Of flies and free will

Posted 20 May 2007 by PvM

On Uncommon Descent Denyse comments on a paper which found how flies have 'free will'. While ID proponents are quick to argue that (Darwinian) evolutionary pathways cannot explain 'free will', I will show, in a future posting, that this is a fallacious argument based on the common appeal to ignorance found amongst ID claims. In this specific case, I will present how science explains the evolution of Levy flight patterns. In the mean time I would like to invite any ID proponent to step forward with ID's best explanation for the existence of 'Levy flight patterns' as found in these fruit flies. Denyse is making a very important concession, namely by dropping the validity of the 'specification' argument. Remember that much of ID's logic is based on the existence of similarities between animals and 'machines'. Denyse admits however that

I am not sure, however, that the researchers have discovered in flies what humans mean by free will. They have discovered something that natural philosophers have always known: Life forms, even simple ones, are not like machines.

Denyse further speculates that

The researchers had expected to find that flies behaved like computers (with natural selection presumably playing the role of the software engineer), but they did not.

Is this what the researchers had expected? Even though Levy distributions have been found to be quite pervasive in biology? She may have been confused by the MSNBC article which stated:

Brembs and his colleagues reasoned that if fruit flies (Drosophila melanogaster) were simply reactive robots entirely determined by their environment, in completely featureless rooms they should move completely randomly. To investigate this idea, the international team of researchers glued the insects to small copper hooks in completely uniform white surroundings, a kind of visual sensory deprivation tank. These flies could still beat their wings and attempt to turn.

However, the news article also mentions that

Specifically, their behavior seemed to match up with a mathematical algorithm called Levy's distribution, commonly found in nature. Flies use this procedure to find meals, as do albatrosses, monkeys and deer. Scientists have found similar patterns in the flow of e-mails, letters and money, and in the paintings of Jackson Pollock, Brembs said.

Source MSNBC However, if Levy distributions are so commonly found in nature, why does Denyse believe that they had expected otherwise? And if a mathematical algorithm can explain it, then what relevance does it have for ID's claims which are based on eliminating chance and regularities (including mathematical algorithms)? As to why ID proponents may be so gullible? I have no idea but they do seem to flock around issues such as 'global warming 'skepticism'' and other pseudo-scientific ventures. Skepticism is abandoned for sake of disagreeing with the scientific data and facts just because most scientists agree. In fact, this kind of behavior is quite predictable :-) Which makes one ask the following question: is the anti-scientific behavior of ID proponents intelligently designed or not or is it just a mechanistic, innate response? As to Augie Auer, we learn from Tim Lambert at Deltoid that

Also former Met Service chief meteorologist Augie Auer, who offers this: Prof Auer said that three quarters of the planet was ocean, and 95 percent of the greenhouse effect was governed by water vapour. "Of that remaining 5 percent, only about 3.6 percent is governed by CO2 and when you break it down even further, studies have shown that the anthropogenic (man-made) contribution to CO2 versus the natural is about 3.2 percent. "So if you multiply the total contribution 3.6 by the man-made portion of it, 3.2, you find out that the anthropogenic contribution of CO2 to the the global greenhouse effect is 0.117 percent, roughly 0.12 percent, that's like 12c in $100. "It's miniscule ... it's nothing," he said.

Actually, humans have increased the CO2 content of the atmosphere by 30%. You would have hoped that a "leading climate scientist", or a climate scientist, or even a plain old scientist of any kind would not have got something so basic so wrong. It's fun to educate ID proponents, mostly because I can predict that much of their claims can be shown to be ill-informed, that seems to be the cost of giving in to the innate response of 'fairness' rather than relying on a more informed approach.

36 Comments

Sir_Toejam · 21 May 2007

Denyse further speculates that The researchers had expected to find that flies behaved like computers (with natural selection presumably playing the role of the software engineer), but they did not. Is this what the researchers had expected?

of course not. her own ignorance of the history of animal behavior as a science "allows" her to produce a complete strawman. of course, the rubes on UD are just as ignorant, so they'll accept her strawman at face value, of course, and nobody will even try to correct her. this is why these people should be considered DAMAGING to the educational process; all they do is reinforce their own ignorance on a daily basis, and spread it to whoever will listen. people like FTK and Denyse are incapable of realizing the damage they cause. Others, like Dembski, DO, but they don't care so long as there's a payday involved somewhere along the line. I point to articles like Denyse's when I want to educate people as to why this shit incenses anyone who knows better.

snaxalotl · 21 May 2007

I'm disturbed by the paper's implication that a mechanism that doesn't have a consistent output for a consistent input is non-deterministic. This is the difference between parallel and serial logic: parallel hardware, with no feedback loops, produces the same result for the same input. Serial logic can exist in so many states that the relationship between input and output is completely unobvious, but its behavior is still deterministic. Denyse has a problem in this regard too - apparently unaware that "computer behavior" is heavily state dependent: the relationship between input and output can me made arbitrarily unapparent with more memory. Having made that quibble (maybe I'm deterministic - discussion of determinism tends to push my buttons), it's no surprise that small random variations would be amplified into large asymmetries of behavior, because game theory frequently tells us that the best strategy is a random strategy. And of course it's also no surprise when Denyse doesn't know what she's talking about.

Stevencnz · 21 May 2007

With respect to Augie Auer:
His former employer, Metservice New Zealand, published a message stating:
The MetService Chief Executive, John Lumsden, said today that the World Meteorological Organization (WMO) latest report confirms the global warming trend.

"The WMO preliminary report on the global climate for 2006 has just been released and it confirms that 2006 is set to be the sixth warmest year on record, continuing the trend of global warming. We are certain of this observation and would like to point out that the views recently made public by Augie Auer in relation to climate change are his own, and in no way do they reflect those of MetService."

Augie Auer was MetService's Chief Meteorologist until late 1998, but has had no association with the organisation since then.

MetService Chief Meteorologist, Neil Gordon, added that MetService's position on climate change is consistent with the United Nation's Intergovernmental Panel on Climate Change (IPCC) 2001 report.

"An increasing body of observations gives a collective picture of a warming world and other changes in the weather and climate system," said Dr.Gordon. "During the 20th century, the global average surface temperature has increased by about 0.60C. In the past four decades, temperatures have risen in the lowest 8 kilometres of the atmosphere. Snow cover and ice extent have decreased. Global average sea level has risen and ocean heat content has increased. Snow, wind and rain storms are costing us more."
http://www.metservice.com/default/index.php?alias=newsreleases&pr=775

a recent interview had him rubbishing someones ideas about the moon dictating weather then defending non-human climate change. Idiot interviewer let him get away with it too.

Within NZ Fed Farmers is also known as:
Deniers of Climate Change, and human origins of same
Ignorers of effects of Climate Change (e.g. drought)
Undermine science to counter it ('Fart tax', actually a fee to fund research that they agreed to)
People who kick up a stink whenever there is a climate event (e.g. drought) and demand govt money in aid

Steven

Frank J · 21 May 2007

They have discovered something that natural philosophers have always known: Life forms, even simple ones, are not like machines.
— Denyse

Poor Michael Behe. First he admits that the designer might no longer exist, and that ID is no more scientific than astrology. Now it seems that all his efforts to compare life to machines were a waste of time.

entlord · 21 May 2007

Fruitflies exhibit free will? Neither Calvin nor Luther considered such a thing, though maybe Edwards did, with his "spiders suspended over a cauldron" sermon.
Do we next start searching for a soul for flies? It seems ID is going off the edge of the table again with this one.

Richard Simons · 21 May 2007

They have discovered something that natural philosophers have always known: Life forms, even simple ones, are not like machines.

So we'll hear no more about Paley's watch, computers and outboard motors?

Prof Auer said that three quarters of the planet was ocean, and 95 percent of the greenhouse effect was governed by water vapour.

He needs to learn the difference between a response variable and a forcing variable. The amount of water vapour in the atmosphere responds quickly to the temperature whereas the amount of CO₂ responds very slowly. So a small additional amount of CO₂ increases the temperature by a little, which increases the amount of water vapour, which increases the temperature, and so on until a new equilibrium is reached. This is extremely basic climatology (it must be, otherwise I would not know it). Either he is ignorant about the whole area or he is lying.

CJO · 21 May 2007

Life forms, even simple ones, are not like machines.

She means, of course, "...not like machines, except when it's convenient to pretend they are, and then, boy howdy, does that look like one complemicated thingamadooby!!!"

bernarda · 21 May 2007

At Slate there is an interview with Daniel Dennett on the concept of "freewill".

Sometimes the interviewer seems rather dense.

http://meaningoflife.tv/video.php?speaker=dennett&topic=freewill

Jedidiah Palosaari · 21 May 2007

It seems that the authors of the study are pursuing the fallacious idea of "free will". They think that just because the fly has no external cues, it's collection of ganglia can now make it's own decisions. All they've proven is that the fly's actions are predetermined by the structural make-up of it's brain- something long argued by biology. Not everything is evironmental; some is genetic. But everything is biologically predetermined.

John Krehbiel · 21 May 2007

OK, so "Darwinism" can't explain free will.

Once again, the superstitious demand that science explain something that might not even exist!

Frank J · 21 May 2007

So we'll hear no more about Paley's watch, computers and outboard motors?
— Richard Simons

Of course we'll still hear about them. IDers never met an inconsistency they didn't like.

ega · 21 May 2007

Free will?

I didnt know it had been proven to exist.

Mike Elzinga · 21 May 2007

These pronouncements from the Discovery Institute are becoming so repetitive and predictable that one has to wonder if anyone at DI has free will.

Bjoern Brembs · 22 May 2007

Oh, and obviously, many thanks for helping us defend our science so publicly and vehemently!
Bjoern

Moses · 22 May 2007

Gee, and to think that God already knows who was going to heaven has already written their names in the big book...

Popper's ghost · 23 May 2007

Specifically, their behavior seemed to match up with a mathematical algorithm called Levy's distribution

Is O'Leary such a moron as to not understand that "machines" operate according to algorithms, not "completely randomly"? That's a rhetorical question.

Popper's ghost · 23 May 2007

Actually, it's not just O'Leary who is a moron, but whoever wrote that MSNBC piece, and Brembs as well if it accurately reflects his views: "Brembs and his colleagues reasoned that if fruit flies (Drosophila melanogaster) were simply reactive robots entirely determined by their environment, in completely featureless rooms they should move completely randomly." What the heck are "reactive robots entirely determined by their environment" ... unless "environment" includes their internal state, in which case it's a truism? Something moving "completely randomly" isn't determined at all, let alone by its environment. This statement is stupid and ignorant on numerous levels, and to say that someone "reasoned" this way is to misuse the word.

Popper's ghost · 23 May 2007

Arrgh, I should have read the MSNBC piece first. Charles Quoi is clueless, and has no business writing about science, and certainly none writing about "free will"; his writing about it is incredibly wrong-headed, and many of the people he quotes seem rather wrong-headed as well, although it's difficult to tell how much is their confusion and how much is Quoi's.

On the other hand, if one's actions were completely determined by outside factors such that no alternative existed, no one would hold that person responsible for them.

This is not "on the other hand" at all -- what a stupid false dichotomy! Of course no one's actions are completely determined by outside factors; one's actions are a function of the interplay of internal and external states. And if we decide that this interplay is wholely deterministic, does that mean we no longer hold a person, or a machine, or an inanimate object, or a process, not "responsible"? No, of course it doesn't -- mosquitos, viruses, hurricanes, earthquakes, are all responsible for their damaging effects. We hold things responsible for their causal consequences so as to properly direct our attention and to avoid consequences we don't want. If we conclude that a mass murderer doesn't have free will, does that mean that we should let them wander about freely? No, of course not; only someone incredibly stupid or incredibly intellectually dishonest would make such a suggestion.

Popper's ghost · 23 May 2007

I'm disturbed by the paper's implication that a mechanism that doesn't have a consistent output for a consistent input is non-deterministic.

Indeed; this failure to grasp very basic computational concepts is disturbing, and makes most of the paper worthless.

This is the difference between parallel and serial logic

Uh, no, it's not. Inconsistent outputs is the obvious and natural consequence of the fact that the output of a device is a function of the device's inputs and the device's state. Of course a complex "parallel" device like a human brain has a constantly shifting state, and thus identical inputs will not produce identical outputs, but even very simple "serial" systems can produce outputs that depend on internal state. Here are some simple systems: for(;;) { n = read(); print n+1; } // output is purely "reactive" to external input for(i = 1;; i++) { n = read(); print i; } // output depends entirely on internal state for(i = 1;; i++ ) { n = read(); print n ^ i; } // output depends on both external input and internal state All biological systems are like the latter. And most computer systems are too. The only reason that we are even familiar with systems that are "deterministic" in the sense of always producing the same output for the same input is because we go out of our way to design programs that behave that way, either because we are trying to solve a formally defined problem or because we want the system to be completely consistent and predictable -- and by "predictable", I mean by a casual observer, not in the Laplacian sense. Of course, living organisms with predators are "designed" not to have that sort of predictability. And programs that are intended for cryptography, or game playing, or simulation, or statistical algorithms to solve otherwise NP-complete problems, or computer art, or theorem generation, or artificial intelligence, or anything else calling for novelty or randomness, are likewise inconsistent and unpredictable (in the pragmatic, not Laplacian, sense), while being completely and utterly determined by {inner state + external inputs}.

Popper's ghost · 23 May 2007

Once again, the superstitious demand that science explain something that might not even exist!

Science is doing quite a good job of explaining the illusion of free will. See, e.g., The Illusion of Conscious Will and other books by Daniel Wegner, Professor of Psychology at Harvard University.

Popper's ghost · 23 May 2007

It is critical to emphasize at this point that the processes leading to behavioral indeterminacy may very well be deterministic: indeterministic output of deterministic systems is a well-known phenomenon [29].

If it's critical, then why so muddy it by misusing terminology, and preceding this "critical" observation with talk about noise: "According to this latter view, individual behavior is fundamentally indeterministic (not fundamentally deterministic but noisy) and precise prediction principally (not only technically) impossible"? Since behavioral unpredictability (not "indeterminacy") is well-known to be producible from deterministic systems, it isn't significant whether there is true noise involved, or whether the observed behavior is "fundamentally" (in principle) indeterministic vs. precise prediction being only "technically" (that is, in practice) unpredictable. From this, indeed critical, observation, it follows that all behavior that we deem "indeterministic" or a consequence of "free will" could be produced by a completely deterministic machine following a well-defined algorithm, destroying the metaphysical underpinnings of these concepts and demolishing the arguments of not just IDists but dualists and human- or animal- exceptionists of all stripes. There is nothing that humans or fruit flies do that robots containing computers executing algorithms, without added noise, could not do, including acting autonomously, willfully, and partially unpredictably. To "term the hypothesis that brains are deterministic input/output systems with added noise the 'robot-hypothesis' " is extremely misleading.

Bjoern Brembs · 24 May 2007

I'm not sure why my very first comment didn't get posted (the one before my thanks to Panda's Thumb), but here's another try at a different comment.

Of course, living organisms with predators are "designed" not to have that sort of predictability.
— Popper's Ghost

Unfortunately, only too few people claim that. Otherwise it would be easier for me to get funding. We are the first to actually show it experimentally. Maybe it'll be easier from now on.

From this, indeed critical, observation, it follows that all behavior that we deem "indeterministic" or a consequence of "free will" could be produced by a completely deterministic machine following a well-defined algorithm, destroying the metaphysical underpinnings of these concepts and demolishing the arguments of not just IDists but dualists and human- or animal- exceptionists of all stripes. There is nothing that humans or fruit flies do that robots containing computers executing algorithms, without added noise, could not do, including acting autonomously, willfully, and partially unpredictably.
— Popper's Ghost

Indeed. People (again, too few) have been thinking along those lines. We are the first to have experimental evidence from live, intact animals without environmental feedback loops leading to potential artefacts. It is really quite telling that the ID people have picked up on this story the way they have.

To "term the hypothesis that brains are deterministic input/output systems with added noise the 'robot-hypothesis' " is extremely misleading.
— Popper's Ghost

"Robot" in the common sense means a mindless machine that gets stuck in corners and only follows external guidance. In our paper we explicitly state that a new generation of robots should incorporate our findings. I talked to robotics specialists before, to make sure I was portraying current robotics state of the art properly in the paper. Actually, one of my esteemed colleagues re-ignited my interest in such analyses by telling me: "oh, their behavior is so predictable, they're like perfect little robots!" when he was standing in front of his poster explaining his latest results to me. So I went out to go and see just how predictable flies are. More to the point, Greene and Cohen (2004) expect neuroscience to be able to predict an individual human's behavior to about 95% accuracy (Phil. Trans. Royal Soc.). We show you can't even do that for flies. Thus, in brief, while nobody in their 'right' mind should be surprised by what we found, it is clear to me that only few neuroscientists are in this particular mind :-) Mainstream neuroscience has a completely different perspective, see references in the beginning of our paper and our extended press release. Cheers, Bjoern

Popper's ghost · 24 May 2007

My experience with neuroscientists, cognitive scientists, and AI researchers, and their literature, is very different from what you describe. As for "robot specialists", for all I know you're talking about people building automated assembly lines, or battlebots. "a mindless machine that gets stuck in corners and only follows external guidance"?? Good grief. That's not merely immensely ignorant, it's self-contradictory -- why would externally guided robots get stuck in corners? Are the people running the controls mindless? Anyone who has watched battlebots on TV knows how idiotic this claim is. The only time those robots get stuck in a corner is when some other robot has blocked it or disabled it. The robots that we expect to get stuck in corners aren't externally guided ones, but rather those that are internally guided by insufficiently sophisticated algorithms -- like my Roomba, which gets stuck all the time.

As for Greene and Cohen's paper, I don't see where they make such a claim. The closest they come is "At some time in the future we may have extremely high-resolution scanners that can simultaneously track the neural activity and connectivity of every neuron in a human brain, along with computers and software that can analyse and organize these data ... At some further point this sort of brainware may be very widespread, with a high-resolution brain scanner in every classroom. People may grow up completely used to the idea that every decision is a thoroughly mechanical process, the outcome of which is completely determined by the results of prior mechanical processes", which isn't close at all. Aside it being a hypothetical, not a prognostication, it doesn't imply anything about the predictability of behavior -- again, determinacy isn't predictability.

Your experiments and findings are useful, but it's a pity that the paper is so poorly written and the implications so badly misinterpreted by both Choi and O'Leary.

Bjoern Brembs · 24 May 2007

My experience with neuroscientists, cognitive scientists, and AI researchers, and their literature, is very different from what you describe.
— Popper's Ghost

Then please let me know when they have a job offering, because from the feedback I received it seems all the ~60-70 faculty jobs I've applied to over the past 2 years belong to the strict input/output camp. If neuroscientists are supposedly so well-versed in spontaneous activity, I also wonder why a PubMed search for "spontaneous behavior" in the title only gives 67 entries, most of which in obscure journals. I probably just used the wrong keywords... In my personal experience, the colleagues you describe of course exist, but they are, until now, in a clear minority. To back this up, you have only to read Marcus Raichle's recent papers in Science, Nature and Annu. Rev. Neurosci. They would not be so highly published if everybody already was aware of their main message. Unfortunately, there's no real statistics on this, so I must rely on circumstantial evidence. I personally wished these concepts will soon be as commonplace as you suggest, my life will be a lot easier then. :-)

Popper's ghost · 24 May 2007

it doesn't imply anything about the predictability of behavior --- again, determinacy isn't predictability

In fact, their paper contains this footnote:

We do not wish to imply that neuroscience will inevitably put us in a position to predict any given action based on a neurological examination. Rather, our suggestion is simply that neuroscience will eventually advance to the point at which the mechanistic nature of human decision-making is sufficiently apparent to undermine the force of dualist/libertarian intuitions.

So what the hell are you talking about? Are there two Greene and Cohen, Phil. Trans. Royal Soc. (2004) that contradict each other?

Popper's ghost · 24 May 2007

I personally wished these concepts will soon be as commonplace as you suggest, my life will be a lot easier then. :-)

They are quite familiar to anyone who has ever played a computer game. The claim that deterministic systems always map the same inputs to the same outputs is absurd, and a belief that this view is widely held among neuroscientists is probably due to a radical misinterpretation or misreading, such as is evident in your misreading of Greene and Cohen. And yes, "spontaneous" is the wrong word, and does not identify a valid distinction.

Bjoern Brembs · 24 May 2007

So what the hell are you talking about? Are there two Greene and Cohen, Phil. Trans. Royal Soc. (2004) that contradict each other?
— Popper's Ghost

It appears so. They clearly describe the Mr. Puppet experiment as if they believe it was possible. This thought experiment clearly involves predicting individual behavior (that Mr. Puppet was going to write a set of angry letters to his father at a certain age) and that one could do this with fairly high amount of accuracy. They go on to write: "So what is the real difference between us and Mr Puppet?" and they answer it right away: "Thus, it seems that, in a very real sense, we are all puppets. The combined effects of genes and environment determine all of our actions." (p1780) This is utter nonsense and no disclaimer can nullify it. This is like saying "physics determines the weather". Physics constrains weather, as you will agree. Of course the weather, as anything, is based on the laws of physics. That's a far cry from determined, though (unless you want to debate the use of the term "determine"). At least the actions of flies are not as determined as Greene and Cohen claim human actions to be. They have constraints, of course, but they're not determined in the strict sense of Greene and Cohen. I totally agree with you: exact individual determinism as in the Mr. Puppet example is not possible (at least in flies), as you so rightfully point out and Greene and Cohen don't seem to agree with us. So, if you read it carefully, you'll notice that their disclaimer is only about how far neuroscience will come, not what is in principle possible. The article clearly shows that they believe that Laplace-like prediction of individual behavior is in principle possible. The disclaimer only covers their rear-ends as to how far they think neuroscience can get towards there in their lifetime. If publication statistics don't even make you wonder whether your personal sample may potentially be blissfully biased, don't take my word, take Marcus Raichle's. Or maybe explain to me why most behavior is referred to as "response"? We write in our paper: "This does not necessarily imply that the same stimulus always elicits the same behavior, but that each behavior is a response to a stimulus: "Indeed, so pervasive is the basic assumption of this model that it is common to refer to any behaviour as a 'response' and thus by implication [...] assume that there must be an eliciting stimulus." [6]" The quote is from renowned Cambridge psychologist (who, of course, agrees with us on spontaneous behavior) Tony Dickinson. BTW, we obviously also don't claim that in input/output systems the same stimulus always must elicit the same response (so much for misreading others, you don't seem to be immune to that. My guess is nobody is). I wonder how such distinguished researchers (Greene, Cohen, Raichle, Dickinson) manage to publish basically opposite articles concerning something which according to you is so commonplace (I agree with you - it should be!) in the leading journals? Maybe there is a lot more controversy about the main function of brains than you think? I've now cited two authors (Dickinson and Raichle) from major peer-reviewed journals who claim (as I do) that stimulus-response (input/output) is the main model in psychology/neuroscience and one (Greene) who just takes it for granted and asks what does the law have to say about something so obvious. I'm waiting for three authors in similarly ranked publications arguing that it is at least equally commonplace to refer to the brain as an output/input system. Respectfully yours, Bjoern

Glen Davidson · 24 May 2007

I didn't read the paper, and only skimmed the comments. What I have read are summaries in various science magazines and journals, and they typically portray the results as being "chaotic" rather than either "random" or "deterministic" in the most naive sense of the word. "Free will" doesn't come up in several of the summaries.

Since there's nothing new about brains having chaotic phenomena in them, and it has long been suspected (observed, but with limited control of circumstances) that our brains are configured to avoid stereotypical responses in many circumstances (presumably through chaotic, or perhaps some random, input), there seems to be nothing new here, besides a better understanding of just how fruitflies deviate from a strict input/output mode of movement. It's merely an experiment confirming what we have long known, that organisms are hardly strictly predictable, and that we have a variety of physical phenomena which could explain this fact.

These results would be expected from evolutionary competition, as well. Of course animals trying to escape predators, or capture prey, would survive better without their movements being predictable from external inputs, so that there would be a mismatch between what "you do" and what your foe (prey, etc.) is expecting you to do. Unfortunately for "you", your opponent also utilizes (apparently) chaotic phenomena to prevent you from predicting what it will do. Are you listening Paul Nelson?

This sort of "free will" is thus predicted by evolutionary theory, provided that physical phenomena will support such unpredictability---which they do quite nicely. It does not predict metaphysically meaningful freewill, which is not at all what we see.

I suspect that "free will" comes up in the paper for two reasons. One is that the authors are naive about the concepts involved in claims of "free will". And the other is that claiming to have found "free will" in flies makes people take notice of otherwise fairly well understood unpredictability in animal movements.

Too bad for the IDists that this kind of "free will" is shared both by flies and by humans. They typically don't credit flies with anything like human "agency" (most will extent "agency" at most to dogs and the like), and portray them as complex mousetraps. True, O'Leary finally said what we've been saying all along, that organisms aren't like present-day human-made machines (oops, the analogy which makes up the entire rationale of ID blew up just then), but that was only because anything that vaguely sounds like it supports their claims is to be seized and forced into their service.

Glen D
http://tinyurl.com/35s39o

Glen Davidson · 24 May 2007

Just to clarify, I believe that the "free will" of the fruitflies is thought to have more to do with search strategy in the investigated behaviors, than with predator-prey strategies.

It's just that the classic, and probably more obvious, benefit to having unpredictable aspects in one's behaviors is exemplified by the hare dodging the hound, where there are predictable aspects and unpredictable aspects, the latter particularly involving timing.

Either way, the apparently chaotic aspects of these behaviors seem to have evolved for fairly obvious reasons.

Glen D
http://tinyurl.com/35s39o

Popper's ghost · 24 May 2007

More to the point, Greene and Cohen (2004) expect neuroscience to be able to predict an individual human's behavior to about 95% accuracy (Phil. Trans. Royal Soc.).

I understand why you're having trouble getting money or a job -- you're a liar.

Popper's ghost · 24 May 2007

It appears so. They clearly describe the Mr. Puppet experiment as if they believe it was possible. This thought experiment clearly involves predicting individual behavior (that Mr. Puppet was going to write a set of angry letters to his father at a certain age) and that one could do this with fairly high amount of accuracy.

Prediction of behavior with "fairly high amount of accuracy", at the level of writing angry letters to one's father at a certain age, is clearly nothing like saying that the same inputs always lead to the same outputs. But they do not claim that we will ever be able to do such a thing, as explicitly stated in the footnote I quoted. Here's what they actually write about "Mr. Puppet":

the idea that one could, in principle, produce a person with a particular personality and behavioural profile through tight genetic and environmental control is plausible.

That's nothing like what you claim they said. To say that it's plausible that one could, in principle, "produce a person with a particular personality and behavioural profile", is radically different from "Greene and Cohen (2004) expect neuroscience to be able to predict an individual human's behavior to about 95% accuracy" -- they state no such expectation. They said nothing about what they expect neuroscience to be able to do -- they posed a thought experiment, for the purpose of exploring what effect acceptance that people are mechanisms would have on moral judgments. This "95%" is taken from hypothetical testimony by a hypothetical lead scientist in this fanciful in principle thought experiment. Of such such scenarios as occur in The Boys from Brazil, where the exact same inputs are generated to produce the exact same behavioral outputs, they say "This is obviously a fantasy".

Let us suppose, then, that a group of scientists has managed to create an individual--- call him 'Mr Puppet'---who, by design, engages in some kind of criminal behaviour: say, a murder during a drug deal gone bad. The defence calls to the stand the project's lead scientist: 'Please tell us about your relationship to Mr Puppet. . .' It is very simple, really. I designed him. I carefully selected every gene in his body and carefully scripted every significant event in his life so that he would become precisely what he is today. I selected his mother knowing that she would let him cry for hours and hours before picking him up. I carefully selected each of his relatives, teachers, friends, enemies, etc. and told them exactly what to say to him and how to treat him. Things generally went as planned, but not always. For example, the angry letters written to his dead father were not supposed to appear until he was fourteen, but by the end of his thirteenth year he had already written four of them. In retrospect I think this was because of a handful of substitutions I made to his eighth chromosome. At any rate, my plans for him succeeded, as they have for 95% of the people I've designed. I assure you that the accused deserves none of the credit.

And yet you ascribe this "95%" to Greene and Cohen as an expectation of what neuroscientists will actually be able to do. Apparently you thought no one would read the paper (anyone who wants to see just how badly Mr. Brembs has misrepresented it can simply google "Greene and Cohen"). I gave you an opportunity to backtrack, to say you had misread the paper or confused it with a different one, but no ... Science doesn't need liars like you, sir.

Popper's ghost · 24 May 2007

BTW, even taking the fantasy thought experiment world as of Greene and Cohen as their actual expectations, Mr. Brembs gets it completely wrong. He states "Greene and Cohen (2004) expect neuroscience to be able to predict an individual human's behavior to about 95% accuracy", whereas their fantasized lead scientist says "my plans for him succeeded, as they have for 95% of the people I've designed". That 95% in successful project outcomes, not "accuracy" in predicting an individual's human behavior. It has nothing to do with predicting a person's behavior to 95% accuracy, it has to do with building them to spec so that they will exhibit desired behaviors at a gross level, such as committing "a murder during a drug deal gone bad" -- conditional on "every significant event in his life" being scripted -- not even every detail, or anything close. How does this relate to the paths of flies in featureless rooms? It obviously doesn't. Despite their "spontaneous" behavior, we can safely predict that, if we go after them with a fly swatter, they will swerve to evade it.

Henry J · 25 May 2007

Re "we can safely predict that, if we go after them with a fly swatter, they will swerve to evade it."

That reminds me of something I noticed while growing up - flies would get harder to swat each year. Previously, if a fly landed somewhere, one could walk to the other end of the house for a swatter, walk back, and the fly would still be sitting there. Swat.

A few years after that - any motion anywhere in the room, and that fly was suddenly headed somewhere else.

Yeah, I know that's anecdotal, but it still seems illustrative. We've been selecting flies for increased paranoia.

Henry

Science Avenger · 25 May 2007

Henry I've wondered the same thing with regard to mosquitoes. When I was a kid, killing flies was a challenge, whereas killing mosquitoes was just a matter of effort. They were slow. Over the years it seems that the mosquitoes have gotten better at evading us. I know its an incredibly short span of evolutionary time, but it's also a very subtle change. Is it possible they are evolving human evasion instincts?

David B. Benson · 25 May 2007

Henry J & Science Avenger --- Are you sure you are not just slowing down? :-)

Henry J · 24 June 2007

Slowing down? Slowing down? Where's your evidence for that? (Well, I mean besides it taking me a month to answer that comment...)

Henry