A Comparison:
General Intelligent Design Versus Restricted Design Theory

Robert A. Herrmann Ph. D.
12 NOV 2000. Last revision 10 JUN 2017.

In this article, I contrast the design theory I originated in 1979, with that of the more recent notions of design as championed by members of the Discovery Institute (DI) in Seattle WA. The original 1979 mathematical model for intelligent design, at that time, was called "general design theory" and now is denoted as the General Intelligent Design model (GID-model or simply GID). The DI form of design theory is termed as Restricted Intelligent Design theory (RID) by me and I hope all of humanity. This article discusses intelligent design as it is related to physical entities and their behaviors. The GID-model is an "intelligently designed" or intelligent agency interpretation of the General Grand Unification model (GGU-model), where two types of "intelligence" are investigated. Relative to the GID-model, it is determined, independent from the patterns produced and any so-called physical laws, that the physical procedures themselves that produce or alter the behavior of a physical-system are intelligently designed and that the production of and alterations in the behavior of a physical-system are intelligently designed via definable deductive procedures. GID-intelligence is scientifically measurable, while RID intelligence is not measurable.

RID is a statistical or descriptive approach used in an attempt to establish surmised intelligence, where the word "design" is the nounal form. For RID, intelligence is implied by investigating the "purposefulness" for specific designs. The term RID includes both the design notions of Dembski and Behe. If certain "patterns" are detected by this RID approach, then it is simply asserted as fact that such patterns are intelligently designed because they serve a describable purpose. The a RID assertion that a pattern (a design) reveals an "intelligent" intent only pertains to a few patterns that correspond to what humans might consider as existing for intelligent "purposes". But, GID-intelligent design is entirely unrelated to the "purpose" notion and is entirely related to measurable modes of logical deduction. Only the articles that do not appear on this website are listed in the "References." Rather than write an exhaustive study, for this immediate article, only a few of the more significant aspects of the two theories are contrasted. I note that chance is involved with the Behe notion. Before making a tentative RID "design" assertion, one must eliminate any known evolutionary "chance" mechanism as a cause. GID-intelligent design is not dependent upon any actual physical law. But it does satisfy such laws if they are fact. In this case, this is factual evidence for GID-intelligence design.

1. (If needed, a GGU-model glossary of terms can be found at Glossary.) In this article, the term "natural" means "physical." The term "GGU-model," usually signifies the "Complete GGU-model," a combination of the GGU-model schemes and the GID-model. Two significant features of the GGU-model are that it is falsifiable and testable (9, p. 73). Neither or these necessary aspects of the basic scientific method hold for RID.

(I) The mathematical GGU-model operators used to produce a physical universe can be interpreted as force-like or physical-like processes that actually produce natural-system behavior, alter natural-system behavior or characteristics, create natural-systems from more fundamental natural-systems and produce the "patterns" displayed within our natural world.

(II) (GID) The mathematical operators can be "interpreted" as representing intelligent actions that produce, and sustain or guide all natural-system behavior. All operators used reveal a signature that identifies them as designed by a GID-defined intelligent agent. (It is possible to consider portions of this interpretation as but "apparent intelligence." This notion is part of the "self-reference" universe. ". . . the universe gives birth to consciousness, and consciousness gives meaning to the universe.")

(III) Statement (II) corresponds to totally separate theological interpretations. These interpretations yield identifiable sources and sources that have describable purposes for each of the intelligently designed patterns displayed through application of (I).

The GGU-model is concerned entirely with establishing that the notions investigated and statements made are scientifically rational. Once an individual chooses an interpretation, then the evidence for the scientific acceptance of this interpreted is examined. As always occurs with a "General" theory, in all of these cases, the mathematical structure can be "truncated" in that various predicted results can be asserted as extraneous. However, one GGU-model result unifies all scientific physical theories and gives a specific method to obtain the best possible unification. This is a standard solution to the General Grand Unification problem and needs to be accepted by the majority of the world's scientific communities as the proper solution. Of course, this standard solution may be considered as purely physical in character and not related to GID.

Relative to the GID-interpreted GGU-model, the basic hypotheses are verified billions of times a day. Further, this model is so highly scientific in character that whenever an individual conducts a scientific experiment and the experiment verifies an inductive or deductive statement, then this also verifies the GID interpreted GGU-model.

2. The Dembski-Behe form of restricted design cannot be interpreted as in (I) and is, if applicable at all, restricted to but a few natural-systems. Dembski's design notion he terms specified-complexity (3). In an answer to criticism of his approach (10), Dembski states in (4) "specified complexity is not just one of several ways for reinstating design in natural sciences - it is the only way." This is an absurd statement if Dembski means "intelligent" design since he has not defined "design" explicitly. Although the GID-model implies RID, where applicable, Dembski's statement does not include the types of intelligence discussed by the GID-model. It is made in a flawed attempt to justify his theory.

It is general practice among some members of philosophic or scientific communities, in order to boaster the significance of a theory, to make unfounded assertions in such a way that only their theory applies. This Dembski does, in this case, by asserting "design" and rejecting any "design" notion if RID does not apply. Dembski's portion of RID is applied only to certain patterns of behavior that satisfy his Filter requirements. Further, his design notion is restricted to the level of human or materialistic intelligence and does not imply the notion of a "higher" intelligence. In GID (9), intelligence refers to agency (actions) and whether the actions are GID-intelligently designed. Any patterns produced by GID agency are, necessarily, GID-intelligently designed patterns. There are many "chaotic" natural-system behavior patterns that are not RID designed, but they can be duplicated by human modes of intelligence. They are, of course, GID-intelligently designed. (Note: In various articles, when it is clearly understood, GID-intelligent design is simply denoted by the words "intelligent design.")

Dembski states, "The fact is, however, that it takes considerable knowledge on our part to come up with the right patterns (specifications) for eliminating chance and inferring design" (4). The complete facts are, as pointed out in (10), that one needs an infinite amount of knowledge to be convinced, by application of Dembski's Filter, that something is intelligently designed via RID. One most first reject the Chance hypothesis. This requires one to ". . . know what probability distribution(s), if any, were operating to produce the event." (10, p.41) Without this knowledge the Chance hypothesis cannot be rejected. However, there are infinitely many such probability distributions that could apply and even infinitely many that have a maximum probably of less than Dembski's magic number now placed at about 1/[10120)] with variations. Indeed, the RID notion of "chance" requires the notion to capture all probabilistic knowledge, objective or subjective, relevant to the declaration of design by "chance." As shown in (9), major contributors to physics require one to accept that all patterns of natural-system behavior are controlled by such Chance hypotheses even though the exact distributions have not, as yet, been discovered.

3. Throughout modern materialistic science "chance" or "general randomness" is a major feature for all natural-system behavior. If you "eliminate" all relevant behavior that is highly probable and if it is possible to show that the entire "chance" category is highly improbable as a cause (which depends upon complete knowledge), then you can assert, in certain cases, that the pattern is "designed" if the pattern also satisfies within nature an independent humanly comprehensible "purposeful" pattern. Intelligence is only inferred since the claim, in both the Dembski and Behe approach, is that such patterns indicate a type of intelligent "pre-design." But, as shown in 10 below, this conclusion is cosmology dependent. Further, for the Dembski approach, most members of the evolutionary community argue that one cannot assign the "highly improbable" status to biological entities due to the presence of unknown constituents during their evolutionary development and "natural selection" is capable of producing by mere chance the same design that serves the exact same describable purpose.

Relative to his approach, Behe himself accepted evolutionary processes as the major cause for biological development, with the exception of those few that meet his irreducible complexity notion. The evolutionary community has eliminated a vast amount of uncertainty in biological evolution by establishing redundancy as a common feature of living organisms. They also admit that the few examples of Behe's complexity notion have yet to be satisfactory "explained" by redundancy or any other mechanism. They also have no doubt that they will be able to explain this "illusion" after appropriate research. Irreducibly complex biological objects within nature can still be viewed probabilistically, but this intuitively seems highly improbable, when our present knowledge is applied. It appears that many known irreducibly complex patterns also satisfy Dembski's notion of specified-complexity. As mentioned, this notion is cosmology dependent.

There is the notion of "general or absolute randomness" used within philosophy. This notion is the stance that there is absolutely no relation, describable in any language, which would "predict explicitly" such behavior although the patterns can be described. Of course, one can often assign a subjective probabilistic measure to such a notion. For those that make this claim, there is a 100% chance that the behavior is "generally random." From the discussion in this section, one should conclude that any RID assertion that something is intelligently designed is tentative. But since intelligence is defined in GID, it has been explicitly or implicitly established by mathematical analysis that any specific "chance" event that is probabilistically predicted to occur, or that satisfies any possible statistical distribution as well as any natural-system perturbation relative to predictive behavior is not only designed by an intelligent agent but each such event is guided in its presentation by an intelligently designed action.

There exists within the GGU-model, various intelligent designed signatures being displayed by application of the operators. GID results now show that any claimed general random behavior displays a specific a higher-intelligence signature (9a). Further, all such behavior serves a definite purpose. There are thousands of examples where human intelligence is applied to produce designs that serve only one purpose. The design simply "pleases" the designer. The designs may even be judged by some as mere "random" collections of entities. From the viewpoint of GID, if various levels of specifically described "mental" processes are applied to obtain any patterns of behavior for any physical objects, then independent from any individual's assumptions as to the significance of the patterns, they are defined as being intelligently produced and, hence, the patterns are intelligently designed.

4. (i) (Natural Law.) Obtaining a statement for a "natural law" whether from observed empirical evidence or by other means requires a human logic-system. (ii) (Scientific Theories.) A highly intelligent scientist using human mental processes constructs an exceptionally complex theory that predicts a pattern for natural-system behavior. It is verified by laboratory experiment that such a predicted pattern is highly probable. (iii) (General Applications of Natural Laws.) In order to make predictions for the highly probably physical behavior of a natural-system, it is easily shown either through computer simulation or directly that the application of human deductive processes is mimicking the step-by-step application of the natural laws that yields the predicted pattern. (For a specific illustration, see the chance.pdf file in chance.zip.) (iv) The vast majority of scientists accept that the behaviors described or predicted by natural laws and accepted physical theories correspond to reality.

GID states that (i), (ii) and (iii) satisfy the requirements for intelligently agency, whereas (i), (ii) and (iii) are rejected as examples of RID intelligent agency. I point out that (i), (ii), (iii) and (iv) yield observable evidence for the GID-model. [A more detailed discussion about the necessity for considering categories (i), (ii), (iii) as indicating intelligent design can be found in this article.]

5. Natural-system behavior or the patterns produced by such behavior that are classified as a restricted design are, as mentioned, also classified as GID-designs. Thus, the set of all restricted designs forms a subset of the set of all GID-designs. Most aspects of natural-system behavior and the patterns produced by such behavior do not satisfy the criteria for restricted design. But all such behavior does satisfy the requirements for GID-design. Hence, the set of all restricted designs forms a proper subset of the set of all GID-designs.

6. GID and restricted design theory make no fundamental appeal within their hypotheses to Scripture. Further, it's rather important that restricted design theorists, such as Dembski, admit that their theory does not imply a cause. "Now it's true that simply knowing that an object is complex and specified tells us nothing about its causal history. Even so, it's not clear why this should be regarded as a defect of the concept. It might equally well be regarded as a virtue for enabling us neatly to separate whether something is designed from how it was produced. . . the claim that they provide no causal story is true but hardly relevant - causal stories must always be assessed on a case-by-case basis independently from general statistical considerations" (4). These statements show another weakness of the RID approach compared with GID-model.

GID identifies specific intelligent "causes" or sources via various interpretations. Unfortunately, many individuals do not know what the word "interpretation" signifies and the fact that additional interpretations need NOT be made. In RID, an intelligent agent that may have produced a design is only surmised externally and characterized by such statements as Behe's relative to his notion of a design being irreducibly complex. He writes, ". . . intelligent agents are the only entities known to be able to construct irreducibly complex systems" (2, p. 156). But this is a human observation and, thus, the intelligent agents being implied by irreducible complexity can be no more intelligent than the human being who makes the observation. Moreover, the phrase "only entities known" is tentative and can be overturned by further investigation.

What constitutes a description for the features of an event in RID seems to require a specific language (10). But, future or different languages can invalidate this claim. GID is not dependent upon the language used to describe an event. GID does use language as a foundation, but the language is a general language that represents all of the modes of human perception as such perception can be re-produced through "virtual reality" techniques. As shown in (9), there are generally highly probable and apparently random patterns of behavior that humans intelligently design for specific purposes. Since the general patterns produced are highly probable, the RID community cannot detect that these patterns are designed. But, the GID-model begins with a definable notion of intelligence and shows that one can accept as empirically established the statement that "all such natural-system behavior and the patterns produced are designed by intelligent actions."

Science is usually interested in the "cause and effect" notion. For the GGU-model, the cause and effect notion is used as a means for us to predict future cosmology dependent physical behavior. As pointed out above, there is nothing in the language of specified complexity that allows for a type (I) interpretation. Often, interpretation (III) comprises a combination of (I) and (II). Using type (I), one has a description as to "how" such designs may be produced. Assume that the "nothing" in Dembski's statement "nothing about its causal history" truly means just that, "nothing." Nothing certainly means that a mere change in interpretation would not identify a "cause" or source. Indeed, if one interprets RID in theological terms, then apparently RID applies equally well to any of the numerous "gods" that have "creator" attributes (6). Indeed, Dembski even states that RID applies to "creation" by an intelligent alien life-form.

For the truncated GGU-model, interpretation (I) does not imply a single source. It does not imply that intelligent actions are the general cause. If you remove the truncation, interpretation (I) does yield a single intelligent source H, the theory-creating source. As a theorist, I admit that the GGU-model does tell us a great deal more about an implied single source, when interpretation (I) and (II) are applied. No case-by-case investigation is necessary. Many of my published articles and those that appear on this website show how predicted intelligent actions can, via interpretation, be attributed to a single source. This can lead to a theological interpretation of (II) (i.e. (III)) that models numerous Scriptural, strictly understood, statements that describe the attributes of the Scriptural Godhead and thus this implies, by such an identification, a single Scriptural source as well. I will never hide this fact, although it may cause some to reject the GGU-model based only upon the fact that it does have a type (III) interpretation. However, rejecting the GGU-model or the GID interpretation based upon such a stance is nonsensical since almost all scientific theories, and especially cosmologies, have theological interpretations.

7. Of course, there remain similarities between RID and the GID-model. In (5), we find that restricted ". . . design is not young earth creationism. . . Design as a theory holds to neither a day-age, nor a gap, nor an apparent age interpretation of Genesis." GID-intelligent design shares these qualities, but only under the (II) interpretation. The GGU-model with a combined interpretation actually upholds the "rationality" of each of these theological interpretations and hundreds of others. However, the mere rationality of a statement does not make it factual.

8. Behe is concerned with his form of design theory being classified as part of creationary science. "Scott refers to me as an intelligent design 'creationist,' even though I clearly write in my book Darwin's Black Box (which Scott cites) that I am not a creationist and have no reason to doubt common descent" (1). Then Dembski fears being considered a fundamentalist. "Nonetheless, design theorists are frequently accused of being, if not fundamentalists, then crypto-fundamentalists." He believes that restricted design theorists should be placed " . . . squarely within the mainstream of American evangelicalism" (4). (I point out that recently Dembski has removed himself from all RID activities.) I probably fail the entrance requirements for admittance into the Discovery Institute (and similar organizations). Indeed, my GID-model and all of my other writings are, apparently, being ignored by anyone associated with this Seattle WA "think tank" as well as those associated with the Templeton Foundation. They are most certainly being ignored by articles on the Internet. The obvious facts are that I'm often discriminated against based solely upon religious grounds.

9. Dembski's second book (7) seems intended to popularize his notions on specified complexity and its role relative to one aspect of "information theory." However, as shown in (8), all specific information (11) is actually designed by the same type of operators that are interpreted as in (I), (II), or (III).

10. Using RID to mathematically infer design for "nature," requires exceptionally small probabilities that the design is by chance, assuming all chance behavior is known. But, the validity of such small probability requirements depends upon the cosmology one chooses. Many members of the scientific community are selecting a multiple-universe (multiverse) cosmology that essentially considers a potentially infinite set of distinct "universe-like" objects and we dwell in one of them that has, by absolute chance, been "fine-tuned" to support our type of life-forms. It's even speculated that the multiple-universe cosmology universe-like objects can be controlled by a different set of parameters or even completely different sets of processes. Such collections of universes also satisfy the GGU-model and satisfy GID on various levels.

For multiple-universes, care must be taken when applying probability-models that appear to predict behavior within our universe and appear to be fixed models for particular scenarios. Further, the probability notions presented by these models are consistent with the frequency definition and the actual trials can be performed within each universe. Assume that there is a fixed collection P of such models that apply for each universe and each universe develops under the same type of probabilistic processes that appear to govern our universe. The existence of even infinitely (denumerably) many disjoint universes of this type would have no affect upon probability-models in collection P. The existence of such universes does not affect the frequency interpretation within a universe. These are objective probabilities.

Significantly, there are many probabilities that apply to special physical scenarios, which are not members of P and that are rather subjective in character. This is especially the case with conditional probabilities, where the "conditions" can occur only a few times within one universe. For these special physical scenarios, the conditional probability can be interpreted for the multiple-universe model via the frequency notion of experimental repetition. If the number of universes is exceptional large, then this leads to the conclusion that the physical events being modeled would almost certainly occur within one such universe even if they are highly improbable within a single universe cosmology. Although the actual cause could be attributed to the conditions and the notion of "random trials and a random cause" some could conclude that the occurrences of such special physical events would be evidence for the existence of multiple universes and cite a multiple-universe cosmology as an additional physical cause for the events.

A collection of infinitely many disjoint universes (technically, at least, denumerably infinite) each containing a finite amount of material and each developing under the same general parameter-guided probabilistic types of physical laws and physical theories (i.e. libraries) is constructed in (9, p. 161). Indeed, from the viewpoint of properton theory, assigning to each collection of the needed universe-generating propertons a specific "phase" can generate these disjoint universes. In this case, the universes would not interact with each other and each can develop under perceived but distinct "physical laws." Suppose that a specific Q-event behaves, relative to the collection of universes, like a single trial Bernoulli event and the probability is approximately 1/k for each. Then the Strong Law of Large Numbers states that there is a trial number m such that if X is the number of Q-events that occur, then |X/m -1/k | < 1/k. Hence, the Q-event must occur at least once during the "m" trials. Thus, no matter what the probability will be for a special physical event to occur within a single universe cosmology, the event will occur within, at least, one universe for an infinite collection of universes of this type. Since any finite collection of universes removed from the infinite collection leaves an infinite collection, a simple induction proof shows that there is a potentially infinite set of such universes U. Of course, this result also holds for a finite collection of Q-events occurring within a single universe.

Using this additional information, letting Q = chance, you cannot state that the event will not occur by chance within our universe, since it will occur by chance in a potentially infinite collection U and there can be no physical science information that our universe is not a member of U. Hence, assuming that the RID methods are viable, then physical science cannot know whether the event is or is not a chance event since physical science cannot determine whether or not our universe is a member of such a U. This means that to apply RID a purely philosophic stance must be taken with respect to an accepted cosmology. RID and complexity arguments are not independent from cosmologies.

However, GID-intelligent design is independent from such selected cosmologies. Although not all aspects of the GID-model need apply absolutely, one aspect that does apply completely is that it is rational to state that each multiple-universe cosmology and the behavior of each constituent contained in each of these universes display GID-intelligent designs. [As discussed in (9), there are also arguments that any design inference is only apparent (i.e. natural design) and is based upon the evolution of human consciousness.]

11. One of the most significant differences between the GID-model and RID is related to interpretation III, a theological interpretation. Any RID claim of intelligent design does not directly imply that the design requires a "higher intelligence." For this reason, RID does not counter most forms of Biblical distortion. Indeed, many distinct theologies will not contradict RID notions. That is, RID does not point to a specific member of a large collection of theologies.

On the other hand, one theological interpretation for GID-intelligent design yields the rational statement that all natural-system behavior is intelligently designed by a "higher intelligence" as such an intelligence is Biblically described. Further, among other results for this interpretation, the GGU-model implies that all natural-systems within our universe are, in all respects, created without the use of any material objects and our universe is created for definite Biblical purposes. This interpretation also counters Biblical distortions and shows, for example, the complete rationality for all of the strictly interpreted creationary statements written in Genesis 1, as well as, the Genesis Flood descriptions.

12. There are many other startling differences between these two notions. One major difference as previously mentioned is the concept of the "purposeful design." Such a design is one that has been "guided by a definite aim." RID requires a description for the "definite aim," the use, etc. for a design or behavior, and then by comparison restricted design theorists claim that such a restricted design could only come about through the means of intelligent agency. However, the "aims" refer to a few materialistic outcomes. All of the identified restricted designs that satisfy such described aims are also GID-designs. But, as shown there are purposeful general GID-designs that do not satisfy restricted design purposes. Often, the many differences depend upon your choice of (I), (II) or (III).

In terms of interpretation (II), you'll find at this URL a summary of many, but not all, of the results obtained from the GID-model. This is a summary relative to the intelligent design of a physical universe. For an illustrative discussion of the results obtained from the GID-model, please see this URL. If you have any questions that pertain to this article, you might receive an answer by addressing your question to e-mail, where the "subject" is to be only the one word "GID." This is a limited e-mail server and it will accommodate only 25 e-mails per day. If you don't receive an answer, this may be the reason and you should try again at a later date.

(1) Behe, M. J. Intelligent Design is Not Creationism. Response to "Not (Just) in Kansas Anymore" by Eugenie C. Scott, Science (May 2000), Science online (July 7, 2000), Discovery Institute WWW Site, Under "Articles By Michael J. Behe."

(2) Behe, M. J. Self-Organization and Irreducibly Complex Systems: A Reply to Shanks and Joplin, Philosophy of Science, 67(2000):155-162.

(3) Dembski, W. A. The Design Inference -- Eliminating Chance Through Small Probabilities, Cambridge University Press, Cambridge (1998).

(4) Dembski, W. A. Another Way To Detect Design? Preliminary reply to review by B. Fitelson, C. Stephens, and Elliott Sober of The Design Inference in the Sept. 1999 issue of Philosophy of Science, "How not to detect design". (June 28, 2000), Discovery Institute WWW Site, Under "Articles By William A. Dembski."

(5) Dembski, W. A. What Every Theologian Should Know about Creation, Evolution, and Design, (April 1, 1996), The Princeton Theological Review, Discovery Institute WWW Site, Under "Articles By William A. Dembski."

(6) Kern, L. In God's Country, (December 14, 2000) Houston Press.

(7) Dembski, W. A. Intelligent Design: The Bridge between Science and Theology Intervarsity Press, (1999).

(8) Herrmann, R. A. Information Theory, Consequence Operators and the Origin of Life, C. R. S. Quarterly, 36(3)(1999):123-132.

(9) Herrmann, R. A. Science Declares Our Universe IS Intelligently Designed Xulon Press, Fairfax VA, 2002.

(9a) Herrmann, R. A. http://arxiv.org/abs/quant-ph/0112037 Probability Models and Ultralogics.

(10) Fitelson, B., C. Stephens, E. Sober, How not to Detect Design (Review of Dembski's The Design Inference - Eliminating Chance Through Small Probabilities.) Philosophy of Science (Sept. 1999).

(11) Gitt, W. Information and Entropie als Bindeglieder diverser Wissenschaftszwiege PTB-Mitt. 91(1981):1-17. (This is the first in a series of articles by Gitt relative to this notion. The need for specific information can be traced to earlier writers such as K. Steinbuch in 1968.)

Click back button, or if you retrieved this file directly from the Internet, then return to top of home page. If you retrieved this file while on my website, then return to top of home page.