Tom English, in his recent post on this blog, has argued that:
Distinguishable entities operating identically by simple rules can form structures high in specified complexity
This is, of course, an example of a long standing critique of specified complexity. However, the critique is nonsense. The critique fundamentally misunderstands specified complexity.
Specified complexity makes no claims as to the probability of the outcomes or entities under consideration. It assumes that you have some other way of assesing that probibility under that given hypothesis. The only role of specified complexity is to justify rejecting a hypothesis if it renders the observed outcome overly improbable.
If we were to assume a uniform random distribution of the stones at Chesil Beach, then it would indeed have high algorithmic specified complexity. However, what that entitles us to conclude is merely that the uniform random distribution and any explanations that would imply a uniform random distribution are incorrect. However, if we look at the actual probability distribution implied by the given explanation, we find that the observed distribution of pebbles is not particularily improbable, and thus would not have high algorithmic specified complexity.
Claiming that specified complexity requires calculating probabilities according to a uniform probability is a long standing straw man. Back in July, in Jonathan McLatchie still doesn’t understand Dembski’s argument, Joe Felsenstein claimed that while since 2005 Specified Complexity has required calculating actual probabilities, prior to that it assumed the uniform.
I have repeatedly documented that this isn’t true, specified complexity has always been of the form I describe it above:
- http://www.evolutionnews.org/2016/01/beating_a_dead102071.html
- http://www.evolutionnews.org/2015/12/specified_compl101321.html
- http://www.evolutionnews.org/2013/08/elephants_sold075441.html
- http://www.evolutionnews.org/2013/04/information_pas071201.html
I would understand it if critics took issue with my documentation of the history of specified complexity and argued that I had misrepresented it. I would also understand if it critics took issue with the logic of specified complexity, arguing that even if life was astronomically improbable under Darwinian mechanism, Darwinism still might be true. I would also understand if critics simply dismissed specified complexity as irreleavent because they thought that Darwinian proccesses rendered life probable.
Instead, critics misrepresent specified complexity and its history. They are not interested in engaging the actual arguments put forward, and instead prefer to spend their time engaging straw men devised in their own minds.
Hi Dr. Ewert. Are you going to stick around for the to and fro, or is this a drive-by?
Nice to see you Winston. Hope you are well.
Agree.
I would understand it if critics took issue with my documentation of the history of specified complexity and argued that I had misrepresented it. I would also understand if it critics took issue with the logic of specified complexity, arguing that even if life was astronomically improbable under Darwinian mechanism, Darwinism still might be true. I would also understand if critics simply dismissed specified complexity as irreleavent because they thought that Darwinian proccesses rendered life probable.
That sounds like an admission specified complexity is useless as a metric for determining “Design” in biological life because you have no way to know the actual probability distribution of any biological entity. Looks like ID has failed yet again.
“Specified complexity makes no claims as to the probability of the outcomes or entities under consideration. It assumes that you have some other way of assesing that probibility under that given hypothesis. The only role of specified complexity is to justify rejecting a hypothesis if it renders the observed outcome overly improbable.”
IOW, it’s completely redundant.
@Winston Ewert
A bit busy now but would love to discuss this with you.
Let’s focus not on the Origin Of Life but on later adaptations.
A. May I assume that you agree that Specified Complexity can be applied to such cases?
B. What was the purpose of William Dembski’s Law of Conservation of Complex Specified Information (which was in his 2002 book and also in his later 2007 edition). It seems to be intended to show that CSI (or SC) cannot be achieved by ordinary processes such as those of evolution, unless you already have it there from the start. Can we please go over this step by step? If you can show that Dembski’s argument works, I will happily climb down in public, loudly.
C. If CSI was never intended to be calculated using a “chance” process such as mutation, but required in its calculation that you take into account processes such as natural selection, then as I think you have agreed, once you have shown that the probability of an adaptation this good is incredibly small, you have argued against its being achieved by natural evolutionary processes. And adding on a declaration that SC (or CSI) is present adds nothing to this. I think that you acknowledged this once before in a UD post. Right?
I don’t have a polite response in me at the moment. The disconnect from a little thing I like to call “reality” is overwhelming. I’d better leave this to Joe for now, and try again tomorrow.
Well, Winston…
I wrote:
You quoted just the first sentence, and attached it to, apparently, beach-stones of widely ranging composition and size. Do you think that’s a sensible thing to do? I gave a link to my original post on the ASC “conga lines” formed by hermit crabs. Did you, perchance, read the (very brief) thread, to see what more I wrote? I think that if you had investigated what I was referring to, you would not have dumped on me the boilerplate in your post. (Yes, I do recall seeing it before.)
I opened “Chesil Beach” with this:
Rather than address directly my observation that you rebranded specified complexity as a measure of meaningful information, you simply assert something different:
Really!? So you have not been characterizing algorithmic specified complexity as a measure of meaningful information? You have not given the title “Meaning of Meaning” to the chapter in which you present ASC, in the forthcoming Introduction to Evolutionary Informatics?
The abstract of “Algorithmic Specified Complexity in the Game of Life” does not begin, “Algorithmic specified complexity (ASC) measures the degree to which an object is meaningful”? You did not define a scheme for encoding patterns, and treat it as a model of the stochastic process generating patterns? Would you say that you assessed probability (as you put it above), or made it up conveniently? As I recall, your justification (in the discussion) was something along the lines of “Well, it worked.”
The title of your most recent publication is not “Measuring Meaningful Information in Images: Algorithmic Specified Complexity“? You did not conveniently define a stochastic image-generating process in terms of the PNG image format?
Show me where I said it does. I have a clue or two or three about the notation
indicating that the probability distribution 
is a parameter of the measure.
Show me where I’ve suggested that the probability distribution on objects should be uniform. I in fact proved a result more fundamental than your “algorithmic specified complexity is rare” theorem, for arbitrary discrete probability measures. So how in the world are you going to pin the uniform on me?
I believe that I have the right, when applying ASC, to choose the distribution
as arbitrarily as you have chosen 
in your published applications. You most definitely are not in a position to lecture me about realistic modeling of the object-generating process.
The fact of the matter is that you’ve proved only one (weak) property of ASC. So you live or die by example applications. Do you and your colleagues ever look for negative examples? All you ever publish are positive examples: “Look how it works!” And if someone else does pretty much the same as you do, but says instead, “Look how crazy it is,” then your immediate response is “deceptive counterexample.” Do you really believe that there are no valid examples of ASC behaving quite unlike a measure of meaningful information? Why would you believe that? You do not have a body of formal results to justify your claims about ASC. So you really ought to suspect that there are examples that don’t pan out the way you’d like. Furthermore, you ought to be trying to find them for yourself. This adversarial bit you have learned in the context of ID is utterly pathological. It is not at all the way that scholarship actually works.
I have in the pipe some applications of your methodology in “Measuring Meaningful Information.” And you are going to scream bloody murder. There is nothing in your paper that precludes my using ASC as I have. I repeat, you have no justification for believing that there are no negative examples, and the reason is that you have not derived a body of mathematical results characterizing the measure. (In fact, your “definition” of ASC is ill-formed. Dog willing, and the Creek don’t rise, I’ll complete a post that explains the error, and shows how to fix it.)
Lest people be confused: Winston Ewert was raising two issues: whether Tom had properly applied Algorithmic Specified Complexity, and whether I had misinterpreted Dembski’s argument about Specified Complexity. So there are two arguments going on here at the same time.
Tom did claim that Chesil Beach is high in algorithmic specified complexity, right?
His argument is that his way of determining ASC is just as valid as Winston’s, and that this tells us something meaningful about specified complexity?
The difference is Ewert claimed the ASC somehow dropped when you took into account the natural sorting action of the waves which are observed to produce complex orderly formations. Where in any description of ASC (or plain old SC) has any IDer every taken into account the observed natural feedback processes of evolution which are observed to produce complex orderly formations in biological life?
Winston Ewert:
[Quote]
The only role of specified complexity is to justify rejecting a hypothesis if it renders the observed outcome overly improbable.
[End quote]
William Dembski discussed another role, though.
[Quote]
The presumption here is that if a subject S can figure out an independently given pattern to which E conforms (i.e., a detachable rejection region of probability less than alpha that includes E), then so could someone else. Indeed, the presumption is that someone else used that very same item of background knowledge — the one used by S to eliminate H — to bring about E in the first place.
[No Free Lunch, p. 75]
http://www.antievolution.org/cs/sc_implicit_design_conjectures
Correct. I heard Bob Marks say that meaning is in the eye of the beholder in a (video of a) presentation at the ASA conference, perhaps two years ago. So I was not surprised when I saw that the last section of the chapter “Meaning of Meaning” in Introduction to Evolutionary Informatics is titled “Meaning is in the Eye of the Beholder.” (My post enabled readers to see Chesil Beach through the eyes of a contemporary of Charles Darwin. The passage that I quoted is loaded with teleological language. I thought it was funny to attach that perspective to a claim that the beach is high in ASC.)
Given the “definition”
the appropriate response to a vague claim like mine is to require that I define terms.
What is the sample space
that includes Chesil Beach (
)?
on 
?
(eye of the beholder)?
What is the probability distribution
What is the context
I get the impression that Marks has moved on, and that Ewert is still trying to save the old claims of Dembski.
The post is in fact more a response to you than to me. It’s fine by me if he continues the discussion with you.
If. So far he hasn’t come back.
I may make a separate post on Algorithmic Specified Complexity, as I have some questions as to what, if anything, it has to do with detecting Design in adaptations of organisms. In the meantime, if Ewert shows up again, I would like to ask him to know what the answers are to questions A, B, and C that I asked him, above. (The reason I labeled them that way is that if you use 1, 2, and 3 and label paragraphs by them WordPress deletes the numbering and smooshes the paragraphs together. Which is irritating.
Joe Felsenstein,
1. backslashes
2. before
3. periods
4. suck, but work.
Entered as adjacent lines (no blank lines between the points).
Can anyone explain why Tom thinks that separate shells, that have no bearing with other shells other than they are all shells, is an example of specified complexity?
The shells are not attached to each other, they don’t work together, each shell is a completely separate entity from another shell. So which shell is Tom talking about, the entire set he calls a group of shells? What groups them together other than their proximity? Do shells 2000 miles away become part of this set?
Isn’t Tom’s set a completely man-made set?
Because given the definition proposed by Dembski et al, the pattern the shells exhibit when hermit craps line up to exchange shells, fits the definition.
Things don’t have to be somehow chemical or physically in contact and share electrons through chemical bonds, to be capable of forming a pattern that exhibits so-called specificed complexity.
To pick an example, the individual letters on a page are not connected (unless you’re writing in cursive, but then the words aren’t connected), nevertheless a page of text written by an intelligent author, even you would presumably agree shows high levels of specified complexity? So things “being attached to each other” is not a prerequisite for specified complexity.
No, just the ones that sit in the line arranged by Hermit crabs when they exchange shells. They’re sorted from biggest to smallest. That’s the pattern.
The actions of Hermit crabs.
Do shells 2000 miles away become part of this set?
No, unless there’s a 2000 mile line of shells arranged by Hermit crabs (or some other mechanism) from biggest to smallest.
Yes it is, in the same way Dembski, Ewert and Mark’s definition of specified complexity is, at bottom, totally in the eye of the beholder. That’s the whole point.
I like it when ID supporters learn about ID.
If we ever see a response from Ewert it will probably be over at UD.
That seems at odds with the feelings articulated in the OP:
I’m sure Ewert intends to engage with the actual arguments put forward in this thread as a response to his OP. Otherwise, why say such a thing at all?
Rumraket,
I am pretty sure that a visual pattern is not the definition of specified complexity. The shells have zero function as they sit.
We’ll have to wait and see, and I honestly hope I’m wrong.
Why say such a thing at all? For attention.
That’s effectively one of the criticisms, that according to the definition as supplied by Dembski et al, mere arrangement would correspond to the definition of specified complexity.
The arrangement of shells has a strong survival function for the Hermit crabs phoodoo. They arrange themselves from smallest to largest, then exchange shells in short order (smallest takes 2nd-smallest shell, 2nd-smallest takes 3rd smallest and so on), thus saving time looking for new shells, and expend very little time out in the open and vulnerable to attack on their soft bodies.
But the point is, the arrangement of stones at Chesil Beach does not (have any function), yet with the definition of specified complexity we would not be able to distinguish the stones at Chesil Beach from Hermit Crab shell lines. They both have high levels of algorithmic specified complexity because they both show the same arrangement (pattern). By that token, a mindless inanimate process can produce specified complexity, because it did so at Chesil Beach.
Now, ID proponents believe mindless inanimate processes can’t produce high levels of “algorithmic specified complexity”, and evolutionists believe it can.
So if ID proponents are right, they need to come up with another and better definition of algorithmic specified complexity than the one supplied by Dembski et al, because that one detects high levels of algorithmic specified complexity in patterns produced by mindless inanimate processes (such as sorting rock-sizes by the mechanical actions of waves on a beach).
No one can teach anything to phoodoo. Not evolution, not ID, no nothing.
Looks like Winston and johnnyb have both bailed on defending their “specified complexity” brain fart. How vey Creationist of them. 🙂
I suspect any response with be at Evolution News and Views, where commenting is not allowed. I would be delighted to be proven wrong.
Winston Ewert has some serious issues to grapple with, so I’d give him a bit more time. Here is his opportunity to clarify how the argument from Specified Complexity works, and what it does and does not do. I call his attention to the three questions I asked above.
Rumraket,
Dembski (he didn’t come up with the term specified complexity) says in part that patterns are specific when they “display a large amount of independently specified information.”
What independent information is in the pebbles or the shells? None.
Ewert, Dembski, and Marks say nothing about detachable specifications in their formulation of algorithmic specified complexity. You seem not to accept that they have rebranded ASC as a measure of meaningful information. Their four publications on ASC are accessible through the “Publications” page at the website of the Evolutionary Informatics Lab. Have a look for yourself, if you doubt what I’ve said.
Dembski’s specified complexity is not Orgel’s specified complexity. Dembski eventually got around to acknowledging Dawkins as his source, but without going into details (see the interview of Dembski at The Best Schools). The notion that seemingly designed objects are “statistically improbable in a direction specified not with hindsight” comes from Chapter 1 of The Blind Watchmaker. What Dembski set out to do, without saying as much, was to establish that the property identified by Dawkins was not just the appearance of design, but instead “a reliable marker of intelligent design.” (The quotes are from memory, but I think I’ve got them right.)
Now you seem to be mixing in FIASCO (I don’t know the currently preferred permutation of the letters). Shudder.
If you look at the most recent publication of Ewert, Dembski, and Marks, “Measuring meaningful information in images: algorithmic specified complexity,” you will see that they in fact claim to measure meaningful information on the pixels in the image, not the object in the image. You should agree with me that it’s a big error on their part, rather than launch into stupid denial of the possibility of error by the champions of the ID movement.
Winston Ewert seems to disagree. According to him, not only is there information in the arrangement of shells, it’s meaningful.
Tell you what phoodoo, I’m going to assume Winston Ewert understands the ASC term he developed with Dembski and Marks, better than you do.
Tom English,
If you want to say that the defintion of specified complexity is not clear enough to you, fine.
But coming from an evolutionist, that is a pretty un-self-aware complaint.
I don’t see anything specific or meaningful about the shells at all. Pixels that add up to an image which conveys information and meaning however, that has specific complexity. The shells put together add up to nothing.
Quite a difference.
With apologies, I’m posting the first installment of my introduction to evolutionary informatics. I’ve had it in the pipe for way too long. We’ve played the waiting game with Ewert repeatedly, never to avail.
Not being a mathematician, that seems a bit risky. It sounds a bit like the claim you can distinguish a data stream produced by an algorithm from one produced by a not algorithm, whatever that is supposed to mean.
We’re all glad you’re here to perform this bare assertion without any kind of substantive demonstration that there in fact isn’t anything specific, or meaningful, about the shells, or that images have “specific” complexity, or that the shells add up to nothing.
One wonders why Dembski, Ewert and Marks bothered to write multiple essays and publications, not to mention entire books on the subject when phoodoo can sort of just ballpark it on intuition alone. Phoodoo states his opinion, nothing more required the matter is settled. Pack you bags guys, we’re done.
Rumraket,
I am pretty sure the only information they contain is that they are shells, and that the shells have probably been moved by something- perhaps crabs.
In that sense, any analysis of them would tell you exactly what you would expect. When shells get moved, something probably moved them.
Specified complexity reveals the truth.
Well you’re sure, Phoodoo, so you can leave us be. I once asked why you didn’t contribute at UD, but now I think I know. You couldn’t get past the math-based CAPTCHA.
Perhaps the, er, crabs do phoodoo?
They have meaning for the crabs, for sure.