The Argument from Intelligence

The crux of Stephen Meyer’s argument in Signature in the Cell can be succinctly summarized in this syllogism:
1. DNA information is complex specified information (CSI)
2. All complex specified information is generated by intelligent beings
3. Therefore, DNA information was generated by an intelligent agent

The first sentence is amply documented in his book following the definitions laid out over the years in ID books by Dembski and others. The second statement deserves a little more attention.
Meyer emphasizes the close relationship between DNA information and information such as computer code that is generated by human intelligent agents. On page 386 he states “Although a computer program may be similar to DNA in many respects and dissimilar in others, it exhibits a precise identity to DNA insofar as both contain specified complexity or specified information.” Throughout the book, Meyer repeatedly notes that specificity of information can be of two types, either deriving its specificity from functionality or else from meaning. Thus, there are at least two distinct subsets of information within the category of CSI. Accordingly, on page 359, Meyer titles one of his subsections “Two Types of Specifications.”
On page 109 Meyer states that DNA information is of the functional type. “Although DNA does not convey information that is received, understood, or used by a conscious mind, it does have information that is received and used by the cell’s machinery to build the structures critical to the maintenance of life. DNA displays a property–functional specificity—that transcends the merely mathematical formalism of Shannon’s theory.” Computer code is information that is used by a conscious mind. Hence, while a computer program and DNA information may be identical in being specified complex information, they differ in the type of specificity. This is significant with regard to its source.
Meyer presents many examples throughout the book to show that CSI can only come from an intelligent source. He might have strengthened his case by adding a deductive argument to his inductive approach. One of the characteristic traits of intelligence is the ability to do abstract reasoning. Abstractly-designed information, the “meaning” type of CSI, is therefore uniquely derived from an intelligent source. Had he done so, he might have noticed, however, that “functional” CSI is not in the group uniquely sourced from intelligence. Indeed, all of the examples that Meyer cites relate to information that derives its specificity at least partially from meaning and not solely from functionality. The latter depends on the physical and chemical structure of the system conveying the information while the former, which includes systems with a combination of meaning and function, depends on abstraction beyond the concrete reality.
As Meyer pointed out, DNA information is functional information, distinct from the meaning type of information. Since only the latter is known to be uniquely sourced from intelligence, it cannot be inferred that DNA information could only arise from an intelligent agent.

12 comments to The Argument from Intelligence

  • William Powers


    I don’t have the book, although I’m thinking of purchasing a copy. So I hope I don’t say anything that is adequately covered by the book.

    In this case, my question is simple. As I understand the relationship between DNA coding and the specific protein sequence that it codes to is contingent. There is no physical necessity in the relationship. Somehow this code was encoded, where here encoding means that a contingent physical link was established. We could start asking irreducibly complex kind of questions at this point. But we’ll hold off doing that to try to get at what Meyer is talking about.

    The relationship I described above is identical to a Morse Code decoder. The dots and dashes have been physically linked to letters. So far we haven’t introduced meaning. The meaning comes when an intelligent being reads the decoded message. Is Meyer saying that until that happens (and it may never happen) the specification is “functional.”

    Assuming this makes some sort of sense, I wonder whether if we ran across a Morse Code decoding machine we would conclude it was designed? Or is the reluctance to judge is designed an attempt to limit false positives?


    • Randy Isaac

      You are correct that Meyer is saying, appropriately I think, that there is no physical necessity in the specific sequence of nucleotide bases, at least from the perspective of the chemical structure. We’ll defer discussion of its function for future posts.
      As for the Morse code, the high level meaning isn’t available until it is decoded and then one can see that it is specified information. At a more fundamental level, however, abstraction occurs long before that. The very association of dots and dashes with letters is an abstraction, and indicative of an intelligent source. Indeed, even the existence of letters is an abstraction. There is nothing inherent about the physical structure of dots and dashes, or even of the shape of characters we use in our alphabet, that requires them to be associated with the function of conveying information of letters of the alphabet. In other words, it is all intelligently designed.

      • William Powers


        I agree with what you have said, although how you know that is the entire ID issue.

        But I don’t think you answered what I was trying to get at: Is the Morse Code Decoding machine an appropriate analogy to DNA and all its machinery?

        You seem to be saying no. But it’s not clear to me why you say so. Remember if some alien civilization found a Morse Code Decoder in the desert somewhere they wouldn’t have the privileged kind of information that you have.


        • Randy Isaac

          I’m sorry, I didn’t finish responding to all your questions. The Morse Code Decoding machine has some similarities to DNA decoding but for the purposes of inferring the source, it is not an adequate analogy. The key difference is that Morse code is fundamentally based on abstraction while DNA information is based solely on functionality and not on abstract meaning. No privileged information is needed for an alien civilization. The difference is that functionality without meaning depends on the physical and chemical characteristics while abstraction depends on its meaning and not on its physical properties.

          • William Powers


            I don’t mean to beat a dead horse, but I’d like to pursue this just a bit further.

            I’m still trying to grasp the difference between functionality and abstraction. I understand that a sentence encoded in some alphabet and semantic has a meaning that is independent of its medium (encoding). Whereas, functionality, you say, depends upon its physical realization.

            Functionality is not a one-to-one relationship. Chairs can function as a chair, and yet vary considerably in material and design. Yet two identical physical arrangements will serve the same function. We might say that function supervenes on the physical arrangement.

            It is not clear how this is different from the relationship between meaning and its physical expression. It seems that meaning too supervenes on the physical medium (whether it be sound, or letters, etc.). Yet the same physical expression produces the same meaning.

            So let me try again. Functionality depends upon its physical characteristics, you say. Depends in what way? Without a given set of physical properties, it cannot function properly. The same appears true of say the written word or audio system. The meaning is conveyed through a certain medium, but it order to do so properly the physical properties of ink and paper, sound and medium must have certain physical properties. Meanings cannot be conveyed by humans under water.

            I’ve backed myself into a hole. I can’t find a distinction. Maybe you can help.

            I note in ending one thing about my Morse Code Decoder. If I had never seen a Morse Code Decoder before, all I could tell about it is that it converts sequences of dots and dashes into thin lined drawings (we call them letters, but that’s more than I would know). The interesting thing about this process is that it appears to have no function. This is not true for the DNA machinery. The coding appears to serve as a part of a larger process, we call it protein synthesis. The Morse Code Decoder appears to have been ripped from a larger context. And this is interesting because in saying this we seem to be saying something about meaning. We don’t expect to find disconnected and “incoherent” processes. We expect to find “wholes.” In a sense this is the engine behind Paley’s watch. In the case of the Morse Code Decoder it appears to be “fatherless.” It’s existence causes us to infer that it has a “father.” This “father” is the missing chain of being in a cosmological argument. It may not be a personal creator, but we require that it have a progenitor, a creator of some kind.


            • Randy Isaac

              Bill, it’s not a dead horse and these concepts need to be thought through carefully. Thank you for persisting. I can’t do justice to all your questions in this reply and I have planned to devote several subsequent posts to addressing many of these questions. I’ll make a few observations here but go into more detail later.

              Yes, there can be redundancy in so far as different physical configurations can induce the same effect. That’s not the mark of meaning or abstraction.
              Yes, there are certain requirements for a physical system to convey information of any kind. That will become clearer as we discuss Shannon’s theory of information.
              For now, let me just repeat the basic point I want to make, which is that the only real evidence of an intelligent source of information is abstraction. The purely functional type of complex specified information may or may not be from an intelligent source.

              It isn’t always easy to ascertain abstraction in a system and it may remain ambiguous. One possible test would be to test for invarience with respect to the basic definition. If dashes become dots and dots become dashes throughout the entire system, including the encoder and the decoder, what happens? The message still comes through. In other words, there’s nothing inherent about the dash and the dot that requires them to have a specific meaning. It’s an abstract assignment of meaning. But indeed, the physical interaction between the signal and the decoder must be set up as designed. The presence of abstraction is evidence that Morse code transmission is from an intelligent source. The physical configuration is necessary for all types of information, whether from an intelligent source or not.

              Stay tuned.


  • Richard Blinne

    MIT is running a series on explaining Shannon here.

  • David Wallace

    “The crux of Stephen Meyer’s argument in Signature in the Cell can be succinctly summarized in this syllogism:
    1. DNA information is complex specified information (CSI)
    2. All complex specified information is generated by intelligent beings
    3. Therefore, DNA information was generated by an intelligent agent”

    Randy you are correct in that you summarize the explicit argument made in SIC. However, as I have read Meyer’s book I have become impressed with another message that seems more implicit.

    That message is based on the

    1. sheer intricacy of the human cell and the details of how proteins are made and then new dna etc
    2. state of research and thinking on OOL at the time Meyer’s did his research for SIC: as far as I can see they have not progressed very far and I understand why people are proposing the multiverse as a possible solution.

    These observations lead one to think that the probability of any wholly naturalistic solution is extremely low and approaches zero, what I intuitively think of as 10**-100. In other words unless God acted in what has been described as the governance aspect of his sovereignty that the first cell probably would not have begun the path to life as we know it.

    What I would like to do is try to separate concerns as much as possible between the ID conclusions and arguments and the more general message that natural factors leading to the first cell seem improbable. The first concern is what Randy is getting at and many here including myself are not comfortable with the ID argument as science and there appear to be numerous threads addressing that concern. Since I can’t make a post on this blog and this post seems relatively inactive I’d like to kick off a discussion here on this implicit message of SIC.

    First two disclosures:
    A. I have not read very much about front loading ideas except as Mike Gene wrote on the asa mailing list. But from what I have read I do not find them compelling but that may just be my lack of study.

    B. Since my high school education occurred on two continents and three locations I managed not to receive any formal education at all in biology, thus all I know on the subject I have learned as result of reading books on evolution and the odd Sci Am that I pick up. I would far rather read about physics, cosmology, and mathematics than biology.

    I agree with other commentators that the ID portion of the book is rather wordy and repeats arguments that have been made in prior ID books. However, that said I found the segment of the book dealing with what I’m referring to as the implicit message very good and extremely helpful although someone with training in the field likely found those portions rather boring.

    I have two questions:

    I. Is Meyer’s overview of how the cell functions and replicates accurate? I expect so as I don’t read much in the blog sphere about people like Dawkins or PVM trashing that part of the book.

    II. Is the description of the state of the research in OOL correct at the time the book was written? Have there been great leaps in understanding since the research was done and OOL is now a more or less solved problem, in other words has something equivalent to Charles Darwin’s “On the Origin of Species” been published in the technical literature?

    (Note to editors I have created this post in wordpad and copied and pasted the contents into the comments window. On other blogs I have lost comments when I hit submit or have paged up in the window and seen something interesting and then navigated away from the blog thus loosing all my comments, at which point I give up in frustration. I checked things in notepad and hope no rtf formatting gets copied over. I also find it much easier to write my thoughts when I can see a decent sized window.)

    • Randy Isaac

      Dave, I agree that the state of research into the origin of life is a most fascinating aspect of Meyer’s book. The intent here is to work our way systematically from the information part to that OOL research. I know you are eager to get there and there are a lot of interesting things to discuss. As a glimpse into what I hope to discuss then, my feeling is that his description of the biochemistry of living cells is generally accurate but not so close for OOL research. More specific discussion on this is found in Darrel Falk’s blog .
      As for the probability calculations, I think most OOL researchers would say Meyer is analyzing the wrong processes. In general he goes from a random, disconnected set of building blocks to a fully functional protein. But OOL folks are looking for intermediate precursors where each step is quite probable. More on that later.

  • David Wallace

    I had started to read Falk’s piece but gave up and just skimmed it when I got to the bit on research that came out while SIC was in process of being written and published. I need to go back and read his post and the comments more thoroughly.

    One of their problems at Biologos is that it is not possible to subscribe to their blogs or to comments so I find them hard to follow. Have written in a suggestion to them…

    Also I do not know much about the people at BioLogos since Francis Collins became the Surgeon General. The impression I get from reading other sites on the net is that they are more into naturalistic evolution-OOL than anybody I have seen here with the exception of PVM who dropped out quite a while back from the email list. Maybe I have gotten a bad impression as the people now running BioLogos are not on my radar screen at all.

    • Allan Harvey

      Sounds like you have gotten a mistaken impression of BioLogos. There are 3 main people there now. Darrel Falk is a biology Prof. at Point Loma Nazarene (author of the excellent book “Coming to Peace with Science”). Karl Giberson is a scholar in religion and science on the faculty at some Christian school in the East (Gordon?). Peter Enns is a top evangelical Old Testament scholar, wrote the fine book “Inspiration and Inspiration”, was on the faculty at Westminster seminary until he was pushed out for not being fundamentalist enough.

      They do stick up for evolution as God’s means of creating, which of course causes some negative reactions. But they are right there with Francis Collins in terms of faith commitment. In terms of the old ASA list, think more along the lines of George Murphy or Loren Haarsma or Keith Miller or Terry Gray, definitely not PVM.

      To see some of their evangelical commitment, go under Projects, Workshops on their website and see material about the 2009 workshop they co-organized with Tim Keller.

  • David Wallace

    “As for the probability calculations” I tend to discount that part of the writings of ID folks thus when I quoted a probability it was my gut feeling based upon the intricacy and complexity of his description of the cell workings. As I wrote on the email list I no longer read Dembski’s books where he is talking about statistics and information theory. I’m sure Rich could point out difficulties better than I can.

    I know humans parse things better than machines but I’m stumped on “As a glimpse inot what I hope to discuss then, my feeling is that his description of the biochemistry of living cells is generally accurate but not so close for OOL research”

    I think you are telling me to wait a bit but I’m not sure. If so then I am fine to hold off. I need to read all the asa posts/comments on your blog re information again, and then try to respond. However I am not an expert in information theory and any way I think Shannon or Chatin/Kolg… algorithmic theory is not what they should be talking about as Dembski does in his books and now SIC does to some degree.

January 2010
« Dec   Feb »