What is more important in determining an (individual) organism’s phenotype, its genes or its environment?
Any developmental biologist knows that this is a meaningless question. Every aspect of an organism’s phenotype is the joint product of its genes and its environment. To ask which is more important is like asking, Which is more important in determining the area of a rectangle, the length or the width? Which is more important in causing a car to run, the engine or the gasoline? Genes allow the environment to influence the development of phenotypes.
-Tooby and Cosmides, emphasis theirs. It occurred to someone on the Less Wrong IRC channel how good this is an isomorphism of, “You have asked a wrong question.”
That sounds like less of a wrong question and more of a “right question with surprising (low prior) answer”. As far as the asker knew, the answer could have turned out to be, “Genes produce the same organism phenotype across virtually all environments, so genes are more important because changing them is much more likely to change the expressed phenotype than changing the environment.” (and indeed, life would not be life if genes could not force some level of environment-invariance, thereby acting as a control system for a low-entropy island)
I don’t see what’s wrong with answering this question with “neither [i.e., they’re equal], because they jointly determine phenotype, as independent changes in either have the same chance of affecting phenotype”.
An example of a wrong question, by contrast, would be something like, “Which path did the electron really take?” because it posits an invalid ontology of the world as a pre-requisite. The question about phenotypes doesn’t do that.
Grognor: chelz: is the area of a rectangle more the length, or the width?
Thewidth. Changing the width makes a bigger change in the area than changing the length does. (By convention, the width is defined as the smaller of the two dimensions of the rectangle.)
You have resolved the question to the nearest available sane question but that isn’t the answer to the question itself and does not make the question valid.
Come to think of it I am somewhat dubious with answering “is the area of this 1km by 1m rectangle more the 1km or the 1m?” with “the 1m”. That just doesn’t seem right.
Since my sibling reply got voted up a lot, I want to follow up: it seems that not only is the question not wrong, the “dissolving” answer is itself wrong, or at least very misleading. (Naturally, I have to tread cautiously, since I’m not an Expert in this area.)
As I said in my other reply, the defining characteristic of life is its ability to maintain a low-entropy island against the entropizing forces of nature. So there must be some range of environments in which an organism (via genes) is able to produce the same phenotype regardless of where its environment falls within that range. In effect, the genes allow the phenotype to be “screened off” (d-separated, whatever) from its environment (again, within limits).
A thing that truly allows the environment equal influence in its final form as the thing itself (as suggested by the T&C answer) is not what we mean by “life”. It’s the hot water that eventually cools to a temperature somewhere between its current temperature and that of its initial environment. It’s the compressed gas molecules in the corner of a chamber that eventually spread out evenly throughout the chamber. It is, in short, not the kind of self-replicating, low entropy island we associate with life, and so has no basic units thereof, be they genes or memes.
So there must be some range of environments in which an organism (via genes) is able to produce the same phenotype regardless of where its environment falls within that range.
The organism needs to successfully thrive and reproduce within that range. Sometimes this means tailoring its phenotype to the environment it finds itself in.
But imagine a world in which environment truly was more determining than genes. Every animal born in a swamp would be a frog (no matter what its parents were) and every animal born in a tree would be a bird. Perhaps coloration or some other trait might be heritable — blue birds who move to swamps give rise to little blue tadpoles — but the majority of phenotypic features would be governed by the environment in which the organism is born and develops.
In our world, all we know about X is that it is a phenotypic feature, then we should expect it is more likely to be stable under different environments than to be stable under different genotypes. Features must owe more (on the aggregate) to genes than to environment. If it were otherwise, then we would not talk about species! We know we are not in the swamp-birds-have-tadpoles world.
When people talk about genes vs. environment, they usually aren’t really talking about all features. They’re usually talking about some particular, politically interesting set of features of humans …
I disagree. Say that many members of the royal family have hemophilia. Is this due to genes, or environment? If it is genes, you can try to track down who has the gene, and not marry any of those people to your current monarch. If it is something in the royal water supply, you can track that down. If you say “It’s both!”, you are unlikely to solve the problem.
This applies to pretty much every case where people argue whether something is genes or environment. The claim that you can’t call some things mainly genetic and some things mainly environmental is, we know with a very high degree of certainty, false. In most cases, the motivation for this claim is, I think, to avoid the unpleasant possibility that the answer is “genetic”.
I never actually considered this viewpoint. But you know, the Tooby and Cosmides quote attacks the false dichotomy of “Everything is genetics! It’s all programmed from before you’re born!” and “Blank slate! Absolutely nothing is determined by anything other than experience!” both of which are nonsense.
But it also doesn’t support the false third option, “You found a genetic basis for autism? Racist!”
For certain traits, you cannot break things down as a ratio of genetics:environment. For example, myopia appears to be a genetically-based trait, but it also appears to be expressed much more frequently when children learn to read (which is why such a disasterous trait for a hunter-gatherer wasn’t eliminated in the tens of thousands of years before literacy). In other words, the phenotype (nearsighted) is both entirely genetic and entirely environmental.
Some things are mostly genetic. Some things are mostly environmental. Some things are a mix of both. But currently, you are supposed to say that everything is both genetic and environmental (or be labelled a racist). And that is false.
Some things are mostly genetic. Some things are mostly environmental. Some things are a mix of both. But currently, you are supposed to say that everything is both genetic and environmental (or be labelled a racist). And that is false.
Everything is genetic and environmental. If you look low enough down.
The fact is, humans share lots of genes with each other.
Example: Suppose I tell you, “What about language acquisition? I’m sure that if I speak better Italian than Nick Bostrom and he speaks better Swedish than me, our genes have f* all to do with that.” You could answer that it’s our genes which shaped our brain in such a way that we could have picked up a native language in the first place, and a chimpanzee (or a human with major neurological problems) wouldn’t have learned Italian or Swedish even if raised in the very same environment. But when more than 99% (I guess) of the world human population would have been able to learn whichever natural (or sufficiently natural-like) language they had been raised in, such an objection wouldn’t be very useful.
On the other hand, while genes require environments in a given range to be expressed (you couldn’t raise a person to be the same as me on Mars, even if he were my identical twin brother), certain features are expressed pretty much the same way throughout the range of environments where one could survive. The probability that John’s blood type is AB+ given that he’s alive and that his identical twin brother’s blood type is AB+ is pretty close to 1, wherever John was raised.
Hence, I’d just say that language is environmental and blood type is genetic. Anything else is useless nitpicking, akin to saying that I shouldn’t say that the C and C# keys on a piano are white and black respectively because even the former does absorb some light and even the latter does scatter some.
Say that many members of the royal family have hemophilia. Is this due to genes, or environment? If it is genes, you can try to track down who has the gene, and not marry any of those people to your current monarch. If it is something in the royal water supply, you can track that down. If you say “It’s both!”, you are unlikely to solve the problem.
If it is both then you are more likely to solve the problem by saying “It’s both” than by saying “it’s one!”
-Tooby and Cosmides, emphasis theirs. It occurred to someone on the Less Wrong IRC channel how good this is an isomorphism of, “You have asked a wrong question.”
That sounds like less of a wrong question and more of a “right question with surprising (low prior) answer”. As far as the asker knew, the answer could have turned out to be, “Genes produce the same organism phenotype across virtually all environments, so genes are more important because changing them is much more likely to change the expressed phenotype than changing the environment.” (and indeed, life would not be life if genes could not force some level of environment-invariance, thereby acting as a control system for a low-entropy island)
I don’t see what’s wrong with answering this question with “neither [i.e., they’re equal], because they jointly determine phenotype, as independent changes in either have the same chance of affecting phenotype”.
An example of a wrong question, by contrast, would be something like, “Which path did the electron really take?” because it posits an invalid ontology of the world as a pre-requisite. The question about phenotypes doesn’t do that.
The conversation in question.
The width. Changing the width makes a bigger change in the area than changing the length does. (By convention, the width is defined as the smaller of the two dimensions of the rectangle.)
You have resolved the question to the nearest available sane question but that isn’t the answer to the question itself and does not make the question valid.
Come to think of it I am somewhat dubious with answering “is the area of this 1km by 1m rectangle more the 1km or the 1m?” with “the 1m”. That just doesn’t seem right.
Hmmm...
“No.”
Is that better?
Yes! :)
Only if you’re augmenting/cutting by a fixed length.
If you’re using a proportion (e.g. cut either the length or the width in half) then they’re equivalent.
I could also meaninglessly answer that the length is more important, as it will always be equal or bigger.
the key to finding a wrong question is finding that the answer doesn’t help the person who asked it.
Since my sibling reply got voted up a lot, I want to follow up: it seems that not only is the question not wrong, the “dissolving” answer is itself wrong, or at least very misleading. (Naturally, I have to tread cautiously, since I’m not an Expert in this area.)
As I said in my other reply, the defining characteristic of life is its ability to maintain a low-entropy island against the entropizing forces of nature. So there must be some range of environments in which an organism (via genes) is able to produce the same phenotype regardless of where its environment falls within that range. In effect, the genes allow the phenotype to be “screened off” (d-separated, whatever) from its environment (again, within limits).
A thing that truly allows the environment equal influence in its final form as the thing itself (as suggested by the T&C answer) is not what we mean by “life”. It’s the hot water that eventually cools to a temperature somewhere between its current temperature and that of its initial environment. It’s the compressed gas molecules in the corner of a chamber that eventually spread out evenly throughout the chamber. It is, in short, not the kind of self-replicating, low entropy island we associate with life, and so has no basic units thereof, be they genes or memes.
The organism needs to successfully thrive and reproduce within that range. Sometimes this means tailoring its phenotype to the environment it finds itself in.
Of course.
But imagine a world in which environment truly was more determining than genes. Every animal born in a swamp would be a frog (no matter what its parents were) and every animal born in a tree would be a bird. Perhaps coloration or some other trait might be heritable — blue birds who move to swamps give rise to little blue tadpoles — but the majority of phenotypic features would be governed by the environment in which the organism is born and develops.
In our world, all we know about X is that it is a phenotypic feature, then we should expect it is more likely to be stable under different environments than to be stable under different genotypes. Features must owe more (on the aggregate) to genes than to environment. If it were otherwise, then we would not talk about species! We know we are not in the swamp-birds-have-tadpoles world.
When people talk about genes vs. environment, they usually aren’t really talking about all features. They’re usually talking about some particular, politically interesting set of features of humans …
I disagree. Say that many members of the royal family have hemophilia. Is this due to genes, or environment? If it is genes, you can try to track down who has the gene, and not marry any of those people to your current monarch. If it is something in the royal water supply, you can track that down. If you say “It’s both!”, you are unlikely to solve the problem.
This applies to pretty much every case where people argue whether something is genes or environment. The claim that you can’t call some things mainly genetic and some things mainly environmental is, we know with a very high degree of certainty, false. In most cases, the motivation for this claim is, I think, to avoid the unpleasant possibility that the answer is “genetic”.
I never actually considered this viewpoint. But you know, the Tooby and Cosmides quote attacks the false dichotomy of “Everything is genetics! It’s all programmed from before you’re born!” and “Blank slate! Absolutely nothing is determined by anything other than experience!” both of which are nonsense.
But it also doesn’t support the false third option, “You found a genetic basis for autism? Racist!”
For certain traits, you cannot break things down as a ratio of genetics:environment. For example, myopia appears to be a genetically-based trait, but it also appears to be expressed much more frequently when children learn to read (which is why such a disasterous trait for a hunter-gatherer wasn’t eliminated in the tens of thousands of years before literacy). In other words, the phenotype (nearsighted) is both entirely genetic and entirely environmental.
Some things are mostly genetic. Some things are mostly environmental. Some things are a mix of both. But currently, you are supposed to say that everything is both genetic and environmental (or be labelled a racist). And that is false.
Everything is genetic and environmental. If you look low enough down.
The fact is, humans share lots of genes with each other.
Example: Suppose I tell you, “What about language acquisition? I’m sure that if I speak better Italian than Nick Bostrom and he speaks better Swedish than me, our genes have f* all to do with that.” You could answer that it’s our genes which shaped our brain in such a way that we could have picked up a native language in the first place, and a chimpanzee (or a human with major neurological problems) wouldn’t have learned Italian or Swedish even if raised in the very same environment. But when more than 99% (I guess) of the world human population would have been able to learn whichever natural (or sufficiently natural-like) language they had been raised in, such an objection wouldn’t be very useful.
On the other hand, while genes require environments in a given range to be expressed (you couldn’t raise a person to be the same as me on Mars, even if he were my identical twin brother), certain features are expressed pretty much the same way throughout the range of environments where one could survive. The probability that John’s blood type is AB+ given that he’s alive and that his identical twin brother’s blood type is AB+ is pretty close to 1, wherever John was raised.
Hence, I’d just say that language is environmental and blood type is genetic. Anything else is useless nitpicking, akin to saying that I shouldn’t say that the C and C# keys on a piano are white and black respectively because even the former does absorb some light and even the latter does scatter some.
The notion of heritability clears up this issue a bit, as it screens off genetic similarities in the population.
Indeed. (On the other hand, people often grow up in the same region as their parents...)
So perhaps heritability should have a counterpart that screens off common environmental factors...
And everything scatters some of the incident light and absorbs some. But this doesn’t mean we should never call anything “black” or “white”.
If it is both then you are more likely to solve the problem by saying “It’s both” than by saying “it’s one!”