Chapter 3 Appendix
Analysis of the Teleological Argument
University of Rochester, Rochester, New York
15 March 1993
The motivations of creationists are often more philosophical than scientific. In particular, creationists are often motivated to claim that evolution cannot be true because it fails to provide a satisfactory account of how the universe can display as much order and complexity as it does – one must, instead, resort to the hypothesis that such order and complexity is a result of intelligent design. Historically, the apparent existence of design in the universe has provided the basis of one of the most convincing arguments for the existence of God. This argument has, appropriately, come to be known as the "design argument," the "argument from design," or the "teleological argument." This article is devoted exclusively to an analysis of some of the more common manifestations of the argument from design.
Version #1 (cf David Hume’s Dialogues Concerning Natural Religion):
1. Both artifacts and natural objects display orderly adjustment of constituent parts to each other.
2. Artifacts have this property as a result of being designed.
3. Of two classes of things, x and y, if the members of x and y share some property p1, and members of x have p1 as a result of having a property p2, then, probably, members of y have p1 as a result of having p2.
4. Therefore, natural objects display orderly adjustment of constituent parts to each other as a result of being designed. (Natural objects are designed.)
5. If natural objects are designed, then some god designed them.
6. If some god designed natural objects, then God designed them.
7. If God designed natural objects, then God exists.
8. Therefore, God exists.
Criticism of Version #1
Let us grant premises 5, 6, and 7 (although premise 6 is tenuous, at best). The real work of the argument is contained in premises 1, 2, and 3. Let us consider each of them in turn.
Premise 1: "Both artifacts and natural objects display orderly adjustment of constituent parts to each other."
The first thing to consider about this premise is the exact meanings of "orderly" and "adjustment." To find this premise acceptable, we need to know just what it is for a configuration of constituent parts to be orderly. In particular, it is crucial that the proponent of the design argument not beg the question by allowing "orderly" to mean "arranged by an intelligent designer," and "adjustment" to mean "relation caused to obtain by an intelligent designer." Sometimes "orderly" means "arrangement in accordance with a rule or law." Again, if we are to say that rules and laws apply to natural objects, we must be careful to recognize that "rule" and "law" can only be taken metaphorically as applying to natural objects, unless we already know that some sentient rule-maker or lawgiver is responsible for the lawlike properties of objects. Perhaps "orderly" simply means "not-random." This may be acceptable, so long as we keep firmly in mind that randomness is largely a function of what we expect to find given our knowledge about the properties of the things we are investigating.
In addition, the term "natural object" is somewhat vague. Does the class of natural objects include every non-human-made object in the universe? Do rocks count? It would probably be better, then, to replace the talk about natural objects with what creationists take to be an unquestionably designed class of objects, viz., living things.
Premise 2: "Artifacts have this property as a result of being designed."
We will grant this (where an "artifact" is a human-made object).
Premise 3: "Of two classes of things, x and y, if the members of x and y share some property p1, and members of x have p1 as a result of having a property p2, then, probably, members of y have p1 as a result of having p2."
This premise has serious problems. Consider this case:
1. Working light-bulbs and stars both have the property of emitting light.
2. Working light-bulbs have this property as a result of being powered by electricity.
3. [Same as for the design argument.]
4. Therefore, stars have the property of emitting light as a result of being powered by electricity.
A little thought about premise 3 shows it to be wildly implausible. Different types of objects often have the same properties for different reasons. Premise 3, therefore, provides extremely poor inductive support for premise 4. It may be objected that premise 3 could be modified as follows so as to make it acceptable:
3′. Of two classes of things, x and y, if the members of x and y share some property p1, and members of x have p1 as a result of having a property p2, and p2 is the only way known of producing p1, then, probably, members of y have p1 as a result of having p2.
Premise 3′ will rule out the light-bulb/star example since in order to preserve validity we will have to add another premise to the effect that electricity is the only known way of producing light-emission – and such a premise would be manifestly false.
Unfortunately, 3′ will also defeat the design argument (or, at best, render it hopelessly question-begging) since to preserve validity a premise will have to be added to the effect that intelligent design is the only way known of producing orderly adjustment of constituent parts of objects. Scientists, however, have no reason to accept such a premise, and extremely good reason to reject it. And yet there is a version of the design argument that relies on a premise very much like the claim that design in the only way known of producing objects whose parts display orderly adjustment of constituent parts:
Version #2 (cf Alvin Plantinga’s God and Other Minds):
1. Everything in the universe that has the property of having parts displaying orderly adjustment of constituent parts (hereafter "order") and which is such that we know whether or not it was designed was, in fact, designed.
2. Living things display order.
3. Therefore, living things were probably designed.
Criticism of Version #2:
Premise 1: There are several important features of premise 1. First, it says, in effect, that it is possible that every object that displays order is designed. Second, it says that every ordered object which is such that we know whether or not it was designed was, in fact, designed. This is because these objects are the products of human artifice. Third, it is implied by this premise that there are no ordered objects in the universe which we know were not designed – because every ordered object is possibly ordered because of having been designed. The question, of course, is whether this is a plausible premise. There are several reasons to think that it is not. In the first place, the premise seems to be an instance of the more general principle:
Everything having property p1, and which is such that we know whether or not it has p1 as a result of having property p2, does in fact have p1 as a result of having p2.
To decide whether or not this principle is, in general, acceptable, it may be helpful to see whether there are other instances of the principle that are more obviously implausible. I suggest the following:
1x. Every book having the property of being known about by my father and which is such that I know whether or not it has the property of being known about by my father as a result of having the property of being read by my father has, in fact, been read by my father.
We might imagine that I ask my father whether he knows about the book The Grapes of Wrath, and he says he does. I then ask him how he knows about it and he tells me that he read it. I now ask him whether he knows about the book Tom Sawyer, and he says that he does. But I do not ask him whether or not he knows about Tom Sawyer as a result of having read it. In fact, let us imagine that, unbeknownst to me, the only book my father has read is The Grapes of Wrath. I now ask him whether he knows about a number of books, and in each case he tells me that he does. Of course, the only book which is such that I know whether or not he knows about it as a result of having read it is The Grapes of Wrath. The inductive inference that my father knows about Tom Sawyer, for example, as a result of having read it, is not particularly strong unless I also have good reason to think that the only way my father is likely to have come to know about Tom Sawyer is as a result of having read it.
The case is similar in the case of the design argument. For the inductive inference to be strong that living things are designed, the defender of the argument must have good reason to think that the only way in which living things are likely to have acquired the property of orderliness is as a result of having been designed. It is far from clear, however, that this is the case. The evolutionary hypothesis provides an alternative explanation of how living things might have come to have the property of displaying order.
In addition, we have excellent reason to reject the initial claim that all ordered objects which are such that we know whether or not they were designed, in fact were designed – we only need to be licensed to claim with respect to one ordered thing that it was not designed. I submit that there are numerous examples of ordered, non-designed objects. Snowflakes, for example, surely count as ordered objects, but they are not ordered as a result of being designed – they are ordered as a result of the micro-properties of ice crystals. The same may be said of living things. Living things are ordered as a result of their micro-structure – particularly their genetic makeup.
Clearly, the defender of the design argument can claim that the only way in which living things (or snowflakes, for that matter) could possibly have come to have such micro-structures is as a result of having been designed. But this is false. There are other possible explanations for the origination of genetic structures. Admittedly, the defender of design may claim that such alternative explanations are wildly improbable. This suggests another version of the design argument:
1. It is extremely implausible that living things could have come to exist by mere chance.
2. Living things exist.
3. Therefore, probably, living things did not come to exist by mere chance.
4. If living things did not come to exist by mere chance, then living things were designed.
5. Therefore, living things were designed.
Criticism of Version #3:
Premise 1: It is very difficult to understand how this premise is justified. What is the basis for the claim that a chance-origination of living things is implausible? More significantly, why should it be any more plausible that a supernatural being exists who is responsible for the creation/design of living things?
Premises 2 & 3: Let us grant both of these.
Premise 4: There are good reasons to reject premise 4. This premise, in effect, says that there are only two possibilities for the origination of living things, viz., design, and mere chance. This is, however, a false dilemma. Again, consider a snowflake. Snowflakes do not spring into being by mere chance – by a spontaneous "poofing" into existence ex nihilo. Rather, they arise as a result of the chemical and molecular properties of water as it freezes. Likewise, it may be that life arose as a result of the chemical and molecular properties of various naturally occurring substances. As various substances bonded together, they acquired new properties and therefore new and different ways in which they could interact and bond with other substances. Over millions or billions of years such combinations may have given rise to simple self-replicating structures, and eventually to DNA. Thus even if there was some chance involved in the genesis of living things, such chance need not have been mere chance – the possibilities were restricted by the objective (and lawlike) properties of the chemicals and molecules in interaction with one another.
1. If living things were designed by a perfect being, then living things should display evidence of perfect design.
2. Living things do not display evidence of perfect design.
3. Therefore, living things were not designed by a perfect being.
Analysis of this argument:
Premise 1: This just seems to be a consequence of what we mean by "perfect being." This being, God, is supposed to be perfect in knowledge (omniscient), so he must know how to design perfect organisms, and perfect in power (omnipotent), so he is capable of creating perfect organisms, and perfect in intent (omnibenevolent), so he wills to create perfect organisms.
Premise 2: It is somewhat difficult to evaluate this premise unless we have a clearer understanding of what features constitute evidence of perfect design. Presumably, a perfectly designed feature of an organism will be a feature which displays both economy and functionality with respect to the use of that feature. In other words, a perfectly designed feature of an organism will be one which it is such that it is impossible to conceive of any alternative design for that feature which would be better than the one it actually has.
I submit that there are clear examples of sub-optimal design in nature. There is the notorious panda’s "thumb," for example, which is not really a thumb at all – it is a wrist bone which has been adapted to allow the panda to grasp its food. Then there are apparently useless features of some organisms, such as the human appendix (to say that it is possible that it serves some unknown function is no good reason to believe that it really does serve such a function), or perhaps more clearly, pelvic bones in whales.
The design argument is unsound. It fails to make it even probable that living things were designed.