Is homosexuality genetic?

[arcticpenguin]

But do you really think he's going to come back with nothing, or something completely laughable?

I'm interested in what he's going to come back with (man, where is this faith coming from?). I can wait.
Adam

If/when that happens, I can deal with it, and if he's got legit evidence I will acknowledge that. But claiming "I have proof but can't post it now" is behaviour indistinguishable from several forum contributors whom I do not respect and will not name.

Cheers,

 

[spoonhandler]

BillHoyt - although there's still debate about the mechanism, Stephens and Fillmore have a pretty good theory involving FGF-2, VEGF and IGF-I, but more importantly, the downstream element, alpha v beta 3 integrin, a cell survival signal for newly formed endothelial cells. I was looking into the possibility that thalidomide might be a good anti-angiogenic agent to use in renal cell carcinoma and found their paper very interesting. Some of my lines of inquiry followed their theory and confirmed the possible importance of beta 3 integrin in angiogenesis and/or tumour growth.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=10799645&dopt=Abstract

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=10661908&dopt=Abstract


And no, thalidomide is probably not effective as a single agent against renal cell carcinoma and could in fact very bad indeed. Bugger. Yes, I published this negative result.

 

[Ace_of_Sevens]

On handness:

From the article linked to above, it does seem to be genetically linked, but it's fairly common for identical twins to have different handedness, so it's not straight-forward.

 

[corplinx]

What I want to know also is if front-loaded studies are causing meme propagation. Most people have heard about "some study saying homosexuality is genetic". Are studies that "look for a genetic cause of homosexuality" that get reported in the press causing these memes?

I say this because I was perturbed by an article I read in a paper a few weeks back where critic reviewing the "Rawhide Kid" comic said that homosexuality was genetic and that was that. He went on to say that if "god made them that way" that it was "for a reason" and that we should accept it.

Of course, he can probably show me about as much evidence for genetic homosexuality as he can for proof of a deity. I just wonder why so many people believe there is proof of this genetic link out there.

 

[Ace_of_Sevens]

I'd also argue that justr because somethign is genetic doesn't mean it's okay in a moral sense. IIRC, genetic links have been shown for alcoholism and violent tendencies as well.

Not saying that homosexuality is wrong, just that this is a seperate issue from its cause.

 

[BillyJoe]

[Handedness] does seem to be genetically linked, but it's fairly common for identical twins to have different handedness, so it's not straight-forward.

Perhaps they're identical mirror images?

 

[Skeptical Greg]

Thalidomide's effect is not a result of genetic damage. Thalidomide suppresses functional pathways of proteins involved in formation of new blood vessels, the downstream effect of which is malformation.

However, your point, that damaged genes can lead to errors in assembly, stands.

Thanks for the correction on thalidomide...


The information that followed from others was very educational..

 

[arcticpenguin]

On handness:

From the article linked to above, it does seem to be genetically linked, but it's fairly common for identical twins to have different handedness, so it's not straight-forward.

My understanding is that if the cilia are spinning in the 'usual' direction, the result will be right-handedness. If the cilia are not spinning, then the result is a coin flip.

Tiime to bring out the anecdotes: I have 5 brothers. 3 of us are left-handed, 3 are right. All of my 4 sisters are right-handed.

 

[DrChinese]

I'd also argue that justr because somethign is genetic doesn't mean it's okay in a moral sense. IIRC, genetic links have been shown for alcoholism and violent tendencies as well.

Are you sure? I know there have been a number of studies of statistical correlations. If there is a genetic link, which gene is it? What is the test to determine an individual is at elevated risk for alcoholism?

I think this is what corplinx was getting at. Everybody "recalls" hearing it somewhere. Then you look, and poof! It's gone.

And please, don't bother showing me studies that "suggest" a link. Scientific evidence only, please. A study done by a scientist does not make it scientific evidence.

 

[Dymanic]

And please, don't bother showing me studies that "suggest" a link. Scientific evidence only, please. A study done by a scientist does not make it scientific evidence.

Insisting on evidence is a good thing and all, but are the requirements for what does and does not constitute evidence ever met with absolute certainty?

Is it inappropriate to speculate at all simply because of the way speculations have of getting transformed into fact as they enter the popular media?

If a genetic link (to anything) exists, won't evidence of that tend to emerge somewhat gradually? Aren't we most often working with small pieces of a puzzle that tend to present themselves one at a time? I mean, how often are we actually able to flick on a spotlight, suddenly suddenly revealing in stark detail all the mysteries of some previously unexplained phenomenon?

 

[DrChinese]

Insisting on evidence is a good thing and all, but are the requirements for what does and does not constitute evidence ever met with absolute certainty?

If a genetic link (to anything) exists, won't evidence of that tend to emerge somewhat gradually? Aren't we most often working with small pieces of a puzzle that tend to present themselves one at a time? I mean, how often are we actually able to flick on a spotlight, suddenly suddenly revealing in stark detail all the mysteries of some previously unexplained phenomenon?

Studies which purport to show "evidence" of a link by virtue of the statistical relationship they document are not really science. It is a lot like doing part of the work and not the rest of the work, because the rest of the work is too difficult.

Did you know that there is a positive statistical relationship between life expectancy and owning life insurance? (I.e. people who have life insurance live longer than those who don't, I'm not making this part up.) This would imply that life expectancy is related to your insurance salesman, wouldn't it? Obviously not. You can intuitively see that such a relationship is spurious.

The reason it is spurious is that no cause-effect relationship is offered by me to allow the base hypothesis to be tested. A hypothesis which states that the correlation is fundamental is avoiding the scientific requirements of theory development. Yet how is my evidence "suggesting" a link between life expectancy and the purchase of life insurance any different than other such studies which "suggest" a genetic basis for this or that?

The correct sequence would be to a) hypothesize a casual relationship; b) determine a test which could falsify that hypothesis; and then c) perform that test. If such was done, one would quickly determine that there are many easier ways to falsify the hypothesis (a) than to perform a study which loosely supports the hypothesis (c). Most statistical studies take the easy way out because measuring a relationship - usually without bothering to create a random sample or gathering a sufficiently large sample size, etc. - is pretty easy and it makes for great headlines in peer journals.

Certainly I agree that theory development takes time. But shortcuts - however well intentioned - do nothing but lead us in the wrong directions. There are statistic relationships between literally millions of putative factors - all totally spurious - yet which purport to show a statistically significant correlation as being "interesting", "worthy of more study", "highly suggestive", etc. It is sometimes difficult to see that measuring the relationship is not really science.

 

[Dymanic]

Did you know that there is a positive statistical relationship between life expectancy and owning life insurance? (I.e. people who have life insurance live longer than those who don't, I'm not making this part up.) This would imply that life expectancy is related to your insurance salesman, wouldn't it? Obviously not. You can intuitively see that such a relationship is spurious.

The reason it is spurious is that no cause-effect relationship is offered by me

Does it have to be you that offers the cause-effect relationship? Suppose that it was you who discovered the statistical correlation, and it was someone else who offerred the explanation that life expectancy and life insurance are related because the individuals who have the foresight and the resources to purchase life insurance also tend to be more consistent about seeking health care and living healthy life styles? Is an explanation like that (even though it may actually be the truth) ever going to be completely testable--or falsifiable? Even when we are pretty doggone sure about something, we aren't ever absolutely sure, are we?

It is sometimes difficult to see that measuring the relationship is not really science.

Seems like it's at least pre-science or something. I see what you're saying, and I don't disagree, but I don't see a big problem with saying, "These are some of the things we think may deserve a closer look". It might even turn out that somebody else already has some data--how else are you going to know?

 

[DrChinese]

Does it have to be you that offers the cause-effect relationship? Suppose that it was you who discovered the statistical correlation, and it was someone else who offerred the explanation that life expectancy and life insurance are related because the individuals who have the foresight and the resources to purchase life insurance also tend to be more consistent about seeking health care and living healthy life styles? Is an explanation like that (even though it may actually be the truth) ever going to be completely testable--or falsifiable? Even when we are pretty doggone sure about something, we aren't ever absolutely sure, are we?

Of course it doesn't have to be me. And someone else could have postulated the cause-effect relationship. But that requires a causal mechanism too, and that is what is missing in such studies. So in the case of the life insurance, a specific, testable mechanism would be put forth for review. I could test it, anyone could test it. It should have sufficient detail to be falsifiable.

In your hypothesis about life insurance, you are actually stating that there is no causal relationship between the measured variables. That is because the relationship is spurious. It is certainly possible that X is related to Y, and X is related to Z. But that does not mean that Y is casually related to Z in any way at all.

A simple example of that is the sex of two siblings. The sex of each siblings is determined by the father (X or Y passed down). Yet there is no correlation between the sex of the siblings themselves. You would never be able to use statistics to figure out what was going on unless the causal mechanism was set forth. Then you would know exactly what to test. Everything else is spurious.

You ask whether there are cases in which such studies might be useful, like pre-science. Sure, there are cases in which the study is useful. I don't dispute that. But you really won't know that tell later, will you? My position is simple: the study is not the fundamental science. You need more - a lot more - to make it science. Theory + Experiment = Science.

 

[Dymanic]

In your hypothesis about life insurance, you are actually stating that there is no causal relationship between the measured variables. That is because the relationship is spurious. It is certainly possible that X is related to Y, and X is related to Z. But that does not mean that Y is casually related to Z in any way at all.

A simple example of that is the sex of two siblings. The sex of each siblings is determined by the father (X or Y passed down). Yet there is no correlation between the sex of the siblings themselves. You would never be able to use statistics to figure out what was going on unless the causal mechanism was set forth. Then you would know exactly what to test. Everything else is spurious.

That makes perfect sense to me now. Thank you for taking the time.