Decentralized Memory

Alex Knapp at Heretical Ideas fires a shot at intelligent design:

You know, there’s always one thing that I think more people should pay attention to with regards to intelligent design as an “alternative” to evolution, and that’s the fact that one of the primary things that attracts people to the argument are the little things in the universe that appear to be conducive to life. One argument commonly mentioned in support of ID is that the position of Earth’s orbit appears to be “fine-tuned” in order to allow for life to emerged. Other arguments invoke things like the gravitational constant, the tilt of the Earth’s axis, etc.

But you’d think one thing that opponents of ID (like myself) should invoke more often are the examples of downright shoddy design put forth by the creator of the universe. The design of the human testicles, for instance. Here we have one of the primary reproductive organs located in a sac outside of the body and very, very vulnerable to damage. Why? Because human sperm can’t be produced at normal human body temperature! That doesn’t seem like intelligent design to me.

All true, and these point have been made before by the likes of Gould, Dawkins, et al. However, I think he makes a mistake with this:

Another example of shoddy design is human memory. It has no organization, no reliable system of retrieval, and is prone to both error and degradation. Think about how many catastrophes, both major and minor, have resulted on behalf of shoddy human memory. This was certainly not beyond the means of the creator–after all, we lowly humans can create computers with more reliable memory setups than the brain.

Regardless of what you think of inteligent design, it's possible that our memory system could be an example of good design. Knapp laments the lack of "organization," but I think our decentralized memories are a strength. While certainly "prone to error," they function very well. But more importantly, its dispersed nature (with no one "memory center") helps protect it and keep it intact. If a "well-designed" brain, by his definition, did exist, it would take just one traumatic accident, or just one mini-stroke, to wipe out a good part of our memories.

Even if Alex would suggest a well-designed brain should have dispersed memory centers but more organized connections, these connections would be subject to the same risk of injury and major memory loss (while possibly to a lower exent) after one unfortunate event.

Update: Alex Knapp quickly replies:

Even if you think that the organization is fine, you have to question the integrity of memory itself. Afterall, one of the major discoveries in criminology/psychology in the past 20 years has been that the memories of the eyewitness of an event are poor and pretty fundamentally unreliable.

I should have completed my argument on the first attempt.

Alex's point is well taken, but I have heard convincing arguments that there is a reason for this: if we had perfect memories, we'd remember every awful thing that ever happened to us. That would make it hard for us to get things done: get past our fears, put disappointments behind us, take risks - all things we need to be able to do to survive.

I agree with Knapp's general point that there are many examples of poor "design" - we just to be careful before we accuse nature of this.

Share this

Scientific American did an

Scientific American did an article a few years ago on a better designed human. Among the changes were to a redesigned knee, a better load-distribution system for the trunk (aka the spine), more distributed redundancies so that *A* liver might fail, instead of *THE* liver might fail. Etc., etc.

This was just fine-tuning the existing design (akin to taking a Model-T to a Porsche) and did nothing about a complete rethinking of the problem (akin to changing your Model-T for a helicopter).

Of course the theists will just posit that the design is perfect because what look like bugs to us customers are actually features to the designer. In other words, failure and inefficiency are part of the designer's plan. Remind me to hire a new designer.

Finally, the unspoken assumption is that the designer is YHVH, and not the Olympians or Chthulu or some previously unsuspected candidate. Even if intelligent design can be demonstrated, we have no evidence as to the identity of the designer. YHVH is assumed for cultural reasons, not because of independent evidence.

"I am actually not sure

"I am actually not sure whether I believe there are a bunch of universes with nobody asking the question or whether a universe can only exist if it will produce intelligent life. I can think of mechanisms for both."

The problem with this question is, is it meaningful to speak of something as "existing" if:

1. No one can, even in principle, observe it?
2. It cannot, even in principle, affect anyone?

If it can affect someone/be observed, then by definition it is part of the multiverse/universe, and not properly a "separate universe." That's when stuff like this starts to hurt my head.

There are other explanations

There are other explanations for why the Earth and the universe seem fine-tuned to allow for life. Even the number of spatial dimensions (3) is perfect for life. More would be as bad as fewer, as has been shown by Roger Penrose.

There are several forms of the "anthropic" principle, which states that the reason the universe is the way it is is because if it weren't, we wouldn't be around to ask the question in the first place. There are verying strengths of anthropic principle, from the weak, which just says what I said above, to the strong, which says the universe must create intelligent life that is capable of observing it to even exist in the first place.

One can most certainly apply the anthropic principle to Earth, even of they don't apply it to the universe. If there are millions of planets out there, *of course* the one(s) that produce intelligent life are going to be well-suited to life.

I am actually not sure whether I believe there are a bunch of universes with nobody asking the question or whether a universe can only exist if it will produce intelligent life. I can think of mechanisms for both.

A discussion of the flaws in

A discussion of the flaws in our design would be incomplete without the classic example of the eye. Specifically, the retina. Specifically, the fact that the nerves coming out of our light-sensing cells don't come from the back of the cells like you'd expect. They stick out front, *right in the path of incoming light*, then go around and back to the brain.

There is no sense in this, from a design standpoint. But it turns out that if you examine the way in which the eye evolves, there is a path of tiny changes from no eye at all to having an eye with a lens, and that path involves the retina turning over at one point. There is no path from nothing to a lensed eye that keeps the nerves in back.

I've always found this a striking example - a feature that makes perfect sense from an evolutionary perspective and no sense from a design perspective. I think it's described in Dawkin's _The Blind Watchmaker_.

what was Micha's

what was Micha's post
"Blogging while high deep thoughts"

Just a note to the "the

Just a note to the "the earth is in the perfect position for life" ID theory.

Well, doesn't that make sense? After all, it is not as if we inhabit the only solar system with planets in the galaxy. Out of the billions of stars in the galaxy, the likelihood is that there will be many thousands with planets placed appropriately for carbon based life to develop. Given that the odds of life arising in a pure evolutionary system must be fairly low, then the odds that life would only develop on "perfectly placed" planets is going to be very high. From a statistical perspective (how many trillions of stars exist in the universe as a whole, how many have planetary systems, etc.) the argument that perfect placement of the planets in our solar system is meaningless and holds no real scientific validity.

By the way, Alex, I understand your contention. That said, for all its "faults" the human brain has far more computing power, memory and capability than we have yet been able to achieve in silicon computers for a much lower rate of power consumption. We're even "designed" for backwards compatibility much better than human created computers. In my opinion (as a computer engineer) the human brain is a marvelous piece of organic engineering, although that shouldn't be construed to mean that I actually think it was intelligently engineered. Rather, I think the human brain shows the true power of Darwinistic theory. With intelligent design we can't yet come close to creating a silicon brain that can achieve what the human brain can do.

Matt, I'm way ahead of you

I'm way ahead of you on that one, it just sounded funny to me...

When you think about it though, the Christian's thought that we are created in "God's Image" would make it seem like we're the end product, thus negating my whole argument.

Oh don't get me wrong Brian,

Oh don't get me wrong Brian, I know that not many people are going to stop believing in god just because it's parsimonious to do so (would that it were so!). I just don't see the point of anyone bringing it into discussions of science at all, since, as you say, belief in a god is generally prior to any empirical considerations. This whole Creationism thing drives me up a wall for that very reason. People can have their god if they like, just as long as they don't get his chocolate in science's peanut butter.

Nobody is positing God as a

Nobody is positing God as a solution to the scientific question of evolution qua theory.

Believers believe in God prior to any scientific discussion. But as part of a theistic worldview that includes a creator/creation aspect, you find a "place" for God given what we know of the world (His world, to a believer). The exercise depends on the level of tension between the believer's scientific commitments and faith as to how nuanced or divorced from reality it becomes.

But it is not a case where one can simply invoke parsimony and wave God away.

Brad -- But can't we just go

Brad -- But can't we just go for parsimony and eliminate the superfluous hypothesis of God? It's just sitting there supporting nothing and explaining nothing, after all.

Being able to remember bad

Being able to remember bad things doesn't have to mean dwelling on them.

Wouldn't it be better to have memories that more accurately remembered facts, so that our decisions were solely (or would that be soul-ly?) based on character? How is a god supposed to judge those who do evil with good intentions as a result of the bad memory he/she/it created?

Memory imperfection isn't as good as an example as male nipples.

Also note that the

Also note that the intelligent design argument is not logically inconsistent with the "God-directed-evolution" argument.

It could just be that we're still a work in progress. :grin:

Even if you think that the

Even if you think that the organization is fine, you have to question the integrity of memory itself. Afterall, one of the major discoveries in criminology/psychology in the past 20 years has been that the memories of the eyewitness of an event are poor and pretty fundamentally unreliable.

I'm pretty sure the

I'm pretty sure the fallibility of memory is a feature, not a bug. Our brains are optimized for dealing with abstractions rather than specifics. The same wetware that lets you quickly recognize something you've never seen before as "a chair" or "food" can make it hard for you to remember the exact characteristics of any particular chair or item of food. Your brain uses a lossy distributed compression algorithm that makes use of the similarities between the thing you just saw/did/experienced and other things you previously saw/did/experienced.

If it didn't, you'd have to do too much processing to handle basic recognition tasks. Forget worries. Rather, think about the computation it would take to recognize your wife or your child and pick them out of a crowd, if you remembered in specific detail every face you'd ever seen from every angle and lighting condition you'd seen it in, and had to somehow sort through that information to determine whether there was a match. What we seem to have instead of that is sort of a one-way hashing algorithm that takes an input and maps it - quickly - to one of the compressed states we know about, but does so with some possibility of error.

That's an excellent point

That's an excellent point Glen. More generically, it's recognition that all designs are compromised by tradeoffs.

If there is an intelligent designer, why would we assume (given what we know of the universe) that this designer would not have to make design tradeoffs? We don't even know what the designer's objectives might be - so speculation on the efficacy of the design is pointless. Put another way, if you don't understand the full purpose of an automobile and the constraints of its creation, how can you criticize it's design? You don't have a baseline.

ID is a "God of the Gaps"

ID is a "God of the Gaps" argument. What happens is that gap is claimed to exist then it is quickly filled by some enterprising scientist. As I recall, the latest edition of the ID "bible" titled Of Panda's And People points to a number of gaps as a means to question evolutionary theory as we understand it today. As I recall, the text was published in the mid-1990s and since then, quite a number of these unresolved gaps have been filled.

Interestingly enough, Behe has also had to backtrack from a number of his claims re: "irreducibly complex" things. The most famous of these is the mousetrap. Thus goes by the wayside another manifestation of the argument by design.

ID is pseudo-scientific clap trap (junk science in other words) and should be treated as such.

Patri Friedman,

Great example.

BTW, I wasn't trying to

BTW, I wasn't trying to defend ID. I'm a firm believer in the anthropic principle. I'm just saying that the brain seems to do so many marvelous things we haven't much clue yet how to replicate, that I'm willing to cut the "designer" some slack regarding those few features we /do/ know how to replicate. Shakey the Robot can remember things and move around reasonably well, but he's not much of a conversationalist and gets easily confused. 30 years later, most of the "hard problems" in AI are still pretty hard.