The Privileged Planet: Single data points and naive falsification

Now that the DVD for ‘Privileged Planet’ has been announced, it is time to remind the readers of why I believe that the Privileged Planet makes for a very poor scientific argument. Although I do believe that the argument serves well as an apologetic and rhetorical tool, which may help explain why it is given such a ‘privileged’ position at the DDD-V conference or creationist websites.

So why do I believe that the Privileged Planet approach is wrong? To quote Xia-Li Meng at the 2004 ENAR Spring Meeting in Pittsburgh PA on statistics:

“If you have not seen all the data, how can you estimate how much you haven’t seen? But, as statisticians, we can do anything!”

As statisticians we can do anything. Especially when the argument is based on a correlation. Mark Twain had this to say about “lies, damn lies, and statistics”? It’s not that statistical arguments are necessarily lies but it’s a warning that good statistics requires a careful approach to avoid obvious and less obvious pitfalls. I argue that in Privileged Planet the authors have failed to do so.

We can show for instance that ice cream sales cause summer drownings, we can show that wet pavement causes rain or even that the Dow Jones causes changes in the skirt length. In other words, correlation can be used and abused and it requires a careful analysis before one can jump from correlation to causation.

Arguments based on correlation need to carefully quantify the terms of interest. How else can one determine if there exists a statistically significant correlation? The authors of Privileged Planet fail to define quantifiable measures of either habitability nor measurability.

Arguments based on correlation need to deal with causal direction. Does rain cause the pavement to get wet or does wet pavement cause rain? The authors of Privileged Planet fail to explain why correlation should be interpreted as causation.

Arguments based on correlation need to deal with additional correlating variables. A good example is the correlation between ice-cream sales and summer drownings. A careless person may argue that this is evidence of a causal relationship between ice cream sales and drownings. Of course the real ‘cause’ was found in the realization that both factors also correlate with a third factor namely ‘summer’. Summer and ice cream sales correlate, summer and summer drownings correlate.

And finally a correlation requires a sample of more than one. The authors of Privilged Planet however base their argument on a sample of one. More on that later.

What is the Privileged Planet about

The Privileged Planet basically argues that there is a correlation between habitability (is it hospitable to life) and measurability (can one learn about the world around it). Based on this poorly defined concept, the authors are trying to argue that life on earth involves purpose, or Purpose. Not only do the authors fail to present quantifiable measures for habitability and measurability (making any claim of correlation suspicious) but they are also relying on a single data point (earth) to make their claims. But a single datapoint cannot be used to make any reliable statistical inferences.

Mark Isaak notices that Voltaire in “Candide” observed that:

“It is demonstrable,” said he, “that things cannot be otherwise than as they are; for as all things have been created for some end, they must necessarily be created for the best end. Observe, for instance, the nose is formed for spectacles, therefore we wear spectacles. The legs are visibly designed for stockings, accordingly we wear stockings. Stones were made to be hewn and to construct castles, therefore My Lord has a magnificent castle; for the greatest baron in the province ought to be the best lodged. Swine were intended to be eaten, therefore we eat pork all the year round: and they, who assert that everything is right, do not express themselves correctly; they should say that everything is best.” [Voltaire 1759]

Single datapoint? Tell me it isn’t so

David Grinspoon: It is always shaky when we generalize from experiments with a sample size of one. So we have to be a bit cautious when we fill the cosmos with creatures based on the time scales of Earth history (it happened so fast here, therefore it must be easy) and the resourcefulness of Earth life (they are everywhere where there is water).

This is one history, and one example of life. When our arguments rest on such shaky grounds, balancing a house of cards on a one-card foundation, we are in danger of erecting structures formed more by our desires than the “evidence.”

Link to Rare Earth discussion

The authors have argued that they are NOT relying on a single data point since they show various examples of what they claim to be a correlation. But when I pointed out that there are also examples of poor correlation, they responded by stating that:

Of course, as we say several times in the book, we don’t argue that the Earth is optimal for observing every particular type of phenomenon we cite. Rather, it’s optimal in the constrained sense of providing the best overall place for discovery.

so before the authors object by saying that they show several such examples, let it be clear that the authors themselves are arguing for a ‘constrained optimization’ which means that a single observation cannot be independent of another observation but rather that when taking into consideration all aspects of habitability and measurability for lets say the planet earth one has one example of a constrained optimization (in a statistical sense). Otherwise any observation which contradicts the authors’ viewpoint would count as an example against their thesis. But they insist that such observations do not affect their arguments.

Hence the conclusion is simple, earth can be at most a single datapoint of an ‘constrained optimization’.

But things get worse. Think about it, how would we know if a particular aspect of earth makes it a poor platform for scientific discovery? Since we would remain ignorant of our ignorance, we cannot even determine if earth were optimal in some constrained sense. And just when you thought, thing could not get any worse, we now realize that any planet where intelligent life were to develop, the inhabitants would argue that their world is ideally positioned for inquiry.

Maybe a quick thought experiment may clarify this. Let’s assume that intelligent life on Planet X has developed a scientific knowledge similar to our with one major difference, their planet is in the worst position to allow them to detect that Newtonian physics is merely an approximation. Ignorant of this fact, they marvel how their planet is in an optimal position to allow them to measure the world and universe around them. Unaware of the fact that a theory of Relativity exists. I argue that this problem is endemic to the whole approach namely that any intelligence would consider its position in the universe to be optimal for scientific inquiry since science has been able to use whatever is available and remains unaware of that which it is missing.

Naive Falsification?

The following statement by the authors

“Finally, we don’t argue that Earth is unique. Discovering another planet around another star in the Galaxy would be quite compatible with our hypothesis, so long as that planet is genuinely Earth-like. Finding a fundamentally different planet with (native) complex life on it, in contrast, would contradict our argument that the conditions for life and scientific discovery correlate in the universe.”

makes even less sense. Why would finding a ‘non earth like planet’ with complex life contradict their claim of correlation. Since such a planet would show another datapoint in favor of a correlation between life and scientific discovery.

Unless of course the real argument is to show that the Earth has a privileged position but such an argument cannot be used a premise unless the authors want to claim a tautology. On the other hand, from a theological perspective it is much easier to understand why one may consider the earth to be privileged. But what we wish or what we believe by virtue of faith may have little relevance to the scientific accuracy of such a belief.

And that’s where my strongest theological objection to the Privileged Planet can be found. By insisting that the earth is privileged based on some mathematical arguments, the authors miss the obvious, the earth is privileged because that’s where we are living. To suggest that a Creator would create a universe with countless planets and consider only one to be privileged requires a knowledge and understanding of said Creator, beyond our realm of knowledge. In fact, it shows a certain level of hubris.

This reminds me of several Farside cartoons

Half baked No rocket scientists Math phobic worst nightmare

I hope you get the picture.

In other words, the design inference is based on a flawed foundation (single sample) , poor mathematics, failure to control for confounding factors, failure to address the causal directions.

And of course lets not forget that the Privileged Planet does not present any testable hypothesis of Intelligent Design.

One may wonder why the ID movement gives the Privileged Planet such visibility. The real question is, will ID do better in the (near) future? Can we expect a positive ID hypothesis?

Why the Privileged Planet is not a positive hypothesis

Without any understanding of motive, means or opportunity, we cannot really constrain a hypothesis. The hidden assumption of the Privileged Planet seems to be that a ‘Designer’ would ‘create’ an environment not just suitable for life but also for scientific inquiry. But that presumes a lot about a ‘designer’ of whom we have little data allowing us to determine if such an assumption is warranted or just ad hoc. An intelligent design hypothesis without constraints fails to be relevant scientifically. In fact, since ID hypotheses are based on elimination rather than on positive evidence, they are not just irrelevant scientifically but unreliable as well.

Gonzalez is well aware of the ‘weak anthropic principle’ (observer bias). In Home Alone in the Universe (From: First Things 103, May 1, 2000 ) with Ross, Gonzalez makes the following assertion:

It is difficult to quarrel with the simple physical interpretation of the WAP (Ed: Weak Anthropic Principle): it is just a type of observer selection bias. We should not be surprised to observe, for example, that we are living on a planet with an oxygen-rich atmosphere, for the simple reason that we require oxygen to live. The WAP “explains” why we should not observe ourselves to be living on, say, Titan, but it fails to account for the origin of the oxygen in our atmosphere and hence for the rarity of planets with oxygen-rich atmospheres. However, Barrow and Tipler, no doubt motivated by the philosophical CP, have burdened the basic physical interpretation of the WAP with unwarranted philosophical extrapolations. In considering the WAP with regard to the observable universe, they claim that we ought not be surprised at measuring a universe so finely tuned for life, for if it were different, we would not observe it.

In other words, Gonzalez does recognize the risk of philosophical motivations leading to unwarranted extrapolations. And that’s exactly what I argue, Gonzalez and Richards have done in “Privilged Planet”.

The Privileged Planet Part 1: Where Purpose and Natural Law freely Mix Part 1

The Privileged Planet Part 2: The failure of the ‘Design Inference’

The Privileged Planet Part 3: The Anthropic principle


Kyler Kuehn’s presentation which was given at the American Scientific Affiliation 2003 Annual Meeting: A Critique of the Privileged Planet Hypothesis.

Kyler Kuehn shows how the Privileged Planet argument suffers from some very basic flaws, making it quite unsuitable for its intended purpose

Kyler Kuehn’s argument is simple

Primary Thesis Habitability + Measurability do not warrant an inference to design (Divine or otherwise). Secondary Thesis Habitability + Measurability cannot provide warrant for a design inference. Therefore Habitability + Measurability are not the right tools for empirically detecting design

The design argument

Design Detection via William Dembski’s Design Filter (detailed in, for example, Intelligent Design or No Free Lunch)

Measurability provides the detachable specification (the target), Habitability provides the complexity (the arrow).

In other words, the Privileged Planet suffers from much of the same problems as any other design inference based on the explanatory filter.

Kyler then argues

Optimal Measurability is Trivially True I: The Statistical/Chance Argument

Ignorant of what we don’t know we always observe ‘optimal measurability’ but we don’t know if it s a global optimum.

The next objection involves correlations caused by a third variable

Optimal Measurability is Trivially True II:The Argument Based on Physical Law

Kyler argues that constraints on measurabiluty are entailed by constraints on habitability. In other words, the correlation between measurability and habitability is caused by natural law (law like regularity). And in fact the authors admit that natural law is not eliminated by their approach and consider natural law quite consistent with a designer.

Conclusions are thus that

A Design Inference based on the correlation of measurability and habitability is at best only trivially true

Kyler Kuehn, who appears to be an intelligent design proponent ends with advice to proponents of intelligent design but also warns detractors

While Intelligent Design has certainly not been scientifically proven (and in the case of the Privileged Planet, appears to be false), it does warrant serious scientific debate–at the very least, it is interestingly wrong!

Well said, if ID wants to be considered as a scientifically relevant hypothesis it will have to do the hard work needed to present a non trivial example of intelligent design explaining particular observations better than regularity or chance explanations. So far the problem has been that ID has failed to present any explanation of means, motives or mechanisms used by the intelligent designer(s). Without such constraints, ID will remain an interesting possibility but scientifically irrelevant.

The Privileged Planet, like other claims by Intelligent Design before, is gaining a lot of interest among religious organizations. Not surprisingly since it is proposing to have found evidence of Purpose or Design. By wrapping Purpose or Design in a scientific jacket, ID proponents have created a potential falsification.
It is by Faith and Faith alone that we are saved, trying to find ways to support our faith by exposing the Creator to the idea of falsification seems to be contrary to faith and to good science. Even those who doubt that we are saved by faith alone have to deal with the effects of creating a designer whose work or existence can actually be scientifically scrutinized and worse falsified.

In this context we may want to look at the 2nd Commandment. We have a jealous God, making Him subject to falsification seems to be creating just another idol.

From the Privileged Planet website

But is this correlation between the existence of complex life and our ability to make scientific discoveries simply a coincidence or the result of blind chance?

Notice how the authors forget an obvious alternative hinted at by Kyler Kuehn namely: ‘regularity’

And notice how no attempt is made to resolve the many pitfalls with arguments based on correlation. In their eagerness to oppose personal opinions of some people about how ‘mediocre’ earth is, the authors may have swung too far, making their conclusion, just like any conclusion of no-designer, a philosophical rather than a scientific conclusion.

from this link we read:

The authors explain that scientists for years have been studying our world through the looking glass of ideology. “Rather than a search for the truth about nature–based on evidence, systematic study, and the like–science becomes applied naturalism: the conviction that the material world is all there is, and that chance and impersonal natural law alone explain, indeed must explain, its existence.”

Somewhat ironic and in fact an erroneous representation of scientific inquiry. But if the authors believe there is a scientific, non-naturalistic explanation of the ‘Privileged Planet’, they have done little to present it.