Critique of ID logic

From: Keith B Miller (kbmill@ksu.edu)
Date: Tue Jul 11 2000 - 15:27:52 EDT

  • Next message: Doug Hayworth: "Re: Critique of ID logic"

    Below is a post to another listserve that responds to the following list of
    steps to identify intelligent design. This "algorithm" was presented by a
    leader of the Kansas intelligent design network. I think the critical
    analysis of this algorithm presented below is very illuminating.

    Keith

    >I think that it is important to note that one can do exactly what John
    >Calvert (ala Dembski) is saying, and it can have some validity. For example
    >consider the SETI research, which is pretty much based on this very idea.
    >The 'devil' is in the details, however.
    >
    >First lets take the steps 1..3 and analyze them.
    >
    >* 1. Find a pattern of events that is functional, carries a message or has
    >some discernable structure - that reflects "specified complexity." *
    >
    >Since anything that is "functional" or "carries a message" inherently "has
    >some discernable structure", the statement can be reduced to simpler
    >elements. Also a "pattern of events" is really referring to a pattern in
    >physical realization, and isn't specific to "events" in sequence, so we can
    >leave out "events" since every pattern is by that broad usage a pattern of
    >"events". (They consider a lifeform to be a "pattern", so the "events"
    >aspect is superfluous.)
    >
    >Furthermore "discernable structure" is redundant to pattern -- what is a
    >pattern in this context if it is not just a "discernable structure"?
    >
    >All the other words are intended to mislead by giving truly "designed"
    >examples of patterns that are messages and "functional" objects. They are
    >going to start with examples of well understood "patterns" that are
    >"functional", and things that carry a message, and "inductively" show that
    >the 'algorithm' is useful for detecting 'design' for these cases, implying
    >that they are also covering the most general "has some discernable
    >structure" case. (They have probably mislead themselves into believing this
    >is meaningful, except Calvert is a lawyer and lawyers are trained not to
    >have to believe their own arguments, just make them convincing.) So just for
    >clarity we will stick to the most general case in the sentence structure, so
    >as to avoid confusion in the argument.
    >
    >Now elsewhere 'complexity' is defined as something that is complex, not just
    >simple (more or less).
    >'Specified' is defined as anything that we can create a mental specification
    >for, either before or after seeing its realization. (A position on a wall at
    >which an arrow has landed can be considered a "specification", so
    >specification is a most general concept, it only requires that one of us be
    >able to conceive of it either before or after the "event" in which the
    >"discernable structure" is exhibited.)
    >
    >At this point we are left with:
    >
    >* 1. Find some discernable structure - that reflects "specified complexity".
    >*
    >
    >Now lets deal with the rest of the sentence. Why do we need new terms,
    >"specified complexity" to deal with something that already has a simple
    >understood terminology, "discernable structure". If it has structure, surely
    >it has some complexity, so we don't need to furthermore identify its degree
    >of complexity. Since anything we can conceive of we can have a
    >specification, why do we need to refer to the "specified" aspect. Once again
    >"specified" in normal day to day terminology only refers to something this
    >is written out or conceived of prior to its realization in physical reality
    >(e.g. production.) So lets drop the terminology "specified", but remember
    >that we dropped it, so any aspects of our concept that required the need for
    >specification will still apply.
    >
    >Thus we are left with:
    >
    >* 1. Find some discernable structure. *
    >
    >Now as to steps 2 and 3.
    >
    >* 2. Second: Rule out Necessity as a cause of the pattern. *
    >
    >* 3. Third: Rule out Chance as a cause of the pattern. *
    >
    >* If we can do that, we can infer design.*
    >
    >First of all in the context of things in nature that we are going to
    >examine, we can't separate 'chance' and 'necessity' as independent. They are
    >different aspects, just not independent or acting separately. All known
    >physical phenomenon have elements of 'chance' or randomness involved. For
    >example all measurements have experimental error, and those errors are
    >exhibited as various 'chance' aspects (possibly unknown) affect the final
    >measurement. Furthermore all known elements of 'chance' in nature depend on
    >some aspect of 'necessity' or physical process. Here we are to understand
    >that 'necessity' means a physical (or natural) process.
    >
    >Now this is furthermore in a context of considering all things to be a
    >result of either "design", "chance" or "necessity". We will be asked if any
    >"event" can fall outside of the conjunction of "design", "chance" or
    >"necessity", and I am willing to concede that any physical event
    >(discernable structure) cannot occur outside of some combination of those
    >aspects. What I am not willing to concede is that multiple aspects cannot be
    >simultaneously present. (For example a bowler throws a ball, intending or
    >'designing' the knocking down of 10 pins. There is always some randomness in
    >how the ball rolls, and the rolling of the ball is clearly a physical
    >process or occurring by "necessity" once it is released. Therefore these can
    >occur in combination. The real question asked in ID is if they are caused
    >'ONLY' by chance and/or necessity.
    >
    >Now in the evolutionary context, chance and/or necessity is really just
    >another way of saying 'natural process' outside of those which are part of a
    >directly observable case of 'design'. (Example to illustrate: Take a
    >sequence of letters coming over a teletype. If we see the operator typing
    >the letters, we directly observe the 'design' aspect. However if we can't
    >directly observe in any way the operator, we can only examine for the
    >possibility of other processes. Wires rubbing together could produce random
    >teletype signals which would produce random letter sequences, explained by
    >physical causes or 'necessity' combined with the random nature of the
    >physical process.)
    >
    >So statements 2 and 3 together reduce to:
    >
    >* 2/3. Rule out natural processes outside of those by an observable
    >'designer'. *
    >
    >Now back to the evolutionary context, we know we don't have direct knowledge
    >or observation of a 'designer' and are working only to find indirect
    >knowledge, so we can reduce the sentence to:
    >
    >* 2/3. Rule out natural processes outside of those of a 'designer'. *
    >
    >
    >SO the algorithm is equivalent to:
    >
    >* 1. Find some discernable structure. *
    >
    >* 2/3. Rule out natural processes outside of those of a 'designer'. *
    >
    >* If we can do that, we can infer design. *
    >
    >[Now I disagree somewhat with Brian, in that I think "1." is mostly a matter
    >of definitions. What is important is not to change the definitions midway
    >through argument. His methodology really depends on changing the definitions
    >midway, so his case does not rest on the original definitions but on his
    >ability to confuse us when he changes the definitions midway through the
    >argument. I agree that this is just a case of dressed up argument from
    >ignorance, as I will show.]
    >
    >This 'algorithm' works somewhat in the SETI example. We receive and record
    >some RF energy. We try to find some discernable structure. We check to see
    >if we can explain the structure using natural processes. If so, it isn't
    >designed, even if it has regular structure.
    >
    >Now the ruling out of natural processes is very difficult, but when we
    >detect a known natural process we do rule it 'IN' and cast out the case as
    >not a product of 'design'.
    >
    >However we go a step further in SETI, beyond the above algorithm. We don't
    >assume that every pattern received must have an understandable natural
    >process behind it, because of the great number of possible natural processes
    >that we don't have fully explained, and which could create the energy
    >pattern.
    >
    >So we furthermore use models of physical or natural intelligence which could
    >do design of a signal, and transmit that signal. For example someone sending
    >us a message who knows we won't know their language, and furthermore may not
    >even be looking for a language might send out sequences of prime numbers as
    >some sort of simple count. Other characteristics taken from our own examples
    >of natural intelligence can be searched for.
    >
    >So the basic 'algorithm' is useful for SETI, but not without some added
    >considerations.
    >
    >Now how would it be useful for evolution related research?
    >
    >Well we could use the 'modified' algorithm like SETI uses, and look for
    >directly 'designed' aspects. Look for sequences of prime numbers encoded in
    >a gene sequence, which could have been placed by genetic engineering for us
    >to discover at a later eon. Look for other characteristics of 'design' by
    >natural intelligence. Square tomatoes might exhibit such a characteristic,
    >because they are apparently 'designed' to fit a packing crate better, rather
    >than to fit common effects of nature.
    >
    >However the modified 'algorithm' appropriate for SETI is not very
    >appropriate or applicable to the general question of evolution, so we are
    >left with the original 'algorithm'. The only types of intelligence we are
    >searching for in the modified 'algorithm' are ones who act in natural
    >systems and furthermore act according to ways that we recognize as relating
    >to human intelligence, working within the materials of nature.
    >
    >Now since the evolution question deals with much evidence from before the
    >appearance of 'natural intelligence' we know of, the question of a
    >'designer' is somewhat extraneous to the statement in a specific way. I am
    >really saying that the 'algorithm' can be reduced to the following for
    >practical purposes:
    >
    >* 1. Find some discernable structure. *
    >
    >* 2/3. Rule out natural processes. *
    >
    >* If we can do that, we can infer design. *
    >
    >The aspect of the 'designer' is implicit, and doesn't need to be stated. The
    >'algorithm' is still intact in the sense that this is really the way the ID
    >folks want to search for the 'intelligent designer'.
    >
    >So how can we carry out the 2/3 algorithmic step, as stated?
    >
    >Well here is another statement, which I think we can really understand says
    >the same thing:
    >
    >* 1. Find some discernable structure. *
    >
    >* 2/3. Do scientific research to understand the natural processes. Do this
    >research just like it is presently done, and just like scientific research
    >would advance if we were not to consider ID. When and if all reasonable
    >human effort at explaining a structure has been exhausted, then give up. If
    >future ideas might re-invigorate an area of research, don't consider that we
    >have given up for all time, just temporarily until the next good idea or
    >project. *
    >
    >* Remember that to infer design, we first have to have ruled out natural
    >processes. To do so required a great deal of research, probably many many
    >lifetimes. If we have not ruled out natural processes, we can not yet infer
    >design, because this was the statement of the 'algorithm'. We keep
    >discovering more and more details of natural process that can explain
    >evolution, and to stop this part of the search in order to support an
    >argument from ignorance would be foolish. *
    >
    >Now there is a further wrinkle. Remember my earlier point about the
    >combination of chance, necessity, and design all occurring in the same event
    >(e.g. the bowling ball hits the pins)?
    >
    >Well lets consider "theistic evolution", an idea held by many scientists. In
    >this possibility an "event" can simultaneously be a product of God's design,
    >and a result of chance and necessity.
    >
    >Before considering this, however, lets break the analysis up into two cases:
    >The first case is design occurs because of purely naturalistic processes
    >(e.g. the designer is an extra terrestrial or a natural being present on
    >Earth). The second case is a 'designer' beyond natural processes (e.g. God).
    >
    >In the first case we modify the algorithm to only search for natural
    >designers, and this would be the 'modified algorithm' useful in SETI,
    >mentioned above. They have used case analysis to describe this process, so
    >there certainly is no problem breaking this algorithm up into cases. Now to
    >find a natural designer we must be able to find mechanisms of design
    >transfer. If design occurs at many different points along the evolutionary
    >time line by a natural designer, then we should be able to find at least a
    >small amount of direct evidence of this designer, or how the designer
    >effects the transfer of the design to the evolved species. This completely
    >changes the process, and I don't believe that the ID people would be
    >satisfied with this version, so I won't even present it further.
    >
    >Therefore we are only left with the second case, that the possibility of a
    >'designer' outside of natural processes.
    >
    >Lets not do 'science' for a moment, but rather consider a larger endeavor of
    >human activity including science and philosophy and religion, all under one
    >heading. We shall furthermore not limit ourselves to naturalistic processes
    >in this endeavor. In this context the people doing this activity believe in
    >God and in God's creation, but also simultaneously practice the methods of
    >scientific research.
    >
    >I'm just bringing in God because we are told we are to allow for God as one
    >of the possibilities for the Intelligent Designer, and I think this specific
    >case using God is instructive -- we would not always have to limit ourselves
    >to God as the explanation for non-naturalistic processes. Since we are
    >allowing the 'God' possibility for the designer, I believe that that is the
    >most extreme case in terms of its effect on our understanding of the
    >definitions we are using. It definitely has consequences and effects on our
    >definitions of terms, so I want to consider that one case within the general
    >second case of a designer beyond natural processes. Now we will have a new
    >algorithm, specific to this larger endeavor with an explicit statement of
    >the aspect of God as well. We will use some added terminology to reflect the
    >religious aspect in every step.
    >
    >* 1. Find some discernable structure. Now any such structure was caused by
    >God and continues to be upheld by God's present will, because all of
    >creation is God's creation. *
    >
    >* 2/3. Do scientific research to understand the natural processes.
    >Understand that all natural processes are the result of God's will, and
    >continue to be upheld by God's present will. God may have chosen for His
    >creation not to be understandable by human beings. But if we do find
    >understandings, then God apparently intended us to be able to understand
    >nature. Do this research just like it is presently done, and just like
    >scientific research would advance if we were not to consider ID. When and if
    >all reasonable human effort at explaining a structure has been exhausted,
    >then give up. If future ideas might re-invigorate an area of research, don't
    >consider that we have given up for all time, just temporarily until the next
    >good idea or project. *
    >
    >* If we can do that (have to give up for all time on finding an
    >explanation), we can infer design. We have identified cases of God's design.
    >*
    >
    >* (Alternately when we do find explanations for a discernable structure
    >based on natural processes, we have also found examples of God's design,
    >because all of creation is God's design.) *
    >
    >Note that we can eliminate these last two statements, because in either case
    >we identify design. The algorithm doesn't identify God's design, but it may
    >have other valuable uses as a human endeavor.
    >
    >Now let's carry this one step further. When we carry out this larger human
    >endeavor, we will keep separate logs, (mental or physical) of all the
    >aspects that mention religion or God. We do this for the following reason:
    >Others want to work with us who have different views of God and religion
    >than ours, and we have problems communicating when we use our particular
    >view in all of our correspondence. So we leave this aspect out of one
    >account, and just focus on the natural processes. We don't change the larger
    >result, including the religious aspect, we just document that religious part
    >separately if we do document it.
    >
    >So do this larger human endeavor, with a final modification of the
    >methodology, but which includes a religious aspect that some scientists
    >believe, and others differ with. Report the religious aspects as philosophy
    >and religion, not as science.
    >
    >* 1. Find some discernable structure. *
    >
    >* 2/3. Do scientific research to understand the natural processes. Do this
    >research just like it is presently done, and just like scientific research
    >would advance if we were not to consider ID. When and if all reasonable
    >human effort at explaining a structure has been exhausted, then give up. If
    >future ideas might re-invigorate an area of research, don't consider that we
    >have given up for all time, just temporarily until the next good idea or
    >project. *
    >
    >Call this final 'algorithm' or methodology "science", and teach its results
    >to our children.
    >
    >Sincerely
    >Gordon Elliott

    Keith B. Miller
    Department of Geology
    Kansas State University
    Manhattan, KS 66506
    kbmill@ksu.ksu.edu
    http://www-personal.ksu.edu/~kbmill/



    This archive was generated by hypermail 2b29 : Tue Jul 11 2000 - 15:23:01 EDT