False, strawmannish and demonising; to the point where an apology should be in order.

I’m sorry that you took offense, kairosfocus, but you seem to read through martyr-colored glasses. We presented several counterexamples, and you rejected all of them for various reasons. Where are the falsehoods, strawmen, and demonizing?

a –> Provide a case where an avalanche of rocks or equivalent produces a text message similar to “Welcome to Wales” at of course the border of Wales.

Personally, if I witnessed such an avalanche, I would attribute it either to God or to a very clever human illusionist. Perhaps someone dug trenches on the hillside in the form of English text to trap some of the falling rocks. Perhaps somebody performed the trick in a way that I can’t think of.

So a case that you would accept as a FSCI from unintelligent causes, I would reject as such. To each his own.

b –> produce a computer screenful of bits produces a screenful of text, say a passage from Shakespeare — FSCI by chance and necessity — lucky noise. [An update to the Million Monkeys banging away on typewriters]

It’s interesting that you refer to “chance and necessity”, but then give examples of avalanches, noise, and Zener noise. You realize, I assume, that not all chance+necessity processes are statistically random.

We all know that English text, and biological structures for that matter, aren’t produced by statistically random processes. Yet CSI and FSCI examples are virtually always based on a hypothesis of pure noise. (By calculating C*S*B in terms of the size of the config space or required storage space, you’re implicitly assuming a pure noise or random walk hypothesis.)

Sure, Dembski gives lip service to non-uniform hypotheses in his CSI approach, but he doesn’t take any into account in his examples. And in his active info work he explicitly rejects any ultimate natural causes other than uniform noise.

So if FSCI succeeds in ruling out only what has already been ruled out, namely lucky noise, then how is it helpful?

None of these or any comparable challenge has been met, and the reason why not is obvious.

Yes, it is obvious. It’s because avalanches don’t result in English text, Zener noise doesn’t produce Shakespeare, and the question of which phenomena fall under the “chance and necessity” category depends on one’s definitions and philosophical and religious assumptions.

Language of course is inherently ambiguous. It is context of usage that determines meaning objectively. And, all along there have been abundant examples and contexts that should make FSCI plain to all but the willfully obtuse.

It would be nice if we didn’t have to try to fill in the definitional details ourselves based on context and examples. Some of us are non-willfully obtuse and might get it wrong. Let’s look just at the terms functional and contingent.

Is the meaning of functional environment-dependent? Should only one environment be considered? If A causes B, and B is functional, is A also functional? (Note that Dembski claims this to be the case for specificity, or else CSI would not be conserved.)

How about contingent? Is a blotch from a spilled ink bottle contingent? How about the locations of fragments from an exploded bomb? How about the output of a pseudo-random number generator? Does contingency reflect a given observer’s lack of prior knowledge, or ultimate non-determinacy? If everything is ultimately deterministic, as it could be under Bohmian mechanics, then is there no such thing as contingency?

This is of course a complete twisting of recent exchanges in the eye on materialism thread etc.

How so? You attribute the FSCI in novel solutions generated by GAs to the human programmers of the GAs. You attribute the FSCI in ocean tides to those who record and analyze tidal data. You have offered no general method for determining the source of FSCI, but you always give the credit to humans.

You KNOW that the programs in question generate the text by being so programmed.

Certainly, but does designing the program equate to designing its output? If a programmer expresses the domain, structure, and objective function of a problem in code, is he designing a solution? If a natural language program utters a sentence that has never entered the head of the programmer, is it the programmer speaking? If so, then why doesn’t the FSCI in human-designed solutions and prose not get attributed to the cause(s) of humans? Is it because of a philosophical assumption that humans’ decisions, unlike computers’, are contra-causal?

Excerpt: The German defence was through anti-aircraft weapons and day and night fighters. The Kammhuber Line used radar to identify the bomber raids and then controllers directed night fighters onto the raiders. During the battle of the Ruhr, Bomber Command estimated about 70% of their aircraft losses were due to fighters.[10]. By July 1943, the German night fighter force totalled 550.[3] Through the summer of 1943, the Germans increased the ground-based anti-aircraft defences in the Ruhr Area ; by July 1943 there were more than 1,000 large flak guns (88 mm calibre or greater) and 1,500 lighter guns (chiefly 20 mm and 37 mm calibre).[5] This was about one-third of all anti-aircraft guns in Germany. [3] Six-hundred thousand personnel were required to man the AA defences of Germany.[3] The British crews called the area scarcastically “Happy Valley”[11] or the “valley of no Return”.

______________

1] Specs, 174: I am struggling to understand how a demonstrative calculation of the very thing that is being discussed (FCSI in a biological entity) is more distractive than the charge that it is a made-up metric that no one can actually calculate.

The issue is not that to calculate FSCI is distractive — note the strawman — but that precisely because it makes a telling point, strident rhetoric will be marshalled to distract from the point.

The standard tactic is red herrings led away to strawman distortions soaked in oily ad hominems and ignited to cloud, confuse, polarise and poison the atmosphere.

Or, the level of flak and such like goes right up as the bomber approaches the Ruhr; precisely because it is the critical target zone; thus, duly ringed with huge phalanxes of 88′s, searchlights etc.

Or, reverting to the terms of indoctrination: where a system is vulnerable to criticism is where those indoctrinated in it are trained to be the harshest in its “defence.” But, once we can break out os where such programming has anticipated and posed the standard rebuttal, and/or expose the rebuttal in a way that forces real thinking, programming tends to break down. That’s why there is so strong a pattern of distraction, distortion, demonisation and dismissal, never mind the implications of breakdown of civility for key institutions and for the culture as a whole. (On which I am distinctly pessimistic.)

2] herb: in order to compute the functional information using Hazen’s definition, we would need to know the fraction of those 4^27,000 configurations which would yield functionality at some specified level

Durston’s metric uses in effect a survey of the observed variability as an indicator. Cf his peer reviewed paper, here, and published table of 35 values here. (I cite the fact that the paper is yet another peer reviewed document to underscore yet another slander against design theory’s want of factual foundation. That one appears prominently in the infamous Judge jones ruling of 2005.)

In the simple FSCI approach, we need do no such elaborate thing: [1] once we see the scope of the search space, and [2] that the functionality is vulnerable to modest perturbations, we simply brig to bear [3] a scope of search on which the resources of the observed cosmos are hopelessly inadequate to scan any significant fraction of the space of configs. Large islands and even archipelagos can be lurking there, but we are not going to sample enough to have any confidence of hitting a desired shoreline of functionality. (And that is why the latest distractive tactic is to try to extend the shoreline of function into the sea of non-function, creating an imaginary continuous “fitness landscape.” Int he caser of life, for first life we need to complete an irreducible set: a coded blueprint, a corresponding reader and an implementer that step by step caries out the function of self replication. Absent any of these or derangement that breaks mutual integration, and the entity will not work. And, to get to novel major body plans, the changes have to be early in the embryological stages and have to integrate with the function of the organism, while requiring on evidence of novel structures and observed DNA sets, ~ 10′s – 100′s of mega bits of novel information, dozens of times over. Again, well into the range of infeasible undirected search.)

(To underscore, a search of 10^150 states in a space of 10^301 is less than the odds of marking a single atom int he whole cosmos for just one moment [10^-43 s] and then having a time and space travelling spaceship run around at random across all space and time and grab a single atom at random, and voila, it is the right atom at just the right moment.

(NB: There is no need to infer to a single unique functional state, to see that undirected search on forces of chance + blind mechanical necessity is not a plausible mechanism to find such instances of FSCI.)

GEM of TKI

Now, the flagellum has maybe 30 proteins and has multiple functionality, e.g. it self assembles and sets up an acid ion powered outboard motor, with an associated control system so the bacterium can use it to go where it “wants.”

A typical protein uses 300 20-state elements, each coded for by 3 DNA bases.

So, we are looking at 30 x 300 x 3 = 27,000 bases on a raw estimate. 4^ 27,000 ~ 4.17 *10^16,255. We may argue redundancies and mods to proteins all we will and this is not going to go below 1,000 bits worth of storage. And no probability or calculus required.

I agree that the evos will probably raise all sorts of objections in order to derail the conversation, but this basic information is valuable to IDers such as myself who don’t have much training in biology or genetics. I have to admit that in some of these threads I feel as if I’ve accidentally stumbled into a graduate seminar!

Now let me try out your numbers. Because the concept of FSCI has been discussed here quite a bit recently, and also because the definition is relatively simple, I’ll stick with that. Our config space has size 4^27,000, and in order to compute the functional information using Hazen’s definition, we would need to know the fraction of those 4^27,000 configurations which would yield functionality at some specified level—say the level achieved by the E. coli flagellum.

I personally have no idea how to estimate this fraction, but let me make the completely unrealistic assumption that only 1 out of the 4^27,000 configurations satisfies the condition, just to give us a starting point. Maybe someone else can help in revising this estimate. The functional information in these flagella would therefore be -log_2(4^-27,000) = 54,000 bits, well above the 500–1000 bit threshold.

1] Herb, 166: maybe it would be wise for us to go the extra mile and post a complete, self-contained example of an FCSI calculation for some “real” example from biology . . .

Sadly, that would simply end in endless distractive objections.

I am struggling to understand how a demonstrative calculation of the very thing that is being discussed (FCSI in a biological entity) is more distractive than the charge that it is a made-up metric that no one can actually calculate.

False, strawmannish and demonising; to the point where an apology should be in order.

In fact I have given a challenge on FOUR specific test cases, over the years, which can be re-presented as folows:

a –> Provide a case where an avalanche of rocks or equivalent produces a text message similar to “Welcome to Wales” at of course the border of Wales.

b –> produce a computer screenful of bits produces a screenful of text, say a passage from Shakespeare — FSCI by chance and necessity — lucky noise. [An update to the Million Monkeys banging away on typewriters]

c –> Using Zener noise sources to spew definitively random noise across disks produce a text string of 143 characters spelling out a good sentence in English.

d –> We have an Internet full of ASCII text strings known to be produced by intelligences. Provide a clear case of one 143 character text string in English produced by chance + necessity.

None of these or any comparable challenge has been met, and the reason why not is obvious.

6] We could spend weeks discussing the ambiguities in your C, S, and B criteria.

Language of course is inherently ambiguous. It is context of usage that determines meaning objectively. And, all along there have been abundant examples and contexts that should make FSCI plain to all but the willfully obtuse.

But if one is sufficiently determined to object to what is otherwise plain, no ends of zany misinterpretations and twistings can be manufactured.

7] If there is a human somewhere among the causal antecedents, you attribute the FSCI to that human. If there’s no human involvement in the causes, you bizarrely attribute the FSCI to the humans that observe and record the phenomenon

This is of course a complete twisting of recent exchanges in the eye on materialism thread etc.

In short at this point i am reading this as a red herring dragged out to a strawman soaked in ad hominems and ignited to cloud, confuse, poison and polarise the atmosphere.

For shame!

8] we can point to computer programs that produce “contextually responsive English”, but you’ll either claim that the output was created by the programmer, or you’ll say that the it isn’t good enough. If you set the bar at the level of the Turing Test, then it’s true that no computer to date can pass it.

You KNOW that the programs in question generate the text by being so programmed.

in short you are seekig to dismiss the fact.

We do not get to such programs by spewing Zener diode noise across disks, and testing for function then rewarding incremental success until voila we have a text responding program that spews out contextually responsive English text (never mind so good a response that blind tests w3ill be unable to distinguish man and machine.).

instead we use domain experts and programmers to DESIGN such systems.

As well you know.

Double shame!

Man, do betta dan dat, nuh!

___________

This one is ever so sad . . .

GEM of TKI

1] Herb, 166: maybe it would be wise for us to go the extra mile and post a complete, self-contained example of an FCSI calculation for some “real” example from biology . . .

Sadly, that would simply end in endless distractive objections.

How do I know that?

BECAUSE IT HAS ALREADY HAPPENED — the concept of FSCI started in 1973, with the issue of the observed and obvious distinction between biological organised complexity and order like a crystal or randomness like a tar:

“[L]iving organisms [-- thus, observed biofunction --] are distinguished by their specified complexity. Crystals are usually taken as the prototypes of simple, well-specified structures, because they consist of a very large number of identical molecules packed together in a uniform way. Lumps of granite or random mixtures of polymers are examples of structures which are complex but not specified. The crystals fail to qualify as living because they lack complexity; the mixtures of polymers fail to qualify because they lack specificity.” [[Leslie E. Orgel, The Origins of Life: Molecules and Natural Selection, pg. 189 (Chapman & Hall, 1973).]

Now, it so happens that functionally specified organised complexity is a commonplace of engineered systems and of software. [That is why I keep on pointing to the 143+ character ASCII text in contextually responsive English example.]

2] the simple FSCI metric

What I did at first level is to give a rule of thumb threshold and metric, to illustrate the principles behind Durston et al’s FSC and Dembski’s CSI metrics “for the rest of us.”

a –> Is it contingent (so capable of storing alternative configurations)

b –> I sit functionally specific? [e.g. disturbing the working config sufficiently by injecting random white noise will derange function.]

c –> Is it sufficiently complex that the number of possible configs will swamp the search resources of the observable cosmos? [At 1,000 bits we have ten times the square of the ~ 10^150 of Planck-time states that the 10^80 or so atoms of our observed universe would take up in 10^25 seconds, or ~3 *10^17 years. We usually date the observed cosmos at ~ 13.7 * 10^9 y.]

d –> We have in hand many millions of cases of FSCI, and for every one where we know directly the origin story, it is produced by intelligence.

e –> thus we have a strong induction.

So if something embeds an information storage system that is functionally vulnerable to perturbation and stores at least 1,000 bits [actually 500 would really be good enough!] we can be practically assured it is an intelligent artifact; unless some one can show a case of such functionality and complexity coming about by chance processes and blind mechanical forces.

The ducking and dodging you see above is because they are unable to do that.

3] say the bacterial flagellum . . .

I’ll double you on this.

Simplest observed life that is not parasitic on other life can by knockout studies go down to about 300,000 DNA base pairs. Below that, we are told, auto-destruction happens. And recall, living cells are von Neumann replicators: blueprint storage, reader, implementer to self-replicate, defining an irreducibly complex set that implies islands of complex function

Now, DNA — the blueprint storage element — is contingent, and functionally specific; indeed life forms have repair mechanisms for DNA. Observed life is well beyond the 1,000 bit threshold: 300 k bases is 600 k bits, or a config space of ~ 9.94*10^180,617. this swamps the search resources of the observed cosmos.

Now, the flagellum has maybe 30 proteins and has multiple functionality, e.g. it self assembles and sets up an acid ion powered outboard motor, with an associated control system so the bacterium can use it to go where it “wants.”

A typical protein uses 300 20-state elements, each coded for by 3 DNA bases.

So, we are looking at 30 x 300 x 3 = 27,000 bases on a raw estimate. 4^ 27,000 ~ 4.17 *10^16,255. We may argue redundancies and mods to proteins all we will and this is not going to go below 1,000 bits worth of storage. And no probability or calculus required.

Now, I believe such rough and ready calculations will be enough for the ordinary and unprejudiced mind, but the point is that, often, we are precisely not dealing with such.

As to a more sophisticated version, the Weak Argument Corrective no 27 links to the Durston and the Dembski papers and calculations.

4] this document 1) should refer to real biology, not drawing cards from a deck or whatever

The drawing cards from a deck case was actually raised by commonly encountered ID objector Mark Frank over at his blog, and I provided the answer there — he did not expect that. in addition, it is general purpose as Dembski’s CSI metric is general.

Duerston has published a table of 35 values of fuctional sequence complexity across various protein families int eh peer reviewed lite3rature. ter eis a zero concessions policy on such inconvenient facts for too many ID objectors.

[ . . . ]