Large-scale Thalamocortical Model

A fellow student in my lab is basing some of his preliminary work on neuron models devised by Eugene Izhikevich, who published a book last year in which he described a system for modeling the diverse spiking behavior of many types of neurons with an elegant set of equations. In a paper co-authored with Gerald Edelman (of Neural Darwinism fame) they implement a model of the cortex and thalamus and their interconnectivity.

Here are some features of the model:

1) Simulates one million multi-compartmental neuron models of 22 basic types
2) Includes approximately half a billion synapses
3) Macroscopic connectivity is based on data derived from diffusion tensor imaging (DTI) of magnetic resonance image (MRI) scans of the thalamus and cortex
4) Microscopic connectivity is based on reconstruction studies of cat visual cortex
5) Synapses are modified through spike-timing dependent plasticity (STDP)

Now, the thalamus is the part of your brain through which almost all sensory input is routed before being sent to the cortex (the exception is olfactory input). This structure is about the size of the end of your thumb, and it is a place through which nearly all of your input from the world flows (visual, auditory, and tactile). But like most brain areas, exactly what it does is not very well understood. For example, we know that it does not function merely as a relay station. There is extensive feedback from the cortex back to the thalamus, creating a thalamocortical loop. Why would the cortex need to send information back to the thalamus if it's just a relay station? Some theorists have proposed that the thalamus is something like an active blackboard, maintaining a constantly updating sketch of the world. Others have proposed that the loop is a way of keeping recent events in a kind of short-term buffer.

Whatever the case, the Izhikevich and Edelman model does not simulate the stream of sensory input from the world. So how does anything happen in the model? Well, initially it is quiescent. The modelers get it going by causing random neurons to spike, which effectively jump-starts ripples of activity throughout the system.

One interesting finding is that the brain state is very sensitive, so much so that the alteration of the spiking activity of a single neuron radically alters the global firing patterns throughout the model within less than half a second. This seems a bit counterintuitive. We might expect that the brain is robust to small changes. After all, neurons can be fairly noisy (that is, they don't always fire reliably), and they also tend to die off. So either their model is overly sensitive to small perturbations (i.e. the butterfly effect) or this really is a reflection of the sensitivity of real neural systems. Either way, it's an interesting result.

One last comment...the authors state that they "started with the thalamocortical system because it is necessary for human consciousness." In discussing the paper with another student, I mused about the ethical ramifications of this kind of simulation. I seriously doubt that a simulation of a million neurons evoked anything like consciousness when randomly jump-started, but then, consciousness is a very poorly-understood phenomenon. I told the other student that I was reminded of Johnny Got His Gun, an anti-war novel in which a soldier is wounded such that he loses all senses but touch, all his limbs, and most of his face. He tries to communicate by banging out Morse code with his head on the hospital bed.

"But this thing doesn't even have a head to try to bang out Morse code," I joked, before I realized just how creepy that sounded. So like I said, I seriously doubt we have to worry about the ethics of simulating consciousness at this point. There are about 100 billion neurons in a human brain and about 20 billion in the cortex, while this model uses one million neurons. But it's something to at least ponder, and definitely something to consider more as models become more and more sophisticated.

E. M. Izhikevich, G. M. Edelman (2008). Large-scale model of mammalian thalamocortical systems Proceedings of the National Academy of Sciences, 105 (9), 3593-3598 DOI: 10.1073/pnas.0712231105

Supercomputer Hyperbole

Wired has an article about the latest iteration of supercomputers breaking the petaflop barrier. That means they can carry out just over a quadrillion (1,000,000,000,000,000) floating-point calculations per second. That's a lot. And it's a significant milestone. But this seems like a bit much:

"The scientific method has changed for the first time since Galileo invented the telescope (in 1509)," said computer scientist Mark Seager of Lawrence Livermore National Laboratory.

Look, this is a quantitative, not a qualitative, change. Some are making the claim that the ability to model and simulate at greater and greater levels of detail will allow for qualitatively different ways of doing science. But I don't think so.

Breaking the petaflop barrier is more like building a bigger and better telescope that allows us to see farther and clearer, and to see things we've never seen before, but it's not like the invention of the telescope in the first place.

Lack of Religion and Belief in Hoo-doo (and more quotes by G. K. Chesteron)

Via The Frontal Cortex, here's a story about the results of a survey done by Baylor University researchers which apparently shows that people who believe in mainstream Christianity are less likely to believe in other supernatural phenomena than people who "never worship".


I'd like to find a link to the survey results broken down by question, but the Baylor site is kind of crappy (I think they want you to buy the new book that talks about the results). I can find the actual survey, but the results are either not there or buried among the links. If anyone else finds them, let me know.

This summary from the Wall Street Journal makes me curious for more information:

The Gallup Organization, under contract to Baylor's Institute for Studies of Religion, asked American adults a series of questions to gauge credulity. Do dreams foretell the future? Did ancient advanced civilizations such as Atlantis exist? Can places be haunted? Is it possible to communicate with the dead? Will creatures like Bigfoot and the Loch Ness Monster someday be discovered by science?


The answers were added up to create an index of belief in occult and the paranormal. While 31% of people who never worship expressed strong belief in these things, only 8% of people who attend a house of worship more than once a week did.

This is interesting, if true. What it might be indicating is that people who reject organized religion might be doing so for reasons that are not entirely rational (e.g. rebellion against parents, bad experiences in church or with clergy, etc.). And it might indicate that people have a strong affinity for hoo-doo, so that if they're not getting it from a mainstream source, they seek it in alternative sources.

But I'd like to see the individual survey results, because something smells fishy. The numbers are actually the opposite of a Gallup poll from the same time frame:

Participants were presented with a list of ten potential paranormal beliefs:

• extrasensory perception (41 percent of the respondents acknowledge belief in this item)
  
• haunted houses (37 percent)
  
• ghosts (32 percent)
  
• telepathy (31 percent)
  
• clairvoyance (26 percent)
  
• astrology (25 percent)
  
• communication with the dead (21 percent)
  
• witches (21 percent)
  
• reincarnation (20 percent)
  
• channeling spiritual entities (9 percent)
  

These results are statistically relevant across lines of "age, gender, education, race, and region of the country," according to Gallup. There is, however, some difference between Christians and non-Christians: the former group scores a 75 percent likelihood of belief, while the latter scores 66 percent. But both groups, as these statistics demonstrate, have a paranormal-positive majority.
(emphasis mine)

These latter results accord more with my own personal experience. The majority of people I talk to about such things tend to believe in ghosts, precognition, ouiji boards, and so on. Not so much Atlantis, Bigfoot, and the Loch Ness monster, but definitely categories dealing with "the spirit world". I simply don't buy the Baylor results that only 8% of regular church-going Christians believe in paranormal things. I think there's some conflation of the results going on here, and there might be some real issues with their methodology.

Anyway, G.K. Chesterton's character Father Brown apparently said: "It's the first effect of not believing in God that you lose your common sense, and can't see things as they are."

And Chesterton speaking as himself said: "When people stop believing in God, they don't believe in nothing - they believe in anything."

Um...right. I'd agree with the premise that if people reject religious thinking on any grounds other than rational ones, they're likely to latch onto pseudoscience and supernaturalism of another flavor. The nice thing about hoo-doo is that it's usually relatively simple to understand the basic concepts, and otherwise shrouded in mystery. People are typically lazy, and prefer a simple, bad hypothesis about the way the world works to one they have to work a bit harder at that may be closer to the truth.

Anyway, Chesterton was full of crap (again). Some people who reject religion open their heads to floodgates of nonsense, but not the ones who reject religion on grounds of reason and common sense in the first place. And if the Gallup poll results above are the more reliable, then Christians tend to be the more gullible in general.

Do We Stand On the Shoulders of Giants, or Kick Them to the Curb?

Laurie's reading The View From the Center of the Universe, and she pointed out to me last night that the authors like to hate all over Thomas Kuhn, saying he is responsible for a commonly-held, mistaken view of science. Namely, that scientific theories constantly replace one another, or to put it in terms of the title of this post, we kick previous thinkers to the curb instead of building upon what they've done. Kuhn laid out his ideas in The Structure of Scientific Revolutions, and the book has apparently gotten similar criticism since it first came out. But is this really what Kuhn was saying?

The fundamental issue is his notion that new theories, such as the heliocentric model of the solar system, are incommensurable (or incompatible) with old theories, such as the geocentric model of the solar system. But what about cases such as physics or evolutionary biology?

Did Einstein's ideas replace Newton's, or augment them? Well, kind of both. In the theoretical framework of Newton, matter is immutable. In the Einsteinian framework, matter may be transformed into energy. In a Newtonian framework, there is the concept of absolute time...the universe has a central clock, if you will. In the Einsteinian framework, there is no concept of absolute time or absolute simultaneity...whether or not two events occur at the same instant depends on the frame of reference from where the observation is made. So yes, in one sense the Einsteinian theoretical framework is not compatible with Newton's.

On the other hand, we obviously didn't toss Newton's work in the trash can. Many of the fundamental precepts he laid out are still valid in a particular frame of reference. In this sense, Einstein's ideas didn't replace Newton's, but augmented them.

So in a way I think the issue may just be a semantic one, and really not all that interesting. Kuhn was not a relativist, thinking we just bounce around between explanations. In the postscript to the 2nd edition of SSR, he writes:

Later scientific theories are better than earlier ones for solving puzzles in the often quite different environments to which they are applied. That is not a relativist's position, and it displays the sense in which I am a convinced believer in scientific progress.

I also think Kuhn would acknowledge that the achievement of a new paradigm is dependent upon all the work that has come before. Einstein was an anti-establishmentarian, and developed many of his ideas by rejecting some of the most entrenched assumptions in physics at the time. But his theories would have been impossible to develop without all of the work that had come before him.

Maybe Kuhn would disagree with this, and I'm misunderstanding him. If he is, then he's saying something much more radical than what I realized. But I don't think so.

Is There a Conflict Between Religion and Science?

So Rev. Malcolm Brown, the head of the public affairs department for the Church of England, has suggested that a formal apology be issued to Charles Darwin by the Church of England. Wouldn't do him much good now, but I suppose it could be accepted on his behalf by all the working biologists.



Here's the actual statement by Rev. Brown in full. Some schools of thought might say that when a religious institution makes such an overture, the polite thing to do is to accept it graciously. Unfortunately, his statement is full of nonsense. Ostensibly he wants to mend the fences, but throughout the essay he continually suggests that science needs to reach out to religion in order to ensure protection against the moral implications and abuses that might arise from theories like Darwin's. This is the old chestnut from Einstein: "Science without Religion Is Lame, Religion without Science Is Blind".

Here's what I'd consider the most relevant section of his essay:

Darwin was, in many ways, a model of good scientific method. He observed the world around him, developed a theory which sought to explain what he saw, and then set about a long and painstaking process of gathering evidence that would either bear out, contradict, or modify his theory. As a result, our understanding of the world is expanded, but the scientific process continues. In science, hypotheses are meant to be constantly tested. Subsequent generations have built on Darwin’s work but have not significantly undermined his fundamental theory of natural selection. There is nothing here that contradicts Christian teaching. Jesus himself invited people to observe the world around them and to reason from what they saw to an understanding of the nature of God (Matthew 6: 25–33). Christian theologians throughout the centuries have sought knowledge of the world and knowledge of God. For Thomas Aquinas there was no such thing as science versus religion; both existed in the same sphere and to the same end, the glory of God. Whilst Christians believe that the Bible contains all that we need to know to be saved from our sins, they do not claim that it is a compendium of all knowledge. Jesus himself warned his disciples that there was more that he could say to them and that the Spirit of truth would lead them into truth (John 16: 12–13). There is no reason to doubt that Christ still draws people towards truth through the work of scientists as well as others, and many scientists are motivated in their work by a perception of the deep beauty of the created world. Nevertheless, it is worth remembering that scientific theories can be overtaken in their turn even as old ideas prove to have an enduring quality. Most of us get by with some version of Newtonian physics and understand little of Quantum Theory. Newtonian ideas suffice for most of our everyday needs – but we now know that we can’t push them too far as there is plenty that they do not adequately explain.

See? No conflict! The Bible tells people to study the world around them, so science and religion are compatible.

Or not.

As I've noted before, the core difference between science and religion, the central element that brings them directly into conflict, is the way in which they deal with epistemology, or how we know things. Religious knowledge is derived primarily from ancient texts, religious authorities, and subjective experience. If you ask a believer in god why they believe in god, they will usually give one or a mix of the sources above.

Science, on the other hand, strives to eliminate human bias as much as possible through the checks and balances of peer scrutiny. Authority counts for very little, and so does subjective experience. You experienced a first-hand insight that your theory of gravitation is true? Great...now where's the evidence? I'm much more in line with the philosophy of science described by Kuhn than by the more popular one of Popper. Popper was all about falsifiability. A good hypothesis has to be falsifiable. Kuhn articulated a continuum, as opposed to Popper's binary view of the validity of a hypothesis. You come up with a hypothesis and you try to find evidence that either strengthens or weakens it, rather than outright falsifying it. Either way, the core element to both philosophies is the development of hypotheses and application of evidence to their merit.

This is why science and religion are fundamentally at odds with one another. These two ways of trying to understand the world are neither complementary or compatible. You either base the way you think the world is on the best evidence that fits the best ideas, or because a religious authority said it was that way or you just feel it to be the case.

So I guess we should accept the apology on behalf of Darwin and go about the daily process of trying to understand the world, but I don't see the need for science to reach out to religion, as Rev. Brown suggests. Religion doesn't provide any insights into science other than how not to do things.

Magpie Mirror Self-Recognition

We talked quite a bit about mirror self-recognition in my primate cognition class last semester. From what I understand from the readings and discussions, mirror self-recognition (the ability to recognize that the thing in the mirror is you, and not another member of your own species or something completely different) has only reliably been demonstrated in chimpanzees, orangutans, and humans. For reasons not very well understood, other apes such as gorillas, don't pass the mark test, where a colorless, odorless mark is placed on a location on the subject's body which can only be seen in a reflective surface, and the subject makes self-referential movements in reaction to such a reflection. No other primates, such as monkeys, are known to pass such a test either.

There have been claims of animals such as dolphins and elephants passing the mark test, though we read some of this research, and the evidence seems relatively weak. Which is one reason why it's interesting that a new study has indicated that magpies pass the mark test.

Here's a pic of a magpie with one of the test marks, a yellow pip placed under the beak:


And here's a video of a magpie trying to get the mark off with its claws and by rubbing itself against the bottom of the enclosure, and then checking itself in the mirror to make sure the mark is gone:



One of the necessary controls for an experiment like this is to do a set of experiments with marks that don't change the appearance of the bird. In this research, they used a black mark placed in the same location, which is not visible in the mirror because it blends in with the bird's feather color. They didn't get the same reactions with the black mark as they did with the colored marks. The reason you'd do a control like this is to make sure that the bird is reacting to something like the smell of the mark or how it feels on their body, but purely to the visual information they're getting from the mirror.

From what I've seen, this research looks pretty solid. If so, what does it say about the similarities between great ape and magpie cognition? One idea that was talked about in my primate cognition class about why humans, chimps, and orangs pass the test had to do with conscious knowledge about their body movements, since they all spend a significant amount of their present or recent evolutionary history in the tree canopies, while other large primates don't. The idea is that they have to be much more aware of where their limbs and torso are positioned at all times, or they will fall and suffer terrible injury or death (while other large apes live mostly on the ground, and smaller primates don't suffer as much from such mistakes). I wasn't really buying this argument, but if it's true, how would it gel with the magpie results, assuming they're solid?

I don't know. I'm not sure anybody does at this point. My guess is that mirror self-recognition may be a very specialized cognitive skill that may arise as a by-product of other cognitive skills, but may really say little to nothing about more abstract reasoning or higher cognition. I'd also guess that we'll see a lot more animals pass such tests in the coming years, but time will tell.

Reference:
Prior et al. Mirror-Induced Behavior in the Magpie (Pica pica): Evidence of Self-Recognition. PLoS Biology, 2008; 6 (8): e202 DOI: 10.1371/journal.pbio.0060202

Why Haven't Wheels Evolved (or Have They?)

In a mailing list discussion, the topic of the evolution of wheels came up. Basically, if wheels are relatively simple and extremely useful devices, why haven't biological organisms evolved them?


The Straight Dope pretty succinctly covered the various responses to this question in 1998.

Response #1: It's impossible to gradually evolve a wheel

Biological features generally evolve gradually, and a quarter or half of a wheel would not only be useless, it would probably actively hurt the fitness of an organism. Cecil says the problem with this argument is that some organisms already have evolved wheels (which we'll get to in a minute), which doesn't really address the actual issue. The gradualist argument has been famously used for all sorts of complicated structures, such as the vertebrate eye. What good is a quarter or half an eye? Well, Richard Dawkins does a great job of demolishing this objection (see The Blind Watchmaker). Half an eye is extremely useful. Even if all it does is let you distinguish light from dark patches, it gives you a selective advantage over members of your own species who can't see a damn thing at all.

Another famous example is the wing. What good is a proto-wing? Ask the organisms that have one (e.g. a flying squirrel). Even if you can't fly, gliding can be really useful, especially if you live in the upper tree canopies. And often in evolution, a feature that was used for an entirely different purpose gets co-opted to serve a different function. My favorite example of this is the hammer, anvil, and stirrup, the little bones in your ear that let you hear. Those were once jawbones in a primitive ancestor, and they got recruited for hearing.

So I don't buy the argument that the impediment to evolving wheels is the impossibility of the evolutionary trajectory. Many organisms have much more complex, labyrinthine structures that were evolved gradually.

Response #2: They've already evolved!

Cecil points out the mother-of-pearl moth, Pleurotya ruralis, which will curl into a ring structure, head-to-tail, if attacked, and roll backwards at about 40 cm/s, much faster than its crawling speed. Wikipedia has a nice entry on other examples, including organisms that use gravity to locomote via rolling, such as the web-toed salamander and the golden wheel spider, and those that are self-propelled, such as the mother-of-pearl moth, the mantis shrimp, and the scaly anteater (shown here).


But the objection is that these aren't true wheels, in the sense of the artifacts, with axles. Cecil points to the bacterial flagellum, which is a rotary mechanism that bacteria use to whip a tail-like structure and move through liquid environments. While the structure is more like a wheel than the multi-cellular organisms above, the function is less like a wheel, which we normally associate with locomotion on a hard substrate.

While I think these examples are interesting, I don't think any of them demonstrate the existing evolution of a biological wheel.

Response #3: It's impossible to evolve an organic wheel because of biological constraints

Cecil's response is a nice explanation:

A more complex creature couldn't evolve the wheel. Every time the thing turned, the nerves and blood vessels serving it would get hopelessly twisted." Science writer Stephen Jay Gould makes essentially this argument in his book Hen's Teeth and Horse's Toes.

But this may not be an insurmountable obstacle. A flesh-and-blood wheel might use the umbilical hookup found on some merry-go-rounds. Tape one end of a piece of ribbon to a tabletop and the other to the bottom of a compact disc. Turn the CD over so that the ribbon drapes over the side. Now move the CD so that it "orbits" the ribbon clockwise, at the same time rotating the disc clockwise, two rotations per orbit. (Not the easiest thing to explain without diagrams, but think of it as an IQ test.) The wheel turns, but the ribbon doesn't twist. Would it be easy for a living wheel to evolve something along these lines? Maybe not, but who's to say it's impossible?

Right. I don't buy this argument either. Nature has found all sorts of ingenious workarounds for seemingly insurmountable engineering problems. I don't think the problem is the inability for the system to supply nutrients to a wheel-like structure.

Which leaves us with...

Response #4: Wheels are great for roads, but they suck for natural terrain

This seems like the most likely explanation. Natural terrains are chock full of divots, rocks, crevices, and so forth. When it rains, wheels get stuck in the mud. And for those terrains that are nice and flat (e.g. ice), wheels wouldn't work all that well.

An ancillary argument that I haven't seen expressed is that limbs typically serve more than one function, besides locomotion. Whether animals are on the top or bottom of the food chain, they tend to use their limbs as weapons, either to claw or kick. Wheels probably wouldn't work quite as well for these secondary functions.

Also, the ability to fold limbs inward gives most vertebrates the ability to tuck in their limbs to protect them from attack, and also the ability to assume different postures. Wheels likely wouldn't tuck or fold as easily as limbs.

So I think the best answer is the simplest: Selection pressure in natural environments on earth does not favor wheels, and it's that reason, and not constraints on physiology or the evolutionary process, that they have not evolved.

Of course, all of the constrains, the environment, and all selection pressures can be manipulated in a virtual environment. I think it would be very cool to set up an artificial environment and test an encoding system in a situation that allowed for wheeled structures to evolve and study how it happens.

Thinking in Deep Time

So we made it to DC in one piece. I'm here, and I'm tired, but after a good night's rest I promise to try to do a decent job of blogging the CogSci conference. I've got my packet and the schedule looks like it has some interesting stuff.

Anyway, I was just reminded of one of the dumb arguments creationists use, which is when they concede that they believe in microevolution (i.e., small changes) but not macroevolution (i.e. speciation). Sure dogs and trees and mosquitos change over time...but they're still dogs and trees and mosquitos.

To this I say: learn to think in longer time spans, people.


Sure mountains and canyons and lakes erode and change shape over time. But they're still mountains and canyons and lakes, right? I mean, who has ever seen a mountain actually thrust up from the earth's crust? That must mean that mountains and canyons and lakes were created exactly as they are, right? Because we've never seen one form before our very eyes. Because one has never been created in the laboratory.

As with geological formations, so it is with biological species. It takes a while to evolve from a single-celled individual to a complex, bilaterally-symmetrical, multicellular organism. But guess what? That's what happened.

Anyway...off to a find a decent restaurant in the area.

Hitchens Discovers Blind Salamanders While Watching TV

This article, in which Christopher Hitchens thinks he has stumbled across some sort of creationist silver bullet, is pretty funny. He was watching an episode of Planet Earth, and they came to a part about cave-dwelling animals that had lost such traits as pigmentation and eyes. This is a very old example of how species evolve, not only by increasing in complexity, but as in this case, simplifying.


I guess this is Hitchens first encounter with such an idea, though you'd think if he were fairly well-versed in evolution this wouldn't come as some kind of epiphany. He dashed off an email to Richard Dawkins, and got this reply:

Vestigial eyes, for example, are clear evidence that these cave salamanders must have had ancestors who were different from them—had eyes, in this case. That is evolution. Why on earth would God create a salamander with vestiges of eyes? If he wanted to create blind salamanders, why not just create blind salamanders? Why give them dummy eyes that don't work and that look as though they were inherited from sighted ancestors? Maybe your point is a little different from this, in which case I don't think I have seen it written down before.

If Hitchens' point is that vestigial traits are evidence for evolution, then of course it's been written down...about a million times. I'm not sure what the heck Dawkins is talking about. If his point is different from this, I don't know what it is.

So as for being some kind of novel argument against creationism, I'm afraid it's as old as the debate itself. It's a good argument...but it's nowhere near new. And it's certainly not a silver bullet. Creationists can always give some version of "if god wanted to make a salamander with useless eyes, then that is what he did." This is on par with saying that the devil buried dinosaur bones, but its silliness doesn't keep people from employing it.

Realism

I finally finished reading Walter Isaacson's biography of Albert Einstein, and it was extremely good. I highly recommend it.

Isaacson does a good job of explaining Einstein's resistance to the Copenhagen Interpretation of quantum mechanics, the idea that some or all elements of the universe are in an indeterminate state that is resolved by the act of observation.


Einstein rejected this interpretation throughout his life and to his death. He held to the tenet of realism, the idea that the world is in a definite state, even in the absence of observation. Most people have heard of Schrödinger's cat, but don't realize that the ideas for such a paradox came mostly from Einstein, whose own example was a keg of gunpowder hidden from observation which paradoxically exists in an exploded and unexploded state.

But apparently the vast majority of physicists currently hold to the Copenhagen Interpretation, with a smaller number holding to a Many Worlds explanation, with a very small minority still in Einstein's camp (I can't seem to find the stats on this right now, though).

Now Einstein never said that quantum mechanics was wrong. He just said it was incomplete.

Typically we use statistics to describe the degree of certainty we have about the state of the world. If a dealer shuffles a legal deck of cards and deals them into two piles of equal size, what is the probability that the ace of spades is in pile #1? It's 0.5, right? This is a description of the extent of our knowledge about the system before observing the relevant variables. If we pick up the piles and flip through them, then we have certainty about the location of the ace, and a statistical description is no longer needed.

What's weird about quantum mechanics is that it asserts that the statistic description is the complete one. In other words, the ace is half in one pile and half in the other. It exists in an indeterminate state, simultaneously in both places, until an observation takes place.

So what Einstein was arguing was that a statistical description was still a description of the limits of our knowledge about the system, while his opponents argued that the statistical description was complete description of reality.

I'm afraid I find it very hard to swallow the Copenhagen Interpretation. I should probably defer to the majority of experts in the field, but I'm afraid I can't...at least until I get a satisfactory explanation why the statistical description should be interpreted as complete, and not as an approximation.

There are a number of experiments which are meant to verify the Copenhagen Interpretation, including the Double Slit Experiment and the Beam Splitter Experiment. Every account I've seen describes the results as "weird", but fail to give a satisfactory account of what is actually going on (at least to me...maybe I'm just being stubborn).

One of the main problems I have with an observer-defined reality is the same problem Einstein had. What constitutes an observer, or an observation? Is a piece of recording equipment an observer? Einstein asked about a mouse. If you show a mouse the readout on a piece of machinery measuring the location of a particle, but don't look yourself, does this resolve the indeterminacy of the system?

And what about the state of the world before life arose? There was a time when there were no observers. Was the world in some kind of constant state of flux? If matter really is in a completely different state before observation, then how did such a fluctuating state give rise to life in the first place?

Another thing I haven't heard explained to my satisfaction is exactly what the act of observation is supposed to do to a system. Here's another example: active and passive sonar. When something like a submarine uses active sonar, that means they generate their own sound waves, and read the information that bounces back from objects the waves bump into. With passive sonar, the submarine doesn't generate its own waves, but relies on sounds that are already bouncing around in the water.

Now I could see how active observation would alter the state of the system being observed. You're injecting a new dynamic into the system when you're bombarding it with sound waves or photons or any other active process. But how exactly does passive observation affect a system?

Maybe there are answers to these questions, and I'm just not smart enough, or haven't read enough, to wrap my head around them. Maybe nobody knows.

For now, though, I'll remain in an indeterminate state.

Evolution of Robotical Organisms

During the poster session, I met a very nice fellow by the name of Peter Krčah, who had a poster and some videos displaying his work, which is an elaboration of work done by Karl Sims back in 1994.

The work involves evolving the morphologies of artificial creatures along with neural networks that control locomotive behaviors such as walking and swimming. Peter's website is here.

Here's a video demonstrating some of his evolved individuals:



In turn, this sparked some ideas that I've been working through relating to encoding and evolving 3D morphologies. I think it's cool work, but I definitely think it can be refined into more powerful and sophisticated representations. The creatures move in a way that evokes biological movement, but their morphologies are still very blocky and artificial.

Anyway, go have a look around Peter's site. It's fun stuff.

PZ Myers' GECCO Talk

Update: Myers has posted his notes and slides.

PZ Myers gave his keynote address at GECCO on Tuesday morning, July 18th. Fortunately (or maybe unfortunately), he didn't breathe any fire, sacrifice any children by plucking out their heart, or butt-rape any consecrated crackers on stage.

Instead he gave a pretty darned interesting talk entitled "Developmental Perspectives for Understanding Evolution," which was basically an overview of evo-devo, with mention of some interesting recent experiments. The main argument of the talk, and evo-devo in general, is that you have to understand development to understand evolution.

One of the early slides had these two quotations:

"Everything is the way it is because it got that way." --D'Arcy Wentworth Thompson, 1917

"Evolution is the control of development by ecology." --Van Valen, 1973

Myers then went on to talk about how development is a hierarchical process that is essentially a series of binary decisions. Cells divide and differentiate, and when they do so they make choices that determine, and thus restrict, their fate. This whole bit reminded me of the most interesting part of the panel talk at last year's GECCO with Lewis Wolpert, Richard Dawkins, and Steve Jones. I didn't attend, but I have a link to the video here. The bit I found most interesting was when they were arguing whether or not it was possible to evolve a mouse with wings (which also has a wierd synergy with parts of Myers' talk, as we'll see later).

Anyway, then he talked a bit about epigenetics and the role of the environment in development, and talked about how the patterns that form during development are not directly encoded by the genes, but are an emergent effect of the process of development. He stressed that this doesn't mean that genes aren't important, just that the environment is important as well.

Then he talked about the opposite concepts of plasticity, which he defined (short version) as the ability of an organism to react to an internal or external change, and canalization, which are constraints which restrict the amount of change. As for the concept of evolvability, the capability of an organism to change, he said he didn't really believe in it as a distinct concept.

A discussion of toolbox genes then followed, in which Myers talked about how biologists used to assume that the difference in the DNA between organisms like a fly and a whale would be enormous, qualitatively, but as it turns out, about 80% of the genes in such animals is carrying out very similar functions, like determining where on the head the eyes are going to be constructed. The PAX-6 gene is a toolbox gene for use in building eyes, and both flies and mice (and by implication many other organisms) use this same gene for doing the same thing. Citing Putnam, Myers said that only about 15% of the genes are unique to an organism, in terms of their function.

Then he talked about genetic assimilation and accomodation (which Myers actually talks about at length here). He also mentioned the Baldwin Effect.

At this point in the talk he started to give some concrete examples, which was a nice way to illustrate a lot of the theory he'd been discussing. He talked about how a species of sheep that lives in a colder environment may, over time, evolve a thicker coat of hair. Or, perhaps if the climate is variable, the sheep evolves such that in warmer conditions it grows at thinner coat, and in colder conditions, it grows a thicker coat. The difference is between a kind of hard-wired trait, and one that is plastic, and can adapt to the changing environment.

Another example was from a 2006 paper by Suzuki and Nijhout, Evolution of a Polyphenism by Genetic Accommodation. Basically, the experimenters were able to artificially select for higher plasticity in the larva of the tobacco hornworm, Manduca sexta. They were able to manipulate the evolving population such that after only a short number of generations, individuals were able to change their color depending on temperature (go read Myers' post on it for all the details).

Then he brought up another cool example of the role of regulatory genes, from work by Chris Cretekos involving the differences in development between mouse and bat forelimbs. Myers' blogged about this work too. Go check it out. There are cool pics of bat embryos. Basically the researchers were able to remove regulatory genes that enhance the expression of the gene PRX1, which controls the growth of forelimbs. They spliced this regulatory sequence from the bat into the mouse, and measured the effects. Myers warned us that we weren't going to see images of mice with bat wings, and not to be disappointed. I was, a little. Instead, the mice grew longer forelimbs. Nothing like a bat's, but still longer. They also deleted this sequence in the mice and measured the result, and found that there was little to no effect, which suggests a fair amount of redundancy in the regulatory genes that mediate such kind of growth.

In conclusion, he reiterated that:

• Multicellular organisms are coherent arrays of patterned elements
  
• Organization is the product of interactions between genes, cells, and the environment
  
• Both evolution and development are dynamic processes, unfolding over time, not static lists of parts
  

The Q&A followed. I didn't take very good notes for that part, but one thing I noted was that he admitted up front that he wasn't that familiar with the field of evolutionary computation, which is fine. Though I wish he had attended a couple of sessions, especially those related directly to the topic of his talk. There was a tutorial on Generative and Developmental Systems that I really would have liked to have heard his reaction to. But in general I would have liked to have heard his reaction to some of the approaches and encodings, even if he wasn't familiar with the intricate details...just a critique of the overall approach.

But it was a good talk...better than most keynotes I've seen at GECCO in the past. In studying and trying to model cognition, I've developed a more profound respect for the role of time, and of trying to understand the extent to which certain cognitive functions have evolved to be highly plastic and which are less adaptive during our development. Myers spoke directly to these kinds of issues, and gave a very informative overview of many important concepts in evolutionary biology today.

Then he pulled a live baby out of his briefcase and bit its head off.

The Blended Mouse

In several of the talks here at GECCO, the issue of complexity has come up. In one talk I just listened to, the speaker discussed three general measures of complexity:

1) Shannon Entropy, or degree of uncertainty
2) Kolmogorov complexity, or the amount of description needed to fully specify something
3) Functional complexity, or how complex its interactions with the environment are

He argued that functional complexity should be kept separate from other measures.

But here's one strange example. Let's say you take a mouse. It has a high degree of regularity in its morphology. It's bilaterally symmetric, hierarchical, and highly modular. Now, if you took that mouse and dropped it in a blender, so that its molecules were distributed randomly in space, by the first two measures above, that blended mouse would be more complex than the intact mouse. However, its functional complexity would be much lower (i.e., a live mouse can run around and do lots of things that a blended mouse can't).

For the things we're interested in either understanding or engineering, I think excluding functional complexity is probably wrong.

Live From Atlanta

I made it to Atlanta in one piece. I'm here for the GECCO (Genetic and Evolutionary Computation Conference). The drive was about 10 hours. Fun. There was some congestion in Baton Rouge, but other than that it was just a long-ass, boring drive. I had Isaacson's biography of Einstein on audiobook to keep me company, and I'll probably finish it on the ride home on Wednesday.

I stopped by the Pharyngula gathering on the way in. I didn't stay long. I was exhausted from the drive. But I did get to meet PZ in person and talked to a few interesting people, including the guy who runs this site: What's the Harm? It's a site that tries to catalog accounts of people who have suffered financial ruin, bodily harm, or death at the hands of pseudoscience and religion. I haven't really had a chance to look at it much, so I can't really comment on it.

I'm staying in student housing at Georgia Tech, which is a big step up from the hostel I stayed in when I attended GECCO in 2006 in Seattle. I at least have a private room here.

I've only had a chance to attend part of a workshop session this morning, so not much to comment on yet. GECCO is the first conference I ever attended, but my research interests and focus have moved relatively far from the focus of this conference, so this will probably be the last one I attend. Still, we'll see how it goes.

TAAH Flashback: Plant a Fucking Flag

Originally posted on Thinking as a Hobby, 2/3/2003 (I just finished watching the series When We Left Earth, so this was on my mind):

Check out this blog from James Lileks (via Instapundit, whose page over the last few days has been essential reading on the Columbia disaster).

NPR had an interview with one of those people who think we should not send people into space, but rely entirely on robots. As I pulled into the parking lot at the mall he casually asked 'what can a man do on Mars that a robot cannot'?

PLANT A FUCKING FLAG ON THE PLANET, I shouted at the radio. Pardon my language. But. On a day when seven brave people died while fulfilling their brightest ambitions, this was the wrong day to suggest we all stay tethered to the dirt until the sun grows cold. Are we less than the men who left safe harbors and shouldered through cold oceans' After all, they sailed into the void; we can look up at the night sky and point at where we want to go. There: that bright white orb. We're going. There: that red coal burning on the horizon. We're going. And we're not sending smart toys on our behalf - we're sending human beings, and one of them will put his boot on the sand and bring the number of worlds we've visited to three. And when he plants the flag he will use flesh and sinew and blood and bone to drive it into the ground. His heartbeat will hammer in his ears; his mind will spin a kaleidoscopic medley of all the things he'd thought he'd think at this moment, and he'll grin: I had it wrong. I had no idea what it would truly be like. He'd imagined this moment as oddly private; he'd thought of himself, the red land, the flag in his hand, and he heard music, as though the moment would be fully scored when it happened. But there isn't any music; there's the sound of his breath and the thrum of his pulse. It seems like everyone who ever lived is standing behind him at the other end of a vast dark auditorium, waiting for the flag to stand on the ground of Mars. Then he will say something. He might stumble on a word or two, because he's only human.

But look what humans have done. Again.

Neil deGrasse Tyson on Religion and Science

Neil deGrasse Tyson is an astrophysicist and director of the Hayden Planetarium at the American Museum of Natural History. In 2005 he wrote an excellent essay entitled "The Perimeter of Ignorance" in which he talks about how historically, many famous scientists invoked god when they reach the limits of their own understanding. He begins:

Writing in centuries past, many scientists felt compelled to wax poetic about cosmic mysteries and God's handiwork. Perhaps one should not be surprised at this: most scientists back then, as well as many scientists today, identify themselves as spiritually devout.

But a careful reading of older texts, particularly those concerned with the universe itself, shows that the authors invoke divinity only when they reach the boundaries of their understanding. They appeal to a higher power only when staring into the ocean of their own ignorance. They call on God only from the lonely and precarious edge of incomprehension. Where they feel certain about their explanations, however, God gets hardly a mention.

However, he recounts the anecdote of when Laplace presented his work to Napoleon that explained aspects of the movement of bodies in the solar system that Newton's work had not sufficiently explained:

According to an oft-repeated but probably embellished account, when Laplace gave a copy [of his manuscript] to his physics-literate friend Napoleon Bonaparte, Napoleon asked him what role God played in the construction and regulation of the heavens. "Sire," Laplace replied, "I have no need of that hypothesis."

Tyson gives other examples of this phenomenon, and then talks about it in its extreme, virulent forms, such as the organized efforts of Intelligent Design.

Another practice that isn't science is embracing ignorance. Yet it's fundamental to the philosophy of intelligent design: I don't know what this is. I don't know how it works. It's too complicated for me to figure out. It's too complicated for any human being to figure out. So it must be the product of a higher intelligence.

What do you do with that line of reasoning? Do you just cede the solving of problems to someone smarter than you, someone who's not even human? Do you tell students to pursue only questions with easy answers?

There may be a limit to what the human mind can figure out about our universe. But how presumptuous it would be for me to claim that if I can't solve a problem, neither can any other person who has ever lived or who will ever be born. Suppose Galileo and Laplace had felt that way? Better yet, what if Newton had not? He might then have solved Laplace's problem a century earlier, making it possible for Laplace to cross the next frontier of ignorance.

Science is a philosophy of discovery. Intelligent design is a philosophy of ignorance. You cannot build a program of discovery on the assumption that nobody is smart enough to figure out the answer to a problem.

Here's a video of Tyson talking about these topics at Beyond Belief:



I plan on reading Tyson's Death by Black Hole soon. I'll post review(s) as I do.

Information is Never Lost?

Here's the latest installment in my apparent inability to understand the technical concept of information.

In an interview with Leonard Susskind, in which he recounts a famous dispute with Stephen Hawking, Susskind says:

Stephen Hawking once said something about black holes that apparently upset you. What was it?

Stephen said that when a bit of information falls into a black hole it is permanently lost to the outside, despite the fact that he also proved that black holes evaporate and eventually disappear. That claim touched off a crisis in physics, a clash of basic principles like no other since Einstein was young.

The problem that upset me is that the most basic principle of physics—the principle that underpins everything including classical physics, thermodynamics, quantum mechanics, energy conservation, that physicists have believed for hundreds of years—is that information is never truly lost. It can be scrambled beyond recognition, but it is never completely erased.

Hawking's claim was outrageous, but he had very good reasons for it. So good that it took more than two decades to figure out why he was wrong. And the question led to a tremendous paradigm shift in the way we think about space, time, matter, and bits of information.

But if information is "scrambled beyond recognition" isn't it lost? Maybe there's a difference between theoretically lost and practically lost.

What happens when a library burns down? Is Susskind saying that all the information written on all the pages is theoretically reconstructible from the ashes?

And Stephen Pinker is no mathematician or physicist, but I liked the definition he gave in How the Mind Works:

Information is a correlation between two things that is produced by a lawful process, as opposed to coming about by sheer chance. We say that the rings in a stump carry information about the age of the tree because their number correlates with the tree's age. The older the tree, the more rings it has. And the correlation is not an accident, but is caused by the way trees grow. Correlation is a mathematical and logical concept. It is not defined in terms of the stuff that the correlated entities are made of. Information itself is nothing special. It is found wherever causes leave effects.

So if information is the remnant of a causal relationship, then how is it that information is never created? Aren't there new cause/effect relationships happening all the time? In Pinker's tree example, isn't novel information being created as the tree grows and creates more rings? If information is simply being transformed in this case, what is the transformation?

Why Two Things Never Really Happen at the Same Time

I'm still reading and enjoying Walter Isaacson's biography of Einstein. He hasn't yet gotten to the Copenhagen interpretation of quantum mechanics and Einstein's famous quotation "God does not play dice", but you can see some foreshadowing of why he might find it so distasteful.

In describing Einstein's theory of Special Relativity, Isaacson uses the following example (let's see if I get this right...if you're interested in the subject you should definitely browse around and make sure I didn't screw up this explanation). Let's say Joe is standing at point X when two lightning bolts strike at A and B:










The light from both strikes reaches Joe at precisely the same time, so he perceives the events as occurring simultaneously. So we can safely say that lightning struck the two places simultaneously, right?

Wrong.

Bob is traveling at a very high, constant velocity toward point X. He is directly on X when the lightning strikes occur. However, the light from the lightning strike at A will reach him sooner than the light from lightning strike B because of his motion toward A and away from B. So from his frame of reference, the lightning will strike A slightly before it will strike B.

Thus, there is no such thing as absolute simultaneity. You can only describe two things happening at once in relative terms. What determines whether two things happen at the same time depends on how fast an observer (or measurement device) is traveling relative to them.

I think I got that right.

And I think this is part of the reason why Einstein had such a problem with "spooky action at a distance", the idea that two particles that are entangled could affect one another simultaneously regardless of how far apart they were. As far as I understand, this phenomenon has been tested in the lab, at very short distances, but there are criticisms. However, if two particles were entangled, and one stayed on earth while the other was transported to the moon, and then the spin of one was modified and it affected the spin of the other...that would certainly be strong evidence.

I'm looking forward to Isaacson's account of Einstein and quantum mechanics.

Teach the Controversy

Links to these t-shirts have been making the rounds for a while, but they are cool.


























































Yeah, I want one. My favorite is the devil burying dinosaur bones.

Engineering Biofuels

Fareed Zakaria interviews Craig Venter, the leader of the successful private initiative to map the human genome, about his current efforts to engineer microorganisms to produce fuels.

Zakaria: How are you going to create the fuel of the future?
Venter: We think multiple fuels of the future are going to come out of biology, by manipulating the genetic code of simple organisms to convert things like sugar or sunlight or carbon dioxide into fuels that people are very familiar with, like diesel fuel and gasoline.

...

How close are you to creating an organism that can produce fuels in this way?
We think the first fuels are maybe one to two years away. We're definitely thinking in terms of years, not decades.

You always have to take predictions with a grain of salt, but this guy has gotten results before. I hope they really are making good progress.