[cross posted from UD Siding with Mathgrrl on a point, and offering an alternative to CSI v2.0, special thanks to Dr. Liddle for her generous invitation to cross post]
There are two versions of the metric for Bill Dembski’s CSI. One version can be traced to his book No Free Lunch published in 2002. Let us call that “CSI v1.0”.
Then in 2005 Bill published Specification the Pattern that Signifies Intelligence where he includes the identifier “v1.22”, but perhaps it would be better to call the concepts in that paper CSI v2.0 since, like windows 8, it has some radical differences from its predecessor and will come up with different results. Some end users of the concept of CSI prefer CSI v1.0 over v2.0.
It was very easy to estimate CSI numbers in version 1.0 and then argue later whether the subjective patterns used to deduce CSI were independent and not postdictive. Trying to calculate the CSI in v2.0 is cumbersome, and I don’t even try anymore. And as a matter of practicality, when discussing origin-of-life or biological evolution, ID-sympathetic arguments are framed in terms of improbability not CSI v2.0. In contrast, calculating CSI v1.0 is a very transparent transformation going from improbability to taking the negative logarithm of probability.
I = -log2(P)
In that respect, I think MathGrrl (who’s real identity he revealed here) has scored a point with respect to questioning the ability to calculate CSI v2.0, especially when it would have been a piece of cake in CSI v1.0.
For example, take 500 coins, and suppose they are all heads. The CSI v1.0 score is 500 bits. The calculation is transparent and easy, and accords with how we calculate improbability. Try doing that with CSI v2.0 and justifying the calculation.
Similarly, with pre-specifications (specifications already known to humans like the Champernowne Sequences), if we found 500 coins in sequence that matched a Champernowne Sequence, we could argue the CSI score is 500 bits as well. But try doing that calculation in CSI v2.0. For more complex situations, one might get different answers depending on who you are talking to because CSI v2.0 depends on the UPB and things like the number possible primitive subjective concepts in a person’s mind.
The motivation for CSI v2.0 was to try account for the possibility of slapping on a pattern after the fact and calling something “designed”. v2.0 was crafted to try to account for the possibility that someone might see a sequence of physical objects (like coins) and argue that the patterns in evidence were designed because he sees some pattern in the coins somewhat familiar to him but no one else. The problem is everyone has different life experiences and they will project their own subjective view of what constitutes a pattern. v2.0 tried to use some mathematics to create at threshold whereby one could infer, even if the recognized pattern was subjective and unique to the observer of a design, that chance would not be a likely explanation for this coincidence.
For example, if we saw a stream of bits which some claims is generated by coin flips, but the bit stream corresponds to the Chapernowne sequence, some will recognize the stream as designed and others will not. How then, given the subjective perceptions that each observer has, can the problem be resolved? There are methods suggested in v2.0, which in and of themselves would not be inherently objectionable, but then v2.0 tries to quantify how likely the subjective perception can arise out of chance and then it convolves this calculation with the probability of the objects emerging by chance. Hence we mix the probability of an observer concocting a pattern in his head by chance and then mixing it with the probability an event or object happens by chance, and after some gyrations out pops a CSI v2.0 score. v1.0 does not involve such heavy calculations regarding the random chance an observer formulates a pattern in his head, and thus is more tractable. So why the move from v1.0 to v2.0? The v1.0 approach has limitations which v2.0 does not. However, I recommend, that when v1.0 is available to use, use v1.0!
The question of post diction is an important one, but if I may offer an opinion — many designs in biology don’t require exhaustive rigor as attempted in v2.0 to try to determine if our design inferences are postdictive (the result of our imagination) or whether the designed artifacts themselves are inherently evidence against a chance hypothesis. This can be done using simpler mathematical arguments.
For example, consider if we saw 500 fair coins all heads, do we actually have to consider human subjectivity when looking at the pattern and concluding it is designed? No. Why? We can make an alternative mathematical argument that says if coins are all heads they are sufficiently inconsistent with the Binomial Distribution for randomly tossed coins, hence we can reject the chance hypothesis. Since the physics of fair coins rules out physics as being the cause of the configuration, we can then infer design. There is no need in this case to delve into the question of subjective human specification to make the design inference in that case. CSI v2.0 is not needed to make the design inference, and CSI v1.0, which says we have 500 bits of CSI, is sufficient in this case.
Where this method (v1.0 plus pure statistics) fails is in questions of recognizing design in a sequence of coin flips that follow something like the Champernowne sequence. Here the question of how likely it is for humans to make the Champernowne sequence special in their minds becomes a serious question, and it is difficult to calculate that probability. I suppose that is what motivated Jason Rosenhouse to argue that the sort of specifications used by ID proponents aren’t useful for biology. But that is not completely true if the specifications used by ID proponents can be formulated without subjectivity (as I did in the example with the coins) 🙂
The downside of the alternative approach (using CSI v1.0 and pure statistics) is that it does not include the use of otherwise legitimate human subjective constructs (like the notion of motor) in making design arguments. Some, like Michael Shermer or my friend Allen MacNeill, might argue that we are merely projecting our notions of design by saying something looks like a motor or a communication system or a computer, but the perception of design is owing more to our projection than to an inherent design. But the alternative approach I suggest is immune from this objection, even though it is far more limited in scope.
Of course I believe something is designed if it looks like a motor (flagellum), a telescope (the eye), a microphone (the ear), a speaker (some species of bird can imitate an incredible range of sounds), a sonar system (bat and whale sonar), a electric field sensor (sharks), a magnetic field navigation system (monarch butterflies), etc. The alternative method I suggest will not directly detect design in these objects quite so easily, since the pure statistics are hard pressed to describe the improbability of such features in biology even though it is so apparent these features of biology are designed. CSI v2.0 was an ambitious attempt to cover these cases, but it came with substantial computational challenges to arrive at information estimates. I leave it to others to calculate CSI v2.0 for these cases.
Here is an example of using v1.0 in biology regarding homochirality. Amino acids can be left or right handed. Physics and chemistry dictate that left-handed and right-handed amino acids arise mostly (not always) in equal amounts unless there is a specialized process (like living cells) that creates them. Stanley Miller’s amino acid soup experiments created mixtures of left and right handed amino acids, a mixture we would call racemic (a mix of right and left-handed amino acids) versus the homochiral variety (only left-handed) we find in biology.
Worse for the proponents of mindless oirgins of life, even homochiral amino acids will racemize spontaneously over time (some half lives are on the order of hundreds of years), and they will deanimate. Further, when Sidney tried to polymerize homochiral amino acids into protoproteins, they racemized due to the extreme heat and created many non-chains, and the chains he did create had few if any alpha peptide bonds. And then in the unlikely event the amino acids polymerize, in a soup, the amino acids can undergo hydrolysis. These considerations are consistent with the familiar observation that when something is dead, it tends to remain dead and moves farther away from any chance of resuscitation over time.
I could go on and on, but the point being is we can provisionally say the binomial distribution I used for coins also applies to the homochirality in living creatures, and hence we can make the design inference and assert a biopolymer has at least -log2(1/2^N) = N bits of CSI v1.0 based on N stereoisomer residues. One might try to calculate CSI v2.0 for this case, but me being lazy will stick to the CSI v1.0 calculation. Easier is sometimes better.
So how can the alternative approach (CSI v1.0 and pure statistics) detect design of something like the flagellum or DNA encoding and decoding system? It cannot do so as comprehensively as CSI v2.0, but v1.0 can argue for design in the components. As I argued qualitatively in the article Coordinated Complexity – the key to refuting postdiction and single target objections one can formulate observer independent specification (such as I did with the 500 coins being all heads) by appeal to pure statistics. I gave the example of how the FBI convicted cheaters of using false shuffles even though no formal specifications for design were asserted. They merely had to use common sense (which can be described mathematically as cross or auto correlation) to detect the cheating.
Here is what I wrote:
The opponents of ID argue something along the lines: “take a deck of cards, randomly shuffle it, the probability of any given sequence occurring is 1 out of 52 factorial or about 8×10^67 — Improbable things happen all the time, it doesn’t imply intelligent design.”
In fact, I found one such Darwinist screed here:
Creationists and “Intelligent Design” theorists claim that the odds of life having evolved as it has on earth is so great that it could not possibly be random. Yes, the odds are astronomical, but only if you were trying to PREDICT IN ADVANCE how life would evolve.
http://answers.yahoo.com/question/index?qid=20071207060800AAqO3j2
Ah, but what if cards dealt from one random shuffle are repeated by another shuffle, would you suspect Intelligent Design? A case involving this is reported in the FBI website: House of Cards
In this case, a team of cheaters bribed a casino dealer to deal cards and then reshuffle them in same order that they were previously dealt out (no easy shuffling feat!). They would arrive at the casino, play cards which the dealer dealt and secretly record the sequence of cards dealt out. Thus when the dealer re-shuffled the cards and dealt out the cards in the exact same sequence as the previous shuffle, the team of cheaters would be able to play knowing what cards they would be dealt, thus giving them substantial advantage. Not an easy scam to pull off, but they got away with it for a long time.
The evidence of cheating was confirmed by videotape surveillance because the first random shuffle provided a specification to detect intelligent design of the next shuffle. The next shuffle was intelligently designed to preserve the order of the prior shuffle.
Biology is rich with self-specifying systems like the auto correlatable sequence of cards in the example above. The simplest example is life’s ability to make copies of itself through a process akin to Quine Computing. Physics and chemistry makes Quine systems possible, but simultaneously improbable. Computers, as a matter of principle, cannot exist if they have no degrees of freedom which permit high improbability in some of its constituent systems (like computer memory banks).
We can see the correlation between a parent organism and its offspring not being the result of chance, and thus we can reject the chance hypothesis for that correlation. One might argue that though the offspring (copy) is not the product of chance, the process of copying is the product of a mindless copy machine. True, but we can further then estimate the probability of randomly implementing particular Quine computing algorithms (that makes it possible for life to act like computerized copy machines). The act of a system making copies is not in-and-of-itself spectacular (salt crystals do that), but the act of making improbable copies via an improbable copying machine? That is what is spectacular.
I further pointed out that biology is rich with systems that can be likened to login/password or lock-and-key systems. That is, the architecture of the system is such that the components are constrained to obey a certain pattern or else the system will fail. In that sense, the targets for individual components can be shown to be specified without having to calculate the chances the observer is randomly formulating subjective patterns onto the presumably designed object.
That is to say, even though there are infinite ways to make lock-and-key combinations, that does not imply that emergence of a lock-
and-key system is probable! Unfortunately, Darwinists will implicitly say, “there are infinite number of ways to make life, therefore we can’t use probability arguments”, but they fail to see the errors in their reasoning as demonstrated with the lock-and-key analogy.
This simplified methodology using v1.0, though not capable of saying “the flagellum is a motor and therefore is designed”, is capable of asserting “individual components (like the flagellum assembly instructions) are improbable hence the flagellum is designed.”
But I will admit, the step of invoking the login/password or lock-and-key metaphor is a step outside of pure statistics, and making the argument for design in the case of login/password and lock-and-key metaphors more rigorous is a project of future study.
Acknowledgments:
Mathgrrl, though we’re opponents in this debate, he strikes me a decent guy
NOTES:
The fact that life makes copies motivated Nobel Laureate Eugene Wigner to hypothesize a biotonic law in physics. That was ultimately refuted. Life does copy via a biotonic law but through computation (and the emergence of computation is not attributable to physical law in principle just like software cannot be explained by hardware alone).
Sal,
Maybe that’s just sloppily written, but if you have 500 flips of a fair coin that all come up heads, given your qualification (“fair coin”), that is outcome is perfectly consistent with fair coins, and as an instance of the ensemble of outcomes that make up any statistical distribution you want to review.
That is, physics is just as plausibly the driver for “all heads” as ANY OTHER SPECIFIC OUTCOME.
You must have been trying to make a different point than how this came out. Are you trying to contrast an “all heads” outcome as a matter of chance with a competing hypothesis that the coins are “not fair” after all, as you stated they were, and that human or other interference has “rigged” the outcome?
So how is this not a rewording of Uprightbiped’s argument?
There are many configurations that have half-heads, half-tails, and many more that approximately have that ratio.
But “all coins heads” is statistically special from an operational standpoint.
Consider that a homochrial protein will racemize over time. Even from a physics and chemistry standpoint, there is a measurable evolution in the distribution of all left-handed to a racemic mix. So it is not just an artifact of our making a homogenous configuration special, it has physical consequence over time that can be empirically measured.
To say that what I offered is sloppy thinking, then one can say that noticing the difference between homochiral and racemic states is just an artifact of our intellectual prejudices.
The coin analogy has obvious relation to homochirality in biology.
Regarding the special state of all coins heads or homochirality, within statistical physics, there is the notion of system equilibrium. If the coins are periodically subject to disturbances, they will reach an equilibrium condition where the ratio tends to be “racemic”. Same with homochiral biopolymers becoming racemic.
If one wants to criticize the notion of equilibrium states being an artifact of our prejudices, that’s fine, but it’s inconsistent with operation practice in physical science. A “racemic” mix of heads and tails is an equilibrium state, “all coins heads” is decidedly far from equilibrium in an environment that admits disturbances to flip the coins.
Assuming you ignore chemistry.
Sal,
You’re missing eigenstate’s point.
Read it again:
Sal
Nice to see you posting and putting forward an independent view.
You right that Dembski CSI V2.0 was trying to avoid the subjective element in specifications. In attempting to do that he came up with a definition of “specified” that was hard to understand and when understood turns out to be incompatible with what most IDists are saying.
However he was right that CSI 1.0 was subjective – even in your coins example. Why is 500 heads 500 bits? Is the specification “all heads” or “all coins the same” in which case 500 tails is also part of the target and it should be 499 bits.
In any case both definitions suffer from the eleP(T|H)ant in the room. Why is the only acceptable chance hypothesis a fair coin (which of course does not exist in reality)? Maybe it is a coin with 2 heads. Maybe it is being thrown by a mechanism that is so accurate it is almost bound to keep on landing on the same side.
All coins heads is statistically improbable relative to “half or about half-coins heads”. This is very consistent with the notions of deciding if a system is a state of statistical equilibrium.
Start with “homochiral” coins and then randomly flip individual ones over time, the statistics will gravitate toward a certain equilibrium condition of about half-coins heads.
One might say this is all a matter of perception that we decided to lump algorithmically complex representations against the algorithmically simple — but well, that same criticism could be used to criticize other operational fields in chemistry and physics where one could say the equilibrium states are just an artifact of a convenience description.
But even if eigenstate’s criticism is true, the operational convention can be used to detect design in certain man-made configurations like a set of all coins being heads. So his point is moot with respect to detecting certain man-made designs.
Sal,
Eigenstate was a) pointing out that your description of the problem is sloppy, and b) offering an alternative description that better captures your intent, as far as he can discern it. To wit:
I was going to write an OP on chirality. Still might.
It is simply not a big problem for Origins – though it is a problem for a ‘peptides-first’ scenario. Who says origins is necessarily a matter of peptide synthesis?
But an even more fundamental problem for peptides-first is simply that of repeat specification. Homochirality is fixed by that; not an essential precursor of it. It matters not a jot that some alpha-acids have a hydrogen where the others have a side chain, and vice versa, if you do not have a means of bonding the acids in a repeatable manner anyway.
The fatality of D acids to protein function is something of a Creationist fantasy. They simply have a hydrogen atom where L acids have a side chain. In that respect, they function like Glycine. Of course, where L acids have a hydrogen atom, D acids have their side group. This can strain the chain, but no more necessarily-fatal than any other substitution. And there is no reason to suppose that the space of peptides comprised of a mixture of acids is any less function-rich than that of the L acids alone. But once you fix on one orientation, everything else – biosynthesis, metabolism – fixes that way too.
When you have a means of distinguishing side-chains (necessary for repeat specification, and for biosynthesis useing stereospecific macromolecules) it will pick a consistent enantiomer as a matter of course – it must grab the carboxyl group in one ‘hand’, the amino group in the other, and if the side chain is ’round the back’ (ie a D acid) it is a discard, just as an acid with a side chain on the ‘right’ side but with the ‘wrong’ conformation.
Sal,
When you “assert a biopolymer has at least -log2(1/2^N) = N bits of CSI v1.0 based on N stereoisomer residues”,
you are assuming that p(A & B) = p(A) x p(B),
as does any ‘bit-counting’ version of CSI/FIASCO/etc.
This is wrong.
As petruska notes, you are ignoring chemistry.
And history.
This is the eleP(T|H)ant in the room.
Sal is assuming – as do all ID/creationists – that molecular assemblies are random.
I am still waiting for any ID/creationist to do a simple high school level physics/chemistry calculation; namely, scale up the charge-to-mass ratios of protons and electrons to kilogram-sized masses separated by distances on the order of meters. Calculate the energies of interaction in joules and in megatons of TNT. Then fold in the rules of quantum mechanics.
After completing this exercise, justify the routine ID/creationist ploy of using coins, letters, tornados in junkyards, junkyard parts, and battle ship parts as stand-ins for atoms and molecules.
In the last fifty years I have not seen an ID/creationist that can even start this calculation, let alone understand its implications. They all ignore it.
Frozen accident?
Allan Miller,
Exactly.
ID/creationist arguments can’t even explain ferromagnetism.
Allowing for that won’t help, and just compounds the error, Sal. Think about where 500 coin flips with this pattern would place the outcome in the statistical distribution:
THTHTHTHTHTHTHTHTHTHTHTHTHTHTHT….
That’s right, you’d have 250 heads, 250 tails, and the tails/heads distribution would place this run of 500 dead in the center of your bell curve. The distribution is perfectly in “equilibrium”, going by your judgment. And yet, it’s algorithmic complexity is just one quantum greater than all 500 being heads (the algorithm just has to repeat “TH” 250 times rather than “H” 500 times).
“THTHTHTH…” is literally right next door in terms of (low) algorithmic complexity to “HHHHHHHH….”, and yet it moved from the far end of the bell curve to the middle.
I think that should be sufficient to show the error of your thinking, here. Statistical aggregates and averages are NOT proxies for complexity, and can’t be. Indeed, if you are familiar with history of Kolmogorov-Chaitin theory, you’ll know that this is largely *why* algorithmic complexity is needed, why it was developed.
This is just as forceful in an “ideal gas” analysis as it is with coins. If you were to sample a volume of gas and found a homogeneous temperature, say, but in your measurement of that temperate (which is, after all, a statistical average), found that the temperature of 20degC was derived from molecules of which were 50% had a kinetic temperature of 30degC and the other 50% had kinetic energy corresponding to 10degC, you’d have a nice even overall temperature – thermal equlibrium — assuming the “hot” molecules were evenly distributed among the “cold” molecules, but you would nevertheless have a stupendously improbable configuration you are observing. And by “improbable” here, I mean far outside the middle of the bell curve of distributions based on the algorithmic complexity of the configuration.
Which is just to say that decisions on equilibrium or no are not dispositive for your purposes. You’re correction/clarification recommits the same error. An observer is correct to say that a gas-state is at or near equilibrium just by observing it to be so. That state, though, does not entail high-complexity (or, by the same token, low complexity) configurations. It doesn’t tell you anything either way, and you are mistaken to proceed on that basis with any heuristic you build on it.
Here’s a question for you: if you did 10,000 trials of 500 coin flips and graphed the heads-vs-tails ratios, you’d expect to get the familiar bell curve distribution, right? Now, if you did 10,000 trials and instead made a graph of the algorithmic complexity for all those trials, what would be the shape of that graph, in your view?
I agree that there are multiple versions of CSI but I do not agree with Sal’s classification. The original CSI (let’s call it CSIa) has probabilities generated by some simple process like random coin tosses, or mutation if the sequences are DNA. Then there is a scale of some sort, which the user gets to specify. My favorite is fitness, if the sequences are genotypes in a simple model. The SI is -log2(P), where P is the probability of being this far or farther out on the scale, and CSI is declared if the SI is greater than 500 bits.
Then one gets to argue whether CSI can be achieved by natural selection, or whether some conservation law prevents that (spoiler: it doesn’t). The issue is at least not obvious, as natural selection is not examined when computing SI.
Note that the scale is not obvious: in the coin-tossing case we might make the scale one of closeness to a prime, for example. The number of Heads is only one way to make the scale, as people have pointed out to Sal here.
The newer definition, CSIb, (or the more newly perceived definition if you take the position that this is what Dembski meant all along) is that we calculate the probability of being this far out on the scale when the sequence is created by a natural process, including natural selection. As Elizabeth has emphasized in her “EleP(T|H)ant In The Room” posts, the SI for this is not only vastly harder to compute, but it actually renders the whole use of CSI redundant, because you only declare CSI to be present after you conclude Design did it.
I don’t understand what Sal means by CSIv2.0 — I suspect it is not the same as this CSIb.
I’d say probably. Both RNA and protein are stereospecific. If (as I argue) RNA came first, each ribozyme could only work with one isomer. If you have a ribozyme that,say, glues a methyl group in place of one of Glycine’s hydrogens (Glycine is not chiral), it has two possible sites. The specific replacement was probably random, but subsequent derivatives that work with more elaborate side chains are more likely to retool the existing methyl site than to start afresh on the other side. And ribosomes are a further potential selector of acids with a consistent position for their hydrogen atom (and hence the side chain), even if a racemic mix was the original input.
One is left needing an explanation for the chirality of RNA itself, but here I think the key is complementarity.
I analyzed Dembski’s last version of CSI — I’d call it CSI-RIP, but you folks keep digging it out of the grave — in Chapter 1 of Design by Evolution:
http://boundedtheoretics.com/ID_chapter_excerpt.pdf
Rereading Section 5 through the end of Section 5.1 for the first time in years, I think I did a good job of describing what Dembski was trying to do. Salvador definitely has gotten it wrong. Of course, what Salvador really wants to do is to stick with what Dembski modified and abandoned.
Salvador has invested a lot of himself in his rhetorical nuggets. He’s been playing with those 500 coins since the ARN days. If he’d actually paid attention to the paper he cites above, he’d know that Dembski increased the “universal probability bound” from 2^-500 to 2^-400. I’ve told Salvador that repeatedly, but he evidently cannot bring himself to talk about just 400 coins. Four or five years ago, I called him on his ill-taken quine metaphor, pointing out that biological cells do not describe themselves. He backed down at the time. Evidently “quine” is too cool to part with, though, no matter how inappropriate.
An issue for anti-evolutionists is that they aren’t comfortable with evolutionary thinking, so understanding what evolution can do – exploration of all the explanatory possibilities before settling on a favoured hypothesis – is somewhat alien.
The lock-and-key metaphor is a typical example. It is irreducible complexity again. Examples of interlinked systems abound – signal-receptor, enzyme-substrate, gender complementarity … we look at the system, functioning in the modern way, and cannot conceive of a means by which such interdependence could possibly evolve.
Yet, to take one example, enzyme-substrate specificity is a piece of cake. If there is a selective advantage in catalysing a certain reaction, doing it badly when your competitors do it not at all will lead to an ‘edge’. Doing it better yet gains a further advantage, and peptide chains are an excellent means of achieving that, by gradual retooling of an active site to ‘wrap around’ the substrate, or to favour conformational changes aiding reaction kinetics, or control through the binding of cofactors etc etc etc. What you end up with is optimised, and somewhat resistant to further amendment.
Signal-receptor, too, is not obviously a problem. A molecule produced by a cell need not, initially, have any ‘purpose’, or function, or may have a function wholly unrelated to signalling. But if the molecule has a particular affinity for a particular location in the cell – the hydrophobic plasma membrane, for example – that’s where it’ll go, thanks to physics. More accurate targetting and reception can build upon that clumsy mechanism. Once a method becomes embedded in the cell cycle, its core becomes non-amendable, because something else has become dependent on recognisable motifs that were, originally, unimportant. And, indeed, such structural and tagging molecules are among the most highly conserved of all, because of multi-system co-evolution ‘pinning’ them in place. A favoured analogy is http:// – nonessential, but try routing through the web without it. It might appear that conservation is due to some fundamental constraint – this is the only way it could possibly work – but it is perfectly possible for constraint to emerge, due to subsequent multiple interdependencies, and such highly conserved systems are found across the tree of life, eg SRP, a ribonucleoprotein (with my ‘RNA-first’ bias, I find that a tantalising potential legacy from an even older life form).
It doesn’t need digging! It comes out all by itself!
Zombie-CSI!
Depending on what you mean by “statistically special from an operational standpoint”, this statement of yours is either (a) categorically false, or else (b) true in a sense which does not support the inferences you want to draw from it.
Statistically speaking, any one individual sequence of N coin-tosses is exactly and precisely as ‘special’ as any other such sequence. We can arbitrarily define classes of coin-toss sequences, such as, just for grins, “500-coin-toss sequences which contain exactly 15 heads” and “500-coin-toss sequences which contain only heads”; we can calculate how many different coin-toss sequences belong to a given arbitrarily-defined class; we can calculate how likely it is that a given coin-toss sequence will fall into a particular arbitrarily-defined class.
And after we’ve done all that… none of it makes any one individual coin-toss sequence any more ‘special’ than any other individual coin-toss sequence. As far as the statistics are concerned, an all-heads 500-coin-toss sequence is exactly and precisely as ‘special’ as an all-tails 500-coin-toss sequence, and each of those two 500-coin-toss sequences is exactly and precisely as ‘special’ as a 500-coin-toss sequence that contains 127 heads, etc ad nauseum.
I suspect that Sal will argue that the distribution of tosses with a given number of heads follows the binomial distribution.
If we call Comb(N, m) the number of combinations of N things taken m at a time, and if the probability of a head on any given toss is p with the probability of a tail being q = (1 – p), then the probability of m heads in N tosses is
Comb(N, m) p^m q^(N – m).
The probability peaks at half heads and half tails if p = ½.
But this is where the problem lies. Atoms and molecules do not behave like coins or strings of letters. Using a binomial distribution or some uniform distribution for atoms and molecules is completely inappropriate. There are very strong energies of interaction among atoms and molecules; even “neutral” ones.
The problem to which you allude is the assertion that a given outcome is special. That is the “Lottery Winner Fallacy” that assumes that a particular outcome is the target of the random selection. It doesn’t acknowledge that any other outcome might also be “special” in some other way.
So Sal is making the usual ID/creationist implications that coins, letters, junkyard parts, or any other inert objects are stand-ins for atoms and molecules, and that a specified complex arrangement of these is the target of a random sample out of an idea gas of these inert objects.
We are expected to argue on their territory.
You may well be right about that. But binomial distribution or no: As far as statistics is concerned, any one individual sequence of N coin-tosses is exactly and precisely as ‘special’ as any one other individual sequence of N coin-tosses. The fact that coin-toss sequence X belongs to an arbitrarily-defined class XC, which has a different number of members than a different arbitrarily-defined class YC to which coin-toss sequence Y belongs, does not alter the unrelated fact that neither of the coin-toss sequences X or Y can rightly be said to be more ‘special’ than the other one.
I’m confused. You have a sequence of all heads and the preferred explanation is goddidit?
I agree that’s a much bigger problem for Sal and other IDists, but first things first, right? Sal is getting wrapped around his own axles bumbling about on *his* territory. There is so much in OP that begs for critique, but really, the focus I think should be on the elementary problems; Sal is confused and incoherent in is conceptualizing complexity and specification just as a pure matter of coins, never mind trying to use that as a bridge to gas microstates and thermal equilibrium (or no).
Sal’s not been back for a while, but I can’t see any progress coming from even taking up your valid objections; they are intractable so longs as this confusion about complexity and specification persists, just at the coin level.
If I were arguing Sal’s case for him, I would put it this way:
Given that we observe a sequence of 500 heads, which explanation is more likely to be true?
The obvious answer is (b).
In the case of homochirality, Sal’s mistake is to leap from (b) directly to a conclusion of design, which is silly.
In other words, he sees the space of possibilities as {homochiral by chance, homochiral by design}. He rules out ‘homochiral by chance’ as being too improbable and concludes ‘homochiral by design’.
Such a leap would be justified only if he already knew that homochirality couldn’t be explained by any non-chance, non-design mechanism (such as Darwinian evolution). But that, of course, is precisely what he is trying to demonstrate.
He has assumed his conclusion.
cubist,
Yes; I was elaborating your point and suggesting that Sal would switch concepts just to mud wrestle.
Getting probabilities wrong is another common characteristic among ID/creationists. The same problem applies to poker hands for them. Every hand is equally probable; it is convention that decides which hands are “most desirable.” Once that convention is decided, then there is a hierarchy of “desirable” hands that win.
In the case of 500 tosses of a fair coin, all permutations of heads and tails are equally probable; however, all combinations are not.
But ID/creationists make the same mistake repeatedly when calculating sequences of letters in a list. If the list is N characters long, they will tell you that the “CSI” of that list is log to base 2 of 26^N (assuming 26 ASCII characters in this example; they may like to use more if it makes the CSI bigger).
The calculation is absolutely meaningless when referring to atoms and molecules. However it also forgets that every permutation of any letter that is repeated two or more times is the same list. E.g., if the letter “a” shows up in several positions in the list, all permutations of that letter produces the same list.
Every permutation of identical atoms or molecules in a specified configuration produces the same configuration.
ID/creationists will not only use inert objects as stand-ins for atoms or molecules, they will always over count in order to get a CSI as large as possible; so they don’t care about permutations of identical things.
I have watched ID/creationist arguments for nearly 50 years now. They avoid physics and chemistry like the plague. The only physics they refer to is the second law of thermodynamics; and they have mangled that in exactly the same way it was mangled by Henry Morris. The only math they seem to know is taking logarithms of probabilities of lists of characters.
All of them – Dembski, Sewell, Abel, Cordova, Meyer, Behe, … – have had far more than enough time to study and get the scientific concepts right; yet they get them wrong every time, even at the high school level.
ID/creationist arguments are designed for endless mud-wrestling in order to leverage “legitimacy” from the scientists they try to lure into debates. And it is their constant, grass roots political efforts aimed at a naive public that makes it impossible to ignore them. If left unchecked, they will flood the public with junk science.
eigenstate,
I have to admit that I get extremely bored with such elementary errors repeated endlessly over a period of nearly 5 decades.
You may have noticed that Sal used a story about FBI agents investigating cheating with a deck of cards. The key point that Sal glossed over was the repetition of the events and a video comparing the outcomes.
First he talks about 500 heads without saying anything about repetition. Then he throws in repetition and the binomial distribution as sufficient for being able to assume design. Presumably he means 500 heads repeatedly; or some low probability combination repeatedly, or for that matter any specified sequence repeatedly.
But the problem is not only that coins flipped randomly are not appropriate stand-ins for the behaviors of atoms and molecules; he wants to use repeated flips as a comparison to a sequence of events that has been tried only once here on this planet.
As several people here have already noted, that’s the eleP(T|H)ant in the room; we have no idea what the probability distribution is for the origin and evolution of life. How can anyone claim to assert that an event is improbable when they have seen only one instance of it? How do you “flip those coins” again and again?
However, knowing something about chemistry and physics changes the picture considerably for anyone who takes the time to learn these subjects. There is good reason to believe that the origin and evolution of life are natural phenomena even though we have at the moment only one instance of them. We know what atoms and molecules actually do.
ID/creationists defend themselves with their self-imposed ignorance of chemistry and physics. That way, anything any expert says is merely an “opinion.”
You’re right. I’ve been reading UD since late 2005, what more can I say than that to underscore your point? That’s not five decades, but it’s long enough to understand that the core ID apologetics are not corrigible.
I agree with the rest of your response to me. It’s hard to understate the prominence of Lizzie’s eleP(T|H)ant in all of these debates. It’s a ubiquitous thread of misunderstanding that runs through nearly all of this. I keep thinking — let’s focus on that, that’s the real barrier to a more productive debate. Not a more agreeable or less intense debate, just less banal, tedious, stupid.
It would be great to be at a point where you could engage the debate on the level of stereochemistry, I fully agree. But so long as we can’t even get past the fetish for the Lottery Fallacy, it’s just talking amongst ourselves, really.
Having just typed that out, I guess that’s not such a bad thing.
Mike,
In Sal’s defense, even one-off events can be identified as improbable under certain hypotheses. If I roll a fair ten-sided die nine times and come up with my Social Security number, then a very improbable event has occurred, even if I don’t repeat the experiment.
I think Sal is correct to reject ‘homochirality by chance’. His mistake is in thinking that ‘homochirality by design’ is the only alternative.
No. I am saying that the specific 500-coin-toss sequence which consists of 500 heads is a 1:2^500 longshot, just like every other specific 500-coin-toss sequence.
It’s true that us humans tend to assign greater significance to a 500-heads coin-toss sequence than we do to a 250-heads-and-250-tails coin-toss sequence, granted. But the probability that a specific coin-toss sequence will come up, is not affected by the level of significance that we humans assign to that sequence.
Likewise, it is true that the number of 500-coin-toss sequences which fit the description “250 heads and 250 tails” is a lot greater than the number of 500-coin-toss sequences which fit the description “all heads”, and therefore a 500-coin-toss sequence which fits the former description is a great deal more likely to come up than is a 500-coin-toss sequence which fits the latter description. But, again, the probability of any one specific coin-toss sequence X coming up, is not affected by the total number of other sequences which share a description with coin-toss sequence X.
Therefore, contrary to Mr. Cordova’s assertion, a 500-heads coin-toss sequence is not “statistically special from operational standpoint”.
keiths,
How did you know it was your Social Security number? Almost everybody else looking at that number wouldn’t have a clue. If you developed some form of aphasia or alexia, you wouldn’t know either.
Which ID/creationist looked at the “roll of the dice” for the origins of life and evolution and was able to conclude that it was a “very improbable event” from a “one-off experiment?”
I think one can see the implicit sectarian religious beliefs in ID/creationist claims even though they try to hide their sectarianism to get around the law. It may not be as crude as Ken Ham’s “Were your there?” shtick; but they seem to believe that their deity was there from what they read in their holy book.
I suspect that belief has something to do with their claims that a one-off event was improbable; for them, it has to be.
Mike,
I’m not arguing that my SSN is objectively more significant than any other 9-digit number. They are all equiprobable, of course.
My point is that for anyone to roll their SSN (or another personally significant 9-digit number) is highly unlikely, because the number of personally significant 9-digit numbers is low for each of us while the space of possible 9-digit numbers is large.
If someone sits down and rolls their SSN, then either:
1. The die was fair, the rolls were random, and they just got lucky. It’s pure coincidence.
2. Something else is going on.
Just as my SSN is significant to me, homochirality is significant to biology. Homochirality has advantages over racemicity. To borrow Salvador’s coin metaphor, the number of ways you can be homochiral is tiny — just two, either all heads or all tails — while the number of ways you can be racemic is huge: 2^500 – 2.
Since biology is homochiral, either:
1. It’s purely by chance. The coins were fair, the flips were random, and we just happened to get all heads or all tails.
2. Something else is going on.
I think most of us, including Sal, would agree that #2 is far more likely than #1.
Sal’s mistake is in thinking that design is the only possibility that fits into category #2. By leaving out evolution, he is effectively assuming his conclusion, as I argued above.
…or the existence, somewhere between OoL and LUCA, of a molecular filter arbitrarily but consistently ensuring that only molecules with a hydrogen atom at position X relative to the carboxyl and amino groups could be processed. Which is pretty much a given – if you have binding sites for the C- and N- groups, you’ve pretty tightly constrained where the side chain can fit, and if it’s on the other side, it’s not a substrate. The monotonous strings of ‘heads’ found throughout Nature point very strongly towards such a filter. 🙂
keiths
I don’t think it is significant to biology – it’s only us that notice a pattern. There are about 500 possible amino acids, any of which may conceivably float by a catalyst; for any given degree of specificity, the group is subdivided into ‘substrates’ and ‘the rest’. If there is any side-chain specificity at all, the isomer will fall into ‘the rest’, because its side chain (to any biological molecule that can distinguish that site) is hydrogen.
I don’t think there is a fundamental advantage over racemity in the amino acids. The modern chiral picture is more likely due to constraint than specific advantage. The fact that all modern biological amino acid isomers are variations on the same ‘left-handed’ theme is probably due to the fact that the enzymes that operate on them are commonly descended. You might start with a simple methylation (alanine). As the group expands, amendment to deal with different side chains will take place at the same site. If that filter isn’t enough, once protein synthesis gets going they must all pass through the asymmetric ribosome. There is the odd organism that makes D acids, but none that incorporate them in protein.
I doubt that there was ever an organism making ‘racemic proteins’, not because they’d suffer disadvantage, but because once an organism takes the left or right fork, its subsequent evolution is constrained. There was just one coin toss.
Yes; I understood your point.
The reason it is significant is because of the background knowledge you have not only about that particular number, but also because you already know about the 10-sided die and how it works.
We don’t yet have that kind of knowledge about the origins of life. We are fairly sure that rerunning evolution would not produce the same results.
What we do know is how atoms and molecules behave. As Allen points out, a particular chirality is most likely a matter of chance. Once a particular set of reproducing molecules falls in that direction, the rest is automatic after that.
Without the specific recipe(s) in hand, we don’t know the sequence of events yet. Given the fact that this planet is already teeming with life, we may not be able to find instances of the origins of life taking place here on Earth because of the dominant forms that already exist. This is not completely ruled out; but the search for life on other planets and moons offers an opportunity to get some handle on probabilities.
There is a handle on probability of a sort that we do have. We get that from the energy range in which life on Earth exists. That energy range is roughly in the range of liquid water. That corresponds to an energy range going from ½kT = 0.01 electron volts to about 0.02 electron volts. (Note: Air molecules at room temperature have average kTs on the order of 1/40 of an electron volt.)
Note that the temperature range between hypothermia and hyperthermia is much narrower than that.
Metals such as iron melt at something like 0.2 eV. Chemistry takes place at energies on the order of 1 eV.
There is a reason to be concerned about climate change. Life as we know it requires a very narrow window in order to exist. Even if the formations of the molecules of life (chemistry) took place in energy cascades that life cannot survive, the products of those energy cascades have to get shuttled into environments much like the one in which we find ourselves.
Some bacteria, tardigrades, cockroaches, and other life forms can withstand wider extremes in energy. We as humans have been successively selected into an increasingly narrow energy niche.
While these facts don’t produce a detailed probability distribution, they do tell us something about the relative probabilities of the existence of things and in which energy ranges certain kinds of phenomena exist.
What you’re saying is that the explanation for the apparently-improbable homochirality we see is that we’re seeing a fixed two-headed coin in action. Of course we will only see a sequence of heads. It’s a certainty, not improbable at all, not now, not since the “fair” coin was swapped out long before LUCA. There was only one “fair” toss, at the beginning, and if that toss had come up “tails” that coin would have been swapped for a two-tailed coin, but it didn’t; it came up heads and was swapped for a two-headed coin. Do I have that right?
Sad, this whole coin metaphor is misleading by its very existence. We invoke the metaphorical coin, and peoples draw in the concept of the mint, the press, the coin designer. I think it arises even in people who want to use neutral imagery, but it’s irresistible to the IDers who are prejudiced in favor of a designer claim. Maybe it would be better if we could invoke the animal bones people used to throw for games of chance, or the yarrow stalks people drop to tell fortunes. Something less tied to concepts of design and manufacture …
Allan,
That’s interesting. I was under the impression that a) homochirality among nucleotides was necessary for DNA to form the double helix, and b) that the homochirality of amino acids followed from the homochirality of the nucleotides.
Am I wrong about (a)?
Or are you saying that that a) and b) are independent, and that (apart from evolutionary constraints) we could just as easily have homochiral DNA along with racemic amino acids and proteins?
Eigenstate: “Statistical aggregates and averages are NOT proxies for complexity,”
I never said they were as a general rule. Only in the specialized case I provided.
All coins heads is statistically improbable compared to half-coins heads or approximately half-coins heads. The expectation percent of heads is half, not 1. Extreme deviations from that expectations ought to attract attention, and that is consistent with operational practice.
With the HTHT…. pattern — taking every other coin should also yield an expectation of .5, but it doesn’t, it yields a value of 1 and that should get attention as well.
I never said, or implied statistics are a proxy for Kolomogorov complexity, only in very specialized cases might statistics give an indication of Kolmogorov complexity. You attributed an argument to me which I didn’t make.
See: A 22 sigma event is consistent with the physics of fair coins?
No one familiar with Sal’s character will be surprised to learn that he shamelessly quotemines eigenstate in his new OP:
SSDD: a 22 sigma event is consistent with the physics of fair coins?
Here’s the full eigenstate quote. The parts that Sal omitted are in bold:
Eigenstate carefully put the key phrase in ALL CAPS so that Sal couldn’t miss it. Sal chose to omit it anyway. What eigenstate said is correct, of course, and Sal is wrong to challenge it, especially in such a dishonest way.
All heads is no more improbable than any other specific sequence of results. Every specific sequence has a probability of 1 in 2^500.
The point that Sal is making in the OP is a different one: that it’s more probable to get a sequence that contains some combination of 250 heads and 250 tails than it is to get all 500 heads. Eigenstate would not disagree, of course.
I was wondering about bringing the nucleotides in …
There are two levels of stereochemistry: that of simple molecules such as amino acids and nucleotides, and that of macromolecules (which happen to be built from the first). The 3D character of the first derives from the tetrahedral bonding around carbon – four different atoms can be arranged as the points of a tetrahedron in 2 different ways. The 3D character of the second derives from the fact that a chain builds up into a kind of ‘wire-frame’ when folded. That frame can in principle be bent to any shape. Where substrate molecules come in ‘mirror-image’ forms, that shape will only work on one. You could make a wire-frame glove, but it would only fit your left (or right) hand. You want a pair, you have to start again.
You can make a biological catalyst out of either nucleobases (ribozymes) or amino acids (enzymes). Each one can develop specificity, by changes to the ‘wire frame’, but that specificity itself will exclude isomers of a chiral molecule along with everything else, regardless how it is itself composed.
A and b are independent, yes (well, there is a possibility of stereochemical coupling at the monomer level due to something of biological history, but the polymer level seems more likely). You could (given a suitable construction system) make ‘wire-frame catalysts’ out of any proportions of L and D acids – but you could even do so out of any proportions of D and L nucleotide monomers, and indeed from mixtures of the two chains – ribonucleoproteins. There is no chemical principle that cares unduly about chirality – it simply means there are two ways of arranging four different atoms in a tetrahedron.
But the wire-frame catalysts you make will all be stereospecific. If you use them to make monomers for their own construction, they will be homochiral, as it is very unlikely that two parallel systems would evolve. Once you have one, what use is another?
Obviously, anything other than a homochiral nucleic acid chain is non-replicable, so this forces a constraint on ribozymes. And evolution of a larger amino acid library from a smaller will most likely produce acids of the same type, since the same ‘machinery’ is likely to form the basis of the extension, giving consistency from contingency rather than necessity.
Of course, the problem of the origin of the chirality of nucleic acids remains. But these select themselves. – both for homochirality and for base complementarity. Because the double helix is the most stable form, and only homochiral nucleotide chains will form this structure, and only certain bases among the many possibilities will pair in the necessary complementary manner, a complex mix of L and D oligonucleotides with various bases attached can (in principle) be ‘purified’ by complementary hybridisation, from which the most stable complementary chains would emerge. They would be homochiral, and only contain certain bases from the many possibilities – those that complement.
Absolutely! Similar thinking lies at the basis of Hoyle’s Fallacy, and many others. It’s appropriate for binary choices, and illustrations of probability distributions, but it soon loses traction, though people continue to run like cartoon dogs on its slippery surface! 🙂
At some point, the racemic mixtures produced by non-biological chemistry were filtered, picking only ‘heads’ (could just as eaily have been tails). This historic coin-picking-up device would only work on one face, or on rows of coins (among a random mix) stochastically happening to all show the same face, or became a coin manufactory that could only turn them out one way up. Or … it’s stereochemistry!
I’m aware, incidentally, or will be made aware by stc or Upright Biped, that the actuality of an ‘origin of macromolecules’ requires some means of ‘specification’. All very well saying you can make a protein catalyst from any amino acid component whatsoever, and that it is enantioselective by default, but it needs to be made more than once, and for more than one generation, else you haven’t made ‘Life'(TM).
But this problem has nothing to do with chirality per se. They are missing a trick – if you want to over-egg the probability calculations, don’t just add in the isomers; add in every last possible configuration that could possibly be called an amino acid! A soup of such molecules could produce 500^50 different polymers just 50 units long. That’s 10^135. It’s not just a coin-toss every step, it’s a roll of a 500-sided die! And I will continue to smile serenely like Buddha, because I subscribe to neither the ‘proteins-first’ nor the ‘prebiotic soup’ versions of OoL.
My point was it didn’t apply in the particular case you did provide.
You have not grasped the problem I’ve raised yet. In your OP, you said you considered “500 fair coins all heads”. My objection was a very elemental one: BY DEFINITION, if you are dealing with “fair coins”, there’s nothing to be suspicious about, and can’t be: they are just fair coins. If, on the other hand, you are looking at a process which produces one of two values (e.g. {“tails” | “heads”}), but the probabilistic dynamics for those outcomes are not known, or understood, THEN you have a very different problem to deal with. And of course, that’s where things go off the rails when this kind of simple probability view gets applied to biology — proteins are not fair coins.
See above. That can’t be true if you are dealing with fair coin flips. If you want to allow that you’re not dealing with coin flips known to be fair, then game on. As you expressed it in your OP, there’s no warrant for any attention of the kind you are interested in.
I am the one who said those probabilities are not a proxy for K-C complexity, not you, Sal. And in the particular you case you gave, the statistic doesn’t indicate what you think it does, which is why I offered the “THTHTH…” pattern, to show that a “perfectly equilibrium” pattern with the same K-C simplicity as “HHHHHHH….” doesn’t give you the indication you seek,
I’ve not attributed this to you in any general sense: “statistics are a proxy for Kolomogorov complexity”. Read my comments again to verify this if you like. I’ve pointed out that your “specialized case” doesn’t indicate what you think it does.
Sal, would you agree with this statement:
A. Every actual outcome from 500 fair coin flips is a 22-sigma event?
Let’s start there.
You are conflating probabilities for a specific single outcome with “classes of outcomes”. If you group all specifications that fall within 40%-60% heads into a class, you are going to have a great many more qualifying patterns than the “all heads” class which is a singleton, a class with just a single instance. For any given run of 500 flips, the odds of that outcome being in the “40-60% heads” class are far greater than being in the “100% heads class”.
And still, “HHHHHH…” is just as likely as ANY OTHER SPECIFIC OUTCOME, every 500 flip test you do being, necessarily, a 22-sigma event, right?
I can’t respond at UD, having been been banned there — thanks for posting your response over there, Sal! — but want to address this from Neil Rickert on that thread.
The order necessarily matters when you are considering any form of CSI — both the “C” and the “S” in CSI depend on ordering. Again, the example I gave to Sal, the “THTHTH…” pattern of alternating heads and tails. If sequence doesn’t matter, it’s indistinguishable from the large number of other possible patterns that also have exactly 50% heads. That is, if the order doesn’t matter, then CSI isn’t a relevant concept in the first place. CSI is an attempt to analyze the internal structure of a data set, and thus is predicated on internal relationships (order).
Right. I understood where Sal was trying to go in my first comment, which was why I characterized his OP as “sloppy”, and explicitly allowed that his goal was something other than what got typed out.
Taking Sal’s point to be what you are suggesting here (which I’m fine with), then we get past sloppy flirting with the Lottery Fallacy on Sal’s part to the obfuscating effects of his proposal (which he calls “CSIv1.0 + statistics”). Here’s Sal:
Sal is right that calculating the probabilities for “specialness” for various sequences (again, not “statistical clusters”, but specific sequences — the Champernowne sequence is unusual only as a specific sequence, and is “in equilibrium” statistically in terms of 1s and 0s) is difficult, intractable I’d say.
But that’s not the fail point for CSIv1.0 + statistics. The fail point is that statistical aggregation and clustering BLINDS any CSI analysis, or more precisely, any coherent and quantitative analysis of internal structures. The only place where a statistical analysis aligns with specification analysis is at the extremes, on “border conditions” — “all heads”. In the case of the “all heads” extreme (from a binomial distribution view), you do get a coincidental specification, but is the nature of that probability extreme.
Once you start using a “binomial distribution” view for evaluating phenomena, even if we are just dealing in coin flips or card shuffling (neither of which we should declare a priori as “fair”), you are hosed in terms of pattern detection. Because your inputs don’t consider internal structure.
The only way such a statistical filter like Sal proposed for the coins is at the trivial extremes — “all heads”, and that help is just coincidental. Which is why it becomes obfuscating and confounding in all the other cases. The “THTHTH…” pattern gets an “A+” from Sal’s heuristic for being in “equlibrium”, as a “pattern of non-interest”. And yet, it’s no different than “HHHHH….” in terms of ordering and structuring, just repeating a 2-bit sequence over and over, rather than a 1-bit sequence.
After dishonestly quotemining eigenstate, misrepresenting his views, and failing to correct the misrepresentation when it was pointed out, Sal is now playing the martyr:
Sal,
I suggest you correct your dishonest OP first. Then you can worry about whether any of us agree with you on any particular aspect of the math. (It so happens that we do, and have already said so in this thread.)