Header 1

Our future, our universe, and other weighty topics


Saturday, April 23, 2016

Firing Squads and Habitable Universes: The Fallacy of Appealing to an Observer Selection Effect

There exist numerous cases of what look like very strong fine-tuning in our universe. Both fundamental constants and natural laws are arranged in a way that allows for us to exist. It seems that the probability of all of these favorable conditions existing by chance is incredibly low. It has been argued that the probability of you existing in a universe as fine-tuned as ours is like the chance of you surviving a firing squad (having 10 or more soldiers firing their rifles at you at close range). If you survived a firing squad, it is argued, you should assume there was some purpose involved in this, and that it wasn't just a lucky accident.

But there is a line of reasoning which attempts to remove any philosophical implications from such a situation, a line of reasoning used by those who prefer to believe that our universe is just an accident. The person who uses this line of reasoning claims that the fine-tuning of our universe can by explained by an “observer selection effect.” The reasoning goes like this:

What type of universe should we expect to observe? Why, of course, it is only a universe as fit-for-life as our universe, because that is the only type of universe that could have observers. So there is an “observer selection effect” such that all observers find themselves in universes like ours, and it is not surprising that we exist in a universe like this.

The same type of reasoning can be used in regard to the analogy of the firing squad. The reasoning goes like this:

In regard to that firing squad analogy, you should not be surprised to find yourself alive after facing a firing squad. This is because there is an “observer selection effect” which guarantees that all people who make observations after facing a firing squad are those who survived the firing bullets.

Let's examine such reasoning in detail, first examining the simpler case of the firing squad.

The Case of the Firing Squad

When reasoning involving an observer selection effect is used in regard to firing squads, this “observer selection effect” is an example of what is called a red herring. A red herring is an argumentative device in which a person introduces some irrelevant or less relevant consideration, perhaps to distract you from considering a more relevant consideration. It may be true that only alive persons could observe themselves as survivors of a firing squad, but such a fact doesn't make you the slightest bit more likely to survive a firing squad. So this “observer selection effect” claim is just a red herring.

But in regard to statements such as, “You should not be surprised to find yourself surviving a firing squad, because that's the only thing you could observe,” is there any validity to such reasoning? No, there isn't. In this case, it must be remembered that the case of a non-observation (no observation of anything, because you're instantly killed) is both a distinct possibility and a high likelihood. So we should not make the mistake of assuming that some observation must necessarily occur.

Here are some examples of correct statements and incorrect statements.

Incorrect statement: Of course, after you finish high school, you'll end up in Harvard with a nice dorm. Everyone knows a school like Harvard can afford nice dorms.
Correct statement: No, I almost certainly will not get such a dorm, because I almost certainly won't get into Harvard.

Incorrect statement: You can look forward to marrying a gorgeous starlet, because pretty much the only type of people who get to be movie starlets are gorgeous people.
Correct statement: No, I almost certainly won't marry such a starlet, because its way too improbable that I would marry anyone who is a starlet.

Incorrect statement: You shouldn't be surprised to survive a firing squad, because the only observation you could have after facing such a squad would be to observe that you survived.
Correct statement: No, I should be supremely surprised to survive a firing squad, because if I faced such a squad, it would be almost certain that I should be instantly killed and have no further observations.

The most persuasive (but devious) way to appeal to an observer selection effect is to present a loaded question, a type of question that makes an unfair assumption, as in the famous question asking: when did you stop beating your wife? The loaded question in the case of the firing squad might be: what should you see around you after facing a firing squad? Such a question is trickily phrased in a way so that the answer of survival is already answered. But such trick questions can always be answered though careful replies, as below.

Incorrect statement: What should you expect to observe after facing a firing squad? Only that you had survived, because otherwise you could make no observation.
Correct statement: You should not expect to observe anything at all after facing a firing squad, because you should be immediately killed by its bullets.

Another way to clarify the situation of the firing squad is to do what we may call an exhaustive possibility analysis. When we do such an analysis, rather than just considering two possibilities, we will try to consider every possibility. We can then consider the likelihood of each possibility.

Exhaustive Possibility Analysis for Firing Squad Situation
Possibility Likelihood
Possibility 1: You immediately die when the bullets kill you, and have no further observations. Very, very likely (unless your soul survives your death).
Possibility 2: You survive for a while, with light wounds Very unlikely
Possibility 3: You survive for a while, with very heavy wounds that will very soon cause you to die. Unlikely, but much more likely than Possibility 2.
Possibility 4: You survive with no wounds, because all of the bullets luckily missed by pure chance. Incredibly unlikely
Possibility 5: You die quickly, but observe your dead body when your soul floats out of your body. Debatable likelihood, but probably far more likely than possibility 4

When we consider all of these possibilities, it becomes clear that any type of “you should not be surprised to find yourself surviving a firing squad” reasoning (based on an observer selection effect) is utterly invalid, particularly if “survive the firing squad” means to end up in pretty good shape when the firing squad is finished. The most likely possibility is that you should instantly be killed by the firing squad, and have no observations after hearing the firing of the guns. The second most likely possibility is that you should be very heavily wounded after facing the firing squad, and have only a pitifully short observation before dying. Possibility 4 (being alive and not wounded) is extremely unlikely both in the full group of possibilities and also in the subset of possibilities that include some type of observations by you after facing the firing squad. It is not at all correct to suggest that some type of observer selection effect will make it likely that you will observe yourself in a good state after facing the firing squad. 

 

The Case of Habitable Universes

Now let's look at the case of habitable universes. The reason why I spent so much time discussing firing squads is that the situation in regard to habitable universes has a strong similarity to the firing squad situation. Here are the similarities: dying instantly in the firing squad is similar to a universe that is uninhabitable; being heavily wounded by the firing squad is similar to a universe that is just barely habitable; and surviving the firing squad without any wounds is similar to a universe that has no shortfalls in regard to habitability, which is the type of universe we find ourselves in.

Before trying to do an exhaustive possibility analysis, let us consider the type of “observer selection effect” arguments made in regard to habitable universes. Below are some examples, along with corrections.

Incorrect statement: We should not be surprised to find ourselves living in a finely tuned habitable universe, because the only type of universes that have observers are finely tuned universes.
Correct statement: We should be very surprised to find ourselves living in a finely tuned habitable universe (under assumptions of randomness), because it is vastly more likely that our universe should have been uninhabitable and not allowed us to exist as observers.

Incorrect statement: The fine-tuned nature of our universe is just as we should expect, because such conditions are prerequisites of our existence.
Correct statement: Under the assumption of randomness, the fine-tuned nature of our universe is incredibly surprising and improbable, given that there is nothing necessary about our existence.

Now let's try to do an exhaustive possibility analysis regarding types of universes and what type of observers (if any) they might have. This will include some interesting possibilities that are often overlooked.

Exhaustive Possibility Analysis for Observers in Random Universes
Possibility Likelihood
Possibility 1: The possibility of a random universe that is uninhabitable, and has no observers because life is impossible for one reason or another. Very likely under assumptions of randomness
Possibility 2: The possibility of a universe such as ours, with no serious shortfall in regard to habitability Very unlikely under assumptions of randomness
Possibility 3: The possibility of a barely habitable universe having some serious shortfalls in regard to habitability, but one in which observers are just barely possible. Very unlikely, but much more likely than Possibility 2, because the list of conditions that must be met for a barely habitable universe is much shorter than the list of conditions that must be met for a universe such as ours (as argued here). 
Possibility 4: The possibility of a universe in which biological observers cannot appear, but some other types of observers exist – perhaps souls or spirits, or minds of pure energy. Hard to estimate this likelihood
Possibility 5: The possibility of a universe in which biological observers cannot naturally appear, but one that might be observed by visitors from other universes






Hard to estimate this likelihood, but it might be allowed by exotic “wormhole” possibilities

`When we consider all of these possibilities, it becomes clear that any type of “you should not be surprised that our universe is like this” reasoning (based on an observer selection effect) is utterly invalid, particularly if “a universe like this” means a universe about as habitable and life-friendly as our universe. Under assumptions of randomness, the most likely possibility is that a universe should be uninhabitable and lifeless. The second most likely possibility is a universe that is just barely habitable, with conditions much harsher than ours (such as a universe with no stars or very few stars or a universe in which either carbon or oxygen was rare). Such a universe is probably at least a thousand times easier to achieve by chance than a universe such as ours (for reasons explained here).  Possibility 2 (a universe as hospitable to life as ours) is extremely unlikely both in the full group of possibilities and also in the subset of possibilities that include some type of observers. It is not at all correct to suggest that some type of observer selection effect will guarantee that observers only exist in a universe with a level of fine-tuning as great or almost as great as ours. 

habitable universes
 

The most devious trick of those who evoke an observer selection effect is to phrase questions such as “What type of universe should we expect to be living in?” This is a loaded question, one that has a particular assumption built in to it (like the famous loaded question which asks when did you stop beating your wife). The very phrase “should we expect to be living in” presupposes a universe with an observer.

Incorrect statement: What type of universe should we expect to be living in? A universe like the one we do live in, for in no other universe can there be observers.
Correct statement: You are asking a loaded question if you ask what type of universe should we expect to be living in, because the phrase “to be living in” presupposes habitability. Avoiding such a loaded question, we should ask: what type of universe should we expect our universe to be? Under the assumption of randomness, the answer is: an uninhabitable universe in which no observers ever existed. And in the very unlikely case that our universe happened to be habitable, the most likely case by far would be that it should be just barely habitable, since the requirement list for such a universe is much shorter. Such a barely habitable universe would be much less life-friendly than ours.

Observer Selection Effect” Reasoning at the Casino

Let's imagine a hotel casino where there's a special room called the Big Gamble. You have to pay $10 to get into the room, where you find a giant laser above your head. There's a lever you pull to try your luck. After you pull the lever, there is a 99.9999% chance that the laser will instantly incinerate you, reducing you to a charred cinder. But there's one chance in a million that you'll get a jackpot of 5 million dollars. Let's imagine a conversation between a gambler and a casino employee who gets a commission on all the people who try the Big Gamble.

Gambler: I was thinking of trying the Big Gamble, but I'm afraid it will probably just get kill me.
Casino Employee: Go ahead, take a chance!
Gambler: But I just can't see myself winning the 5 million dollars. That would be too surprising.
Casino Employee: Well, I can see you're just ignoring the “observer selection effect” here.
Gambler: What's that?
Casino Employee: Well, it works like this. If you don't get incinerated, and win the 5 million dollars, it won't be surprising at all. Because the only person who could have an observation after pulling the lever is a person who survived, and won the 5 million dollars. So you won't be surprised to have survived, because it's the only kind of observation you could have after pulling the lever.
Gambler: So it's not such a long shot that I'll win the 5 million?
Casino Employee: No, when you look at it as I just discussed, it won't be surprising at all.
Gambler: Okay, I'll make the bet.

The casino employee has used the classic argumentative technique of the red herring, something that distracts you from concentrating on what you should be concentrating on, and diverts you into thinking about some irrelevant distraction. In this case, whether or not you will be surprised by still being there after trying the Big Gamble is irrelevant. What is the relevant consideration is the probability of you winning. The reasoning of the casino employee is utterly sophistical and fallacious, and when similar “observer selection effect” reasoning is used on a cosmic scale when arguing about universes, it is just as irrelevant and misleading as this casino employee's sales pitch.