Design arguments point to the need of a designer to explain certain complexities. More recently, iterations of these arguments have focused on the apparent fine-tuned conditions of our universe to habit intelligent life. Jonathan Weisberg suggests that the discovery of cosmological fine-tuning for the existence of intelligent life is “irrelevant to debates about design,” (1) given the fact that we already know that there is intelligent life. It seems Weisberg is right to say that these new empirical findings do not provide direct evidence for the “design hypothesis,” but it is far from “irrelevant.” This new data may be useful insofar as disconfirming competing theories to the “design hypothesis,” which subsequently makes the surviving “design hypothesis” more probable.

Let us begin by establishing Weisberg’s critique of the “design hypothesis,” also called the cosmological fine-tuning argument. The argument says that “certain parameters in the laws of physics and the initial conditions of our universe are fine-tuned so as to allow for the existence of intelligent life.” (1) Such fine-tuning is much too improbable to explain by chance, thus calling for a “designer.” Weisberg calls the data pointing to fine-tuning “New Datum” (3); more specifically, the facts in physics which tell us “our universe is fine-tuned so as to allow for the existence of such life.” (1) He contrasts this with “Old Datum”: namely, the fact that “we have known for a long time that there is complex, intelligent life.” (1) Weisberg’s critique is that a designer may have chosen to create intelligent life without fine-tuning, or a world “whose laws whose conditions and parameters do not need such careful setting.” (2) The key premise is that there does not seem to be any preference between a world that is finely-tuned verses a world that is not finely-tuned. So given our Old Datum, the New Datum is not positively relevant; that is, the New Datum does not affect the probability of the hypothesis for the existence of a designer.

This critique is made clearer by outlining its defeater. If there were some reason that a designer necessarily had to create a fine-tuned world, then New Datum would be positively relevant for the designer hypothesis. But there does not seem to be any reason to think this (apart from maybe theological grounds). Thus, given Old Datum, New Datum does not make the “design hypothesis” any more probable.[1]

I am sympathetic to Weisberg’s general line of thinking, as well as his replies to some proposed objections, but I think it is wrong to jettison the New Datum and to think it has nothing to add to the designer argument. I suggest that even if the New Datum does not directly support the designer argument, it might indirectly support it by eliminating the alternative hypotheses. My support to this idea is derived from normative claims – I mean, not specifically committed to any epistemological baggage, rather, what is putatively more prudent for the pursuit of knowledge. This will become clearer as my argument unfolds.

The first premise is that we should hold theories or hypotheses[2] that are most probable. This is fairly uncontentious, although there is some dispute on what exactly probabilities are; yet, like Weisberg’s critique, this critique also “applies no matter which conception of probability is favored.” (2) This is a common practice in science: we hold the theory (say) that the moon is a rock because it is more probable than alternative theories, like the moon being made of cheese. It seems obvious that in the pursuit of knowledge or truth, in the normative sense, people prefer more probable theories over less probable theories.

However, it is not always clear which theories are more probable than others. For instance, is the theory of evolution more probable than the special theory of relativity? It is not so obvious. Thus an added stipulation is needed for the premise: it is appropriate to compare theories that are mutually exclusive[3] and within the same discourse. Theories within the same discourse have some common unifying idea. To illustrate, the theory of geocentricism and the theory of gravitation do not have a common unifying idea, but geocentricism and heliocentricism have the common unifying idea of an astronomical orbital center. However, it might also be the case that even with competing theories it is not so obvious which is more probable. This is often seen in cases like non-standard models of the universe – the respective theories have their strengths and weaknesses, but it is uncertain how even order them by plausibility. Still, assigning probabilities or ordinal values is not a concern in this argument, so we can move on.

Moving on to the second premise: eliminating or disconfirming a theory from a set of competing (or mutually exclusive) theories (within the same discourse) affects the probability of the other standing theories. Suppose there was a set of three theories (within the same discourse and mutually exclusive): (1) the sky is blue, (2) the sky is red, and (3) the sky is yellow. Let us begin with an added stipulation to clarify the premise – namely, that this set is jointly exhaustive – which we will subsequently drop after further explanation. Thus, we assign all the theories an equal probability of 1/3. If we disconfirm a theory (say, theory (2)) and it is no longer within our set of theories, the other two theories consequently become more probable (theory (1) and theory (3) now have a probability of ½). This idea carries over to scenarios where the theories are not jointly exhaustive. Now imagine some arbitrary probabilities assigned to each theory: say, theory (1) has a 70% chance of being true, theory (2) has a 40% chance of being true, and theory (3) has a 60% of being true. If we again disconfirmed a theory, the surviving theories would look more probable. This might be because evidence that disconfirms one theory acts to support another, but this is not always the case; perhaps, a more plausible explanation is that the surviving theories look more probable just in virtue of the fact that they were not disconfirmed. Since, by the first premise, we should hold to theories that are most probable, any processes that helps us do this, like eliminating competing theories, is a useful process. Colloquially, we call this the “process of elimination,” and it is a widely used tool for reasoning, particularly in the sciences. In any case, I think we have good reason to think that eliminating a theory is useful, and consequently any evidence that does this is also useful.

Let us return now to Weisberg’s argument. Perhaps we may grant that New Datum does not directly support the designer hypothesis, but the New Datum can indirectly support it by disconfirming competing theories. Imagine a theory that says the parameters necessary for intelligent life are huge that it is hardly a surprise that intelligent life exists; furthermore, by parsimony, a designer is cut out of this picture. This theory appears to be a feasible competing theory to the “design hypothesis;” yet, New Datum disconfirms this theory, and (as Weisberg shows) remains neutral on the “design hypothesis.” Nevertheless, by the second premise of my argument, the New Datum still affects the probability of the “design hypothesis” by disconfirming this competing theory. If one has an issue with New Datum affecting the “probability,” it can be said that, more minimally, New Datum positively affects the general tenability of the “design hypothesis.”

Weisberg’s general argument sounds right: given the fact that there is no preferable choice a designer would take between a fine-tuned universe verses a not fine-tuned universe harboring intelligent life, and Old Datum, New Datum does not provide direct support for the “design hypothesis.” I also think he is right that there is no good explanation to think a designer had to necessarily choose a fine-tuned world, unless we appeal to theological considerations. However, I think New Datum does provide support to the “design hypothesis” in a roundabout way. New Datum can disconfirm other competing theories, which then makes the “design hypothesis” a more viable option. By and large, New Datum is at the very least not “irrelevant.”

Source

Weisberg, J. (2010). A note on design: What’s fine-tuning got to do with it?Analysis, 70(3), 431-438.

---

[1] Weisberg gives a formal presentation of the critique by making use of Elliot Sober’s “likelihood formulation.” (4) The design argument becomes the probability of the New Datum (N) given the probability of the Design Hypothesis (D) and Old Datum (O) is greater than the negation of the Design Hypothesis: “P(N|D^O)>P(N|~D^O).” Given the Likelihood Principle: “if P(E|H)>P(E|~H) then E supports H over ~H.” Therefore, “N supports D over ~D (given O).”

[2] “Hypothesis” and “theory” will be used interchangeably because the nuances in their definitions have no bearing on the argument, plus the illustrations are made clearer.

[3] It might be inappropriate to call theories “mutually exclusive,” but I mean that there is a contradiction between some proposition(s) entailed by theories, which cause conflict between theories.