Complex specified information (CSI), also known as "specified complexity", is a concept that is often used by creationists, especially those of the Intelligent Design variety. Their argument, briefly, is that anything that demonstrates CSI can only have arisen thru intelligent design, and not thru naturalistic processes like evolution.

Creationists persist in making this claim despite their continued failure to support it with evidence of any sort. Indeed, they have shown themselves unable to even demonstrate how CSI can be calculated in given hypothetical or real scenarios. This has not deterred creationists from insisting that CSI is not only a real, quantifiable, entity, but a useful one as well for demonstrating the existence of God.

Here on RationalSkepticism we have been privileged to welcome one the more promising young members of the Intelligent Design Creationism movement, Jonathan McLatchie. He has been generous enough to share one of his articles on the concept of CSI, which can be found here:

http://www.evolutionnews.org/2013/08/tw ... 75771.html

Now, part of that article attempts to provide a definition of CSI, and as you can see it remains a frustratingly vague and ill-defined concept. But Jonathan then goes on to provide what he considers to be examples CSI, one of which follows:

The second problem with objecting to specified complexity is that most or all of the arguments for common ancestry are in fact based on reasoning from specified complexity. The discovery of precisely the same improbable evolutionary events (e.g. parallel point mutations or mobile element integration) in multiple lineages begs for explanation. If the complex event in question occurred independently in each of the separate lineages, then it would constitute specified complexity, and thus be suggestive of some kind of teleology. Thus, the most parsimonious explanation from a materialist standpoint -- i.e. that the event in fact only happened once and each of the lineages subsequently inherited the resultant changes from a common ancestor -- is favored.

Now, it is important to note that Jonathan has confirmed that he accepts common ancestry, so he is not intending this as an argument against that. However, he does not seem to have realized the full significance of what he has written there. He is saying that genetic markers of common ancestry between varies lines of descent meet the definition of CSI. And therefore the only conclusion this can lead to is that CSI does not indicate intelligent design. He explicitly states that the presence of these markers in two or more separate lineages would constitue an example of specified complexity. Yet he also admits that "the most parsimonious explanation from a materialist standpoint" is that this CSI exists as a result of natural processes such a mutations, reproduction and heredity. There is no need for an "intelligent designer" at any point of the process. By implication, the only reason one would reject this "most parsimonious explanation" in favour of one dependent on intelligent design is if one has a prior ideological commitment to the rejection of "materialism" and instead feels the need to invoke supernatural explanations.

You will notice this article was posted only a few weeks ago on one of the leading websites of the Intelligent Design Creationist movement, so I'm sure the impact of its ramifications is only just beginning to be felt by advocates of that movement. However, since ID is supposed to be dedicated to the pursuit of science, and is not committed to any religious or other ideologies, then surely it will only be a matter of time before other ID creationists acknowledge that the hypothesis that CSI can only be produced by intelligence has been falsified by one of their own acolytes, and cease to use it as an argument against evolutionary theory.

Congratulations, Jonathan!

If anyone really wants to get into the nitty gritty of the failures of GSI I highly recommend a visit to http://theskepticalzone.com/wp/ where this subject often comes up (just put CSI in the search field). As shrunk correctly points out, and what's particularly hilarious of the whole thing, is that nobody ever seems able to actually calculate it.

It's interesting that champions of the ID project invariably create a quibble-zone of apparent dispute around a particular issue, and point at it and say, as if with a knowing chuckle, "Look, current science just doesn't entirely understand this." In pointing out that science doesn't have a complete answer they obviously wish to imply that there is an obvious and complete answer: Goddidit.

What they appear reluctant to accept of course is that current science is quite happy with not having a complete answer as long as the parts of the answer it does claim to have can be assured, verified, and independently validated; what is verified is known (until something better comes along) and what isn't verified and known invites further consideration, scrutiny and research, with the hope of perhaps pinning a little more knowledge down along the way.

What creationist are using to criticise and denigrate our current understanding is the lack of a completeness, in circumstances where say complete understanding is mandatory and an incomplete explanation is equivalent to knowing next to nothing. Proffering God as a complete answer tawdry, intellectually dishonest pursuit, particularly when ID/creationism cannot support its claims (merely dispute the checkable claims of science), and when it seeks to undermine the parsimonious explanation in favour of the presumed necessity of the magical explanation.

For this reason I think it is nearly always useless to engage the ID/creation proponent directly on scientific detail unless one has first engaged them on their underlying assumption; which is, that only a complete explanation of cosmology or evolution etc that meets with their approval shall be called proper, truthful knowledge.


Just thinking out loud really.

The design /non-design argument is something of a red herring, because it leads to no long-term understanding of what science is about. The primary, and fatal fail of intelligent design is the designer agency posited by ID. Neither ET or a designer gods, are in evidence, and therefore must be tentitivly rejected until such evidence exists, and only then could anyone have a reason to re-open the investigations into the possibility of intelligent design.
The logical research program, and one entirely compatible with methodological naturalism, is to search for natural agencies or algorithms that are apparent from observing natural phenomena. Thus the patterns like snowflakes arising out of physical law or the power of natural selection and other mechanisms to filter out unfit patterns or designoid functionality, and thus leave us with the illusion of the "designed" organism.
That science is not about reality, but rather the testing of models concerning natural phenomena really makes any regard of the DEEP alledged disconnect between designed and designoid processes somewhat moot.
This is not to say that the design concept [in the creationist meaning of the word] should not be challenged, but used as an opportunity to wedge the designoid concept into the creationist mindset. This in turn, may lead to the acceptance that natural, rather than posited supernatural forces, is quite sufficient to give the illusion of pattern and design in both the living and non-living areas of "nature".

In mounting a too vigourous denial of apparent design, we really sacrifice a valuable pedagogical‎ technique, because we offer what might be perceived by the more "intellectually honest' creationist resistence to what they might believe as reasonable. While design [in the Paley sense] is illusion, the designoid concept is not. So if we can convince that design and designoid structure and function can have diverse creative origins, then we are half way there.

Conceptually, the main failure of creationist/Intelligent design "science" is the ad-hoc nature of their modelling technique, rather than the believability of gods or ET as causal forces. Thus when we point out some difficulty or detail in their models [such poor 'design" by a creator-god], they will reply with ad-hoc and independently un-testable apologetics.

This does not deny that many creationist/IDers are intellectually dishonest [either by deliberation or lack of insight], and the vast majority of them are trying to prove ideas that they already believe are true. However, such is their scientific naivity, we should not waste any opportunity that may provide an inroad to evolve their thinking processes into a more scientific mode of thought. This can hardly be done if we press undue emphasis on their mundane or Umwelt belief systems, which after all were the original tools that science had to use in its infancy.
Only as we progressed and learned that naive realism [which is pushing creationists to adhere to the concept of design in nature] did we learn that science is not about truth or reality, but about constructing and testing falsifiable descriptive and predictive models concerning natural phenomina, and tentatively accepting such notions, subject always to review in the light of future discoveries.

To the creationist, god is real. This is the primary delusion in the whole system. It is a naive-realist view, and exactly equivalent to naive-realist views held by scientists for most of the history of science.
We have to sell them methodological naturalism. It may not be about reality, but it works, and works powerfully well. of course, from a metaphysical rather than a scientific perspective, science probaly works because in observing natural phenomena, we are seeing at least enough of reality or true nature to suggest that this is why science is so successful, and hence there is probably no mystery to the "why" of scientific success.

Another problem arises quite simply because all too many of the devotees of creationism simply don't understand what science is about. They come from a background in which mythological assertions are claimed to dictate how reality behaves, and assume that science is in the same business, namely dictating via a priori assertions how reality behaves, and can therefore be summarily dismissed on the basis that science is purportedly erecting "wrong" assertions, deemed to be thus because they don't agree with the assertions of mythology, regarded a priori as purportedly the "right" assertions.

That this view they hold of science is not merely wrong, but a travesty thereof, is the first weapon in our arsenal. Because science isn't in the business of erecting blind assertions, it's in the business of testing assertions to destruction. Any assertions not consonant with observational data are tossed into the bin. Exposing this basic error on the part of creationist devotees is the first step that needs to be taken - namely, informing them that the modus operadi of science is dametrically opposite to that of religion. In short, the religious modus operandi consists of claiming that the products of the televisions inside our head magically dictate how reality behaves, regardless of whether or not reality agrees with this. The supernaturalist regards the assertions of his favourite mythology as purportedly constituting "axioms" about the world, which the world must necessarily conform to. Science, on the other hand, regards assertions, and indeed all of our ideas, as discardable entities, and regards consonance with observational data as the appropriate test to apply, in order to decide which ideas to discard.The moment an idea fails to be consonant with observational data, that idea is discarded, unless that idea possesses utility value with respect to a subset of that observational data. See, for example, the continued use of Newton's ideas where the error in so doing is too small to measure, and far too insignificant to influence the outcome, a topic I've covered at greater length in previous posts.

In short, science asks the following basic questions:

[1] Is a given idea or assertion consonant with observational data? If the answer is "no", discard the idea or assertion.

[2] If the answer to the above is "yes", does that idea reliably predict future observations? If the answer is "no", look for a better idea.

[3] If the answer to the above is "yes", said idea becomes the basis of a scientific theory. Subject of course to revision as more observational data becomes available.

Note that I've exercised scrupulous effort here to centre scientific tests of assertions upon observational data, without asking whether or not that observational data constitutes "reality" in the naive sense. I personally happen to think that if said observational data isn't at least a component of what might be termed "actual reality", then the problems arising with respect to matters of parsimony alone are serious. Not least because we have plenty of empirical evidence to the effect that mere illusions don't work, and I enforce the distinction here between inability to observe relevant data and illusion in the sense that the data is in some sense "misleading", not least because Newtonian physics was a product of the former, revised when that physics bestowed upon us the ability to fill in the observational gaps. For appropriate subsets of observational data, Newtonian physics works very well, which it wouldn't if its ideas were completely wrong. However, scientists turn to different ideas for observational data outside those subsets, and do so not only because those ideas are consonant with that larger body of observational data, but also because those ideas provide an explanation as to why Newton's ideas work so well with the relevant observational subsets.

Since the entire history of science is one of discarding ideas demonstrated not to be consonant with observation, we should be promoting this as one of its virtues. Admitting that a previous set of ideas, a set of ideas that worked so well with relevant subsets of observational data, is no longer consonant with new observational data, and that scientists are working hard to build a better set of ideas that is consonant with both the old and the new, is the great strength of science, one we should be promoting hard, and whilst doing so, we should be noting the contrast between this entirely proper scientific virtue, and the vice endemic to supernaturalism of treating ideas as "sacrosanct" or "beyond question", regardless of whatever vast mountains of observational data render those purportedly "sacrosanct" ideas eminently discardable.

Rumraket wrote:If anyone really wants to get into the nitty gritty of the failures of GSI I highly recommend a visit to http://theskepticalzone.com/wp/ where this subject often comes up (just put CSI in the search field). As shrunk correctly points out, and what's particularly hilarious of the whole thing, is that nobody ever seems able to actually calculate it.

The math there quickly gets too complicated for my level of knowledge, but she makes at least one point that I am able to grasp and which should be much more obvious to the ID Creationists than it seems to be: The argument for design from CSI cannot be made until one figures out a way to calculate the odds of an entity demonstrating CSI arising thru evolutionary processes. That is a gaping hole in the argument, of which the creationists do not even seem to be aware.

Shrunk wrote:

Rumraket wrote:If anyone really wants to get into the nitty gritty of the failures of GSI I highly recommend a visit to http://theskepticalzone.com/wp/ where this subject often comes up (just put CSI in the search field). As shrunk correctly points out, and what's particularly hilarious of the whole thing, is that nobody ever seems able to actually calculate it.

The math there quickly gets too complicated for my level of knowledge, but she makes at least one point that I am able to grasp and which should be much more obvious to the ID Creationists than it seems to be: The argument for design from CSI cannot be made until one figures out a way to calculate the odds of an entity demonstrating CSI arising thru evolutionary processes. That is a gaping hole in the argument, of which the creationists do not even seem to be aware.

Yes, exactly. Dembski attempts to rule out "all relevant chance hypotheses", but in his calculation example he only rules out a fair coin. He simply can't calculate the odds of evolution by natural selection and genetic drift. Nobody knows how to do that.

Rumraket wrote:

Shrunk wrote:

Rumraket wrote:If anyone really wants to get into the nitty gritty of the failures of GSI I highly recommend a visit to http://theskepticalzone.com/wp/ where this subject often comes up (just put CSI in the search field). As shrunk correctly points out, and what's particularly hilarious of the whole thing, is that nobody ever seems able to actually calculate it.

The math there quickly gets too complicated for my level of knowledge, but she makes at least one point that I am able to grasp and which should be much more obvious to the ID Creationists than it seems to be: The argument for design from CSI cannot be made until one figures out a way to calculate the odds of an entity demonstrating CSI arising thru evolutionary processes. That is a gaping hole in the argument, of which the creationists do not even seem to be aware.

Yes, exactly. Dembski attempts to rule out "all relevant chance hypotheses", but in his calculation example he only rules out a fair coin. He simply can't calculate the odds of evolution by natural selection and genetic drift. Nobody knows how to do that.

I can: practically 1. The uncertainty is in the odds of abiogenesis. Once that has happened, evolution is inevitable, because it's what life does.