Free Will and Fruit Fly Behavior

I've been seeing articles popping up all over the place about a recent
PLOS article called Order in Spontaneous Behavior. The majority of
the articles seem to have been following the lead of the Discovery Institute, which claims that the article demonstrates the existence of free will, which they argue is inconsistent with naturalism and darwinism.

The thing is, the paper says nothing of the sort.

The paper did a very interesting study on the behavior of fruit-flies. They basically tethered fruit flies inside of a small cylindrical apparatus, which basically amounts to
a little tiny sensory-deprivation tank. Then they did a series of tests to measure the flies turning behavior over time in the absence of any kind of stimulus.

The resulting data basically amounted to the time periods between turns: the fly flies straight for a while, then turns, then flies straight, then turns, etc. The data consists primarily of the sequence of time periods between turns.

With that data, they proceeded to do a number of really fascinating mathematical analyses to try to understand the random/non-random elements of the turning behaviors.
The most common theories about insect brains (or in fact, animal brains in general) would suggest that they would have found two components that contributed to the behavior: a deterministic one, and a random one.

That's not what they found. The simplest statement of what they found is: the
turning behavior of a fly in a sensory deprivation tank does not match a simple
gaussian random process; nor does it match a simple combination of a deterministic process
with a simple gaussian random process.

Studying the data a bit more, they were able to find something that looked suspiciously like a Levy distribution - which is a form of power-law distribution. So they proceeded to test that - and again found that a random process didn't match their observations even using a power-law distribution.

They proceed through several other possible models and analyses, including some
interesting tests of non-linear chaotic systems - and find a kind of fractal structure to the data, but beyond that, pretty much fail to find any simple random model that really captures their observations.

So what does all that really mean? The papers authors do a great job of stating it
clearly:

We show
here that random noise cannot be the sole source of behavioral
variability. In addition to the inevitable noise component, we
detected a nonlinear signature suggesting deterministic endoge-
nous processes (i.e., an initiator) involved in generating behavioral
variability. It is this combination of chance and necessity that
renders individual behavior so notoriously unpredictable. The
consequences of this result are profound and may seem
contradictory at first: despite being largely deterministic, this
initiator falsifies the notion of behavioral determinism. By virtue of
its sensitivity to initial conditions, the initiator renders genuine
spontaneity (''voluntariness'' [30]) a biological trait even in flies.

That quote is the source of all of the endless babble around the net about this - because it refers to an "initiator" and to the idea of "spontaneity" or "voluntariness". The thing is, read in context, the "initiator" is not what bozos like the DI guys claim it is. They're talking about something in the flies brain that is neither
a perfectly deterministic process like a computer program, nor the result of random stimulus. There is something complex in the interactions of a fly's brain that isn't simply deterministic, nor simply random. But it does exhibit a great degree of order and structure - it's behavior shows the hallmarks of a non-linear dynamic system.

Basically, this means that simple models of animals brains, even applied to relatively
simple scenarios, don't work very well. Assuming that an animal brain will produce behaviors that consist of the combination of some deterministic behavior with some simple random input is, apparently, incorrect. The behavior of an animal is more complicated than that. How complicated? We're not sure. The basic simple models of randomness - gaussian systems, power-law systems, non-linear chaotic systems - all do not produce a behavior with the complex traits that we see in the data collected for this experiment. But the data is also not consistent with any simple deterministic system. So there's something complex going on.

This does not mean that the fly has some kind of "spirit" which is deciding
"Hey, I think I'll twitch now". Nor does it mean that the behavior of the fly
cannot be modeled by any deterministic process with some kind of random input. In
fact, the data suggests exactly the opposite: the strong fractal structure shown for the
data suggests that there is some combination of complex deterministic structure and
randomness. Just that it's not the kind of very simple one that we might have expected from
something as a simple as a fly.

Tags
Categories

More like this

This reminds me of the time Psychology Today published an article about cloned pigs who exhibited different behaviors. Clearly, said Psychology Today, this confirms the existence of a non-material mind since the pigs must be genetically identical in every way.

'Cept that brain development is epigenetic. (That wasn't my explanation, but was pointed out to me by some biologists after I tried to take the David Bohm approach and promptly had my ass kicked by Torbjörn.)

This is way over my head, but are they saying that the flies' turning behaviors were not random but were hard to predict? And in the absence of some kind of input that could explain it, it looked like the flies could generate their own hard to predict turns? And it seems like the ability to generate some behavior that is hard to predict might be an adaptive advantage for the flies as well as the rest of us?

Where does free will come into play in that scenario and how does any of that count against evolution and naturalism? Sounds to me like a nice (if way over my head) application of the scientific method to study natural phenomena.

Mel:

Yes, that's pretty much it. The flys behavior was not regular or deterministic - so it wasn't directly predictable. It also wasn't a function of sensory inputs. And it also wasn't just random.

Where "free will" comes into play is in shallow readings of the paper - where people read since the behavior is neither sensory nor random, it must have an initiator - and immediately glom onto that without understanding what's meant by an "initiator". They see initiator, and they think it means some conscious actor pulling a switch. *If* the fruitflies behavior can only be described in terms of a non-random initiator, and you interpret initiator as meaning a conscious actor, then from this work, you'd conclude that there must be a conscious actor controlling the fruitflies behavior. (And I thought working in Burger King in high school was a boring job!)

The anti-evolution line comes out of that as well. Creationists have this strange idea that science says everything is random. They insist that materialistic things without the intervention of consciousness are always random processes, and that consciousness is a non-material property. If that's the case, then non-randomness in fruit-fly behavior requires a conscious initiator - and a conscious initiator is a non-material agent, which can't be the result of a purely material process like evolution.

Garbage, but that's how they try to present it.

So all the article is saying is that the animal brain (specifically that of Drosophila) response is not simply random, but exhibits a fractal-like structure and suddenly this is evidence for a "spirit"? That's stretching it.

The research is interesting - I'm always interested in mathematical/computational models like neural networks and CAs.

By Brian Thompson (not verified) on 17 May 2007 #permalink

Brian:

yeah, that's pretty much it. It's a really interesting paper - I highly recommend reading it. But the whole spin that's been put on it is just completely unreasonable. What the paper says is that even behavior is not well modeled by simple random processes. The spin that's been put on it, everywhere from slashdot to CNN is that it's somehow a demonstration of free will, the existence of souls, or other crap like that which simply does not follow from the data, at all.

I don't know about you guys, but as far as I'm concerned that paper should have been titled "Evidence for the Existence of the Drosophila Soul". That's because "nonlinear system dynamics" has three big scary words in it, and it's easier to simply call it a soul. Heck, by that logic Torbjörn just became the world's foremost apologist.

Anyway, the real tactic here is to prove that fruit flies have souls, and that way the Discovery Institute can (at long last) finally prove beyond all question that biologists are genocidal maniacs.

With respect to the DI's claim: is that even within the realm of scientifically testable hypotheses? As far as I can tell, a sufficiently inscrutable pattern can be any of (i) a deterministic chaotic system beyond our ken, (ii) an authentically random phenomenon, or (iii) the product of extraphysical consciousness, i.e. "free will".

As I understand statistics, all three of these "look" random, and the only one we can really identify definitively is (i) -- once our analytical techniques and capacity to model systems catches up with the realities driving the inexplicable patterns, the patterns cease to be inexplicable. But I don't even see how we'd tell the difference in practice between (ii) and (iii); a non-deterministic response coming from a Big Bag o'Random Numbers is indistinguishable from a nondeterministic response from the non-corporeal portion of a mind. The question seems more metaphysical than physical, and doesn't strike me as scientifically testable.

Then again, even if we accept the DI's posited position of free will, I still don't see how that invalidates evolution. Free will on the individual scale looks about the same as determinism if you zoom out in time and space. That's how plenty of sciences other than biology work too. Economics, psychology, marketing, whatever... we may not understand how individual minds work and what they do, but discussing populations as a whole, whether or not the individual minds have free will is actually not particularly relevant.

Who woulda guessed that flies have souls. Heaven sure is going to be one busy and congested place.

By Gary Bohn (not verified) on 17 May 2007 #permalink

Thanks MarkCC.

I didn't read the paper, but I would have expected the history of the fly to have a huge effect on the experiment before the experiment was conducted. Simple random + deterministic components sounds too simple to me.

And it seems like the ability to generate some behavior that is hard to predict might be an adaptive advantage for the flies as well as the rest of us?

Yes, apparently Levy distributions shows up in several processes, but especially here from a beneficial search pattern that many animals use. AFAIK it goes: when looking for something, do a random local search, then move (randomly in absence of cues) a bit away, repeat. It can be used for efficiently locating food or your lost car keys.

Where does free will come into play in that scenario and how does any of that count against evolution and naturalism?

I guess earlier researcher assumed simpler animals were quite like the stimulus-response robots of behaviorism; Pavlov's conditioning, Skinner's operant conditioning, et cetera.

Instead they seem to share a basic generator for complex behavior with us. That behavior would presumably be part of what we describe as free will. At least when I'm looking for my keys I feel like I exert free will. :-)

Dustin:

[Sheepishly] Sorry about that.

I try to not get up on such high horses that it hurts much when I fall (or worse, that other persons can't reach me), but at times when I meet one of my pet peeves I can't keep from kicking them. If you get my mixed metaphor. [/Sheepishly]

Not only is the development epigenetic, synapses grows and dies as we learn and forget. In any case, taking the red pill has never worried me, since behavior is emergent from a very large and complex system. If it feels like free will, it is free will, whether it is free, constrained or programmed.

Hmm. Perhaps I am saying that the philosophic question seems backwards. Neuroscientists are trying to explore what free will is instead of worrying if it is free or not. I can get with that.

By Torbjörn Lars… (not verified) on 17 May 2007 #permalink

I believe Jake is right. Free will vs. not free will is not an experimentally testable question.

I think this fact ought to be a hint that it is not a question that matters, and yet people seem to feel it does. I assume they have plans contingent on the outcome; that is, if it's found there is free will then they'll drop out of banking and become an artist, whereas no free will means they should get a different hair-do, or something. (This is supposed to be a way of saying 1) it's not answerable; 2) the answer doesn't make a difference to anything, except how people feel and think about themselves. Furthermore, 3) even defining "free will" in the first place is a tough one. One definition could be that the existence of free will means that the initial conditions of an isolated system don't uniquely specify it's future trajectory i.e. it's not one-to-one evolution, and the particular solution obtained is selected by some irreducible "black box", which by its nature can't be simplified or understood, which we call free will. Except that you could just as well call this "black box" a random number generator, and people seem to feel that a deterministic system with a stochastic input does not count as "free will".)

By El Christador (not verified) on 17 May 2007 #permalink

I wonder if it would be possible to approach this question from the other end with GAs or similar? Include a fly and a generic predator (dragonfly, flyswat, etc), maybe throw in a random source, and see what behaviours allow the fly to dodge most effectively.

Wouldn't a creationist not want to find anything resembling free will in a fly or anything other than our species? Maybe I'm confusing the modern variety with the dumb "All I know is I didn't come from no monkey" variety I grew up around (authentic quote from my childhood Sunday school teacher, by the way), but I'd think some sort of fly adaptation that resembled free will would be disconcerting for them.

Or are they interpreting that study as showing that God is directly intervening in each fly's flight like some kind of fly air traffic controller?

Corkscrew's point is an interesting one. Any fly whose trajectory could be understood by, say, a frog, probably isn't around to be tested. The survivors' default motion behavior may be evolutionarily tuned as a form of evasive action.

But I also wonder about the sensory-deprivation idea; flies' senses work differently from humans'. How do they know the fly wasn't responding to one molecule of some interesting (to flies) scent, or air currents too small to be easily detected by humans?

"In the absence of any kind of stimulus" is an awfully strong phrase.

Sorry about that.

Nah, it's fine. I think I needed it. I'd just immediately pulled out some of Bohm's commentary on the possibility of quantum mechanical effects in the brain without first checking to see if nerve impulses operated in the quantum mechanical regime. I should have known better, since the bulk of my last two years has been devoted to learning quantum mechanics.

The flies might be responding telepathically to the experimenters, the way that lab rats have always done.

I'll bet that several teams of Complex Systems people are hacking together nonlinear emergent simulations to try to get fly-like behavior.

I'd like to suggest that single cells don't have a "combination of some deterministic behavior with some simple random input."

That was part of my 1973-1977 PhD dissertation, as explained in numerous subsequent papers: single cells operate "at the edge of chaos" (to use the Santa Fe Institute's later phrase) due to feedback and branching and hierarchy in their metabolism (even with genetic control).

I was thinking the same thing as Mel about how exactly this supports a creationist (or more specifically, literal or semi-literal Christian) viewpoint. I thought most monotheisms where pretty clear that only humans have souls/freewill (not to say they do or don't have feelings) since if animals did have free will they would be subject to the same morality humans are.

THEN (and this is the important part) I decided to try and look up the DI article saying that. I couldn't find it. If anyone can find a religious article in support of this I'd be interested in seeing it. Otherwise I think the intro paragraph might be a little misleading.

Additionally, while looking for that or any other ID article I did run into this ( http://brembs.net/spontaneous/ ) press release, apparently hosted on one of the authors sites, titled "Do fruit flies have free will?" and with such quotes as

"The results of our analysis indicate a mechanism which might be common to many other animals and could form the biological foundation for what we experience as free will".

by the authors. (shorter, text only version http://www.eurekalert.org/pub_releases/2007-05/plos-dff051007.php )

So I think it might actually be the title and content of that press release, not the DI institute or Marks quoted paragraph which is causing some of the confusion.

Torbjörn: "At least when I'm looking for my keys I feel like I exert free will."

Really? Philosophically, I believe in free will, but I'm fully aware that most of the time I'm just surfing along on my sensory input and previous decisions.

Reading this post, the first thing that occurred to me was that, with further research and analysis, a scientific explanation for this phenomenon might likely be found, and if so, would not those making those claims have to admit that science could then explain what they call "free will" to remain consistent?

By John Morales (not verified) on 17 May 2007 #permalink

I think scientists should be required to use socially relevant bugs in their research. Like mosquitoes. Now if mosquitoes had free will, that would be scary.

#11: "I think this fact ought to be a hint that it is not a question that matters, and yet people seem to feel it does. "

I've started thinking along similar lines, that 'free will', at least as defined by lots of fundie types, is an ill-posed concept that doesn't have meaning. Like it or not, we all interact very closely with our environment, and can't be readily extracted from it. Our 'most free' choices in life are those in which there are no consequences to the choice, or no obvious consequences, like the fruit fly experiment. "The freedom to make irrelevant choices" seems like a pretty sad definition of free will.

Hey, great post and great discussion ver here. Let me just quickly address two issues that somehow struck a chord:
### "In the absence of any kind of stimulus" is an awfully strong phrase.
Research using our setup (see extended press release) over the last 40something years has yielded an incredible amount of insight into the stimuli which guide flight behavior. So we know very well what the fly uses for course control and we have eliminated it. Of course, it is never possible to exclude any unknown stimuli (earth radiation, moon position, aliens, invisible pink unicorns, anything really :-).
However, since these stimuli arrive at the fly at random, the fly's reaction to them should reflect their stochasticity. And this we have excluded as well.
In summary, one would have to make fairly unreasonable assumptions to keep environmental stimuli in the picture as the causing agents for our results.

###
I'd like to suggest that single cells don't have a "combination of some deterministic behavior with some simple random input."
That was part of my 1973-1977 PhD dissertation, as explained in numerous subsequent papers: single cells operate "at the edge of chaos" (to use the Santa Fe Institute's later phrase) due to feedback and branching and hierarchy in their metabolism (even with genetic control).
###

Actually, it seems that our results suggest that fly brains "operate at the edge of chaos" and we actually use that exact phrase in the paper, I think. The reason is that we believe that feedback and branching and hierarchy are deterministic properties of brains. These properties are of course influenced by the genetic make-up of the fly and by its experience. I speculate that it's these factors which influence when, where and to what degree the noise you always have in such a system will have any influence on the overall output of the system (more details outlined here). Hence, a tight interaction between noise and a nonlinear skeleton which leads to the nonlinear signature the S-Map procedure detected.
The ID post over at UD appears to suggest that they don't believe the initiator is a brain function, as we do, but that it is some non-materialistic entity or something.
Again, awesome post, Mark, and great discussion.
Cheers,
Bjoern

I think scientists should be required to use socially relevant bugs in their research. Like mosquitoes. Now if mosquitoes had free will, that would be scary.

Drosophila are relevant... they're one of the best model species out there, and share almost all of their genes with all of the other insects out there.

Anyway, mosquitos are vampires, and so undead. All of them are mindless drones except the master mosquito in Transylvania.

"It also wasn't a function of sensory inputs. "

This seems like a huge hole in their paper. Sensory deprivation is NOT the same thing as "no sensory inputs." You CANNOT perfectly eliminate sensory inputs, and most nervous systems in such situations actually end up creating hallucinatory impressions.

Fun story: when someone was put in the noise canceling chamber at Bell Labs, he came out and told the techs that it was broken: he had heard two distinct sounds in there. The tech smiled and responded: "the high pitched one was your nervous system, and the low pitched on was your circulatory system."

#22: "Anyway, mosquitos are vampires, and so undead. All of them are mindless drones except the master mosquito in Transylvania."

What!? Why does my hometown spend so much money on mosquito spraying, then? Can't we just send an expedition out to Transylvania with a toothpick to kill the master?

That was a really nice experiment. It's simple, repeatable, reliable and generates and can generate an enormous amount of objective data for analysis. Conclusions about fuzzy things like "free will" will always be ill-defined, but their analysis of the data was impeccable.

I'm pretty sure this one will be one of my own personal all time favorite experiments list. With Mendel's pea experiment, that makes two from biology.

By ObsessiveMathsFreak (not verified) on 18 May 2007 #permalink

#22. I'm glad you agree (or at least have tendencies in the agreeing direction) with my opinion that whether or not free will exists is an irrelevant question.

Your further comments suggest that your definition of "free will" is not the one I use, so, for further clarity, I will attempt to say what I mean when I say "free will". Lest you be agreeing with something you don't actually agree with, I mean...

I mean it in the sense of what I believe is also called "hard determinism", that is, 'is the future evolution of the universe already determined?'. The idea being that all phenomena (as far as anyone knows, with the noteworthy exception quantum measurement which still isn't fully understood, AFAIK) are described by deterministic physical laws, in which the state of an (isolated) system given at one time specifies its state at all future times. This incorporates "environmental" effects because you just draw your boundary big enough to include them, making it the entire universe if necessary.

My point being that my interest isn't really in whether we can isolate from an environment or not, that being irrelevant (IMHO), but whether the system is ultimately deterministic in the sense that its state at one time specifies its state at all future times. Or, put differently, 'do initial value problems with that equation of motion (i.e. the true physics of the universe) have a unique solution?'

The key part of the above is that, if the universe does obey such a deterministic law, then people's brains, and their decision-making machinery and so on, are also part of this deterministic machinery, so, someone with enough initial information about the universe, long ago, could (in principle) predict all the decisions anyone will make. Naturally, you wouldn't be able to feel this absence of free will, it wouldn't feel like mind control or anything, because it's not your mind being controlled by an "external" force, it's that your mind is a predictable, in-principle-knowable-in-advance phenomenon. There's no reason you would expect to be able to tell it from the "inside" of a mind.

I think "does free will exist?" is an irrelevant question because knowing the answer doesn't really change anything. I mean, if that's the case, then we've always not had free will and it hasn't been presenting any problems or bothering us, or impinging on our daily lives in any way. So if we discover there isn't free will, then carry on, as you were before, etc. etc. Or, put differently, "it feels like free will, so close enough". It's a good enough illusion, so accept it on its own terms.

By El Christador (not verified) on 18 May 2007 #permalink

#26: I have absolutely no arguments or criticism with your view. Your initial comment just tripped some weird, foofy, philosophical neuron in my brain that's been misfiring for some time.

My impression is that the religious and philosophical types picture 'free will' as some sort of 'external force' calling the shots in terrestrial affairs. My conclusion, though, is that those 'shots' would still be based on terrestrial observations and experience - where else would they come from? In the end, even if you had a free will, the choices 'it' made would be very similar if not identical to the choices some optimized deterministic problem solver would make. The only time your choices would seem to be truly 'free' would be those situations when you lack information to make an informed decision or when the choices have absolutely no significance, in which case it may as well be random.

Anyway, my thoughts aren't that well formulated, but maybe that's the point I'm getting at: when it comes to the philosophical/spiritual impression of 'free will', the concept is not that well defined and can lead to odd, even contradictory conclusions. And if I can't tell someone what it is, and how it differs from a deterministic view, what's the point?

Another thought: even if the fruit fly paper did demonstrate the existence of non-determinism in nature (which, obviously , it really, really, really doesn't), it still wouldn't say anything about the DI's view of 'free will'. Maybe it would demonstrate the 'free' part, i.e. external degrees of freedom in the system, but says nothing about those degrees being guided by a 'will.'

Okay, I'm done rambling - I get myself in so much trouble when I try and philosophize...

Really? Philosophically, I believe in free will, but I'm fully aware that most of the time I'm just surfing along on my sensory input and previous decisions.

Houston, we have a problem. "Free will" is a folk psychology concept, a philosophical concept, and, perhaps, a subject in neuroscience.

I was using the folk psychology concept, since that is what I think it is.

Because I agree that we are really surfing along and always postdicting a model for what we just did. (The modeling after the event is AFAIK established fact.) Seems pretty lame to attribute 'free will' to such a backwards process.

By Torbjörn Lars… (not verified) on 18 May 2007 #permalink

"My impression is that the religious and philosophical types picture 'free will' as some sort of 'external force' calling the shots in terrestrial affairs"

I grew up among the fundies and that's definitely my impression of them. I think the primary reasons they care about "free will" are: (1) it is a back door way of talking about "soul" and (2) their core dogma includes "choosing to accept Jesus."

So when they think they hear science say something about free will, they perk up and say: (1) I told you I'm not just a natural process--I'm a soul! I transcend this icky (sinful) physical world; and (2) Ha! Science just proved what I've been telling you about Jesus all along. No need to actually read the paper. It is just confirmation bias and hyper-religiosity in the clinical sense coming together to maintain what looks from the outside like a pretty delusional way of thinking.

By Anonymous (not verified) on 19 May 2007 #permalink

A cautionary note upon reading the many replies.

Yes, we should assume no supernatural magickry in explaining the apparent puzzles of consciousness, decisionmaking, free will, etc.

Beyond that, there simply is not much understood by science as of yet regarding how the mind works. We just don't know much yet.

So while we can clearly slap away hypotheses of magickal causes for mind activity, there's no justification for throwing our current paltry understanding in as though it's all already been studied & understood, it hasn't, not even close.

what looks from the outside like a pretty delusional way of thinking.

Delusional systems never look like delusions from the inside.

By Caledonian (not verified) on 19 May 2007 #permalink

there was also a study where they had participants choose their left or right hand, and they were able to control which by electrically stimulating the brain. however all participants said they freely choose the hand

By Anonymous (not verified) on 19 May 2007 #permalink

great review of that paper...and interesting discussion...thanks!

It has always seemed to me that the anti-evolutionists simply cannot accept the nature of emergent phenomenon in complex systems within their worldview. As evidenced by Behe's 'irreducible complexity' argument. At least in my (simple) mind it all boils down to this. Life, itself, is an emergent phenomenon...so, perhaps they cannot accept anything along those lines.

Unless the emergent phenomenon is free-market capitalism, which, of course, is a self-organizing system that needs no central authority (i.e., intelligent designer).

Not to vindicate any stupidity in the reporting of this but: there are philosophers who've argued for a non-linear dynamics account of free will. Presumably the idea behind this experiment is that a fly in a sensory deprivation tank will experience random sensory stimulus and the behaviour exhibited will then be a simple product of randomness and the deterministic machinations of the nervous system. I believe the non-linear dynamics approach to free will has something to do with some non-linear systems being mathematically equivalent to probabilistic systems, so you get some of the irreducible uncertainty that's the hallmark of free will. So if that's what they found it's quite interesting. I'm not sure if the result of these experiments point in that direction (or if the philosophy is sound).

Of course, one of the problems with the whole free will debate is defining "free will" in the first place.

I'm kinda disappointed nobody has responded to my challenge to find any anti-evolutionist article claiming this study supports there arguements. Don't get me wrong I don't support creationism (i'm an Athiest with a capital A, although not an evangelical one), but given that a large thrust of this article is that creationist are misinterpreting it I think either an example of them doing so or a retraction or response would be appropriate. If we distort there arguements we are no better than they are.

Finally, on the topic of what is free will, I think this quote from the press release I linked to earlier sums it up nicely.

Our subjective notion of "Free Will" is an oxymoron: the term 'will' would not apply if our actions were completely random and it would not be 'free' if they were entirely determined. So if there is free will, it must be somewhere between chance and necessity - which is exactly where fly behavior comes to lie. "The question of whether or not we have free will appears to be posed the wrong way," says Brembs. "Instead, if we ask 'how close to free will are we"' one finds that this is precisely where humans and animals differ".

I just returned to this very interesting thread and saw that somehow my comment either never got posted or was somehow deleted. I don't have a copy of it anymore, so I can't really re-post it.
The ID comments on our paper are of course ludicrous, but very akin to their stand on evolution so rather consequential to their thinking. Basically all our paper showed was that behavioral *variability* comes from a nonlinear property of the brain and not from system noise. This means there must be an evolutionary advantage for brains to be more variable in their output than they could be, given the principal constraints of the individual neurons' physiology. There are many evolutionary advantages for injecting variability in behavior. What remains is to find out a) how does the fly brain generate this variability, b) is that mechanism related to how we do it and c) what happens if we experimentally block this mechanism from occurring?
There are a number of reasons (too many for this post) why one may feel inclined to speculate that this adaptive injection of unpredictability into almost any behavior could be one of the processes which interact to attribute agency to our actions and thereby facilitate the evolution of a strong subjective notion of "free will".
If this scenario were to actually substantiate, then the *folk notion* of free will would indeed exist as a biological mechanism.
That's my humble opinion and it may be proven wrong by any of my future experiments (if I get funded, which is entirely unpredictable).
Bjoern :-)

somehow my comment either never got posted or was somehow deleted

A pity, really.

And it is always good to see researchers describe their research for the rest of us.

my future experiments (if I get funded, which is entirely unpredictable).

Also, as in all research, the best result is really the unpredicted. :-|

That said, it would be nice if "free will" can be subsumed under biology. In any case I look forward to see a description over the mechanism in question.

And the blocking experiment would be totally cool. From fly to "robo-fly"?!

By Torbjörn Lars… (not verified) on 21 May 2007 #permalink

The "bozos" describing the initiator as free will are the study's authors.

It's uncool that Mark won't correct himself.

By Herb West (not verified) on 21 May 2007 #permalink

undergroundman:
I challenge you to define free will sufficiently rigorously such that you can properly show that Conway's argument is incorrect.

Considering that such a task has never been done in the history of mankind, you've got your work cut out for you.

By Xanthir, FCD (not verified) on 24 May 2007 #permalink

It's not necessary to provide an exact definition of "free will" to demonstrate that Conway and Kochen's theorem doesn't work, because the definitions they use are too loose. As far as I can see, the theorem is intended to prove that "A person's actions are not determined by their history if and only if the actions of a particle are not determined by past history." Nothing in the theorem suggests that this non-determinism is the same sort of non-determinism, however. Their definition covers both the simplest kind of randomness and the most complex sort of free will theories, with no way to experimentally decide between them.

By Matthew L., FCD (not verified) on 25 May 2007 #permalink

I agree with undergroundman and Matthew L., and Xanthir nailed it.

Conway and Kochen showed something, but "free will" it ain't. Conway is a very precise, crisp thinker. Rather than blurry and philosophically clueless, I'd prefer to think that he was giving the theorem a title that was somewhere between an analogy and a joke.

Conway has a strange sense of humor. Surreal numbers are presented in a funny, fictional way. The "Game of Life" is not really pretending to be about biology, or human life, but only a kind of broad analogy about emergent behavior. His combinatorial game "Atomic Weight Calculus" is not really about atomic weight in a Chemistry sense. Many other examples could be given.

Funny guy, at the deepest level of abstract humor. And a genuine genius, sui generis.

From the article Methods:Experiment:Paragraph2:Sentence 3: "Any potential auditory stimuli are uncontrolled and bear no correlation to the fly's behavior."

Does they qualify that anywhere? Is it common knowledge with fruit flies?

By Anonymous (not verified) on 28 May 2007 #permalink

My definition: If people have free will, then their actions are determined not by past material occurrences but by their own metaphysical soul. Free will as people tend to think of it is reliant upon a sort of dualism. Actions are not "non-determined", but determined by the some spiritual soul-substance.

Defining or comprehending free will is not so difficult; it's reconciling free will (and the moral responsibility it implies) with the physical determinism that we see all around us without becoming nihilistic that is difficult.

I had trouble following Conway's proof, probably because I haven't really studied quantum mechanics, but also because he didn't do a very good job at it. I suspect that he's getting senile in his old age.

When it was over someone challenged him to act freely, so he dropped a quarter on the floor. Some of us chuckled and pointed out that his action had been caused by the question, and he just kinda sighed and said that the idea that we are all just automatons is too much for him to handle.

I would have been happier if he had called it the "Non-determinism Theorem" or something. Non-determinism != free will. If I act randomly than I am not choosing my actions.

Defining or comprehending free will is not so difficult

I don't think so - if "free will" could be given an operative definition perhaps neuroscience could study it.

Look at the circumscribed nature of an author's comments here: "this adaptive injection of unpredictability into almost any behavior could be one of the processes which interact to attribute agency to our actions and thereby facilitate the evolution of a strong subjective notion of "free will". If this scenario were to actually substantiate, then the *folk notion* of free will would indeed exist as a biological mechanism."

But I agree that determinism and/or non-determinism isn't the issue between folk psychology and the biology of evolution. (And FWIW, IMO QM combines both, sort of like how it combines continuity with discreteness.)

By Torbjörn Lars… (not verified) on 30 May 2007 #permalink

Blarg. I haven't chimed in much to this subject, but I feel compelled to now because I've been thinking of it for a bit.

I think that an acceptable definition of free will that avoids dualism is that your actions are at least partially unpredictable by simple models but not random.

The definition of 'simple' is up for grabs, and is essentially defined by our understanding and processing power, so that's not great. I would prefer a static definition that doesn't classify things as free or not based on whether or not you're using the latest supercomputer.

However, it does point to some interesting directions. Namely, that behavior must be chaotic (in the mathematical sense) before one can even think about calling it 'free'. This is pretty much what the authors were saying - the flies have 'free will' because their actions are determined by a chaotic (deterministic) process, rather than a simple deterministic or a random process.

One may be able to go further by defining free will in such a way that it involves intractability, in the computer science sense. That way, barring the revelation that P=NP or something of that nature, the boundary between things with free will and those that are deterministic would be static and well-defined.

Of course, there is the very high possibility that all of this would end up imputing free will to naturally chaotic systems like hurricanes and such. ^_^

By Xanthir, FCD (not verified) on 30 May 2007 #permalink

It's sometimes struck me that one of the issues in discussing free will is that we generally use it in two very different senses.

One is the sense that we are not purely mechanistic or deterministic in our behaviour, and that sense seems to me to correspond pretty closely with randomness, though some people prefer a supernatural (non)explanation.

Second is the sense that "You can't tell me what to think", of the "You are not the boss of me" sense. I'm not sure that's even connected to the determinism question.

By Stephen Wells (not verified) on 05 Jun 2007 #permalink