A ways back in the week when I was pretty crunched for time and motivation, Blunt Object ranted about an article on Slate badly misunderstanding genetics and what we can know from it. It’s pretty typical boilerplate biology-is-scary stuff, or at least the part he’s ranting about is; there’s essentially one paragraph of raw stupidity in the middle of an otherwise reasonable piece talking about the implications of fetal genotyping. The relevant paragraph:
What fetal genes might one day suggest about a baby’s eye color, appearance, and intellectual ability will be useful to parents, not insurers. But with costs coming down and insurers interested in other aspects of the fetal genome, a Gattaca-like two-tiered society, in which parents with good access to health care produce flawless, carefully selected offspring and the rest of us spawn naturals, seems increasingly plausible.
Well… no, not really. To put it mildly. If the world worked like this I’d be able to write poetry in Linear A, but merely finding something that does something in particular and making some more or less educated guess as to what it does does not translate into being able to use it for engineering.
The analogy Blunt used was programming, and it’s a pretty good one; I set out to quote it for effect but wound up concluding it really just needs to be read in its entirety. It’s not long, and is resistant to excerpting.
The only issue with his analogy is that it doesn’t even capture the impossibility of pulling off something like the Slate author’s scenario adequately; at least a piece of computer hardware and its programming were produced via a process we could find relatively intuitive. Genomes were produced by evolution, with no one on hand trying even remotely to ensure that the code was efficient or clean, let alone commented. Kludges and elegant solutions exist side by side, sometimes in several different copies, some of which are broken and others of which do subtly different things in each version. Much of the information is if-then instructions and operating instructions, sometimes to provide for cases that are remote or no longer exist. If your computer were equivalent, it would have every operating system and program you had ever used installed at once, with the instructions for which pieces work for what and are active at any given time being completely hidden information. All possible hardware styles and protocols are present as well, and which ones are active or not is equally obscure.
Among the list of what the fetal testing is meant to do: determine Rh-factor, sex of child, presence of Down’s syndrome. Testing for any of these things is not like looking for a line of code in a computer program; it’s like seeing if a hard drive rattles or not when you pick it up, or how many USB ports there are. The number of chromosomes as well as what kind there are at pair 23 is determinable by technology we’ve had since the early part of the twentieth century; it is to genotyping as correctly naming a shape to be a square is to polygonal geometry.
Let’s tackle the first line in the author’s GATTACA scenario piece by piece:
What fetal genes might one day suggest about a baby’s eye color, appearance, and intellectual ability
1. Do you know we currently have no idea how eye color genetics work beyond two genes that happen to handle “blue” or “brown” relatively straightforwardly? Beyond that we know that there are many more genes that affect eye color, that there are two more genes that definitely do something though we’re not positive what, that there might be as many as 16, and that green and gray and hazel are handled somewhere entirely different, but you’re simply not going to know what color a baby’s eyes are going to be based on even its entire genome- because we only know what two genes are going to do and can’t even find the rest to see if they’re present and what they’re going to do.
2. You can know everything about a baby’s appearance that is determined by a single, stand-alone trait that we know about, understand to be a trait influenced by a single gene or at least a manageable handful of them, and know exactly which gene does that. Compared to all factors of a baby’s appearance, the number of traits this describes is teeny tiny. If the driving force of your curiosity is knowing whether a baby boy’s ears will have attached pinnae, you’re in luck*.
3. We don’t currently even know what intellectual ability quite is. We can’t nail down a single test accurately measuring “general intelligence”, all the tests we currently have produce wildly different results from one another, and while we know more or less that there are different cognitive domains and skills, we can’t nail any of them down particularly well either. Worse than that, we understand vaguely that intelligence is more of an emergent property of many systems and skills, but we can’t quantify or measure it well. For something like a car, “speed” is an emergent property with no corresponding part of the car that develops out of nearly every other part of the car- but we can concretely and easily measure speed.
Most of what we know about genetics and intelligence can be summed as this: 1)It seems to be, broadly, heritable, and 2)cognitive impairments are much, much easier to detect and quantify than variations in normal intelligence or extremely high intelligence. This is, in fact, what IQ tests were originally designed for- picking out those sufficiently impaired to need different schooling. We can expect legitimate bioethics issues surrounding the ability to detect those sorts of cognitive impairments caused by developmental disorders that are known and genetically quantifiable- not engineered superbabies versus dull “naturals”.
So, of the author’s three projected super-baby traits, one of them is a simple thing that turns out very much not to be on the genetic end, and two are emergent gestalt qualities we cannot even quantify, let alone reverse-engineer. Provided we develop the ability to directly engineer in the first place, which currently we can’t.
As science-fiction-come-reality scares go, I’m not that impressed.
*Actually I’m lying. This old chestnut of simple Mendelian genetics, as well as sex-linked traits, turns out to involve multiple alleles of opaque effect as well. Surprise!