nature). 

If someone wants my name they can get it pretty easily from my profile. I might not be a scientist in your view but its rather weird that the University still pay me for such well imto my 4th decade and the other scientific academic staff in my School trust me to represent them in numerous formal respects.

Yes I agree there are genetic correlations in many areas but I probably will diagree with Plomin's conclusions and think you are being plain ridiculous stating genes are the main reason someone will end up in jail, from current evidence.

Post edited at 15:20
1
Coel Hellier 07 Aug 2019
In reply to Offwidth:

> If both babies are affected it distorts the data.

No it does not.   If it affects both babies when the twins are identical twins and also affects both babies when the twins are fraternal twins, then it does not distort the data.  That's because the studies are comparing what happens to identical twins to what happens to fraternal twins.   It seems you don't understand the principles of these studies.

> If Plomin says it, you would have quoted him already ...

Amazing how many times you ignore my quotes from Plomin!

> and I've linked alternative critical scientific views on his work from experts in the area (including a scathing review in Nature). 

No, the Nature review was not by an expert in the area, indeed he's not even a scientist, he's a historian.  And many non-scientists do indeed freak out at any mention of genetics and suggestions that genes are important (just as you are doing!).  They are virtue signalling by poo-pooing any departure from blank-slatism. 

Your second link is by the same author.

As for your third link, the Bath blog is not "high profile" and not by an expert in the area and not by a scientist, his connections are with Bath's department of education.

So, no, you have not "linked alternative critical scientific views on his work from experts in the area".

Coel Hellier 07 Aug 2019
In reply to Offwidth:

> ... and think you are being plain ridiculous stating genes are the main reason someone will end up in jail, ...

Feel free to attempt an answer to my post of 14:42.   Just freaking out at my claim is not a rebuttal. 

Pefa 07 Aug 2019
In reply to Coel Hellier:

> They are virtue signalling by poo-pooing any departure from blank-slatism. 

Come on no one thinks that we get no conditions passed down to us from our parents genetically as it is clear for all to see in height, looks and certain hereditary conditions. So that is a bit of a wild statement to make and virtue signalling? Do you really think anyone cares about that. 

From what I read so far in the links you provide in this thread the heritability estimates for psychological illnesses are usually between 30% - 50% at best. Which is significant however environmental conditions therefore could contribute anywhere from 50%- 70% in these cases.Meaning environmental conditions could be a greater influence on these psychological behaviours, no? 

I'm not so sure on his meta-analysis on substance abuse though as there was only one study sited. 

Post edited at 16:34
Coel Hellier 07 Aug 2019
In reply to Pefa:

> Come on no one thinks that we get no conditions passed down to us from our parents genetically as it is clear for all to see in height, looks and certain hereditary conditions.

Well ok, granted, but the "blank slatism" is taken to be about our personalities, our behavioural and psychological traits.  Blank slate ideology asserts that genes have little influence on those.

> the heritability estimates for psychological illnesses are usually between 30% - 50% at best. Which is significant however environmental conditions therefore could contribute anywhere from 50%- 70% in these cases.Meaning environmental conditions could be a greater influence on these psychological behaviours, no? 

Possibly, yes, though it's probably not sense to lump all psychological illnesses together, and likely not sense to lump all the different non-genetic factors into one "environment" either. 

E.g, according to this paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6421104/

"Heritability estimates from sibling data varied from 0.30 for Major Depression to 0.80 for ADHD."  

They also give estimates for autism (0.65), schizophrenia (0.58), biploar disorder (0.52), alcohol dependance (0.41) and others. 

Offwidth 07 Aug 2019
In reply to Coel Hellier:

Maybe you need to explain in small words for us thickos, as I just can't see how criminality linked traits from unrecorded alcohol or smoking effects in the womb can be split from recorded genetic effects, if no fair control sample was set up for that. Yes the effects will be secondary distortions but given the small numbers overall and looking for differential responses these could be important in impact sizes . Yes the old well known papers still clearly show a link to criminality but only with smallish impacts to specific types. Genes (other than being male) are not linked to overall propensity to being jailed in any of these papers, it just increases likelihoods of criminal traits in specific areas. It seems you need to read the source references and reviews again.

If Plomin says somewhere what you said, quote him... its easy... none of the quotes you linked did.

Nature don't publish reviews from random historians on books on such subjects, the author is a historian of biology and medicine (typical of your distortions).  None of the reviews seemed to deny his base results, just his writing, scope and conclusions...Also the links indicate the publisher refused many scientific reviewers an advanced copy.

Another expert one:

https://www.geneticshumanagency.org/gha/blogpost-more-on-robert-plomins-blueprint-nature-defeats-nurture-50-50/

Plomin is a social scientist, why you feel the need to dismiss other social science views in such a way is beyond me.

Hey, I did just discover Dominic Cummings likes him though, so you're in good company.

Post edited at 17:18
Pefa 07 Aug 2019
In reply to Coel Hellier:

Very interesting so in some psychological conditions like ADHD it is 80% heritable and for most others it can be anywhere from 30% to 60% heritable. Which leaves environmental factors to be from 40% to 70% the determining factors in these conditions. 

So in conclusion we still have more or less equal influence of both genes and environment from these studies of Plomin's that you show. 

Post edited at 17:53
Coel Hellier 07 Aug 2019
In reply to Pefa:

> So in conclusion we still have more or less equal influence of both genes and environment from these studies you show. 

Yes, correct, overall the genetic influence on variation in traits is about 50% and the environmental influence is about 50%.

But, what is notable -- and the point I started with -- is that the dominant environmental influence is not what most people think it is.  It is not "shared environment" (= experiences/environment shared by siblings), it is non-shared environment. 

This fact is Number 9 in the "Top 10 findings" primer linked above:

"It is reasonable to think that growing up in the same family makes brothers and sisters similar psychologically, which is what developmental theorists from Freud onwards assumed. However, for most behavioral dimensions and disorders, it is genetics that accounts for similarity among siblings. Although environmental effects have a major impact (see Finding 2), the salient environmental influences do not make siblings growing up in the same family similar. The message is not that family experiences are unimportant but rather that the relevant experiences are specific to each child in the family. This finding was ignored when it was first noted (Loehlin & Nichols, 1976) and controversial when it was first highlighted (Plomin & Daniels, 1987a, 1987b), but it is now widely accepted because it has consistently replicated (Plomin, 2011; Turkheimer, 2000)."

And:

"However, the basic finding that most environmental effects are not shared by children growing up in the same family remains one of the most far-reaching findings from behavioral genetics. It is important to reiterate that the message is not that family experiences are unimportant but rather that the salient experiences that affect children’s development are specific to each child in the family, not general to all children in the family."

Pefa 07 Aug 2019
In reply to Offwidth:

> Another expert one:

That is an excellent explanation for Plomin's big conclusions in his book. 

Coel Hellier 07 Aug 2019
In reply to Offwidth:

> Maybe you need to explain in small words for us thickos, as I just can't see how criminality linked traits from unrecorded alcohol or smoking effects in the womb can be split from recorded genetic effects, if no fair control sample was set up for that.

tl;dr: Short answer: The MZ twins are the control sample for the DZ twins.

Long answer: Take a large sample of twins.  Let's say half are monozygotic (MZ = identical twins), and half are dizygotic (DZ = fraternal twins).

Now let's say some smaller number of those are damaged by their mother's alcohol use.   Most likely, the fraction damaged by alcohol will be the same for the MZ sample as for the DZ sample.  (But even if it's not, that won't be a problem.)

What we are then looking at is differences between twin pairs.  So (to use the example in Plomin's book) we produce a scatter plot, plotting the height of one twin versus the height of the other twin.  Or the weight of the twins, or their GCSE scores, or their location on a political left-right axis, or whatever we're interested in. 

Now, if the twins were StarTrek-Transporter duplicates, identical in all respects, then the plot would be a straight line with each data point from each pair being beads-on-a-string along that line, agreed?   The correlation coefficient would be 1. 

First take the MZ twins. They are genetically identical, so any scatter away from the line on that plot is non-genetic (= environment).      And since that is significant, the correlation coefficient will be less than 1.

But the DZ twins are not genetically identical, so scatter away from the line is scatter from environmental effects plus scatter from genetic differences.  So the correlation coefficient of the plot will be lower than that for the MZ twins.

Then the difference in the correlation coefficients will tell us how much of the scatter (variation in traits) is caused by the genes.

Quoting Plomin: "In TEDS [a study] the MZ correlation for weight is 0.84, whereas the DZ correlation is 0.55. Because DZ twins are only half as genetically similar as MZ twins, the difference is correlations (0.84 versus 0.55) estimates half the heritability in weight. Doubling this difference in correlations puts heritability at 58 per cent".

So, let's suppose some fraction of the same are alcohol-damaged.  If the fraction is the same for both MZ and DZ (as is likely) then it will make no difference. 

But even if the fraction  were not the same it would not affect the heritability estimate.  That's because alcohol damage that affects both twins equally is just going to move them up and down the line, it will not affect scatter away from the line, and thus not affect the correlation coefficients.  

That's because what matters for the heritability estimate is *differences* between the twins, and that's not affected by factors that affect the two equally.

So, to confound the studies, you'd have to invent a scenario in which alcohol damage affected one twin more than the other, making them *less* similar, *and* argue that this mechanism worked differently for MZ twins than for DZ twins, *and* that the fraction of MZ twins affected by alcohol damage was different from the fraction of DZ twins affected.     All of that is not totally impossible, but it's getting pretty unlikely.

Post edited at 18:52
Pefa 07 Aug 2019
In reply to Coel Hellier:

Right so for example say three kids in a family all experience a traumatic say divorce of their parents with them fighting all the time and one kid develops OCD during this period it will be rare for the other kids to also get this. Makes sense as each may cope in different ways, one may shut off all emotion and end up with an early death from being worried all the time and giving their life to everyones needs except their own and the other could end up with a drug habit.The environmental factor as Offwidth posted above in his link is not so easy to measure that's why the one that is easiest to measure is the genetic one but it is still only half the picture. 

Incidentally there is no mention of drug use in these studies only alcohol and there must be 10s of millions of people regularly taking various narcotics in the USA alone

Post edited at 18:41
Coel Hellier 07 Aug 2019
In reply to Offwidth:

> Nature don't publish reviews from random historians on books on such subjects, the author is a historian of biology and medicine ...

But Nature does do a lot of virtue signalling these days.

>  None of the reviews seemed to deny his base results, just his writing, scope and conclusions...

Exactly, they don't deny his science, but they do dislike the book. 

So, having read that Nathanial Comfort blog post you linked to, let me quote from it:

"Plomin is so caught up in his DNA delirium that he says that environmental interventions toward human betterment are futile. Parents, teachers, government officials: Relax, there’s nothing you can do that really makes a difference. DNA will out. Kids will be what they will be, regardless, so don’t waste your money and time."

No, actually, Plomin says none of those things.  He really doesn't.  I've read the book, and he says the opposite.  And Nathanial Comfort nowhere quotes Plomin saying those things, so he "paraphrases".

He continues:

"Plomin’s argument is dangerous because it minimizes those absolutely robust findings. If you follow his advice, you go along with the Republicans and continue slowly strangling public education and ..."

No, actually, Plomin gives no such advice and nothing he says implies such advice.  Comfort is again inventing stuff.

But he continues:

"Plomin is spreading a simplistic and insidious doctrine that says “environmental intervention is futile.”"

That might look like a quote from Plomin, but it's not, it's a "paraphrase" (otherwise known as a porky pie).  Plomin says no such thing, he says the opposite. 

Now, Comfort sort of -- if you read between the lines -- admits that Plomin says none of these things.  Instead he just asserts that other people might say them, and claim they result from Plomin's work (which they don't).    This is not a rebuttal of anything Plomin has actually said or any of his work, it's purely a political objection to ideas that do not follow from Plomin's science and do not follow from his writings either.

1
Pefa 07 Aug 2019
In reply to Coel Hellier:

So when you wrote way upthread-

> The biggest reasons for people ending up in jail are aggressive, violent and impulsive personalities, low IQ, low educational attainment, consequences of substance abuse, and similar things.  And all of those traits have high heritability and are under strong genetic influence!   In comparison, the contribution of factors such as "brought up in a household where criminality is the norm" (referred to as "shared environment" in the studies) is much lower.

Were you mis-quoting Plomin to? 

Post edited at 18:53
Offwidth 07 Aug 2019
In reply to Coel Hellier:

The results show genetically identical twins obviously develop differently for different factors, giving different repeatable standard correlations for them; but similar differences for alcohol effects in utero ( or anything else) could give different correlations for those same factors and although its likely the correlation in response for identical twins will be higher it won't neccesarily be by the same amount, so if you don't control for it, this could cause secondary scatter from a pure genetic variance.

Post edited at 19:19
Coel Hellier 07 Aug 2019
In reply to Offwidth:

> The results show genetically identical twins obviously develop differently for different factors, giving different repeatable standard correlations for them; but similar differences for alcohol effects in utero ( or anything else) could give different correlations for those same factors and although its likely the correlation in response for identical twins will be higher it won't neccesarily be by the same amount, so if you don't control for it, this could cause secondary scatter from a pure genetic variance.

Can you actually explain that?

Coel Hellier 07 Aug 2019
In reply to Pefa:

> Were you mis-quoting Plomin to? 

It was neither a quote nor a mis-quote.

But it is notable that many of the prime factors for ending up in prison (aggressive, violent nature of a sufficient degree to get you locked up; low IQ, low educational attainment; and substance abuse) are some of the things with rather high estimates of heritability.

TobyA 07 Aug 2019
In reply to Coel Hellier:

You completely miss the issue that different types of people are treated differently by different criminal justice systems. That's because of some social label put on them, not any particular psychological trait.

Coel Hellier 07 Aug 2019
In reply to TobyA:

> You completely miss the issue that different types of people are treated differently by different criminal justice systems.

I've not missed it.  Your point is obvious, not that interesting, and irrelevant to anything I've claimed. 

4
TobyA 07 Aug 2019
In reply to Coel Hellier:

How can it not be relevant for asking why people end up in prison? 

I find it very hard to believe that you are really that naive about how criminal justice systems work.

Coel Hellier 07 Aug 2019
In reply to TobyA:

> How can it not be relevant for asking why people end up in prison? 

Stating that criminal laws are different in different countries -- which is obvious and not that interesting -- is not that relevant to issues of what factors determine who ends up in jail within a given country. 

> I find it very hard to believe that you are really that naive about how criminal justice systems work.

I'm not. 

TobyA 07 Aug 2019
In reply to Coel Hellier:

> Stating that criminal laws are different in different countries 

That's not what I said. I said different groups of people get treated differently within different criminal justice systems. Like I asked earlier, do you think African Americans are just genetically more criminal than white Americans? Surely you don't think it's that simple?

I'm running out of data on my phone, I've been following this thread from various campsites across France - so as much as I'd like to discuss this further, it's probably going to have wait until the end of the week when I'm home. I do know how you like to have the last word in these lengthy debates! If I don't reply further, please don't take it as you having convinced me. ;-)

1
birdie num num 07 Aug 2019
In reply to Pete Pozman:

I do feel terror 

Coel Hellier 08 Aug 2019
In reply to TobyA:

> I said different groups of people get treated differently within different criminal justice systems.

There are several issues there:

* That different criminal justice systems have different effects is obvious, but it's not what the people doing the twin studies are interested in, so is not what the studies are designed to do or tell us about.

* Differences between group means are smaller (often a lot smaller) than differences between individuals, and again it's the latter that the twin studies are about.   You'd do different studies if your interest was differences in group means.

* It is not the cause that the causes of individual differences are necessarily the same as the causes of differences between group means.   That's a fallacy.

Thus, for example, it's entirely consistent for it to be the case that the cause of the diversity of a trait among British schoolchildren is largely genetic, and that the diversity of that same trait among Japanese school children is also largely genetic; but that the difference between the Japanese mean and the British mean is cultural. 

Finally, a reminder that no-one has said that any trait is 100% genetic and than environment does not matter.  Despite the misrepresentations of Nathaniel Comfort et al, people are saying that both genes and environment matter a lot.   (Though *which* sorts of environment matter can be crucial.)

So, all of that means that if you want to start discussing differences between group means, then ok, happy to do it, but one can't just simplistically extrapolate and assume from statements about studies of diversity of *individuals*.

Offwidth 08 Aug 2019
In reply to Coel Hellier:

No... I'm not falling into a trap of following your red herring (bottom line I'm just not convinced in utero afects with dependant variables can't alter the differences in correlations from the scatter plots).

The big picture is that demographics of incarceration vary so much from country to country, with very extreme BME imbalances in places like the US, that as usual, your extreme thesis is ludicrous when faced with realities. The critique of the likelihood of being jailed mainly being determined by genes is more easily tackled following Toby's line. Going to jail is a much more complex interaction of nature and nurture than someone's height, and unlike height, depends on morals and free will, and state systems and their cultures. The overall numbers are probably too small to show clearly one way or another, with current twin studies with incarceration rates well below 1% in the west, apart from the US.  It's the same with your arguments on Islam... tens of millions of good people written off based on ideology, cherry picking of statistics, highlighting of problems in failed states and misrepresentation of history (on that subject I was delighted to see the BBC show 'The Art of Spain' celebrating the early Moorish kingdom in Cordova being probably the most enlightened in the world in the 9th and 10th centuries... far ahead of christian Europe.)

Post edited at 14:37
2
Coel Hellier 08 Aug 2019
In reply to Offwidth:

> ... bottom line I'm just not convinced in utero afects with dependant variables can't alter the differences in correlations from the scatter plots).

But you've given no clear reason why not.

> The big picture is that demographics of incarceration vary so much from country to country, ...

My claim wasn't even about variation over countries.  In context, in a thread about a British Home Secretary, and about crime in the UK, it was a claim about crime within a nation similar to the UK.

> ... that as usual, your extreme thesis is ludicrous when faced with realities.

Just using words like "extreme" and "ludicrous" is not even an attempt at rebuttal  (especially as you then avoid actually examining the claim; see my post at 14:42 yesterday). All you're doing is emoting because you dislike opinions not in-line with yours.

> Going to jail is a much more complex interaction of nature and nurture than someone's height, and unlike height, depends on morals and free will, and state systems and their cultures.

Firstly, our morals and wills also depend on genes. Human nature and human psychology derives to a large extent from genes.  Second, as stated repeatedly, the claim is not about country to country variations, it's about what leads to some individuals being incarcerated (when the majority are not) within a given justice system.

>  The overall numbers are probably too small to show clearly one way or another, with current twin studies with incarceration rates well below 1% in the west, ...

Which is merely a made-up dismissal on your part, based on nothing other than you wanting it to be true.  Most experts  would agree with me that strong factors for ending up in jail are having violent, aggressive and impulsive personalities, having a low IQ and/or low educational achievement, and having drug dependencies -- all of which have high heritabilities, showing that they are strongly genetically influenced.  

Again, if you don't think that those are key factors (talking about the UK here) what would you add to or subtract from the list?

>  It's the same with your arguments on Islam...

It's amazing how often you turn the topic to Islam -- when you've also complained of me over-doing the topic!

> tens of millions of good people written off based on ideology, ...

Wrong.  I've not "written off" "tens of millions of good people", all I've done is critique the ideology.   You really do like inventing stuff that you attribute to others, don't you?

> ... the early Moorish kingdom in Cordova being probably the most enlightened in the world in the 9th and 10th centuries... far ahead of christian Europe.)

If you want me to agree that Islamic countries were often ahead of Christian theocratic countries back in those centuries, then sure, yes, I agree.   But it's rather damning with faint praise isn't it?   Do you really have to go back to the 9th and 10th centuries to find things to praise them for?  Nowadays, the Islamic world is way behind the secular Western countries on nearly all indicators. 

1
Offwidth 08 Aug 2019
In reply to Coel Hellier:

I always agreed genetics will play a part in likelihood of incarceration, the ludicrous position you have is claiming it is the main reason, with no evidence other than lots of other contributory factors correlate. Just as I agree plenty of evil has been done in Islam's name, but I think its ludicrous to claim Islam is evil. You have exceeded by far the limits of evidence to push ideology, just like those you often criticise. 

1
Coel Hellier 08 Aug 2019
In reply to Offwidth:

> the ludicrous position you have is claiming it is the main reason, ...

Again, labeling it "ludicrous" is not a rebuttal, and you've not even attempted a rebuttal.

> ... with no evidence other than lots of other contributory factors correlate.

The point is that what are widely agreed to be the biggest factors in who gets sent to jail (and again, you've not attempted to rebut that list) are all traits with rather high heritability and thus genetic influence. 

Then there is the study cited up-thread for which the conclusions are: "For all criminal convictions, heritability was estimated at around 45% in both sexes, ..."

Now, unless you lump all non-genetic factors into one "factor" (which would be a bit perverse), that means that genetic influence is the biggest single element.   And lots of similar studies have been cited up-thread.

So, your "no evidence" claim is just wrong.  And it's vastly more evidence than you've produced. 

> ... but I think its ludicrous to claim Islam is evil.

Again, using the word "ludicrous" is not an argument. And I do indeed maintain that, considered overall, Islam does much more harm than good in the world.  I've expounded at length on why, and you're quite welcome to actually debate the issues if you care to. 

Offwidth 08 Aug 2019
In reply to Coel Hellier:

I know it's not an argument, it's a statement of fact. Even if your statement turns out to be true with future research, stating it now as a fact is unjustified. You just don't care that your wilful exaggeration in science is plain wrong and seem to delight in playing the iconoclast. cb294 shows how you could have behaved well discussing the same data.

I've rebutted with numerous links that disagree with you upthread, as have others. You ignore epigenetics where the story goes full circle and environment influences genes. Being able to state incarceration is mainly due to genes cannot be known according to anything I've seen.  It would have massive implications for criminal justice which is mostly based on intent to commit crime.

Your level of debate on this thread includes accusations of my denialism when you know full well we probably agree with all the base research data; childish ad hominem attacks like saying I'm not even a scientist (despite working as a science academic); and in private emails calling me a 'vile shit'. Why on earth would you need to do all that if my points are not worthy of relevance in debate? Yes I've attacked you, but only for what you have said. Despite your certainty, you still haven't produced a quote from any scientist working in the field backing your specific claim.

Post edited at 19:24
2
Coel Hellier 08 Aug 2019
In reply to Offwidth:

> Even if your statement turns out to be true with future research, ...

Rather a weird thing to say about a claim that you've labelled "extreme" and "ludicrous", no?

> ... stating it now as a fact is unjustified.

Wrong.  It is justified by findings such as: "For all criminal convictions, heritability was estimated at around 45% ...".

My claim was that genes were the biggest factor.  That is then  indeed true, unless one takes everything in one's entire life that is not genetic and calls that a single "factor", which would be a rather weird use of language.

And my underlying point was that genetics is a bigger factor than "shared environment", and that's exactly what that study (and all similar studies) states. 

> You just don't care that your wilful exaggeration in science is plain wrong ...

Sorry, it's what the study says! Read it!

> ... and seem to delight in playing the iconoclast.

OK, guilty, I do admit that I bit of perverse pleasure in twitting left-wing science-denialists who proceed on emotion and ideology rather than evidence.  (Please note, everyone, that that sentence does *not* say that all left wingers are of that ilk!)

> ... cb294 shows how you could have behaved well discussing the same data.

That's pretty amazing, from someone who has repeatedly attacked me personally, just because I hold a different opinion, and who has, on the thread, called me deceitful and dishonest based on a *wrong* *guess* at what did not in fact happen!

> You ignore epigenetics where the story goes full circle and environment influences genes.

Which is adequately accounted for in twin studies since MZ twins act as controls for DZ twins, as I've explained (and also, epigenetic claims are anyhow exaggerated and weak; see up thread).

> It would have massive implications for criminal justice which is mostly based on intent to commit crime.

Again, you seem to have some notion that our "intent" and "will" is not the product of our genes and environment, which is non-scientific nonsense.

> Your level of debate on this thread includes accusations of my denialism ...

Which is true, since you reject the science when it conflicts  with your ideology.

> ... when you know full well we probably agree with all the base research data;

Do we?  OK then, do you agree that "For all criminal convictions, heritability [is] around 45%"? 

Do you agree that: "strong factors for ending up in jail are having violent, aggressive and impulsive personalities, having a low IQ and/or low educational achievement, and having drug dependencies"?

Do you agree that "all of these have high heritabilities, showing that they are strongly genetically influenced"? 

It's notable that you repeatedly avoid getting into the substance of the claim.

> childish ad hominem attacks like saying I'm not even a scientist (despite working as a science academic);

Well, you're more of a technology and computer and engineering guy, and more into the management side of things these days.   Your own website says so.  Not that there's anything at all wrong with any of those things.

> and in private emails calling me a 'vile shit'. 

Which was in response to you calling me deceitful and dishonest because you had made a wrong guess at what did not happen, but instead of first establishing the facts you leapt to the conclusion that I was being deceitful and dishonest, and said so on a public forum, while all the time taking pains to hide behind a pseudonym. 

It's amazing that you don't see anything wrong with that and don't seem to be at all shamefaced, since you seem to think there is something wrong with my justified retort.

> Why on earth would you need to do all that if my points are not worthy of relevance in debate?

It wasn't "your points" (meagre and occasional though they are), that elicited my response.

> Yes I've attacked you, but only for what you have said.

Quite clearly factually wrong!  That really is amazing!  Re-read the paragraph just above and the relevant parts of the thread!

> Despite your certainty, you still haven't produced a quote from any scientist working in the field backing your specific claim.

The paper that cb294 cited gives sufficient substantiation, see opening paragraphs of this comment. 

1
Offwidth 09 Aug 2019
In reply to Coel Hellier:

I'm not convinced the MZ vs DZ methodology works for gene dependant variables and epigenetics. You say it just moves pairs up and down the line but that's an assumption. I think you're convinced by Plomin, and go beyond him; ignoring the warnings of the likes of Adam Rutherford (that I linked early on) who explains why we need to be careful about extrapolating with things like 'warrior genes'. Fundamentally the whole way we look at criminality in the west is wrong if genes dominate and we are largely biological machines who mostly kid ourselves about nurture and free will (in such complex social situations and systems that will get us jailed). Before we throw out that massively important bathwater I really want to be sure there are no babies. I never had the biases that those who deny genetic effects have: I'll just follow the science... criminality IS linked to genes. I am however sensitive to far right wing ideologues who will seemingly grasp at anything to 'prove' bogus genetic links (given the criminal history of the far right with eugenics). Irrespective, I think claiming anything in Science as fact, that is not yet properly confirmed, is wrong. Scientists hypothesise and might even predict what they expect (with suitable disclaimers);  they don't say what IS in advance. In philosophical terms any scientific results provide evidence wrt to the best current models, so some would say nothing is ever a fact.  When the shadow of eugenics is present, I think much greater care than normal is required.

I've had a long career. I'm a graduate physicist who has also worked extensively in materials science, physics and physical chemisty, alongside some of the best in the UK (before we started to 'gut' government and industrial materials research labs in the late 80s). I had the luck to get to 'play with' a huge range of analytical techniques (physical and mathematical) and manufacturing development in a very difficult research area which was, back then, very generously funded.

The left wing bit is another of you regular dishonest exaggerations: I'm a fairly standard social liberal, centrist. 

If you read the whole paper cb294 cited I'd lay odds it's more careful in its claims than you are. You're really clutching at straws if that abstract is the best justification you currently have.

What insults do you mean that I made then... things like 'Poundshop Peterson'? He is famous and just as obsessed with calling centrists leftists and mighty stubborn and, although often right to have concerns, is often wrong in the grandiose spin he applies to them. I thought you might be proud of the comparison, given your similar habits and it applies pretty well to what you say here. 

On calling you dishonest on a specific point  I did something you might want to consider one day... I apologised.... when I realised the mods might just have deleted the thread and not have warned you before, not to use my real name... albeit that's based on trusting your honesty. Plus it looked that way given you had emailed me privately before (end of last year?) with similar insults and had a post removed for using my name around the same time. On the rare occasions my posts got removed the mods explained why and asked me not to do it again. 

1
Coel Hellier 09 Aug 2019
In reply to Offwidth:

> I'm not convinced the MZ vs DZ methodology works for gene dependant variables and epigenetics.

Yet it has been an established and accepted technique on the field for 30 years now.  You need a better reason than you've given if you want to avoid suggestions of denialism.

> You say it just moves pairs up and down the line but that's an assumption.

No, it was a reasoned argument.

> I think you're convinced by Plomin, and go beyond him; ignoring the warnings of the likes of Adam Rutherford (that I linked early on) who explains why we need to be careful about extrapolating with things like 'warrior genes'.

Rutherford warned against over-interpreting links to *single* genes or a few genes.  He is *right* on that.  But that is not what either Plomin or I are talking about.  I've made that explicitly clear repeatedly.  I've repeatedly pointed to Plomin's "Top 10 findings" primer, of which number 3 is explicitly: "Heritability is caused by many genes of small effect".

> Fundamentally the whole way we look at criminality in the west is wrong if genes dominate and we are largely biological machines who mostly kid ourselves about nurture and free will ...

Read up on "compatibilism" for why this is not true.  It *is* the case that our "will", our decision making, *is* determined by some mixture of genes plus environment in all its forms. The implications you draw are there regardless of whether our will is determined by "50% genes, 50% environment" instead of being determined by the blank-slatist "it's all environment".

And yes, you *are* kidding yourself if you don't accept that and instead adopt some non-scientific nonsense about a dualistic "soul" doing the "willing".

> If you read the whole paper cb294 cited I'd lay odds it's more careful in its claims than you are.

It's amazing how you just hope for things like that, I thought you said you wanted to judge on the evidence?

> You're really clutching at straws if that abstract is the best justification you currently have.

Well let's see, my claim was that the biggest factor in who gets sent to prison is genetic, and I then asserted that "shared environment" was a smaller effect. 

And that study says: "For all criminal convictions, heritability [is] around 45%".   So I was right, genes are the biggest factor.

(Unless one wants to group together absolutely everything in ones life that is not genes and call it "one" factor, which is a rather weird use of language; and if that's what you're doing then all you're doing is making a rather sad semantic quibble.)

And you call that "clutching at straws"?  Really??   Pointing to a paper that says that genes account for 45% of the variation, with all other factors then being smaller, is somehow "clutching at straws" when I'm trying to defend a claim that genes are the biggest factor in who gets sent to jail?

> What insults do you mean that I made then ...

Calling me deceitfull and dishonest!  Duh! And "an embarassment to science" and similar stuff.   When, actually, the peer-reviewed literature supports me not you.  

Not that I care that much about online insults, but as I've said I do regard it as cowardly to insult people while hiding behind a pseudonym.  Why not put your name being your insults? And, as I've said, on other threads you've complained about UKC becoming an unfriendly environment!  Do you really lack the self-awareness to realise that you are one of the posters most likely to make things unfriendly (and I'm not just talking about threads involving me)?

4
Coel Hellier 09 Aug 2019
In reply to the thread:

This thread is getting long, so just to reiterate the main point:

My original claim was that genes were the biggest factor in who gets sent to jail, and that "shared environment" (environment that siblings would share, such as being brought up in a home where crime as accepted and normal) was a smaller effect.

From the paper that cb294 cited, twin studies show that the contributions to convictions for violent crime are:

Genes: 45%
Non-shared environment: 31%
Shared environment: 24%.

Numbers are somewhat different for other sorts of crime, but the overall heritability of criminal convictions is 45%. 

So I was quite clearly right.

Meanwhile, Offwidth has indulged in an amazing palava of whataboutery and called my claim "ludicrous", "extreme", and an "embarrassment to science", and asserted that it was going way beyond the evidence, and said that my pointing to that paper is somehow "clutching at straws".   

People are welcome to draw conclusions, and wonder why he doesn't want to put his name to his posts. 

8
abr1966 09 Aug 2019
In reply to Coel Hellier:

I haven't read the whole thread in detail and know only a little about genetics but I do recall previously engaging on a thread with you Coel where you were making huge generalisations and passing opinion off as evidence. I can't recall the detail but it was something linking in to my work as a psychologist and I asked you for specific evidence to support your claim.....which you did not or could not do for the obvious reasons.

1
Coel Hellier 09 Aug 2019
In reply to abr1966:

Your comment rather lacks any substantiation, so is just empty put-downs. 

4
RomTheBear 09 Aug 2019
In reply to Coel Hellier:

> This thread is getting long, so just to reiterate the main point:

> My original claim was that genes were the biggest factor in who gets sent to jail, and that "shared environment" (environment that siblings would share, such as being brought up in a home where crime as accepted and normal) was a smaller effect.

> From the paper that cb294 cited, twin studies show that the contributions to convictions for violent crime are:

> Genes: 45%

> Non-shared environment: 31%

> Shared environment: 24%.

> Numbers are somewhat different for other sorts of crime, but the overall heritability of criminal convictions is 45%. 

> So I was quite clearly right.

The problem is that these studies typically cannot be replicated, used heavily massaged data and p hacked models, have a huge survival bias as only the ones with interesting results are published,  and are conducted by people with a skin deep understanding of probability and statistics.

The reality is they have no effing clue. The whole field of social science is completely shot.

I’ve read the paper, it’s heavily short on details, they are no transparent about the data, but the methodology itself is a joke. Look at these correlations, they are completely meaningless, they don’t even know how to interpret them.

Post edited at 17:01
1
Coel Hellier 09 Aug 2019
In reply to RomTheBear:

> The problem is that these studies typically cannot be replicated, used heavily massaged data and p hacked models, have a huge survival bias ...

Your point is indeed fair enough about large swathes of social science.  But it's not fair about all of it, and the stuff we're talking about here has been replicated multiple times.  That's why the Plomin primer is headed: "Top 10 *Replicated*  Findings from Behavioral Genetics."

From the intro: "A recent concern in psychological science is that many statistically significant findings, including some classic findings, do not replicate [...] In this context, the purpose of this paper is to highlight 10 findings about the genetic and environmental origins of individual differences in behavior that have consistently replicated. On the basis of our decades of experience in the field of behavioral genetics and our experience in writing the major textbook in the field (Plomin, DeFries, Knopik, & Neiderhiser, 2013), we selected these 10 findings because in our opinion they are ‘big’ findings both in terms of effect size and their potential impact on psychological science. These findings are not novel precisely because we have selected results that have been repeatedly verified."

2
TobyA 09 Aug 2019
In reply to Coel Hellier:

> So I was quite clearly right.

Do your siblings/children/parents often state the same thing? I'm certain the urge to state that can't be a result of environmental factors. ;-)

Coel Hellier 09 Aug 2019
In reply to TobyA:

> I'm certain the urge to state that can't be a result of environmental factors. ;-)

You're getting the hang of this! 

Offwidth 10 Aug 2019
In reply to Coel Hellier:

Still going eh?... for the nth time my name is easy to work out on my profile. Also there is a nice bloke with the same user name as my real name and at least one other user with the same name as well... my name is common, but I'm the only user called Offwidth on UKC.

You are right that at a perfunctory level I might seem to be hypocritical in the way I have said forum users should be treated. My main concerns about the friendliness of forums are to do with the way new posters get treated and the under-representation of women and ethnic minorities, compared to the numbers I see in climbing. UKC is too male and too white and intolerent of new users unaware of site memes. I've never had any issue whatsoever about robust debate with regulars; and as an academic inconsistently intolerent to a mainstream religion (Islam), a willful and rude exaggerator, and someone who regards UK liberals as left wing, you fit right in a venn diagram overlap of my some of my strongest debating concerns. Sure I do sometimes get things wrong in terms of tone... I'm human... but I normally apologise when its clear I'm wrong (even with you).

2
Coel Hellier 10 Aug 2019
In reply to Offwidth:

> Sure I do sometimes get things wrong in terms of tone... I'm human... but I normally apologise when its clear I'm wrong (even with you).

Sometimes: "get things wrong in terms of tone"?   Your first two posts on this thread are pure attack-dog (go and re-read them!).  No seeking to have a couple of iterations first, you just go straight in with the insults. 

And you were in the wrong on the science!  The claims I made are in-line with the peer-reviewed literature, as defended, for example, in my post yesterday, 14:23 Fri. 

So why the default attack-dog mode? Is that how you think discussions of science should be conducted? 

It's pretty clear that you simply weren't familiar with this field of twin studies.   And from your comments on "free will" it's clear you've never really thought through issues of criminal justice, given what we know from science, nor read those who have.     Of course there's nothing wrong with that -- there's plenty of areas that I don't know about and have never read up on.  

But your default attack-dog posting towards anyone who doesn't fit your ideological mold, when you are the ignorant one unfamiliar with the science, is not how academics should behave.    Which is why I consider it cowardly for you to ask the mods to delete comments just because I addressed a reply to you with your name. 

Overall, your tactics here (whataboutery about the science; refusal to accept peer-reviewed consensus; claiming that your opponents are politically motivated; and personal attacks on them) are exactly what I face when I go on some US websites to defend the science of climate change.  It's tiresome.  Of course I expect it from them; sadly I've come to expect it also from too many of your ilk.   If you're not actually a science denialist, you're doing a good impersonation of one. 

And as for apologising, you still haven't apologised for calling me deceitful and dishonest when you had simply leapt to false conclusions without first knowing the facts.  (Again, is that how you should behave?)  And instead you've then tried to turn the tables on me by suggesting that I apologise -- for what exactly? 

5
Duncan Bourne 10 Aug 2019
In reply to TobyA:

> Where? When? Are Americans a lot more prone to all of those things as they imprison people at higher rates than I think ANY other country? Do they account for imprisonment rates in apartheid era South Africa? Or in Xinjiang Province now?

I think Coel has a point in the general sense. People with aggressive, violent and impulsive personalities, low IQ, low educational attainment, consequences of substance abuse, etc. are more likely to end up in prison in this country and America as such behaviour puts them at odds with the social mores of the society they live in. The examples you raise constitute an additional layer ie the prejudice of those enforcing the laws and the make up of a countries laws not to mention the social strata of that country where being black automatically puts you at a disadvantage.  

> What evidence could possibly prove that? 


I am not sure what is being asked here are you saying prove to me that violent, stupid, drug takers, and poor people who steal break the law?

Duncan Bourne 10 Aug 2019
In reply to Coel Hellier:

> My original claim was that genes were the biggest factor in who gets sent to jail, and that "shared environment" (environment that siblings would share, such as being brought up in a home where crime as accepted and normal) was a smaller effect.<

I am with Sapolsky on this one:

“You are studying the behavioural effects of a gene in two environments. Someone asks, “What are the effects of the gene on some behaviour?” You answer, “It depends on the environment.” Then they ask, “What are the effects of environment on this behaviour?” And you answer, “It depends on the version of the gene.” “It depends” = a gene/environment interaction.”

Coel Hellier 10 Aug 2019
In reply to Offwidth:

And to add, just noticed:

> My main concerns about the friendliness of forums are to do with the way new posters get treated and the under-representation of women and ethnic minorities [...]  I've never had any issue whatsoever about robust debate with regulars ...

Well, what you complained about on another current thread is:

"The dislikes in particular adds damaging negativity (particularly awful on posts on the starting out forum but unwelcoming in general), Many people tell me they don't like the increased negativity of the site, compared to earlier days,  so they only visit now and don't post or in some cases have given up reading the forums altogether."

Which is clearly a comment about long-term users.  So you're not keeping your story entirely straight.  And, if you don't realise that you are one who adds substantially to the unfriendliness, in that you are one of those most ready to move from a difference of opinion to a personal attack, then you are hugely lacking in self-awareness.   

(And again. that's not just about interactions with me; and plenty of people can disagree with me without descending to personal attacks, such as TobyA for example.)

3
Siward 10 Aug 2019
In reply to Pete Pozman:

A viewpoint this week from Chris Henley QC, chair of the Criminal Bar Association:

"Criminals ‘to literally feel terror’? Really?
 
This week the incoming Home Secretary announced her arrival in the new cabinet of all the talents by promising to make criminals ‘literally feel terror’ at the thought of committing offences. Well, yes. Similar promises were made about making radical improvements to the state of prisons within a year, a little over 12 months ago. The words were fine, the aspiration laudable, and no doubt sincerely meant, but today so dangerous has HMYOI Feltham become that there is currently a stop on any further teenagers being sent there. The ‘fine words’ failed to stem a further deterioration in the unsafe and miserable conditions. Conditions there are described as having 'plunged' over the last 6 months in the latest Chief Inspector’s inspection report
 
So too with the Home Secretary’s words, which are so far off being squared with the reality that it is a challenge not to laugh. At the weekend in the middle of the day in central London it took four minutes for a 999 call reporting an assault on a vulnerable woman to be answered. The voice that eventually came down the line apologised for the delay and their inability to send anyone; its not much of an exaggeration to say there is almost no one there any more. 
 
Reported crime has been rising rapidly over the past 8 years, yet prosecutions have fallen by 45% during the same period; from 895,000 cases prosecuted in the Crown and Magistrates Courts in 2011/12 to just 495,000 in 2018/19. Serious fraud prosecutions are in catastrophic decline. The police are not just struggling to cope, they are not coping. Only 7.8% of reported crime results in a charge within 12 months of the initial report. You will have read the report last week about the serious assault case in Truro which had taken 2 years 9 months to reach its conclusion. The Judge described the delay as a ‘disgrace’, and evidence that the system was ‘beyond the point of collapse’. He felt compelled in the circumstances to suspend the custodial sentences.
 
The police and the CPS are being denied the resources they desperately need to keep us safe. It’s not a pretty picture. However politically expedient the rhetoric of ‘instilling terror’ we are years away from instilling a degree of mild nervousness."

Duncan Bourne 10 Aug 2019
In reply to Siward:

May be she meant she was going to install a huge Bat signal on the Houses of Parliment and have masked vigilantes roaming the streets?

Offwidth 10 Aug 2019
In reply to Coel Hellier

I usually have no idea why someone dislikes nor how long they have been a user here unless in a small minority of cases they say they did it and explain why.  I think dislikes are bad for the site and have explained why many times.

When does something like writing off all of Islam, again and again and again, its more important to me to challenge such regular extremism than worrying about the negativity necessarily involved. It's the worst I've seen here that survives the moderators and makes the site unwelcome for a good proportion of the UK population, and I think its particularly shameful it's coming from an academic.

3
Coel Hellier 10 Aug 2019
In reply to Offwidth:

> When does something like writing off all of Islam, again and again and again, its more important to me to challenge such regular extremism than worrying about the negativity necessarily involved.

I note your complete lack of shame for your attack-dog posting style, even though all I was doing was making claims in-line with the peer reviewed literature, and even though, to anyone who knows about science, it's been obvious that you were wrong ever since, early on, cb294 cited a paper that sufficiently justifies my claim.

I note that you ignore the substance of that and, once again, try to change the topic to Islam, which is pretty irrelevant to the thread.

I note that you sort-of and tacitly admit that your attack-dog posting style on this thread was not about the science, but just a desire to attack me personally owing to stuff I've said on other threads. And that, despite claiming to be judging on the evidence.

And I note your ongoing attempts to disallow any view that sees Islam as harmful by labeling it "extreme" and "shameful".

5
wintertree 10 Aug 2019
In reply to Coel Hellier:

> and even though, to anyone who knows about science, it's been obvious that you were wrong ever since, early on, cb294 cited a paper that sufficiently justifies my claim.

I am explicitly not commenting on the original subject of this thread.  I stated my views and had/have no interest in talking round in circles for another 4 days.

Checking back in, I feel the need to comment on what you just wrote (above).

To many people who know about science, a single paper “proves” nothing.  I’ve seen some absolute tripe published right up to the level of Nature and as a scientist would never use the argument of “here’s a single paper so it’s proved”.  I disagree with your “anyone who knows about science” line on multiple different levels.

Post edited at 16:34
Coel Hellier 10 Aug 2019
In reply to wintertree:

> ... as a scientist would never use the argument of “here’s a single paper so it’s proved”.

Which is why I didn't use the word "proved" and did not claim that the single paper "proved" it was right. 

What I said, quite explicitly, is that that paper "sufficiently justifies my claim", meaning that it justifies it sufficiently to make the claim on a UKC thread, without it being described as "ludicrous", "extreme", "idiotic", a "willful exaggeration", an "embarassment to science", and "way beyond current science" and all the other rather-pathetic non-attempts at non-rebuttal that a certain poster has been too shy to put his name behind. 

By the way, I also would not have made such a claim if it were based only on a single paper, but the point is that findings along these lines have been consistently replicated multiple times, as Plomin's primer and book and other stuff attest. 

Post edited at 17:06
1
wintertree 10 Aug 2019
In reply to Coel Hellier:

> Which is why I didn't use the word "proved" and did not claim that the single paper "proved" it was right. 

> What I said, quite explicitly, is that that paper "sufficiently justifies my claim", [...]

Yes, and in your original claim “But the point is to overcome the refusal of swathes of society to properly address these issues, as exemplified by this thread, and for that a bit of stark language is in order. ​​​​​​“ you appear to me to take this point as proved beyond any reasonable doubt.  Hence my use of “proved” in inverted commas and not claiming you out-and-out said it.  Context.

> [...] meaning that it justifies it sufficiently to make the claim on a UKC thread, without it being described as "ludicrous", [...]

Well, that’s been a matter of debate for the last few days, and I don’t see much of a verdict emerging.  A poster more in the know than me about biology doesn’t share anything like the depth of my concerns over “unknown unknowns” in these studies [*], other posters question the quality and genuine reproducibility of social science studies.  I’ll check back in a few more days to see who gets the last word.

[*] working with a two-organism model system to push to getting five sigma repeatability week-to-week is teaching me a lot about unknown unknowns in even the simplest organisms without genetic spread.  There is a temptation as a physicist to step into the biosciences and oversimplify...

1
TobyA 10 Aug 2019
In reply to Coel Hellier:

With my original concerns about your claim that genetics is the reason that most people are in prison, you don't even seem to have thought about the problem of the difference between the reason why an individual might be in prison compared to the imprisonment rates of certain types of people. You hadn't specified what society or legal system you were talking about at the start so you had to later accept that within different systems the case could be very different as to why certain types of people end up in prison. Ok so then you say you're talking about UK, that's fine but the original claim still seems very difficult to ever prove because even within the UK different types of people get treated differently by the system. Look at the difference between the prison system in Northern Ireland and mainland Britain, even now but clearly 20 years ago, to show how social factors can play major role in accounting for why someone might end up in prison. It seems quite obvious that with  two people who came from very similar backgrounds, went to the same school, had similar families and in one case ends up in prison and the other case not, then psychological traits which are inheritable would clearly be the obvious place to look for the reason of the different outcomes. But people with totally different types of social backgrounds I don't see how you could ever make the claim search genetically inherited traits of them as individuals is the reason why they ended up in prison.

1
Coel Hellier 10 Aug 2019
In reply to wintertree:

> Yes, and in your original claim “But the point is to overcome the refusal of swathes of society to properly address these issues, as exemplified by this thread, and for that a bit of stark language is in order. ​​​​​​“ you appear to me to take this point as proved beyond any reasonable doubt.

Leaving the specific claim aside, I would regard the more general points as "proven" in the sense of being established and replicated science that people should accept.  That could be summarised as done in Plomin's primer, or the ideas that: all psychological traits have strong generic influence, much more so than commonly supposed; "shared environment" seems to be much less influential than commonly supposed; and many of the traits underlying criminality (aggression, low IQ, low educational achievement, drug addiction) are among the traits with the highest heritabilities).

From there, yes, the particular percentages quoted might be subject to unknown biases, and might change a bit, but I think that the broad picture is now pretty well established.  And, as I've said, I really am just pointing to criminality as an example of a much wider picture. Twin studies really have been saying the same thing (in a consistent and replicated way) for 30 years now. 

I remember reading Judith Rich Harris's "Nurture Assumption" 20 years ago now, and I can still remember the experience. It was one of the few times in my life where I've started reading a book having one opinion, and by the end of the book my opinion had completely changed!   Things have moved on since then, so Plomin's book "Blueprint" is a better account as being much more recent, but it is notable that even 20 years later, this stuff has not yet seeped into popular appreciation, even though it is important. That's partly because people react to it like Offwidth and Nathanial Comfort, emotionally and ideologically, rather than on the evidence.

And that's really why I started posting on this thread.

3
Coel Hellier 10 Aug 2019
In reply to TobyA:

> With my original concerns about your claim that genetics is the reason that most people are in prison, ...

Well let me quibble about that wording.  I am not saying that for most people in prison, "the" reason they are there is genetics.  It is never the case that "the" reason is genetics, it is always the case that there are multiple reasons involving a complex interplay of genetics and environment. 

What I have said is that genetics is the "biggest factor", which means that if you assess the variation in a whole population in being "sent to jail", or indeed in any other outcome largely dependent on psychological traits, then genetics would explain the largest portion of that variance, larger than any other single factor you could point to (such as "brought up in a household with a Dad with criminal convictions", or "brought up in poverty in a single-parent household", or "went to a school were drugs were prevalent" or whatever).

> ... you don't even seem to have thought about the problem of the difference between the reason why an individual might be in prison compared to the imprisonment rates of certain types of people.

Why do you need to assert that I've not "thought about" it? 

Of course there will be a lot of factors, including "non-shared environment". The whole point of the studies is that they do a good job of disentangling the importance of different factors. 

> Look at the difference between the prison system in Northern Ireland and mainland Britain, even now but clearly 20 years ago, to show how social factors can play major role in accounting for why someone might end up in prison.

No-one has said that such factors have zero effect!  Repeatedly, in this thread, I have emphasized that no-one is claiming that any trait is 100% genetic.  And, similarly, in his book and in his primer, Plomin repeatedly says that.  

It seems that sometimes critics can only deal with the possibilities "it is 100% environmental" and "it is 100% genetic" so reduce any claim to one of those! 

Yes, cultural issues, such as The Troubles have an influence. The point of the twin studies is that they tell us the relative important of genetic and environmental factors given the range of environments in the study.

That last bit really is crucial, as Plomin repeatedly emphasizes.

>  But people with totally different types of social backgrounds I don't see how you could ever make the claim search genetically inherited traits of them as individuals is the reason why they ended up in prison.

As above, no one thing is ever "the" one reason why anyone ended up in jail.  It is *always* a mixture of genetics and environment.  But the twin studies really can tell us the relative influence of those things averaged over the population and environment being studied. 

Of course the "non-shared environment", which is *always* found to be important in twin studies!, will involve a whole range of "reasons".

4
Duncan Bourne 10 Aug 2019
In reply to Coel Hellier:

Good reasoning here.

I have turned off likes/dislikes so now i have to post my approval/disapproval it is so much better

1
Coel Hellier 10 Aug 2019
In reply to Duncan Bourne:

> I have turned off likes/dislikes

Me too!  I think it's better.

elsewhere 10 Aug 2019
In reply to Coel Hellier:

Does this mean criminals are about 45% innocent victims of their genes?

Plus another percentage for innocent victim of their environment.

Coel Hellier 10 Aug 2019
In reply to elsewhere:

> Does this mean criminals are about 45% innocent victims of their genes? Plus another percentage for innocent victim of their environment.

Those two together would add up to 100%, since there is nothing else.  Which shows that (1) this question needs to be addressed regardless of the role of genes, and (2) the fact that we are products of prior causes does not negate guilt. 

So we invoke notions of innocence and guilt because it is socially useful to do so, because those societal attitudes form part of the environment and thus affect how we behave.

So, to answer the question, we are indeed "victims" of the combination of our genes plus our environment, but we're still guilty when we misbehave. 

elsewhere 11 Aug 2019
In reply to Coel Hellier:

> Those two together would add up to 100%, since there is nothing else.  Which shows that (1) this question needs to be addressed regardless of the role of genes, and (2) the fact that we are products of prior causes does not negate guilt. 

It does when it comes to some medical reasons. I am wondering why genes should  on the bad end of the mad or bad spectrum of diminished responsibility.

Coel Hellier 11 Aug 2019
In reply to elsewhere:

> It does when it comes to some medical reasons. I am wondering why genes should  on the bad end of the mad or bad spectrum of diminished responsibility.

I don't think we can accept that genetic influence amounts to diminished responsbility.  Your genes are not something apart from you, they are you, just as other aspects of you are you. 

Indeed, it is where someone has a genetic disposition to criminality that societal (environmental) interventions to ameliorate that disposition are most needed.

It is fallacious to conclude: "The fact that he is a violent rapist is largely because of his genes, therefore we won't hold him responsible and just let him continue his violent rape spree" -- society would still want to intervene.  Of course, finding out which interventions are appropriate and effective is something that should be discussed. 

Offwidth 11 Aug 2019
In reply to Coel Hellier:

The whole point of my argument is there IS something else. If you overestimate heritability based on false assumptions or bad methodology then genes + environment will have a combined probablity which is greater than 1.

MZ vs DZ relies on assumptions (notably the environment assumption).  The experiments are then statistical methodologies where problems can occur. Raw data is always subject to statistical error given limited numbers (tens of thousands sounds a lot but, like opinion polls, raw data will be subject to overall statistical error roughly of the order of a percentage point) . This method produces scatter plots where claiming much accuracy in the correlations seem generous, and differences between correlations combine errors.

I also rechecked some texts and they seem to say in utero effects can break the first assumption of the method.

A few more links:

The method... those familar with stats should have a look at the example scatter plots (and try not to roll their eyes). Plus an indication other methods show a lower correlation than twin results.

https://en.m.wikipedia.org/wiki/Twin_study

Assumptions and warnings for the method.

http://nuffieldbioethics.org/wp-content/uploads/2014/07/Genetics-and-behaviour-Chapter-4-Quantitative-genetics.pdf

...indicating  why base assumptions might be wrong for MZ and against making sweeping generalistaions from a particular study group.

Another critique of the methodology. 

https://www.madinamerica.com/2013/03/the-trouble-with-twin-studies/

elsewhere 11 Aug 2019
In reply to Coel Hellier:

To run a society, pretending we are responsible for our actions is essential. 

However it seems criminality is significantly an expression of the genes dealt by the lottery of nature and it seems illogical to say we are responsible for that lottery result.

Coel Hellier 11 Aug 2019
In reply to elsewhere:

> However it seems criminality is significantly an expression of the genes dealt by the lottery of nature ...

Yes.

> ... and it seems illogical to say we are responsible for that lottery result.

Yes, we are a lottery result.  But we are still "responsible" for what we do, since it is we who are doing it.    The fact that we didn't have any choice in being who we were does not change that. 

elsewhere 11 Aug 2019
In reply to Coel Hellier:

> Yes.

> Yes, we are a lottery result.  But we are still "responsible" for what we do, since it is we who are doing it.    The fact that we didn't have any choice in being who we were does not change that. 

That's not how it works for an adult with the mental capacity of a child due to the genetic lottery.

If criminality is significantly genetic why is the criminal responsible for that more than they are responsible for their heght?

Coel Hellier 11 Aug 2019
In reply to Offwidth:

> The whole point of my argument is there IS something else.

Other than genes and environment?  Such as?

Anyhow, the things you point to are known and accounted for by those in the field.  Such as:

> The experiments are then statistical methodologies where problems can occur. Raw data is always subject to statistical error given limited numbers ...

Well yes, this is why one uses statistics to analyse the results.  However, as has been emphasized, twin studies have been giving repeatedly replicated results for 30 years or so now.  They are not statistical flukes.

> those familar with stats should have a look at the example scatter plots (and try not to roll their eyes).

What's your problem with them?

> Plus an indication other methods show a lower correlation than twin results.

So what these other methods?  They are methods were one tries to link traits to individual genes (generally a set of set of SNiPs using GWAS).   Then you find something like: If you take the 20 top genes for trait X, then the sum of their influence adds up to less than the twin studies are saying.

But that is now well known and is entirely in line with what Plomin et al have been saying!   Genetic effects are **not** causes by small numbers of genes each having large effects; they are caused by large numbers of genes (thousands) each having small effects.   That's why tests involving a limited number of SNiPs will miss much of the genetic effect.

Again, I have stated this "large number of genes each having small effects" several times in the thread. You're not pointing to anything that rebuts anything that Plomin et al are saying.

> I also rechecked some texts and they seem to say in utero effects can break the first assumption of the method.

Go no then, give some actual cites and quotes from authoritative texts.

> Another critique of the methodology. 

Not by a scientist in the field.  

Post edited at 11:09
1
Coel Hellier 11 Aug 2019
In reply to elsewhere:

> If criminality is significantly genetic why is the criminal responsible for that more than they are responsible for their heght?

Because that's how we construe being "responsible".

If someone has a tendancy to get drunk and beat their wife, we still hold them responsible even if they have genes that make them prone to alcohol addiction and violence. 

But, yes, I do agree that understanding the role of genes will have implications for how many people think about these things.  And it is indeed true that the whole justice system operates on concepts from before anyone knew about genes, and hasn't adjusted the way it conceptualises such things.

That -- the fact that large swathes of society haven't yet assimilated important information about our natures that science has been telling us for 30 years now -- is why I'm posting on this thread. 

(And note, everyone, that "I don't like the implications of this, therefore the science is wrong", is not valid reasoning.)

1
RomTheBear 11 Aug 2019
In reply to Coel Hellier:

> Other than genes and environment?  Such as?

> Anyhow, the things you point to are known and accounted for by those in the field.  Such as:

> Well yes, this is why one uses statistics to analyse the results.  However, as has been emphasized, twin studies have been giving repeatedly replicated results for 30 years or so now.  They are not statistical flukes.

> What's your problem with them?

> So what these other methods?  They are methods were one tries to link traits to individual genes (generally a set of set of SNiPs using GWAS).   Then you find something like: If you take the 20 top genes for trait X, then the sum of their influence adds up to less than the twin studies are saying.

> But that is now well known and is entirely in line with what Plomin et al have been saying!   Genetic effects are **not** causes by small numbers of genes each having large effects; they are caused by large numbers of genes (thousands) each having small effects.   That's why tests involving a limited number of SNiPs will miss much of the genetic effect.

And this is the obvious, massive, flaw here. If genetics effects are caused by thousands of genes, which we can reasonably assume to be the case, then trying to make inferences about the influence of genes is pointless, because of the curse of dimensionality.

Hence why the whole field of behavioural genetics is completely shot from the start. What they are trying to do is mathematically impossible. Every time we use their models to predict individual human behaviour in a controlled environment, it fails. They give you a lot of BS reasons as to why it fails, but the proof is in the pudding: it doesn’t work. Unsurprisingly so.

Post edited at 19:02
2
RomTheBear 11 Aug 2019
In reply to elsewhere:

> If criminality is significantly genetic why is the criminal responsible for that more than they are responsible for their heght?

You’ve alluded to the central problem. Fraudster such as Plomin treat, statistically speaking, psychological traits the same way they would treat a variable like height. It’s fundamentally, fatally, flawed.

Post edited at 19:08
2
Coel Hellier 11 Aug 2019
In reply to RomTheBear:

> Hence why the whole field of behavioural genetics is completely shot from the start. What they are trying to do is mathematically impossible. Every time we use their models to predict individual human behaviour in a controlled environment, it fails.

Can you point to times when behavioural geneticists *have* claimed to predict individual human behaviour?   Because this smells suspiciously like a strawman. 

Yes, exactly, if psychological effects are influenced by thousands of genes, most of which we don't (yet) know about and have little idea about what they do, then we cannot predict individual human behaviour.   (But then no such claim has been made.)

Nothing about that negates the claims that are actually being made, that the combined effect of all of those genes adds up to about 50% of the variation in psychological traits. 

Coel Hellier 11 Aug 2019
In reply to RomTheBear:

> Fraudster such as Plomin treat, statistically speaking, psychological traits the same way they would treat a variable like height. It’s fundamentally, fatally, flawed.

In what way is it "fundamentally, fatally, flawed"?

Pefa 11 Aug 2019
In reply to Coel Hellier:

> This thread is getting long, so just to reiterate the main point:

> My original claim was that genes were the biggest factor in who gets sent to jail, and that "shared environment" (environment that siblings would share, such as being brought up in a home where crime as accepted and normal) was a smaller effect.

> From the paper that cb294 cited, twin studies show that the contributions to convictions for violent crime are:

> Genes: 45%

> Non-shared environment: 31%

> Shared environment: 24%.

> Numbers are somewhat different for other sorts of crime, but the overall heritability of criminal convictions is 45%. 

> So I was quite clearly right.

No you weren't, you need to define shared and non shared environment first. 

Coel Hellier 11 Aug 2019
In reply to Pefa:

> ... you need to define shared and non shared environment first.

They have been defined up-thread.  "Shared"  environment is defined as the environmental similarities that siblings would share.  "Non-shared" environment is environmental factors that siblings would not share. 

RomTheBear 11 Aug 2019
In reply to Coel Hellier:

> Can you point to times when behavioural geneticists *have* claimed to predict individual human behaviour?   Because this smells suspiciously like a strawman. 

I haven’t said they do make this claim. I am saying they can’t make it.

But it seems we agree, their theory doesn’t work in the real world. You can’t use it to do anything useful, or gain any insight. I can easily get similar results with randomly generated data.

It’s not science, it’s bad entertainment.

> Nothing about that negates the claims that are actually being made, that the combined effect of all of those genes adds up to about 50% of the variation in psychological traits. 

That’s another fundamental flaw. Correlation is not additive.

1
elsewhere 11 Aug 2019
In reply to Coel Hellier:

> Because that's how we construe being "responsible".

No it is not. There is in law (and understood generally) such a thing as diminished responsibility. 

Post edited at 20:44
RomTheBear 11 Aug 2019
In reply to Coel Hellier:

> In what way is it "fundamentally, fatally, flawed"?

Height belongs to a class of distribution with defined mean and variance, under which statistical method such as those used in those studies do work. That’s not the case of most human behaviour, or at least they cannot show it. 

Pefa 11 Aug 2019
In reply to Coel Hellier:

> They have been defined up-thread.  "Shared"  environment is defined as the environmental similarities that siblings would share.  "Non-shared" environment is environmental factors that siblings would not share. 

But that doesn't work as one kid can get traumatised by say child physical abuse and the other doesn't or one could get sexually abused and the other 2 don't or any miriad of combinations. I mean this study may seem fine on paper but real life is not paper. 

Post edited at 20:49
Coel Hellier 11 Aug 2019
In reply to RomTheBear:

> I haven’t said they do make this claim. I am saying they can’t make it.

But seeing they're not claiming that they can make this claim, that is not a defect of their theories.

> But it seems we agree, their theory doesn’t work in the real world.

It does work in the real world.  All theories have things that they can and cannot do.  An engineering account of how a lottery machine works could not predict the next lottery numbers.  But that doesn't mean we don't understand how lottery machines work.

> You can’t use it to do anything useful, or gain any insight.

Actually, properly understanding the role of genes in human behavioural traits both gives insight and is useful.

1
Coel Hellier 11 Aug 2019
In reply to Pefa:

> But that doesn't work as one kid can get traumatised by say child physical abuse ...

Which is why the studies are talking about averages over large samples.  We can indeed talk about genetic and environmental influences on average.

No-one is claiming that twin studies can then predict behaviour or outcomes of individuals, of course they can't.

1
Coel Hellier 11 Aug 2019
In reply to RomTheBear:

> Height belongs to a class of distribution with defined mean and variance, under which statistical method such as those used in those studies do work. That’s not the case of most human behaviour, or at least they cannot show it.

You've not shown that a trait such as "height" is any different from a trait such as "being prone to violence" in such regards, nor have you given any reason why they would be. 

2
Coel Hellier 11 Aug 2019
In reply to elsewhere:

> No it is not. There is in law (and understood generally) such a thing as diminished responsibility. 

But it is not generally accepted in law that: "I have genes X and Y, therefore I'm not responsible for the violent assault that I committed while drunk".  Nor is it generally accepted in law that: "I grew up in a violent household where I was regularly beaten, therefore I'm not responsible for the six burglaries that I committed". 

TobyA 11 Aug 2019
In reply to Coel Hellier:

I was listening to Radiolab while driving through France earlier this week and they interviewed a guy who was a physicist with a sideline in genetics, and his company IIRC can predict someone's height from their genome (within an inch I think it was).

It's here - https://www.wnycstudios.org/story/g-unnatural-selection

Looking at the transcript, the guy - Hsu - says 20,000 mutations on the genome affect height and has come up with AI systems that scan a genome and can predict height from that. He wants to do for intelligence too, but it seems hugely more complicated, partly because although over a million people have had their genome read, not many have IQ scores to go with the genomes. But they do seem to have data on levels of education and have used that instead of IQ. They come up with a predictive tool based on that, but it doesn't seem very accurate - the physicist doesn't really explain whether that's because the data they use to show intelligence doesn't really show intelligence, or that the genetics behind intelligence is just a lot more complicated than height? Maybe both I guess.

It was also interesting that the guy from the Social Science Genetic Association Consortium noted that the vast majority of the genome data available to study currently is from people of white European descent, so the predictive power of Hsu's tool falls even further (or just doesn't work he seems to say) if applied to people from other backgrounds.

TobyA 11 Aug 2019
In reply to Coel Hellier:

> Nor is it generally accepted in law that: "I grew up in a violent household where I was regularly beaten, therefore I'm not responsible for the six burglaries that I committed". 

That does seem to be the defence when victims of domestic violence use violence on their tormentors.

elsewhere 11 Aug 2019
In reply to Coel Hellier:

> But it is not generally accepted in law that: "I have genes X and Y, therefore I'm not responsible for the violent assault that I committed while drunk". 

Didn't you say something about the law not reflecting current science? I'm wondering how that new science might fit into diminished responsibility. 

Post edited at 23:19
Pefa 11 Aug 2019
In reply to Coel Hellier:

> Which is why the studies are talking about averages over large samples.  We can indeed talk about genetic and environmental influences on average.

Nah, how can you adjust for that when it's so random? 

> No-one is claiming that twin studies can then predict behaviour or outcomes of individuals, of course they can't.

Sorry I thought you were saying that genes are more important than environment. 

Post edited at 23:39
RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> You've not shown that a trait such as "height" is any different from a trait such as "being prone to violence" in such regards, nor have you given any reason why they would be. 

It is up to the social “scientist” to tell us why on earth violent behaviour, on any real world measure, against the most basic intuition, would be normally distributed like height, before they start unleashing their statistical software package on the data.

Post edited at 08:34
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> It is up to the social “scientist” to tell us why on earth violent behaviour, on any real world measure, would be normally distributed like height.

Maybe it is normally distributed, maybe it isn't.  So what? 

Coel Hellier 12 Aug 2019
In reply to Pefa:

> Nah, how can you adjust for that when it's so random? 

By averaging over it (an entirely normal technique in science).

> Sorry I thought you were saying that genes are more important than environment. 

A rule of thumb is that genes are roughly as important as all environmental factors added together. 

But, it's still the case that **twin** **studies** cannot then predict behaviour or outcomes of individuals; it's not what they are designed to do or what they try to do.  They don't even deal with individuals (accept to average over lots of people), they don't even look into the genes any particular individual has, so of course they don't predict behaviour or outcomes of individuals. You'd need different types of study for that. 

Coel Hellier 12 Aug 2019
In reply to elsewhere:

> Didn't you say something about the law not reflecting current science? I'm wondering how that new science might fit into diminished responsibility. 

For about 150 years now the mainstream scientific opinion has been that humans are the determined product of prior causes.   Whether that is genes or environment or the combination of both doesn't change that. 

And we can't move to a stance of "if there are prior causes that made you as you are, then you are not criminally responsible", since that applies to everyone. It would amount to abolishing the criminal justice system.  We need to hold people responsible for their acts, even though those acts then had prior causes; society cannot operate without doing that.

There is indeed a role for the legal concept of "diminished responsibility", but it needs to be about **temporary** departures from a person's normal behaviour.  It can't be about their long-term, settled personality.  

Coel Hellier 12 Aug 2019
In reply to TobyA:

> Hsu - says 20,000 mutations on the genome affect height and has come up with AI systems that scan a genome and can predict height from that. He wants to do for intelligence too, but it seems hugely more complicated,

Yes, this stuff is developing fast, as techniques for reading people's genomes get better and better.   It's early days as yet, but will soon become mainstream.

Society really needs to be aware of this and to discuss what to do about it.  If we do get to the stage where we can make good and reliable predictions as to how kids are likely to turn out, then we need to decide what sort of interventions to make.**

[**Just because I know some people will misunderstand this: a genes-based prediction of what is likely to happen is a prediction "if environmental factors continue as they are"; it does *not* say that one cannot change the future likelihoods by interventions. As Plomin repeatedly emphasizes in his book, such studies tell us what **is**, and thus what is likely to be the case if we continue much the same; they do not tell us what could be.]

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> By averaging over it (an entirely normal technique in science).

Which works very well under very defined circumstances often found in proper science, but rarely found in the messy world of human behaviour.

if I took two random people and their combined number of conviction for violent behaviour was ten, my best guess would be that one guy did the ten and the other did zero. A social scientist will tell you the best guess is that both did 5 because their model doesn’t work otherwise. It’s bullshit.

> But, it's still the case that **twin** **studies** cannot then predict behaviour or outcomes of individuals; it's not what they are designed to do or what they try to do.  They don't even deal with individuals (accept to average over lots of people), they don't even look into the genes any particular individual has, so of course they don't predict behaviour or outcomes of individuals. You'd need different types of study for that. 

In other word, they don’t work in the real world, it works only on paper, all they can do is make statements about the average without telling us whether the average is meaningless or not. And that’s only one of the problem.

2
RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> Maybe it is normally distributed, maybe it isn't.  So what? 

So what  ? Most of the statistical methods they use don’t work unless this condition is satisfied.

Post edited at 09:04
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> if I took two random people and their combined number of conviction for violent behaviour was ten, my best guess would be that one guy did the ten and the other did zero. A social scientist will tell you the best guess is that both did 5 because their model doesn’t work otherwise. It’s bullshit.

Go on then, quote a social scientist saying that.  Bet you can't.  Nice strawman!

> In other word, they don’t work in the real world, it works only on paper, all they can do is make statements about the average without telling us whether the average is meaningless or not. And that’s only one of the problem.

Your complaints are akin to saying:

"You guys can't predict whether it will rain in Manchester on July 20th in the year 2022 (but can only give a probability), so your models don't work in the real world, it's all bullshit, so climate change is nonsense".

1
RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> [**Just because I know some people will misunderstand this: a genes-based prediction of what is likely to happen is a prediction "if environmental factors continue as they are"; it does *not* say that one cannot change the future likelihoods by interventions. As Plomin repeatedly emphasizes in his book, such studies tell us what **is**, and thus what is likely to be the case if we continue much the same; they do not tell us what could be.]

At least Plomin seems to have the humility to say, at least quietly, that his stuff doesn’t actually work when applied to the real world.

1
elsewhere 12 Aug 2019
In reply to Coel Hellier:

> There is indeed a role for the legal concept of "diminished responsibility", but it needs to be about **temporary** departures from a person's normal behaviour.  It can't be about their long-term, settled personality.  

WHAT? You mean diminished responsibility cannot apply to somebody with a stable irreversible lack of mental capacity? What planet are you on? Do you even know that Down's syndrome does not go away? Do you even know that a brain injury can be permanent? Do you not know there might be no settled personality due to dementia but the lack of capacity is permanent?

You have very strange ideas the range of people in this world.

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> Go on then, quote a social scientist saying that.  Bet you can't.  Nice strawman!

Of course they don’t generally voice it as otherwise it would point to the ridiculousness of their methods. But thats what they would need to say for their model to be applicable.

> Your complaints are akin to saying:

> "You guys can't predict whether it will rain in Manchester on July 20th in the year 2022 (but can only give a probability), so your models don't work in the real world, it's all bullshit, so climate change is nonsense".

Absolutely not. You clearly haven’t understood the problem. It’s not a problem of accuracy.

Post edited at 09:23
1
RomTheBear 12 Aug 2019
In reply to elsewhere:

> WHAT? You mean diminished responsibility cannot apply to somebody with a stable irreversible lack of mental capacity? What planet are you on? Do you even know that Down's syndrome does not go away? Do you even know that a brain injury can be permanent? Do you not know there might be no settled personality due to dementia but the lack of capacity is permanent?

> You have very strange ideas the range of people in this world.

Human beings are much better at judging personal responsibility than social scientists with their bullshit models.

Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> At least Plomin seems to have the humility to say, at least quietly, that his stuff doesn’t actually work when applied to the real world.

Just as climate-change scientists have the humility to admit that, no, they cannot predict whether it will rain in Manchester on July 20th in the year 2022. 

Despite that, their models do actually work in the real world. 

5
Coel Hellier 12 Aug 2019
In reply to elsewhere:

> ... You mean diminished responsibility cannot apply to somebody with a stable irreversible lack of mental capacity? What planet are you on? Do you even know that Down's syndrome does not go away?

Such cases are rather a different issue from the one we were discussing, that is, a normally-functioning adult who is the product of prior causes (as we all are). 

2
elsewhere 12 Aug 2019
In reply to Coel Hellier:

> Such cases are rather a different issue from the one we were discussing, that is, a normally-functioning adult who is the product of prior causes (as we all are). 

So it is utter bollocks to say 'There is indeed a role for the legal concept of "diminished responsibility", but it needs to be about **temporary** departures from a person's normal behaviour.  It can't be about their long-term, settled personality.'

And you don't seem to have thought this through as criminality so obviously revolves around those who deviate and grossly deviate from legal/moral norms or normal functioning.

My question original remains and given that you appear to acknowledge some genetics (Down's syndrome) might justify diminished responsibility why would the uncertain legal/moral boundary of diminished responsibility not be shifted by developments in genetic science?

Coel Hellier 12 Aug 2019
In reply to elsewhere:

> And you don't seem to have thought this through as criminality so obviously revolves around those who deviate and grossly deviate from legal/moral norms or normal functioning.

Not really, criminality is a "normal" mode of human beheaviour.  It has been and is a feature of all human societies, and all societies adopt measures to reduce it.

> My question original remains and given that you appear to acknowledge some genetics (Down's syndrome) might justify diminished responsibility why would the uncertain legal/moral boundary of diminished responsibility not be shifted by developments in genetic science?

Because criminal "responsibility" is not really about prior causes leading to people's personalities and behavioural traits, it is basically about capacity to be deterred.  Where adults are capable of understanding: "This is against the rules, people don't want me to do it, and there is a punishment for it if I get caught, but I'm going to do it anyhow", then we hold them criminally responsible. 

That isn't changed by findings that our personalities are strongly affected by genetics.  Indeed, it is for those people with personality traits inclining them towards criminality that we need the strongest deterrence, and thus the concept of holding people criminally responsible. 

Post edited at 10:39
RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> Just as climate-change scientists have the humility to admit that, no, they cannot predict whether it will rain in Manchester on July 20th in the year 2022. 

> Despite that, their models do actually work in the real world. 

Yes, they do, I can verify experimentally, for example, things like the greenhouse effect. I can set up a controlled experiment, in the real world, and make useful predictions.

They make proper scientific claims which can be falsified.

1
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> They make proper scientific claims which can be falsified.

There are some things they can predict (such as global-average temperature) and some things they don't even claim to be able to predict (individual rain showers in individual locations years into the future).

Pointing to things that a model does not even claim to be able to predict, in order to dismiss the model overall, is a denialist tactic. 

elsewhere 12 Aug 2019
In reply to Coel Hellier:

It seems we currently have medical grounds for diminished responsibility but you have determined that the new science you laud should not inform medico legal thinking about the uncertain legal/moral boundary of diminished responsibility.

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> There are some things they can predict (such as global-average temperature) and some things they don't even claim to be able to predict (individual rain showers in individual locations years into the future).

> Pointing to things that a model does not even claim to be able to predict, in order to dismiss the model overall, is a denialist tactic. 

No, the point is, if your model cannot be tested and used in the real world, then it’s useless. I can generate a 100x100 matrix of random numbers and I am pretty sure to find significant correlation between a few variables and then claim I have a « «model ». It doesn’t mean I’ve discovered anything of any value whatsoever.

If your argument is that these people don’t claim that their model have actually have any bearing on reality, then fine, but then you have to look at it not as science but as a deranged form of entertainment (or rather, a convenient way to keep cushy academic jobs).

Post edited at 11:21
2
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> No, the point is, if your model cannot be tested and used in the real world, then it’s useless.

Yes, the models need to be tested.  And the models can be and are indeed tested. But that means testing what they are *claimed* to be able to do! 

Thus, if a climate-change model cannot predict individual rainshowers years into the future that does not mean it "can't be used in the real world and is useless"

But feel free to continue your denialist red-herrings and whataboutery.

Coel Hellier 12 Aug 2019
In reply to elsewhere:

> It seems we currently have medical grounds for diminished responsibility but you have determined that the new science you laud should not inform medico legal thinking about the uncertain legal/moral boundary of diminished responsibility.

Why would "determined by prior causes, where those causes are to a large extent genetic" be any different from "determined by prior causes, where those causes are to a large extent environmental"?

elsewhere 12 Aug 2019
In reply to Coel Hellier:

> Why would "determined by prior causes, where those causes are to a large extent genetic" be any different from "determined by prior causes, where those causes are to a large extent environmental"?

I see much in common as they are both part of the lottery of life. 

Why should new science not inform medico legal thinking?

Coel Hellier 12 Aug 2019
In reply to elsewhere:

> I see much in common as they are both part of the lottery of life. 

So since "determined by genes" is not that different, in this respect, from "determined by environment", why do you think it would require substantial re-thinking of the "legal/moral boundary of diminished responsibility".

As above, as I see it, that boundary is more about capacity to understand actions and their consequences, and thus capacity to be deterred.

elsewhere 12 Aug 2019
In reply to Coel Hellier:

> So since "determined by genes" is not that different, in this respect, from "determined by environment", why do you think it would require substantial re-thinking of the "legal/moral boundary of diminished responsibility".

Because something relevant may be discovered. If have no idea if the rethink would be substantial as that depends on the discovery. However given diminished responsibility is I assume established for some genetics (Down's syndrome) then I assume new science might result in the recognition of other genetic factors in medico legal thinking.

> As above, as I see it, that boundary is more about capacity to understand actions and their consequences, and thus capacity to be deterred.

So it is really just as you see it and you don't see how new science may change your thinking or medico legal thinking.

Post edited at 12:32
Coel Hellier 12 Aug 2019
In reply to elsewhere:

> So it is really just as you see it and you don't see how new science may change your thinking.

Well, new "unknown unknown" stuff might indeed, in the future, change my thinking.

But what I mean is that I don't see why the sort of now-known science expounded by Plomin etal -- that personality traits have strong genetic influence -- forces any major rethink of "criminal responsibility".

I do think that the science does have other implications that we need to think about, especially as things might develop over 20 years.

For example, do we do genetic screening of newborns?  Suppose such a screening would tell us that a newborn had genes giving him 50 times the usual propensity to end up in jail for violent assault.  Should we do the screening? And then what do we do? What sort of interventions would we then decide on?   Maybe one-to-one support from a carefully chosen mentor to attempt to ameliorate the propensity? 

Society will have to think about that question because, as in TobyA's comment, private firms will start offering this sort of screening whatever. 

Offwidth 12 Aug 2019
In reply to Coel Hellier:

If the twin studies are faulty because the assumptions are invalid, the main cause of being in jail might not be genes nor environment but the choice to commit a crime (followed by successful prosecution) or an unfair or faulty justice system. Genes or environment being correlated to a lesser extent with criminality might well be still be as close to irrelevant as they were previously thought to be before twin studies started.

Nature versus nurture is a fair enough question for something like height where choices are not involved but I'm still not convinced its even sensible for incarceration (despite the smaller correlations than you claim being mostly likely real).

The paper I linked above and I've copied again below gives the best discussion of the arguments involved between the two groups from the critics side that I've been able to find and seems to be fair about the failure of the likes of Plomin to answer the questions from these critics of their methodology (I can't find anything: they like you attack red herrings instead). You dismissing such critisism by scientists saying they are 'not in the field' is just childishly ad hominem, avoiding the point, again something shameful for a scientist who should deal with the arguments, not dismiss the man.

https://www.madinamerica.com/2013/03/the-trouble-with-twin-studies/

4
Shani 12 Aug 2019

In reply:

On economics and crime: 

https://www.econlib.org/library/Enc1/Crime.html

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> Yes, the models need to be tested.  And the models can be and are indeed tested. But that means testing what they are *claimed* to be able to do! 

You don’t test a theory by re-running a flawed statistical analysis on a similar dataset until you get a similar result. It’s not science, it’s post-hoc storytelling.

The way to test is to take your model, make predictions from it, and then test it against the real world.

1
RomTheBear 12 Aug 2019
In reply to Offwidth:

Furthermore, even if you were to make the hypothesis that the assumptions and the method are not flawed (and they are flawed) just look at the outputs of these studies.

The correlations are typically so weak ! These people don’t seem understand that the interpretation of correlation is non-linear.

1
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> The correlations are typically so weak !

Where the correlations are strong they are strong; where they are weak they are weak.  This is why one uses statistics.    There are plenty of strong correlations in this field.

> These people don’t seem understand that the interpretation of correlation is non-linear.

More substance-free derision, so typical of denialists.

Coel Hellier 12 Aug 2019
In reply to Offwidth:

> ... the main cause of being in jail might not be genes nor environment but the choice to commit a crime

Those things are not either/or.  Any "choice to commit a crime" has prior causes, and those are genetic or environmental. 

What else do you think there is?  Randomness?  Dualistic souls?

> Nature versus nurture is a fair enough question for something like height where choices are not involved but I'm still not convinced its even sensible for incarceration

Why not?  Are you really suggesting that a dualistic soul is doing the choosing, unrelated to any material/chemical/neuroloigical process?

1
RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> Where the correlations are strong they are strong; where they are weak they are weak.  This is why one uses statistics.    There are plenty of strong correlations in this field.

Are there, really ? On significant findings ? Most of the time, the numbers are either completely weak (30/40 %, most likely just noise) or moderate (60/70 %, typically  inconclusive).

> More substance-free derision, so typical of denialists.

I can easily prove it to you mathematically, but never mind, keep crying.

But it’s typical of the field when they are faced with any criticism on their methods, which would get a fail on a first year student probability exam. Instead of addressing the criticism, they just accuse the others of being politically motivated, crazy, or deniers.

I do not in fact object at all to the idea that genetics has significant effects on human behaviour.

I’m just pointing out that these people are fraudsters, typically academic who get paid proportionally to the popularity of their papers rather than on the success of their theory when applied in the real world, and the methods they use are a fucking joke and an insult to intelligence.

Post edited at 14:46
3
Coel Hellier 12 Aug 2019
In reply to Offwidth:

> The paper I linked above and I've copied again below gives the best discussion of the arguments involved ...

However, the central claim made by that article, that maybe parents treat identical twins more similarly than fraternal twins, is not a problem for twin studies where the twins are separated at birth and reared apart.

Now yes, such studies have fewer twins than the studies where the twins are not separated, but there is still enough of them. And they show the same thing, which tells us that the issue you point to is not a strong confounder (it might be a weak one).  Indeed, all the different types of studies (not just twin studies but adoption studies and other studies) give consistent and replicated results.

And, further, the assumptions being made can indeed be tested. E.g.:

"Bivariate and trivariate models include several instances in which the EEA [equal environment assumption] can be tested. [...] The number of twin pairs that is needed is no greater than the number typically available in most twin registries. The analysis of spatial ability and aggression [the traits in this study] indicated no detectable violation of the EEA."

https://www.ncbi.nlm.nih.gov/pubmed/16790150

> You dismissing such critisism by scientists saying they are 'not in the field' is just childishly ad hominem, avoiding the point, again something shameful for a scientist who should deal with the arguments, not dismiss the man.

Ho, ho, ho, that's rich, coming from you!

(And yes, he's not in the field and not really a scientist, he's a medical practitioner.)

Post edited at 14:40
1
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> Are there, really ? On significant findings ? Most of the time, the numbers are either completely weak (30/40 %, most likely just noise) or moderate (60/70 %, typically  inconclusive).

You are confusing two distinct things:

(1) Effect size:  What fraction of the variance is explained by a given factor.

(2) Statistical significance: How likely a finding is to be just noise. 

What you seem to be doing is pointing to effect-size numbers as though they were statistical significances.  They're not; with a large sample one can have a high statistical significance for a low effect size.   And note that for most of the findings that Plomin et al point to the effect sizes are not small, and nor are the statistical significances.

> I can easily prove it to you mathematically, ...

Go no then.

cb294 12 Aug 2019
In reply to RomTheBear:

I side with you on most threads, but here you really seem to have no clue about study design and statistics. Stop digging seems to be the traditional recommendation.

CB

1
what the hex 12 Aug 2019
In reply to Pete Pozman:

Slight deviation from the OP and maybe something of a paranoid thought but could the bulking up of the police/justice system be a preemptive move to protect the ruling class from the rest of the population in the event of a no deal brexit?

I wouldn't normally stoop to speculation like this but everything Johnson does seems to be for the benefit of Johnson and I don't see why bulking up the police and prisons is any different?

RomTheBear 12 Aug 2019
In reply to cb294:

> I side with you on most threads, but here you really seem to have no clue about study design and statistics. Stop digging seems to be the traditional recommendation.

> CB

What have I said that is wrong ?

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> You are confusing two distinct things:

> (1) Effect size:  What fraction of the variance is explained by a given factor.

> (2) Statistical significance: How likely a finding is to be just noise. 

> What you seem to be doing is pointing to effect-size numbers as though they were statistical significances. 

I know very well they are not, but the chance of a correlation being the result of noise increases very fast as correlation decreases. So as a rule of thumb, in presence of problems where statistical significance tests are more harmful than helpful, the rule of thumb should be to be extra suspicious of low correlations.

I suggest you simply generate a 100x100 matrix of randomly generated numbers, and look at the number of variable that are correlated over a threshold, keep decreasing that threshold and take a count, and plot it.

> They're not; with a large sample one can have a high statistical significance for a low effect size.   And note that for most of the findings that Plomin et al point to the effect sizes are not small, and nor are the statistical significances.

Statistical significance is not deterministic, it’s stochastic, the overuse of statistical significance tests in the social science literature is another thing that make the bullshit detector bells ringing.

Instead of doing rubbish hypothesis testing incorrectly without knowing to interpret them, they’d be better of going back to the basics of the scientific method, of testing your prediction in controlled experiments.

When you’re hunting for spurious correlation, you’re guaranteed to find some, and that’s what these guys are doing.

> Go no then.

Well a bit of information theory will tell you that, in the ideal case of X and Y being a joint bivariate normal distribution (the kind of ideal case these type of papers typically assume without proof) the mutual information between the two is 

-(1/2) * log (1-p^2)

So the information gained on Y by studying X is about 3.8 times bigger if p=0.99 than if p=0.80, 8.8 times bigger than if p=0.60 etc etc...

As you can see, even in this perfect case, the interpretation of correlation is not linear.

Post edited at 17:01
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> I know very well they are not, but the chance of a correlation being the result of noise increases very fast as correlation decreases. 

Agreed, though that can be countered by increasing the sample size. Anhow, the correlations (effect sizes) of the stuff we're talking about are not small.

> So as a rule of thumb, in presence of problems where statistical significance tests are more harmful than helpful, ...

??? "... where statistical significance tests are more harmful than helpful ..."??

> I suggest you simply generate a 100x100 matrix of randomly generated numbers, and look at the number of variable that are correlated under a threshold, keep decreasing that threshold and take a count, and plot it.

This is why one does not take just the correlation coefficient (effect size), but also considers the statistic significance of it.

> Statistical significance is not deterministic, it’s stochastic, the overuse of statistical significance tests in the social science literature is another thing that make the bullshit detector bells ringing.

The **wrong** use of statistical significance tests is indeed, I grant you, a problem in swathes of the social sciences.  Again, we're talking here about solid, consistent and replicated results.

> in the ideal case of X and Y being a joint bivariate normal distribution (the kind of ideal case these type of papers typically assume without proof)

No, they don't assume that.  Really they don't.

> So the information gained on Y by studying X is about 3.8 times bigger if p=0.99 than if p=0.80, 8.8 times bigger than if p=0.60 etc

Are those supposed to be correlation coefficients? They seem to be, but you've labelled them "p" as though they were probabilities.   Again, you seem to be confusing effect sizes with statistical significance probabilities. 

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> Agreed, though that can be countered by increasing the sample size. Anhow, the correlations (effect sizes) of the stuff we're talking about are not small.

It can be countered with sample size under specific circumstance. In the case of distribution that have high kurtosis, and with undefined mean and variance, such as is likely (or at least possible) to be the case for many human behaviours, the sample sizes you need to counter it are often bigger than the population of the planet.

> ??? "... where statistical significance tests are more harmful than helpful ..."??

Yes, more harmful than helpful, in most cases, do bad in fact that psychology newspapers are starting to ban their use.

> The **wrong** use of statistical significance tests is indeed, I grant you, a problem in swathes of the social sciences.  Again, we're talking here about solid, consistent and replicated results.

As I’ve said, running the same flawed methodology on a similar dataset doesn’t make it better. Moreover, the social science studies that are heavily replicated (there aren’t many) tend to be based on circular arguments, they basically leak the conclusion in the assumptions.

> No, they don't assume that.  Really they don't.

Well if they don’t then they shouldn’t be using the types of method they are using. You can pretty much ditch linear regression and correlation for a start if they don’t, that’s about 90% of the papers going to the bin right there if you don’t make these assumptions. Other problems I haven’t mentioned, correlation is not additive, so it requires strictly no sub-sampling. So already there is a big flaw using it for twin studies. Moreover, it’s not transitive, here you go, you can’t use it for heritability studies. The list of major mistakes goes on...

There are other methods that don’t have these problem, but you don’t see them using them. Either because they don’t know, or more likely, they can’t produce anything interesting without using the rubbish methods.

Noboby gets credit and grants, and nice academic jobs for publishing studies that find nothing.

> Are those supposed to be correlation coefficients? They seem to be, but you've labelled them "p" as though they were probabilities.   Again, you seem to be confusing effect sizes with statistical significance probabilities. 

Yes, I don’t know how to type the Greek letter “rho” on my keyboard so I used p. It’s the correlation coefficient. Not a probability.

The argument had nothing to do with statistical significance, it was simply to show you why the interpretation of correlation is not linear.

Post edited at 18:07
1
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> It can be countered with sample size under specific circumstance. In the case of distribution that have high kurtosis, and with undefined mean and variance, such as is likely (or at least possible) to be the case for many human behaviours, the sample sizes you need to counter it are often bigger than the population of the planet.

I don't see why you're asserting that we don't know the distributions.   Generally, we do. (The issue is then explaining them.)

So, for a set of twins, we can plot height of one twin against height of the other twin.  We then know the properties (including mean and variance) of the dataset that we're trying to explain.

And it works just as straightforwardly with any other traits that we can plot. For example, units of alcohol consumed per week, or self-described location on a political spectrum. 

If your compliant is that any such dataset won't properly sample the extremes of human behaviour, then ok it won't, but then that's not what the studies are about, they're trying to understand normal and typical humans, and sample sizes of a few thousands are fine to sample normal human behaviour.

Pefa 12 Aug 2019
In reply to Pete Pozman:

So psychological outcomes depend equally on environment and genes.I'm skeptical that genes play as big a part myself and see other experts take apart Plomin's work.  

Basically if you raise kids in a horrible way or in a horrible culture that is what you will get but does that mean it will be passed down if they are not genetically predisposed to that? Well yes if the traumatisation/environment continues so genes don't even have a say. 

This measure of violence and crime is so subjective though as in many countries attacking other countries is not seen as a crime when it is. And now we have some utter crackpot writing in one of the Tory papers yesterday saying perhaps we should give the idea of war a chance. Wtf is going on with these genetically/environmentally predisposed ruling class psychopaths ? 

Post edited at 19:03
Coel Hellier 12 Aug 2019
In reply to Pefa:

> And now we have some utter crackpot writing in one of the Tory papers yesterday saying perhaps we should give the idea of war a chance.

It's called hyperbole.  He wasn't serious!

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> I don't see why you're asserting that we don't know the distributions.   Generally, we do. (The issue is then explaining them.)

No, actually, in most cases, you don’t know them. All you can do is estimate them. But you rarely even see this basic step being done. 

> So, for a set of twins, we can plot height of one twin against height of the other twin.  We then know the properties (including mean and variance) of the dataset that we're trying to explain.

Yes, as I’ve said it works beautifully well with height. 

It’s pretty obvious that human behaviour, in most cases, is not the same.

> And it works just as straightforwardly with any other traits that we can plot. For example, units of alcohol consumed per week, or self-described location on a political spectrum. 

Good example, alcohol consumed, an example of human behaviour. The method you use for heights mostly won’t work straight away. If you look at alcohol consumed per week, it doesn’t follow a normal distribution like height, it looks more like a Weibull or a gamma.

That has consequences when comparing group average and on sample size. In that particular case you can actually do something with it but you’re not always so lucky.

Add some kurtosis and quickly there is little you can do in terms of correlation and regression.

> If your compliant is that any such dataset won't properly sample the extremes of human behaviour, then ok it won't, but then that's not what the studies are about, they're trying to understand normal and typical humans, and sample sizes of a few thousands are fine to sample normal human behaviour.

The problem is that in many cases involving social systems, most of the variance is due to exceptions.

Its pretty obvious if I take, for example, cases of violence. If I take a random group of hundred people, and their combined number of convictions for violent behaviour is say, ten, chances are, it’s due to only one or two individuals in the group.

This has SEVERE statistical consequences.

Now, the “trick” pseudoscientists use, it to design pseudo metrics that DELIBERATELY follow easy to use distributions so that they can run their model. The problem is that you end up measuring things that are completely bollocks and have no relationship with the real world interpretation, and it all becomes circular. If you massage the metrics you use until you find statistically significant results, it is not science, I’m sorry.

Physicists, or other real scientists, don’t come up with a new custom definition of distance or speed that has nothing to do with reality in order to make their theory fit.

To make it clear, I have nothing against people telling us stories with data, and some of these stories are compelling, what’s unacceptable is to use unrigorous methods, and sell it to us with the authority of science to push their agenda on us. It’s not science, it’s storytelling.

People like Plomin (and as far as they go, he is not the worse by far) are now advocating rewarding and punishing people based on their performance relative to their genetic ability. That’s the type of BS the absurd logical continuation of their argument takes you to.

At some point we have to realise these people are not scientist and not even experts. Their models perform worse at predicting human behaviour than my grandmother or taxi driver’s intuition. No wonder these people are mostly employed in academia and government. 

2
Stichtplate 12 Aug 2019
In reply to RomTheBear:

> At some point we have to realise these people are not scientist and not even experts. 

Quite. Who's this Plomin bloke anyway?

Robert J. Plomin, FBA (born 1948) is an American psychologist and geneticist best known for his work in twin studies and behavior genetics. A Review of General Psychology survey, published in 2002, ranked Plomin as the 71st most cited psychologist of the 20th century. Plomin earned a B.A. in psychology from DePaul University in 1970 and a Ph.D. in psychology in 1974 from the University of Texas at Austin under personality psychologist Arnold Buss. He then worked at the Institute for Behavioral Genetics at the University of Colorado at Boulder. From 1986 until 1994 he worked at Pennsylvania State University, studying elderly twins reared apart and twins reared together to study aging and is currently at the Institute of Psychiatry (King's College London). He has been president of the Behavior Genetics Association. In 2002, the Behavior Genetics Association awarded him the Dobzhansky Memorial Award for a Lifetime of Outstanding Scholarship in Behavior Genetics. He was awarded the William James Fellow Award by the Association for Psychological Science in 2004 and the 2011 Lifetime Achievement Award of the International Society for Intelligence Research. In 2017, Plomin received the APA Award for Distinguished Scientific Contributions. Plomin was ranked among the 100 most eminent psychologists in the history of science. In 2005, he was elected a Fellow of the British Academy (FBA), the United Kingdom's national academy for humanities and social sciences.

Yeah, why should we listen to him? He obviously knows nowt.... and what exactly are your credentials in this field Rom?

2
Coel Hellier 12 Aug 2019
In reply to RomTheBear:

> Good example, alcohol consumed, an example of human behaviour. The method you use for heights mostly won’t work straight away. If you look at alcohol consumed per week, it doesn’t follow a normal distribution ...

Nothing prevents you from calculating a correlation coefficient, regardless of whether the trait has a normal distribution.   There is a simple formula for calculating the "r" coefficient of a sample that doesn't make any such assumption about the distribution.

> People like Plomin (and as far as they go, he is not the worse by far) are now advocating rewarding and punishing people based on their performance relative to their genetic ability.

Citation to any such quote from Plomin?

> At some point we have to realise these people are not scientist and not even experts.

Another anonymous commenter of unknown background sneering at established and reputable scientists. Not convincing. 

2
FactorXXX 12 Aug 2019
In reply to Stichtplate:

> Yeah, why should we listen to him? He obviously knows nowt.... and what exactly are your credentials in this field Rom?

I think Rom's main credential is that he sees Coel as the enemy and will argue that black is white in an attempt to get one over on him. 

RomTheBear 12 Aug 2019
In reply to Coel Hellier:

> Nothing prevents you from calculating a correlation coefficient, regardless of whether the trait has a normal distribution.   There is a simple formula for calculating the "r" coefficient of a sample that doesn't make any such assumption about the distribution.

Yes, of course, nothing prevents you from calculating the sample correlation, its very easy, it’s just that it’s pretty useless if you cannot use it to make statistical inferences.

> Another anonymous commenter of unknown background sneering at established and reputable scientists.

They might be established, and reputable, but they are not scientists.

RomTheBear 12 Aug 2019
In reply to Stichtplate:

> Quite. Who's this Plomin bloke anyway?

> Robert J. Plomin, FBA (born 1948) is an American psychologist and geneticist best known for his work in twin studies and behavior genetics. A Review of General Psychology survey, published in 2002, ranked Plomin as the 71st most cited psychologist of the 20th century. Plomin earned a B.A. in psychology from DePaul University in 1970 and a Ph.D. in psychology in 1974 from the University of Texas at Austin under personality psychologist Arnold Buss. He then worked at the Institute for Behavioral Genetics at the University of Colorado at Boulder. From 1986 until 1994 he worked at Pennsylvania State University, studying elderly twins reared apart and twins reared together to study aging and is currently at the Institute of Psychiatry (King's College London). He has been president of the Behavior Genetics Association. In 2002, the Behavior Genetics Association awarded him the Dobzhansky Memorial Award for a Lifetime of Outstanding Scholarship in Behavior Genetics. He was awarded the William James Fellow Award by the Association for Psychological Science in 2004 and the 2011 Lifetime Achievement Award of the International Society for Intelligence Research. In 2017, Plomin received the APA Award for Distinguished Scientific Contributions. Plomin was ranked among the 100 most eminent psychologists in the history of science. In 2005, he was elected a Fellow of the British Academy (FBA), the United Kingdom's national academy for humanities and social sciences.

> Yeah, why should we listen to him? He obviously knows nowt.... and what exactly are your credentials in this field Rom?

>

Well, that’s a perfect example of the problem with the field. They just award medals and honours between themselves by the truckload, to make up for the lack of successful applications of their theory in the real world. These people are not checked by reality.

My plumber is an expert, if I pay him to fix a pipe, he can do it, and I can check the result. If he can’t do that, he most likely goes out of business at some point. His success depends on his actual competence, not in his street cred with other plumbers.

For social scientists, and other academia careerists, it’s the reverse, their success depends not on their competence, but on the popularity of their pet theories with their peers.

Post edited at 23:05
Pefa 13 Aug 2019
In reply to Coel Hellier:

> It's called hyperbole.  He wasn't serious!

“Even as satire, this is offensive and extremely not funny to those who have experienced war. Sorry that my sense of humour cannot be stretched to encompass the death toll in Iraq,”tweeted historian Moudhy Al-Rashid at the end of a series of scathing (and not-safe-for-work) comments.

https://www.rt.com/uk/466352-rod-liddle-times-war/

Tell the people of Iraq,Syria or Donbas how frightfully satirical it is or tell the starving kids in Yemen. 

Post edited at 08:49
RomTheBear 13 Aug 2019
In reply to FactorXXX:

> I think Rom's main credential is that he sees Coel as the enemy and will argue that black is white in an attempt to get one over on him. 

No, I am just shouting fraud when I see it.

Note that I don’t dont disagree that genetics probably play an important part in behaviour, I just disagree with the pseudoscientific methods used to try to prove it.

Post edited at 08:46
Coel Hellier 13 Aug 2019
In reply to RomTheBear:

> Yes, of course, nothing prevents you from calculating the sample correlation, its very easy, it’s just that it’s pretty useless if you cannot use it to make statistical inferences.

You can then use it to calculate the difference in the "r" correlation coefficient for DZ twins versus MZ twins, and hence discover how much of the correlation derives from genetic influences. 

Nothing about that requires any assumption of a normal distribution.

1
Coel Hellier 13 Aug 2019
In reply to Pefa:

> Tell the people of Iraq,Syria or Donbas how frightfully satirical it is ...

Faux outrage.  Everyone sensible could tell it was a hyperbolic and not-serious suggestion, and would not be upset by it. I enjoyed the writing. 

Post edited at 09:37
3
RomTheBear 13 Aug 2019
In reply to Coel Hellier:

> You can then use it to calculate the difference in the "r" correlation coefficient for DZ twins versus MZ twins, and hence discover how much of the correlation derives from genetic influences. 

> Nothing about that requires any assumption of a normal distribution.

It doesn’t require normality, but it requires the distribution to have defined moments. And even if you have moments, there are consequences on the significance on r depending on the distribution.

And many very common distribution have undefined moments, or need very large sample size before their estimation become stable.

I suggest you revisit the definition of correlation. It’s defined in terms of moments. Unless you know something about math I don’t, I don’t see how you can make any sense of the sample correlation unless your distribution has defined moments.

of course you’ll always be able to calculate the sample correlation coefficient, but the goal of doing this is to have a good approximation of the population correlation coefficient,  so that we can make inferences. 

Post edited at 09:57
Coel Hellier 13 Aug 2019
In reply to RomTheBear:

> It doesn’t require normality, but it requires the distribution to have defined moments.

You can measure the moments in the data you have -- it is straightforward to calculate the mean, r correlation coefficient, standard deviation, etc, from your sample.   

And again, all one is doing is comparing the DZ sample to the MZ sample.   Most issues with the distribution don't matter, if they're similar for the two samples that one is comparing.     The fact that the MZs act as a control for the DZs makes the technique pretty clean and powerful.

All your bluster really is red herrings, irrelevant stuff that doesn't stop the technique giving reliable and replicated results.

If the complaint is that the sample might not properly sample the extremes of the real-world population, then ok, yes, it won't.  But that's not what the studies are interested in.   The studies are aimed at normal human nature and normal human behaviour as manifest in samples of several thousand people.   And, by definition, the samples are adequate to sample that. 

Stichtplate 13 Aug 2019
In reply to RomTheBear:

> Well, that’s a perfect example of the problem with the field. They just award medals and honours between themselves by the truckload, to make up for the lack of successful applications of their theory in the real world. These people are not checked by reality.

If real world application constitutes your criteria for scientific credibility then do you also regard Stephen Hawking as a 'fake, charlatan or fraud'?  

> My plumber is an expert, if I pay him to fix a pipe, he can do it, and I can check the result. If he can’t do that, he most likely goes out of business at some point. His success depends on his actual competence, not in his street cred with other plumbers.

> For social scientists, and other academia careerists, it’s the reverse, their success depends not on their competence, but on the popularity of their pet theories with their peers.

By 'street cred with other plumbers' and 'popularity of their pet theories with their peers' I take it you mean the process of external peer review which has been the foundation for establishing scientific credibility for almost 100 years.

In this, as in so many areas, I'm sure your judgment is superior to experts in their field of many decades standing, so what do you propose as a method of establishing scientific credibility? A show of hands from a collection of internet randomers? 

I don't take issue with your assessment of Plomin's scientific methods (I'm not interested enough to investigate). What I do take issue with is how often you set yourself up as an expert in so many areas and how readily you declare established experts as idiots and frauds. You don't really discuss complex issues so much as tell us all how you're right and everyone else is wrong (with very few links backing you up) all the while frequently decrying others for a lack of humility!

RomTheBear 13 Aug 2019
In reply to Coel Hellier:

> You can measure the moments in the data you have -- it is straightforward to calculate the mean, r correlation coefficient, standard deviation, etc, from your sample.   

Yes, you can always calculate them, the question whether they adequately describe the population.

> And again, all one is doing is comparing the DZ sample to the MZ sample.   Most issues with the distribution don't matter, if they're similar for the two samples that one is comparing.     The fact that the MZs act as a control for the DZs makes the technique pretty clean and powerful.

Wrong, simply having the two samples following the same distribution is not enough.

As I have explained, they need to have moments, and the sample size required to estimate them depends on those.

> All your bluster really is red herrings, irrelevant stuff that doesn't stop the technique giving reliable and replicated results.

Well that’s typical of social scientist. They consider mathematical rigor a red-herring.

> If the complaint is that the sample might not properly sample the extremes of the real-world population, then ok, yes, it won't.

That’s not the central problem although it is one. If I look at criminal behaviour, for example, most of the variance will be driven by the exceptions. Just looking at the real world around me, so much of it is driven by the exception. A few individuals are extremely wealthy, a few authors write most of the books that are sold, a few people commit all the crimes, etc etc.

> But that's not what the studies are interested in.   The studies are aimed at normal human nature and normal human behaviour as manifest in samples of several thousand people.   And, by definition, the samples are adequate to sample that. 

Maybe, they are, maybe they aren’t, first you have to show that the sample are indeed adequate, which they generally don’t, or do under assumptions that are not verified.

That is, unless your objective is to produce meaningless numbers that don’t explain the real world. That seems to be Plomin’s defence, he always points out that indeed they just describe “what is”. But that’s a poor defence, if the only thing they do is to produce number from which I cannot actually draw inferences or explanations of the real world, you have to question the usefulness of doing this.

cb294 13 Aug 2019
In reply to RomTheBear:

My apologies, from the subsequent discussion I see that you do indeed understand the statistics (even if I disagree with your characterization of biological statistics, in particular with respect to genetics, as fraud)!

CB

Post edited at 10:48
Coel Hellier 13 Aug 2019
In reply to RomTheBear:

> Yes, you can always calculate them, the question whether they adequately describe the population.

I've already addressed this.  Yes, a sample of a few thousand people will not sample the extremes of the human population.  Thus, the sample is unlikely to contain someone as good at football as Lionel Messi.

So what? The studies are about the normal human behaviour adequately sampled by the sample. This limitation is accepted.  But the normal human experience is what the studies are interested in.

> Wrong, simply having the two samples following the same distribution is not enough.

Yes it is. 

> As I have explained, they need to have moments, ...

As stated, you can calculate what you need from the data.  (See 1st paragraph if you're going to make that complaint yet again.)

> Just looking at the real world around me, so much of it is driven by the exception. A few individuals are extremely wealthy, a few authors write most of the books that are sold, a few people commit all the crimes, etc etc.

So what?   The studies can still ask: if one twin is rich, is the other likely to be rich?  Is that more so if the twins are MZ than DZ?  Et cetera.   The method still works fine. 

But, again, if your complaint is that the extremes of behaviour won't be in the sample then ok, they won't. But, as stated, the studies are trying to get at normal humans.  If something is true within samples of a few thousand humans then that is interesting and important and relevant to most of us. 

OK, so it doesn't capture the footballing ability of Messi or the wealth of Bill Gates. So what?  That's a pretty minor and irrelevant complaint.

1
cb294 13 Aug 2019
In reply to RomTheBear:

> If I look at criminal behaviour, for example, most of the variance will be driven by the exceptions. Just looking at the real world around me, so much of it is driven by the exception.

I think that here it is you making the intuitive but unfounded assumption: Any influence of genetics on the likelihood of criminal behaviour is bound to be driven by more general psychological traits as discussed above (e.g. lack of empathy, exaggerated narcissism, psychosis, etc.).

AFAIK, it has be shown solidly that for all (most?) psychological pathologies related to these traits manifestation of clinical disease simply reflects the extremes of the distribution of the trait in the general population as expected under a multilocus genetic model: There simply is no single gene for e.g. intelligence, depression or schizophrenia*.

Thus, the bulk of the variation in the likelihood of engaging in criminal behaviour cannot be solely or even largely attributed to the extremes in the genetic distribution**. 

CB

* The single exception is severe intellectual disability (i.e., it is reasonably easy for a single mutation or an unlucky combination of a few mutations to mess up brain development). 

** as I have posted above multiple times, the deficiencies of our judicial systems mean that our prisons are full of people who would be better off (and society would be better off) with treatment for their underlying psychological problems. In this sense the genetic extremes do indeed contribute, but the bulk of the variance will still come from closer to the centre.

1

This topic has been archived, and won't accept reply postings.