News Bits

by Luke Muehlhauser on November 2, 2011 in News

New Less Wrong posts: Great Explanations and Rational Romantic Relationships, Part 1.

Clear and well-written: The Meaning of Morality.

Mistakes skeptics make when arguing, when citing history, when using philosophy

Jeffrey Jay Lowder’s bibliography on religion and morality.

Just in case anybody was confused about this… there is no “moral judgment” faculty:

moral judgment is not a wholly unified faculty in the human brain, but rather, instantiated in dissociable neural systems that are engaged differentially depending on the type of transgression being judged.

From Twitter:

Previous post:

Next post:

{ 60 comments… read them below or add one }

Garren November 2, 2011 at 10:54 pm

The “‘moral judgement’ faculty” link goes to a lesswrong login page. Direct download available here without login:

http://philpapers.org/rec/PARIMU

(I linked to the same paper yesterday.)

  (Quote)

Luke Muehlhauser November 3, 2011 at 2:00 am

Garren,

Fixed, thanks.

  (Quote)

Reginald Selkirk November 3, 2011 at 2:19 pm

Mistakes skeptics make when arguing

Quoting obsolete philosophy. Many skeptics only quote the great philosophers, such as Karl Popper and Bertrand Russell. The problem with this is that apologists have long formed powerful objections to these men’s arguments. There are countless professional philosophers, such as Richard Swinburne, Peter Van Inwagen, and Alvin Plantinga, who have dedicated their lives to creating strong responses to such skeptical claims.

Just because a philosopher is dead does not make him obsolete, and just because an apologist who prefers to call himself a philosopher has dedicated his life to something does not mean he has succeeded.

Debating has nothing to do with being right or wrong.

At least he got that one right.

  (Quote)

Very Warm Poop November 3, 2011 at 4:53 pm

Just because a philosopher is dead does not make him obsolete

That misrepresents what Greg was saying. In the quoted passage Greg seems to imply that an “obsolete” philosopher is one whose arguments have been met with “powerful objections.” Their discontinued state of breathing has nothing to do with whether they are obsolete in this case.

just because an apologist who prefers to call himself a philosopher has dedicated his life to something does not mean he has succeeded.

To be clear it does not seem that Greg is making that argument either. He instead seems to be saying that some people’s life long efforts have made certain philosophers “obsolete.”

  (Quote)

antiplastic November 3, 2011 at 7:20 pm

Thanks for that quoted passage.

Hearing someone claim that Plantinga has formulated “powerful arguments” tells me I don’t need to waste my time clicking the link and reading through it.

  (Quote)

antiplastic November 4, 2011 at 12:51 am

Oh good gracious it’s worse than I thought.

He thinks “who designed the designer” is a bad argument.

Has all your Yud-fellating taught you nothing about the importance of minimizing description string length? A complex designer takes more bits to describe than the complicated universe he is supposed to have designed. Therefore, a designer is more improbable (by definition). Therefore, a designer is no “explanation” at all.

  (Quote)

Ryan M November 4, 2011 at 5:31 am

Antiplastic,

Does the “Who designed the designer” argument work against theists who believe that its both the case that God is necessary (Either exists in all logically possible worlds or is metaphysically necessary), and that God is simple (Divine simplicity seems to be pretty standard since Augstine/Aquinas)?

I don’t see how it could considering that following divine simplicity it makes no sense to say that God is complex, and following God being a necessary being it makes no sense to say God could have been the sort of thing that could have been created.

I’m an atheist, so I obviously believe that the arguments for theism are failures. But, I still think the “Who designed the designer” argument obviously misses the point of cosmological arguments and shows a great ignorance of the God that is believed by the likes of Swinburne, Aquinas, etc. Am I wrong in thinking that the allegedly necessarily existing simple God is still the type that could be argued needs a creator and is also complex?

Also, I don’t buy into Plantinga’s arguments either. But as far as philosophical arguments go, his are obviously not false and they are indeed sophisticated. I don’t think you will find the average man refuting Plantinga’s Ontological argument during a commercial break.

  (Quote)

Reginald Selkirk November 4, 2011 at 6:41 am

(Divine simplicity seems to be pretty standard since Augustine/Aquinas)

Yes and no. It is pretty standard as a counter-claim to certain arguments, but is not consistently maintained in all theological claims. Consider just stuff you mentioned in your own short post. “Necessary existence” – is that consistent with simple? Able to create at least one Universe – is that consistent with simple? Simple might be something like the force of gravity, but no one claims gravity to be the source of objective morality, or that gravity is a “person” – let alone three persons. And so on.

But as far as philosophical arguments go, his are obviously not false and they are indeed sophisticated. I don’t think you will find the average man refuting Plantinga’s Ontological argument during a commercial break.

Plantinga’s evolutionary argument against naturalism is obviously false. I don’t know what you mean by “sophisticated” – if you mean displays an awareness of what is known of a field of science being used in the argument, then EAAN is not sophisticated. Perhaps the Average Man does not have an understanding of evolutionary theory, but any person who does should have no trouble at all pointing out the inadequacies of Plantinga’s attempt.

  (Quote)

Reginald Selkirk November 4, 2011 at 6:53 am

As for the ontological argument, the concept can be conveyed in a sentence or two. Plantinga’s The Nature of Necessity checks in at 272 pages – all the more verbiage in which to hide the question-begging. and even the Stanford Encyclopedia of Philosophy says:

In more recent times, Kurt Gödel, Charles Hartshorne, Norman Malcolm and Alvin Plantinga have all presented much-discussed ontological arguments which bear interesting connections to the earlier arguments of St. Anselm, Descartes and Leibniz. Of these, the most interesting are those of Gödel and Plantinga; in these cases, however, it is unclear whether we should really say that these authors claim that the arguments are proofs of the existence of God.

Much more length, much less claim. Here’s a fair description of what Plantinga is doing: Polishing a Turd.

  (Quote)

PDH November 4, 2011 at 10:25 am

You can say that God is ‘simple’ just as you can say that human intelligence is simple but that doesn’t mean that it actually is. If we wrote human intelligence out as a computer program that actually worked would that program be simple? Clearly not. Nor would ‘God,’ if the word has any meaning at all, be a simple explanation in those terms. If you want to use the God hypothesis to explain things like where the universe came from then you run into massive problems. For example, why did God create the universe on theism? Because He loves us, of course. But love is an exceedingly complex part of human psychology so maintaining that God is simple necessitates maintaining that God cannot love but then why did He create the universe?

God is supposed to be an intelligent being but intelligence is literally the most complex thing in the known universe. So, you can say that God is not intelligent but then,

a) How is He to be distinguished from a purely natural process?
b) How can we then use this as an explanation of where the universe came from?

If God is just a process that makes universes, well so is eternal inflation. When we talk about God we’re obviously not talking about inflation. We’re talking about an intelligent being with recognisable psychology not some simple, unintelligent process that just does its thing in a predictable, mechanical way.

And there must be some reason to think that God would create the universe if this is to be an explanation of the origins of the universe. There’s no getting away from the fact that theism is the claim that the universe runs on human psychology. What little predictive power it has flows entirely from this.

As for ‘necessary being.’ Nobody disagrees that necessary beings are necessary but atheists fairly obviously don’t believe in the hypothesis that there is a necessary being analogous to God and merely defining him as such does not make said being pop into existence. It means only that the being thus defined has no referent in reality, and we can say this with a degree of confidence that can be described by probability theory.

Because most beings are not necessary and the beings that people generally agree are necessary (numbers, for example) tend to be utterly unlike an intelligent being it is very improbable that there is some new necessary being that we don’t know about. And it’s even more improbable that that being is the God of traditional theism (and still less probable that it is Christian theism, say). So here we have a probabilistic argument against a necessary being.

The probability that there is a necessary being corresponding to the God of theism (or deism, for that matter) is not ’1′ just because if it was necessary it couldn’t not exist! Suppose I define a ‘Snorgle’ as a necessarily existing sandwich. Is the probability of this hypothesis ’1′? It would be absurd to say that it was and even more absurd to say that it was beyond Bayesian probability theory because of its alleged logical necessity. The probability of a hypothesis derives from the logical relationships between it and the rival, mutually exclusive hypotheses that constitute its negation. Snorgle has a logical relationship with ~Snorgle in that they can’t both be right at the same time and there are rather more possible world descriptions that don’t contain Snorgle than there are ones that do, thus the probability of Snorgle is is lower than ~Snorgle.

It’s the probability of the hypothesis that we’re interested in, not the ontological probability of the thing itself existing, whatever that means. Likewise, the complexity of the hypothesis, not the complexity of the entity itself. Theism is not a simple hypothesis.

  (Quote)

Ryan M November 4, 2011 at 12:55 pm

You two are writing far more than you need to.

What my post was trying to do was show that it makes no sense to respond to lets say Thomas Aquinas’ Five Ways by saying “What caused the designer/What caused God”. It makes no sense because the theist’s arguments conclude with a necessary being of some sort (i.e. First mover, the only unchanging being [Pure act]).

Nothing in my post disputes that the given P (There being a necessary causal being) the probability of G (God being the necessary being) is not 1. I think one of the best responses to most cosmological arguments is in fact that the arguments do not give good reason to believe that the necessary being with causal powers must be God or something like God. Way more arguments are needed to show that it is either necessary or probable that the first cause is God. Also, I don’t think a theist would claim that intelligence is necessarily complex. I assume they would argue that intelligence is only complex in the physical world due to the relation of large numbers of parts. Or something like that.

I’m not even sure who you are referring to when you say “It’s the probability of the hypothesis that we’re interested in”. I’m not sure who you are defending, or why.

As for you Reginald, necessary existence and creation don’t seem to be inconsistent with divine simplicity as it has been defined and argued for by the likes of Swinburne and Aquinas. As for Plantinga’s EAAN, if you think it is obviously false then maybe you should write something about it. I think its false, but I can’t seem to find why it is obviously false (If it is).

  (Quote)

PDH November 4, 2011 at 1:11 pm

Ryan M wrote,

You two are writing far more than you need to.

What my post was trying to do was show that it makes no sense to respond to lets say Thomas Aquinas’ Five Ways by saying “What caused the designer/What caused God”. It makes no sense because the theist’s arguments conclude with a necessary being of some sort (i.e. First mover, the only unchanging being [Pure act]).

Nothing in my post disputes that the given P (There being a necessary causal being) the probability of G (God being the necessary being) is not 1. I think one of the best responses to most cosmological arguments is in fact that the arguments do not give good reason to believe that the necessary being with causal powers must be God or something like God. Way more arguments are needed to show that it is either necessary or probable that the first cause is God. Also, I don’t think a theist would claim that intelligence is necessarily complex. I assume they would argue that intelligence is only complex in the physical world due to the relation of large numbers of parts. Or something like that.

I’m not even sure who you are referring to when you say “It’s the probability of the hypothesis that we’re interested in”. I’m not sure who you are defending, or why.

To be honest, when I looked over the thread after posting I immediately thought ‘who am I talking to?’

I think maybe I’d just skimmed Selkirk’s post, assumed the objection to which he was replying was something else entirely based on previous arguments that I’ve had recently and then went off one. Oops!

And my writing wasn’t especially clear or concise, in any case, you’re right. Sorry about that, I’m in a bad mood today.

  (Quote)

Rorschach November 4, 2011 at 4:30 pm

I assume they would argue that intelligence is only complex in the physical world due to the relation of large numbers of parts. Or something like that.

Well, wouldn’t it be functionally complex? Besides, that doesnt make sense. If the mind is a immaterial undivisible substance, then why certain faculties of the mind are lost due to brain injury?

  (Quote)

Patrick November 4, 2011 at 9:15 pm

I’m the Patrick who had the long argument with him in the “who designed the designer” thread.

I’ll happily concede that this response has no force against someone who is more into God as a “necessary being,” or whatever horrific abortion of stolen philosophical language is being applied to God at the moment.

But I’ll argue to the day that I die that if someone says, “all things that begin to exist have a cause,” and then defines “beings to exist” in such a way that it applies to God, then its valid to respond with the “who designed the designer” type critique.

He thinks atheists are making a mistake to not permit theists to incorporate, without reference, massive amounts of philosophical positioning. Its the iron man fallacy: rewriting your opponents arguments to be less stupid.

  (Quote)

Reginald Selkirk November 5, 2011 at 6:18 am

As for Plantinga’s EAAN, if you think it is obviously false then maybe you should write something about it. I think its false, but I can’t seem to find why it is obviously false (If it is).

It’s certainly been done before by others. I have commented on it often when the topic has come up on this blog.

I will mention (in order to elucidate my viewpoint, not to assert authority) that I am a research biologist by profession, so that Plantinga’s account of the implications of evolution strike me as a heaping helping of WTF. The fact that he took a ride on the Intelligent Design Creationism bandwagon does not help him.

  (Quote)

Yvain November 5, 2011 at 8:12 am

I’ve written a response to the objections against “Who designed the designer” here.

  (Quote)

Ryan M November 5, 2011 at 8:31 am

It’s certainly been done before by others. I have commented on it often when the topic has come up on this blog.

I will mention (in order to elucidate my viewpoint, not to assert authority) that I am a research biologist by profession, so that Plantinga’s account of the implications of evolution strike me as a heaping helping of WTF. The fact that he took a ride on the Intelligent Design Creationism bandwagon does not help him.

Nice. I knew you are in some sort of science field, but was not sure which.

  (Quote)

antiplastic November 5, 2011 at 1:13 pm

Does the “Who designed the designer” argument work against theists who believe that its both the case that God is necessary (Either exists in all logically possible worlds or is metaphysically necessary), and that God is simple (Divine simplicity seems to be pretty standard since Augstine/Aquinas)?

In fact not one single, actual, flesh and blood Christian believes in divine simplicity. So in addition to having to sit through an argument which is demonstrably incoherent, the thinking atheist must suffer the double indignity of having it delivered by someone who is being dishonest and not arguing for what they really believe.

Read what PDH said about the multivariate aspects of even simple consciousness and then read the bolded passage in my last post again.

A complex designer takes more bits to describe than the complicated universe he is supposed to have designed. Therefore, a designer is more improbable (by definition). Therefore, a designer is no “explanation” at all.

Note that none of this “assumes materialism”. An immaterial mind still takes billions of bits to describe. Observe that this is orthogonal to “necessity”. A necessary mind still takes billions of bits to describe.

Who designed the designer? One cannot answer this without patheically question-begging special pleading, lying, or both.

But, I still think the “Who designed the designer” argument obviously misses the point of cosmological arguments and shows a great ignorance of the God that is believed by the likes of Swinburne, Aquinas, etc. Am I wrong in thinking that the allegedly necessarily existing simple God is still the type that could be argued needs a creator and is also complex?

Catastrophically. You’ve been bamboozled by obscurantists and professional kickers-up of dust. To be fair, virtually the entire nonbelieving population of the professional Philosophy of Religion community has been bamboozled too.

Also, I don’t buy into Plantinga’s arguments either. But as far as philosophical arguments go, his are obviously not false and they are indeed sophisticated. I don’t think you will find the average man refuting Plantinga’s Ontological argument during a commercial break.

Arguments from WLC or the gnomelike Plantinga are more “sophisticated” in the sense of “having more moving parts”. But like a rube goldberg machine, all this distracting apparatus serves only the function of concealing the handful of very simple-to-explain fallacies at their core. Like the 3-traunched credit default swaps that brought the global economy to the brink of destruction, they are “sophisticated” ways, not of getting at truth, but of hiding from the light of day the fundamental fraudulence of the enterprise.

  (Quote)

Ryan M November 5, 2011 at 3:32 pm

Anti,

Do you literally mean that the set of Christians who believe in divine simplicity is en empty set? That seems nonsensical. I wonder how you could even defend that notion.

So rather than answer my objection that “Who designed the designer” ignores the premises and conclusion of cosmological arguments, you just say I have been confused. Others might take that for a good rebuttal, but I will not.

What is a necessary mind? Are you talking about a mind that exists in all possible worlds?

Lets try this: A cosmological argument attempts to show that due to either logical or metaphysical necessity, there must be a terminus to either motion, change, or contingency. The arguments attempt to show that since it is either logically or metaphysically necessary that motion, change, contingency must end, or that there cannot be an infinite chain of the 3, then there must be a start of the 3. The start of motion would be a non moved thing, the start of change would be an unchanged thing (Pure act), the start of contingency would be a necessary being. The causes to the 3 (Or really, the cause) is supposed to be a necessary being. If the being is supposed to be a necessary being, then it exists in all possible worlds, it was not created, it could not possibly have not exited.

So understanding this, in what way could “What designed the designer” be a possibly good rebuttal to the 3 types of cosmological arguments that conclude with a being that could not possibly be created since it is necessary?

Just wonder btw, what books/papers have you read by theist/atheist philosophers? I’d like to see you (Not write a book rebuttal of theist books), but show how theist philosophers such as Swinburne present invalid or valid but unsound arguments. I honestly think that some of my fellow atheists who complain belittle the arguments of Craig, Swinburne, Plantinga, cannot actually refute their arguments by means of showing them to be invalid, or valid but with false premises. ‘sigh’

  (Quote)

Leon November 5, 2011 at 9:19 pm

Antiplastic said:

An immaterial mind still takes billions of bits to describe. Observe that this is orthogonal to “necessity”. A necessary mind still takes billions of bits to describe.

PDH said:

The probability that there is a necessary being corresponding to the God of theism (or deism, for that matter) is not ’1′ just because if it was necessary it couldn’t not exist! Suppose I define a ‘Snorgle’ as a necessarily existing sandwich. Is the probability of this hypothesis ’1′? It would be absurd to say that it was and even more absurd to say that it was beyond Bayesian probability theory because of its alleged logical necessity. The probability of a hypothesis derives from the logical relationships between it and the rival, mutually exclusive hypotheses that constitute its negation. Snorgle has a logical relationship with ~Snorgle in that they can’t both be right at the same time and there are rather more possible world descriptions that don’t contain Snorgle than there are ones that do, thus the probability of Snorgle is is lower than ~Snorgle.

I think you’re both misunderstanding necessity here. In the language of PDH’s argument, necessity implies that the necessary thing exists in all possible worlds. If Snorgle is indeed necessary, apart from your having defined it so, then “with probability 1” the sandwich exists; i.e., it exists in every possible world. (Actually, it’s more than “with probability 1” if we’re using classical measure theory.)

An example of something necessary in LessWrong-style philosophy might be Bayes’ theorem itself, or the idea that Bayes’ theorem captures “learning” in a “mathematically correct” way (Luke’s language from a LW post I can’t be bothered looking up). Or maybe the law of non-contradiction.

Both Bayes’ theorem and the law of non-contradiction take a nonzero number of bits to describe, yet it doesn’t seem crazy to suggest that they are both necessarily true in some pretty strong sense.

Also, I doubt most (mono)theists would be happy with the idea that their god could be completely described by some finite bitstring.

My main points are therefore:

1. In this context, you two should be arguing that God isn’t necessary.

2. You can’t just turn any proposition you like into a hypothesis for Solomonoff induction without being explicit about your assumptions — in particular, that there is a 1-1 mapping between anything that could possibly be true and strings of bits.

  (Quote)

Kevin November 6, 2011 at 5:43 am

I think you’re both misunderstanding necessity here. In the language of PDH’s argument, necessity implies that the necessary thing exists in all possible worlds. If Snorgle is indeed necessary, apart from your having defined it so, then “with probability 1” the sandwich exists; i.e., it exists in every possible world.

They aren’t missing anything. They are talking about epistemic probability. What you are referring to is the ontological probability. Suppose you roll a fair six-sided die and then cover it. We then ask the question of whether the outcome was a six. The ontological probability is either one or zero (the six is either face up or not). The epistemic probability, given the information that we have is 1/6. Even though the roll is already determined, from our perspective, we can only predict the outcome probabilistically. Even if the roll landed on a six, we still wouldn’t be in a position to believe it since we don’t have the information needed to reach that conclusion. For this reason, there is no need to deny the attribute of necessity on the basis of probability theory since we can still evaluate the claim as probably false, despite the outcome being already determined, with it being possibly true.

Also, accepting the attribute of necessity makes the case for theism much harder to demonstrate since we don’t have a method for distinguishing a necessary being versus a non-necessary being. This would mean that even if Jesus came back, it would still not be conclusive evidence for theism since we would not be able to determine whether he is necessary or not. However, I doubt this would cause any hesitation for believers to claim so, supporting the notion that this wordplay is simply to make theism superficially sound more plausible by making it apart of a special class of objects.

  (Quote)

Reginald Selkirk November 6, 2011 at 8:15 am

Some brief notes on Plantinga’s Evolutionary Argument Against Naturalism (EAAN):

From Wikipedia

EAAN argues that the combination of evolutionary theory and naturalism is self-defeating on the basis of the claim that if both evolution and naturalism are true, then the probability of having reliable cognitive faculties is low.

First of all, are our cognitive faculties reliable? No. People are subject to all manner of perceptual and cognitive illusions. I have made this point repeatedly on this blog, and I think I have Luke convinced.

Consider one example, the Monty Hall problem: this is a fairly simple problem based on probability, and yet many people get it wrong. A large number of those people, including a few professional mathematicians, persist in their wrongness even after it has been explained to them at length.

Consider another example: assuming naturalism, a large majority of people are wrong in believing their imaginary friend actually exists.

Despite what you may have gleaned from your topnotch education watching the Discovery Channel, evolution does not yield perfection. It yields “usually good enough;” it balances trade-offs, and it works from pre-existing material. “Reliable” is not a binary trait; there are varying degrees of reliability. The perceptions and senses of our ancestors were reliable enough to get us here. But they are not perfectly reliable, and they are so in a way which indicates our evolutionary past.


Other problems with Plantinga’s argument:

Assuming evolution is true (as Plantinga does for purposes of his argument), human perception and cognition are very much one with animal perception and cognition. They are tied to our evolutionary history. For example, as an ape, we can see three primary colors, not four as birds do or 10 as mantis shrimp do (including polarised light!). We should consider studies in animal cognition, and there nothing exceptional left that needs explaining.

There are numerous examples of species whose perception and cognition were not perfect. For example, the many species which did not show fear of man when he expanded into their territory, like the dodo bird, and thus suffered extinction. Extinction is not compatible with perfection, it implies something was not created to deal with everything it would face. Due to our shared evolutionary history, such examples cannot be considered apart from the human situation. Anyone with a decent understanding of evolution is going to reject an argument based on human exceptionalism.

  (Quote)

Patrick November 6, 2011 at 8:24 am

The biggest and most obvious problems with Plantinga’s EEAN (and with Plantinga in general, see, eg, reformed epistemology) is that he likes to think of cognition as a black box that just spits out otherwise inexplicable data.

  (Quote)

PDH November 6, 2011 at 9:11 am

Kevin wrote,

They aren’t missing anything.They are talking about epistemic probability.What you are referring to is the ontological probability.Suppose you roll a fair six-sided die and then cover it.We then ask the question of whether the outcome was a six.The ontological probability is either one or zero (the six is either face up or not).The epistemic probability, given the information that we have is 1/6.Even though the roll is already determined, from our perspective, we can only predict the outcome probabilistically.Even if the roll landed on a six, we still wouldn’t be in a position to believe it since we don’t have the information needed to reach that conclusion.For this reason, there is no need to deny the attribute of necessity on the basis of probability theory since we can still evaluate the claim as probably false, despite the outcome being already determined, with it being possibly true.

Yes, this was exactly my point.

  (Quote)

joseph November 6, 2011 at 9:18 am

@Reginald Selkirk

You forgot to mention:
1. How tasty Mantis Shrimp are.
2. That they are also known as “Pissing Shrimp” in Chinese (roughly)

More seriously I was aware they can detect polarised light, but 10 colours! Amazing.

  (Quote)

Reginald Selkirk November 6, 2011 at 10:36 am

link

As impressive as their arms are, the eyes of a mantis shrimp are even more incredible. They are mounted on mobile stalks and can move independently of each other. Mantis shrimps can see objects with three different parts of the same eye, giving them ‘trinocular vision’ so unlike humans who perceive depth best with two eyes, these animals can do it perfectly well with either one of theirs.

Their colour vision far exceeds our too. The middle section of each eye, the midband, consists of six parallel strips. The first four are loaded with eight different types of light-sensitive cells (photoreceptors), containing pigments that respond to different wavelengths of light. With these, the mantis shrimp’s visible spectrum extends into the infrared and the ultraviolet. They can even use filters to tune each individual photoreceptor according to local light conditions.

The fifth and six rows of the midband contain photoreceptors that are specialised for detecting polarised light. Normally, light behaves like a wave that vibrates in every possible direction as it moves along. In comparison, polarised light vibrates in just one direction – think of attaching a piece of string to a wall and shaking it up and down. While we are normally oblivious to it, it’s present in the glare that reflects off water and glass and we use polarising filters in sunglasses and cameras to screen it out.

Light can also travel in a the shape of a helix, moving as a spiralling beam that spins either clockwise (right-handed) or anti-clockwise (left-handed). This phenomenon is called ‘circular polarisation’. Tsyr-Huei Chiou from the University of Maryland found that the mantis shrimp’s eye contains the only known cells in the animal kingdom that can detect it. Our technology can do the same, but the mantis shrimps beat us to it by as much as 400 million years.

  (Quote)

Leon November 6, 2011 at 1:49 pm

They aren’t missing anything. They are talking about epistemic probability. What you are referring to is the ontological probability. Suppose you roll a fair six-sided die and then cover it. We then ask the question of whether the outcome was a six. The ontological probability is either one or zero (the six is either face up or not). The epistemic probability, given the information that we have is 1/6. Even though the roll is already determined, from our perspective, we can only predict the outcome probabilistically. Even if the roll landed on a six, we still wouldn’t be in a position to believe it since we don’t have the information needed to reach that conclusion. For this reason, there is no need to deny the attribute of necessity on the basis of probability theory since we can still evaluate the claim as probably false, despite the outcome being already determined, with it being possibly true.

What do you mean by “ontological probability”? It seems like in the Bayesian-epistemological system, there is no such thing: it is either zero or one in every case. The universe is not sure or unsure about itself; what exists is what exists and what is true is what is true.

You say that there is “there is no need to deny the attribute of necessity on the basis of probability theory”. But what is necessity in a Bayesian framework? And how does it relate to “ontological probability” above? It seems to me that LW-style Bayesian epistemology aims to make “necessity” unnecessary — to “dissolve” it, using the lingo.

All I’m trying to do here is to make your argument more consistent: it should be something like,

1. Necessity is a meaningless concept: we can be relatively sure or unsure about truths/hypotheses, but the necessary/contingent distinction makes no sense.
2. Given the idea of a necessary being is bunk, we can come up with a prior for the God-hypothesis by its complexity, which is of the order of blah TB.

  (Quote)

Kevin November 6, 2011 at 3:14 pm

What part of the dice example did you not understand?

  (Quote)

PDH November 6, 2011 at 3:39 pm

Leon,

I can remain mute on the issue of ontological probabilities. It does sound nonsensical to me but if, for example, certain interpretations of quantum physics turned out to be correct, well it seems to me that I could just say that there are two kinds of probability and the one I’m interested in is epistemic probability.

To be clearer, take three hypotheses:

H1: The universe was created by a process.
H2: The universe was created by an intelligent process.
H3: The universe was created by a logically necessary intelligent process.

H1 is simpler and more probable than H2, which is simpler and more probable than H3.

Note that there are no situations in which H1 is false and H2 and/or H3 are true. If there is a logically necessary intelligent process, well that’s still a process so H1 is still correct. But there are many situations in which H1 is correct and the others are not. There are also situations in which all three false, for instance if the universe was not created at all.

Adding more specificity to a hypothesis – such as specifying that a being be not only intelligent but logically necessary to boot – lowers its probability. More things have to go right in order for it to be correct.

If it’s true, it might or might not be true necessarily. If it’s true. But we don’t know, at least not for sure, that it is true and it seems very unlikely to me that it is.

And yes, of course, all the talk about necessary beings and divine simplicity might be complete gibberish, anyway. This is all just to point out that theistic objections to the ‘Who designed the designer?’ retort are not an automatic win. One has to make a good case for the divine simplicity and logical necessity of God, if one intends to appeal to such things.

  (Quote)

Luke Muehlhauser November 6, 2011 at 5:29 pm

Did anybody like ‘Fable of the Dragon Tyrant’?

  (Quote)

Leon November 6, 2011 at 5:36 pm

Kevin –

What part of the dice example did you not understand?

I did not understand how it was supposed to relate to necessity — it seemed more relevant to the distinction between different interpretations of probability.

PDH –

H1: The universe was created by a process.
H2: The universe was created by an intelligent process.
H3: The universe was created by a logically necessary intelligent process.

Note that there are no situations in which H1 is false and H2 and/or H3 are true. If there is a logically necessary intelligent process, well that’s still a process so H1 is still correct. But there are many situations in which H1 is correct and the others are not. There are also situations in which all three false, for instance if the universe was not created at all.

Mmm. As I understand it, logical necessity is not just an attribute like intelligence. The “logically necessary” part of H3 asserts instead that H3 is a logical deduction from H1. In other words, the “necessary” part claims that there are no situations in which H1 is correct and the others are not. Equivalently (maybe) it is a claim that H3 is exactly as simple as H1, which perhaps explains the motivation for the divine simplicity idea.

My only point is that one needs to dispute this claim.

(As an analogy, I think the cosmological-argument theist presumes that the above hierarchy actually looks like:

H1. Euclidean triangles exist.
H2. Euclidean triangles whose internal angles sum to pi exist.
H3. Euclidean triangles whose internal angles necessarily sum to pi exist.)

And yes, of course, all the talk about necessary beings and divine simplicity might be complete gibberish, anyway. This is all just to point out that theistic objections to the ‘Who designed the designer?’ retort are not an automatic win. One has to make a good case for the divine simplicity and logical necessity of God, if one intends to appeal to such things.

True! But “who designed the designer” is itself a retort, and divine simplicity and logical necessity are normally not retorts, but part of the original argument (if it is well-put).

The original issue was, “how good is ‘who designed the designer?’?”. All I’m saying is, “it’s great, in conjunction with either ‘divine simplicity is gobbledygook’ or ‘I can’t see how God could be logically necessary’”. Or both.

  (Quote)

Leon November 6, 2011 at 5:46 pm

Did anybody like ‘Fable of the Dragon Tyrant’?

I thought it was interesting, but perhaps longer than it deserved to be, and not especially “literary”. It felt like it could have been either more fable-like and less obviously written by a philosopher trying to make a point, or some kind of magical realist novella with well-fleshed out characters and relationships.

As to the point itself, it reveals something I think transhumanists, Christians, and Harry Potter fans have in common (“the last enemy to be destroyed is death”) over and against Steve Jobs and a pretty good proportion of the run-of-the-mill irreligious (“death is what gives life meaning”; “where death is, I am not; where I am, death is not”, “We thank with brief thanksgiving / Whatever gods may be / That no life lives for ever; / That dead men rise up never; / That even the weariest river / Winds somewhere safe to sea.”, etc.).

  (Quote)

PDH November 6, 2011 at 6:27 pm

Leon wrote,

Mmm. As I understand it, logical necessity is not just an attribute like intelligence. The “logically necessary” part of H3 asserts instead that H3 is a logical deduction from H1. In other words, the “necessary” part claims that there are no situations in which H1 is correct and the others are not. Equivalently (maybe) it is a claim that H3 is exactly as simple as H1, which perhaps explains the motivation for the divine simplicity idea.

It claims that this is the case but the claim itself does not have a probability of 1. If I say, ‘there is a cabbage in my fridge,’ and you say, ‘there is a logically necessary cabbage in your fridge’ it is more likely that I’ll be right, even if logically necessary cabbages exist in all possible worlds. Because even if there is a logically necessary cabbage in my fridge, I’ll still be right. It’s still a cabbage. But if there is just an ordinary cabbage, which is much more likely to be the case, I’ll be right again and you’ll be wrong.

I would even go as far as to say that when you claim something is logically necessary on top of everything else you’ve actually quite drastically reduced its probability. That’s a very bold claim. Think how hard it is to establish that something is logically necessary.

It’s a distinction between epistemology and ontology. You’re talking about the territory and I’m talking about the map.

  (Quote)

joseph November 6, 2011 at 9:01 pm

@Reginal Srlkirk

Thankyou! Don’t want to use too much of Luke’s space, but here’s something a bit more in my sphere (though I am not a researcher, just an occassional user):

http://pda.physorg.com/_news89542035.html

http://www.newscientist.com/mobile/article/mg21128262.000-crittervision-what-a-dogs-nose-knows.html

  (Quote)

Leon November 6, 2011 at 10:17 pm

PDH –

I find this confusing, because it seems to me that “logical necessity”-type claims concern how you partition your hypothesis space. They are statements about equivalences or inclusion-type relations among hypotheses, so it seems strange to suggest that they are themselves hypotheses. Another level of meta has slipped in. To do well-defined Bayesian inference, you need a well-defined hypothesis space. I don’t know if it’s productive or consistent to allow inference on spaces of hypothesis spaces (ad infinitum); normally, statements about the hypothesis space are either true or false. Maybe there’s some kind of category-theoretic way of getting around it. :P

It claims that this is the case but the claim itself does not have a probability of 1. If I say, ‘there is a cabbage in my fridge,’ and you say, ‘there is a logically necessary cabbage in your fridge’ it is more likely that I’ll be right, even if logically necessary cabbages exist in all possible worlds. Because even if there is a logically necessary cabbage in my fridge, I’ll still be right. It’s still a cabbage. But if there is just an ordinary cabbage, which is much more likely to be the case, I’ll be right again and you’ll be wrong.

As I understand it, logical necessity applies to statements or hypotheses. Saying an entity is logically necessary is shorthand for saying that the statement that that entity exists is logically necessary, e.g. “there is a logically necessary cabbage” is shorthand for “it is logically necessary that there exist a cabbage”.

You put ‘there is a logically necessary cabbage in your fridge’. To me, the most natural reading is that “(it is logically necessary that a cabbage exist) & (one of them is in my fridge)”. This clearly entails “there is a cabbage in my fridge” and not vice versa, and to that extent I agree with your reasoning.

The difficulty arises when you try to assign a probability to the the sub-hypothesis “it is logically necessary that a cabbage exist”. If this statement is “totally” true, there is a cabbage in all possible worlds — no problem. If it is “totally” false, then it is false for all worlds and we’re fine. But if it has a non-zero, non-one probability, then it is apparently true in some possible worlds, but not others. But if it is not true in some possible worlds, then in some possible worlds, it is not true in all possible worlds — *syntax error*.

To pick up on a potential ambiguity: ‘there is a logically necessary cabbage in your fridge’ could be parsed as ‘there is a logically necessary (cabbage in my fridge)’ (i.e. equivalent to ‘it is logically necessary that there is a cabbage in your fridge’). In that case, when you say “even if logically necessary cabbages exist in all possible worlds”, you should say “even if logically necessary cabbages in my fridge exist in all possible worlds”. If you can assign a probability to the statement “logically necessary cabbages in my fridge exist in all possible worlds”, then it again seems like you’re implicitly making use of an all-possible-worlds space in which some all-possible-worlds have cabbages in your fridge.

—–

If we try to simplify things by going “back to the maths”, then the difficulty is as follows: how can you assign a probability to the statement “hypothesis A necessarily implies [i.e., materially conditionally, not via Bayesian updating] hypothesis B”? (Or even “hypothesis A is identical to hypothesis B”.)

P.S. I think this (pdf) might be related, even if it’s not directly relevant.

  (Quote)

Kevin November 7, 2011 at 4:02 am

But if it has a non-zero, non-one probability, then it is apparently true in some possible worlds, but not others.

This is where you make your mistake. The probability applies to the entire proposition. Saying that it has a 50-50 probability means that the proposition “a necessary vegetable exists” has a 50% chance of being true and a 50% chance of being false, not that it is true in 50% of the possible worlds and false in the other 50%. If we find that it doesn’t exist in one possible world, then that falsifies the proposition; it doesn’t decrease the probability by 1/(n possible worlds).

  (Quote)

Larkus November 7, 2011 at 4:12 am

Did anybody like ‘Fable of the Dragon Tyrant’?

I didn’t like it. I found it too preachy.

I like this talk better:
http://www.ted.com/talks/aubrey_de_grey_says_we_can_avoid_aging.html

  (Quote)

Zeb November 7, 2011 at 5:34 am

My response to the Dragon Tyrant was the same as Leon’s. One could make a fine parable out of it by trimming it to 100-500 words, or maybe a Narnia-style loose allegory by building out the characters. Personally I would rather see a straight argument for fighting senescence than either a parable or allegory anyway.

  (Quote)

Zeb November 7, 2011 at 11:43 am

Oh yeah, that TED talk, I thought this subject sounded familiar to me. That was good, it got me wishing we were working on ending senescence!

  (Quote)

Leon November 8, 2011 at 4:55 pm

This is where you make your mistake.The probability applies to the entire proposition.Saying that it has a 50-50 probability means that the proposition “a necessary vegetable exists” has a 50% chance of being true and a 50% chance of being false, not that it is true in 50% of the possible worlds and false in the other 50%.

That’s true — it doesn’t have to be true in 50% of the possible worlds. But my understanding of possible worlds and personal probability is that our probability distribution is over possible worlds. To say that a proposition has “50% chance of being true and a 50% chance of being false” means that half our probability mass is on possible worlds where it is true, and half is on possible worlds where it is false.

If we find that it doesn’t exist in one possible world, then that falsifies the proposition; it doesn’t decrease the probability by 1/(n possible worlds).

Yes — combined with my above interpretation, this is exactly the paradox. If we assign the proposition any probability other than 1 or 0, then it is true in some possible worlds and not others. But then the proposition is just plain false, so we should assign it a probability of zero. So we can only assign that kind of statement a probability of 0 or 1.

In other words, I think this kind of reflexive, “necessity”-like statement is a problem for a certain interpretation of personal probability/possible worlds — or at least, that the traditional concept of “necessity” is latent in that interpretation. Maybe a LW discussion post is in order, but I don’t know if I have the karma!

  (Quote)

Kevin November 9, 2011 at 3:10 am

If we assign the proposition any probability other than 1 or 0, then it is true in some possible worlds and not others.

This is a mistake. The probability that we are assigning is the confidence level that we have with regards to the proposition. Lets take mathematical equations as an example of necessary truths. Suppose you’re taking a math test and you reach a complicated problem. After completing the problem, I ask you how confident are you with your answer. You’re answer will likely not be either 0 or 1, since in either case, that will mean that you will not even be right by chance or that you will always be correct. You might be very confident, say 95%, or not very confident, say 5%, but it would be unwise to say 0 or 1. This confidence level concerns your ability to reach the correct necessary truth, it does not concern the truth itself.

If we find that it doesn’t exist in one possible world, then that falsifies the proposition

I think I need to explain here. Our tools of observation are not error free, so we can only say with a certain confidence level that we have not found something. This doesn’t reduce the hypothesis to zero, but it reduces the probability to a level where we would say that we can be confident to a very high degree that it is false. For example, one of the teams testing Einstein’s theory of relativity returned a negative result (they did not find relativistic effects). If this was the only team that was sent, this would have reduced our confidence in the theory, but it would not have reduced the probability of the theory to zero. Fortunately, there were two teams and the other one returned a positive result, so they had two conflicting answers so they decided to follow up with another test with more accurate instrumentation. This applies to practically any observation. It just so happens that our eyesight is very reliable, as well as our memory and ability to accurately interpret events, but they are not 100% accurate so we can’t say with certainty that we actually haven’t found X, which leaves the proposition just a slight chance of being true.

Another way of saying it would be to say that the proposition has a zero percent chance of being true, but then put little error bars on it (corresponding to the accuracy of the observation), and each additional negative result decreases the error bars so that they asymptotically approach zero. This means that you’re actually saying the probability of X being necessarily true is between zero and a very, very, very small number. Again, not absolutely certain, but enough to discredit the hypothesis beyond all reasonable doubt.

  (Quote)

Patrick November 9, 2011 at 6:22 am

The general issue is “epistemic” versus “ontological” probability. I think those are the terms… but don’t quote me on that part. You can quote me on the next part though.

Basically, epistemic probability refers to how certain we are of something. Ontological probability refers to a more objective idea of how likely something is to be a particular way rather than another way.

For example, take this equation: 4687543 * 128795 = 603732101685

If that is true, then its “necessarily” true. There are no possible worlds in which that equation is false. And vice versa- if this equation is false, then its necessarily false. No possible world exists where it is otherwise. That’s how you would describe it in modal logic.

But if you haven’t verified the math, odds are that you aren’t certain that its true. You could make an estimate of how likely you think it is to be true, though. Essentially, you’re saying that its epistemically possible that the equation is necessarily true in the modal sense.

That’s the number one bit of confusion, and often the number one swindle, in arguments that use modal logic. Plantinga’s ontological argument trades on this, for example.

  (Quote)

PDH November 9, 2011 at 11:09 am

Leon,

I don’t really have anything to add to what Kevin and Patrick have said, other than to just reiterate the point.

Outside of fuzzy logic it may be that a proposition can’t be 90% true but a person can be 90% sure that a proposition is true.

We can’t be certain about whether or not something is necessary, possible or impossible in modal terms and in many of these cases – such as Plantinga’s ontological argument – that is precisely what we’re debating. But it seems to me that we can still assign probabilities to various outcomes, when using the relevant interpretation of probability.

So I wouldn’t say that the probability mass is divided over the possible worlds (some of which might be impossible, for all we know).

  (Quote)

poban November 10, 2011 at 12:25 pm

Luke, That dragon story is awesome. Even though its a wishful thinking killing that “dragon” would be awesome but what if one is killed by snakes and tigers (lets call it natural disasters) and what if new predators originate?? Many questions arise from that death conquering thing. I think you could come up with better questions without thinking than me with contemplating for hours.

  (Quote)

antiplastic November 10, 2011 at 9:03 pm

Anti,

Do you literally mean that the set of Christians who believe in divine simplicity is en empty set? That seems nonsensical. I wonder how you could even defend that notion.

It is literally incoherent, and hence literally impossible to believe. This is an example of a doctrine which people mistakenly say they believe, when in fact they merely *believe* that they believe it. Belief and mere affirmation are not identical!

Don’t take my word for it, Plantinga argues at length for its incoherence, and if even Alvin thinks something is nonsense that’s rather probative.

Any person who says that that Yahweh in any way has any properties (is a mind, has beliefs, performs miracles, has more than one personality trait) by definition does not believe that he is “simple”. Therefore, the set of Christians who believe it is literally zero.

So rather than answer my objection that “Who designed the designer” ignores the premises and conclusion of cosmological arguments, you just say I have been confused. Others might take that for a good rebuttal, but I will not.

I anticipated it and answered it twice, and you ignored it.

Here it is, again. For the third time.

A complex designer takes more bits to describe than the complicated universe he is supposed to have designed. Therefore, a designer is more improbable (by definition). Therefore, a designer is no “explanation” at all.

So understanding this, in what way could “What designed the designer” be a possibly good rebuttal to the 3 types of cosmological arguments that conclude with a being that could not possibly be created since it is necessary?

As a wise man once said, a complex designer takes more bits to describe than the complicated universe he is supposed to have designed. Therefore, a designer is more improbable (by definition). Therefore, a designer is no “explanation” at all.

I don’t really have much to add that PDH hasn’t already said. I don’t feel you’ve really internalized the bit about the cabbage.

  (Quote)

Yair November 10, 2011 at 10:05 pm

I don’t understand the interest in that neuro-ethics paper.

It’s supposed to imply that moral judgment is done by several distinct systems in the brain. Yet the authors themselves note that they could find no relation between the moral judgment and neural activity (“Also, judgment confidence (high vs. low) was not a significant predictor of brain activity within any system, with the exception of Disgust”). It appears that the experimental set-up is incapable of recording the moral judgment – it only records the emotions associated with it. It doesn’t even indicate whether these cause the moral judgment, or accompany it, or are caused by it!

The so-called “moral” scenarios were carefully designed to elicit only one emotion (Disgusting, Harmful, or Dishonest); presumably, they also selected for the “(morally) wrong” answer. The so-called “neutral” scenarios were carefully designed to avoid any of these emotions (and be “not wrong”). So the real difference between them is that “moral” scenarios involve one emotion, and “neutral” ones involve none. From the get-go the “experiment” is geared up to reveal that each moral scenario is tied to a specific emotion – and lo and behold, it finds just that!

All that the article finds is, basically, that the arousal of these three emotions is associated with different parts of the brain.

It is not a very telling paper, I’m afraid. The conclusions are far ahead of the evidence.

  (Quote)

Leon November 11, 2011 at 6:00 am

PDH, Kevin, Patrick –

Thanks for your clear posts. Again, just a reminder of my original contention: an epistemology based purely on Bayesian inference does not work well with the idea of necessity (or divine simplicity). Because of this, you can’t respond to a cosmological argument that uses either concept by just plugging the God-hypothesis into MML or Solomonoff induction: you must also be clear that you also reject necessity/divine simplicity.

Yudkowsky, for example, is explicitly suspicious of necessity/contingency/modal logic. Antiplastic rejects divine simplicity here:

It is literally incoherent, and hence literally impossible to believe. [...] Belief and mere affirmation are not identical! [...] Any person who says that that Yahweh in any way has any properties (is a mind, has beliefs, performs miracles, has more than one personality trait) by definition does not believe that he is “simple”. Therefore, the set of Christians who believe it is literally zero.

(A quick correction: based on the SEP article Anti linked to, divine simplicity fans claim that God is not a being who “has” properties, but in some sense “is” His properties. Personally I’m not sure if that makes any sense either, but the argument above still misses its target slightly.)

Back to necessity:

Basically, epistemic probability refers to how certain we are of something. Ontological probability refers to a more objective idea of how likely something is to be a particular way rather than another way.

For example, take this equation: 4687543 * 128795 = 603732101685

If that is true, then its “necessarily” true. There are no possible worlds in which that equation is false. And vice versa- if this equation is false, then its necessarily false. No possible world exists where it is otherwise. That’s how you would describe it in modal logic.

—–

Outside of fuzzy logic it may be that a proposition can’t be 90% true but a person can be 90% sure that a proposition is true.

We can’t be certain about whether or not something is necessary, possible or impossible in modal terms and in many of these cases – such as Plantinga’s ontological argument – that is precisely what we’re debating. But it seems to me that we can still assign probabilities to various outcomes, when using the relevant interpretation of probability.

So I wouldn’t say that the probability mass is divided over the possible worlds (some of which might be impossible, for all we know).

I think the confusion here comes from two senses of “possible worlds”; let’s call them 1 and 2.

In Bayesian inference, one could consider hypothesis spaces to be spaces of possible worlds(1). This is especially true for MML and Solomonoff induction, where we represent each hypothesis/cluster of hypotheses as a string/program, each of which is a ((compressed) representation of a) possible world(1) in some sense.

On the other hand, in modal logic, we use possible worlds(2) to talk about necessity, contingency, etc.

My main argument so far has been that if you subscribe to Bayesian-inference-as-epistemology and conflate senses (1) and (2) of “possible world”, you can’t have degrees of belief in necessary propositions without paradox.

But it seems from the above that you guys are actually keeping (1) and (2) separate. I think this leads to some very weird conclusions — for example, that possible worlds(1) are in some sense sets of all possible worlds(2). And it’s not at all clear what “possible world(2)” actually means in the Bayesian-inference scheme of things. A concrete example:

Lets take mathematical equations as an example of necessary truths. Suppose you’re taking a math test and you reach a complicated problem. After completing the problem, I ask you how confident are you with your answer. You’re answer will likely not be either 0 or 1, since in either case, that will mean that you will not even be right by chance or that you will always be correct. You might be very confident, say 95%, or not very confident, say 5%, but it would be unwise to say 0 or 1. This confidence level concerns your ability to reach the correct necessary truth, it does not concern the truth itself.

Here, if we conflate senses (1) and (2) of possible world, we end up with the paradox I argued for earlier.

So let’s assume we keep them separate, and I am only 5% confident in my answer. Then only 5% of my epistemic probability mass lies on hypotheses/strings/possible worlds(1) in which my answer is true in all possible worlds(2). The other 95% of my epistemic probability mass lies on hypotheses/strings/possible worlds(1) in which my answer is true in only some possible worlds(2).

This seems weird to me, which is why I recommend you either take the Yudkowsky route of just rejecting necessity and contingency altogether, or soften your belief in Bayesian inference as the one “mathematically correct” epistemology.

—–

I also thought of way of restating the argument I tried to make here with respect to this example. To say that God is a necessary being is to say that the truth or falsity of his existence is like the truth or falsity of a mathematical equation. But imagine we have a mathematical theorem, and we want to do Solomonoff induction or MML on it. Which hypothesis has the longer bitstring/lower prior: the one that says the theorem is true, or the one that says the theorem is false?

I’m pretty sure the answer here is “neither” — which is why I think that if the idea of God as a necessary being makes sense, then the argument that His bitstring is too long is moot. Similarly, if the idea of divine simplicity makes sense, then the argument that His bitstring is too long is moot.

The solution: the consistent Bayesian should reject both divine simplicity and necessity/contingency/modal logic as incoherent. Problem “dissolved”, if not solved.

  (Quote)

Kevin November 11, 2011 at 7:56 am

So let’s assume we keep them separate, and I am only 5% confident in my answer. Then only 5% of my epistemic probability mass lies on hypotheses/strings/possible worlds(1) in which my answer is true in all possible worlds(2). The other 95% of my epistemic probability mass lies on hypotheses/strings/possible worlds(1) in which my answer is true in only some possible worlds(2).

This seems weird to me

I don’t think this is stated correctly, I think it should be (changes in bold):

“Then only 5% of my epistemic probability mass lies on hypotheses/strings/possible worlds(1) in which my answer is true in all possible worlds(2). The other 95% of my epistemic probability mass lies on hypotheses/strings/possible worlds(1) in which my answer is false in all possible worlds(2).”

So, if we were to complete 20 such problems, each with equal confidence in our answer, we should expect, on average to have one answer line up with the necessary truth and the rest to not. This means that, on average, one answer will correspond with the necessarily true conclusion and the rest will correspond with the necessarily false conclusion. However, this doesn’t mean that answers we think are false are necessarily false. The same holds for the true answers. We are not applying the attribute of necessity to our epistemology, it is applied to the ontology of the claim being considered.

  (Quote)

Ryan M November 11, 2011 at 11:38 am

Anti,

If some proposition P is logically or lets say nomologically impossible it does not follow that S cannot believe P. You have not provided good reason to think that people cannot believe impossible things. By your response, I would need to believe that no one in fact believes in God, since God is impossible. But that is absurd. And as a side note, I am still an atheist like I stated a while ago. So you don’t need to try to show that divine simplicity is nonsense, I already agree.

“A complex designer takes more bits to describe than the complicated universe he is supposed to have designed. Therefore, a designer is more improbable (by definition). Therefore, a designer is no “explanation” at all.” To refute some deductive cosmological argument P you need to show that either the argument contains false premises, or the argument is invalid. How does your response show that Thomas’ Aquinas’ arguments are either invalid or have false premises? By definition your quote would need to if it is used to dispute deductive cosmological arguments like Aquinas’.

I’d like to get a little perspective here:

1. Does anyone else think it is absurd to think that the set of people who believe impossible things is an empty set? It seems by Anti’s thinking that there could not exist a person who believes in concepts that are incoherent.

2. Does anyone else agree that the complexity argument does not refute cosmological arguments, rather at best it shows that God cannot be the designer? So, some cosmological argument P could be sound, but God could not be said to be the necessary being the argument concludes with.

  (Quote)

Patrick November 11, 2011 at 12:49 pm

Leon- Bayes is a mathematical theorem about the relationships between probabilities. It doesn’t have to be wed to a particular means of calculating probability. The key isn’t to come up with some weird definition of “possible worlds” or “bayesian epistemology,” its to get the basic math right and not confuse subjective epistemic probability with objective ontological probability. If you keep that line clean, there’s no paradoxes around at all.

  (Quote)

antiplastic November 13, 2011 at 11:34 am

Anti,

If some proposition P is logically or lets say nomologically impossible it does not follow that S cannot believe P.

Let’s not say that, since it is a ridiculous caricature of what I actually said. (Hint: I never mentioned believing nomologically impossible things. Anyone who is mistaken about whether a certain kind of food is good for their diet holds a “nomologically impossible belief”.)

I said it is impossible to believe logically incoherent things. Not that it is impossible to affirm them, or to behave in ways inconsistent with prior affirmations.

You have not provided good reason to think that people cannot believe impossible things.

To believe something is to have certain coherent behavioral tendencies to draw inferences from it. By definition, while one can behave inconsistently at different times, one cannot behave inconsistently at a single time. Therefore, it is impossible to believe logically impossible things, although it is not impossible to believe that one believes impossible things.

By your response, I would need to believe that no one in fact believes in God, since God is impossible.

No one in fact believes in incoherent descriptions of God, of which there are many. Those who affirm they do are engaged in theological noncognitivism. But there are plenty of perfectly serviceable, non-incoherent concepts of god, and this planet is eyeball-deep in people who believe it.

“A complex designer takes more bits to describe than the complicated universe he is supposed to have designed. Therefore, a designer is more improbable (by definition). Therefore, a designer is no “explanation” at all.” To refute some deductive cosmological argument P you need to show that either the argument contains false premises, or the argument is invalid. How does your response show that Thomas’ Aquinas’ arguments are either invalid or have false premises? By definition your quote would need to if it is used to dispute deductive cosmological arguments like Aquinas’.

What people are not adding is that “who designed the designer” is also not a good response to the argument from fulfilled prophecy, or “why would they die for a lie?”

Who exactly is saying this is supposed to be a silver-bullet retort to every bit of priestcraft one might come across? Not the author of the linked article.

It is in fact not necessary to show which premise in a valid deductive argument is false in order to show that its conclusion is false. It is nice when one can do so, but it is not logically or conversationally required. The notion that this is so comes from a kind of cargo cult mentality towards philosophy which those of a theistic mindset love to promulgate, in which logic-chopping and argument maps are reliable producers of truth. If I show that your conclusion is false, one can infer that there is a false premise, in exactly the same way as a counterexample to a mathematical proof means that one of the steps in the proof is flawed, but at that point one is no longer in a fight, one is in an autopsy.

Think of it this way. If John says the number of planets MUST be seven, because deductive logic shows it reflects the number of Perfect Platonic Solids or something, and Jack says the number of planets MUST be ten, because he has a deductive argument that shows it reflects the ten spheres of the Qabalah, and I look through my telescope and count exactly eight planets, then the conclusion you should draw is that there are eight planets. It is not my burden to “show which premise in their sophisticated philosophical arguments is false”.

It seems by Anti’s thinking that there could not exist a person who believes in concepts that are incoherent.

Please come to the floor to collect your prize!

Does anyone else agree that the complexity argument does not refute cosmological arguments, rather at best it shows that God cannot be the designer?

It is clear that you and I operate under very different definitions of the concept “shows God cannot be the designer” as it relates to the concept “refutes cosmological arguments”.

  (Quote)

Ryan M November 13, 2011 at 2:28 pm

Anti,

I was not caricaturing what you said. Rather, what I said clearly is just including nomologically impossibility rather than talking about it exclusively.

Also, it has been widely believed for a long time that people believe in logically impossible affairs, not to mention physically impossible affairs. i.e. The logical Problem of Evil has historically been used to show that when S believes in the set T which contains ‘P, Q, R’, then S believes in a logically inconsistent set, which would mean S believes in a logically impossible state of affairs. Another example could be that some people believe both that God is perfectly free and that God is not morally free (i.e. Swinburne believes this), but arguments attempt to show that Swinburne’s belief is logically impossible.

I take your planet example to work. If we have good reason to think that God is not the sort of thing that could be necessary, or that modal logic is inherently flawed and misleading, then we need not show cosmological arguments to be invalid or containing false premises. But the “Who designed the designer” does not do this until it has been established that God must also be caused to exist. Prior to establishing that God is not necessary the WDTD response must fail.

  (Quote)

Patrick November 13, 2011 at 3:56 pm

Leaving aside all other issues, I’m not convinced that its actually impossible for someone to believe something that is logically incoherent. Its tough for a single proposition to be logically incoherent- in order not to cohere, you need more than one proposition, so that one proposition can be logically incompatible with another.

So all you’d need in order for someone to believe something logically incoherent is for them to believe X, and for them to believe Y, and for them never to put two and two together and realize that X and Y are logically incompatible.

Essentially, X is a proposition, and Y is a proposition, and the fact that X and Y are incompatible is a third proposition. One could believe X and Y but not be aware of the third.

  (Quote)

Patrick November 13, 2011 at 4:04 pm

Ryan M wrote, “Prior to establishing that God is not necessary the WDTD response must fail.”

This is only partially correct. It depends what the WDTD response is being used to do. If all you intend to do is demonstrate that a particular formulation of the cosmological argument is trash, then the WDTD response can work wonders without the issue of disreputable crap like “necessary beings” ever being mentioned.

  (Quote)

cl November 13, 2011 at 6:23 pm

Ryan M,

You two are writing far more than you need to.

What my post was trying to do was show that it makes no sense to respond to lets say Thomas Aquinas’ Five Ways by saying “What caused the designer/What caused God”. It makes no sense because the theist’s arguments conclude with a necessary being of some sort (i.e. First mover, the only unchanging being [Pure act]).

Actually, it’s you who’s writing more than you need. Years of experience have taught me that those you engage are simply too dense and/or flippant (as in Reginald’s case) to get it. Hence, they think “who designed the designer” is a good argument.

antiplastic gets a pass because he/she is usually spot on but for whatever reason are royally blowing it here. For example,

As a wise man once said, a complex designer takes more bits to describe than the complicated universe he is supposed to have designed. Therefore, a designer is more improbable (by definition). Therefore, a designer is no “explanation” at all.

Pure chutzpah. Naked assertion sustained by intuition masquerading as cogency. Sorry antiplastic, you’re spot on most of the time, but here you’re actually swallowing the snake oil you usually reject. If you wish to disagree, give me one reason why I should agree. Make a cogent argument instead of just lambasting Ryan.

  (Quote)

antiplastic November 13, 2011 at 8:15 pm

I’m being asked for an “argument” why a complicated thing is more complicated than no complicated things? April 1 is a long way off…

  (Quote)

cl November 13, 2011 at 8:47 pm

I’m being asked for an “argument” why a complicated thing is more complicated than no complicated things?

No, you’re not. But I understand if you’d rather play rhetoric.

  (Quote)

antiplastic November 13, 2011 at 9:10 pm

No, you’re not. But I understand if you’d rather play rhetoric.

What, you want me to wipe your ass for you, too? Don’t be so lazy. Use that thing called the internet and see what you can find.

As for me, I already know it’s true and don’t care to convince you or earn your respect.

  (Quote)

Leon November 17, 2011 at 9:40 pm

I don’t think this is stated correctly, I think it should be (changes in bold):

Thanks, those changes look good.

So, if we were to complete 20 such problems, each with equal confidence in our answer, we should expect, on average to have one answer line up with the necessary truth and the rest to not.This means that, on average, one answer will correspond with the necessarily true conclusion and the rest will correspond with the necessarily false conclusion.However, this doesn’t mean that answers we think are false are necessarily false.The same holds for the true answers.We are not applying the attribute of necessity to our epistemology, it is applied to the ontology of the claim being considered.

Honestly, I don’t quite understand this. Could you try explaining it another way? (Feel free not to if you think the thread is dead.) As I understand it, if something is necessarily untrue, it isn’t possible to form a coherent representation of what it would be like if it were true. So claims about necessity necessarily impact on our epistemology.

Leon- Bayes is a mathematical theorem about the relationships between probabilities.It doesn’t have to be wed to a particular means of calculating probability.The key isn’t to come up with some weird definition of “possible worlds” or “bayesian epistemology,” its to get the basic math right and not confuse subjective epistemic probability with objective ontological probability.If you keep that line clean, there’s no paradoxes around at all.

I agree that Bayes is a mathematical theory about the relationships between probabilities, but it certainly does provide a means to calculate probabilities. Also, when we’re talking about “God’s bitstring”, there are a large number of assumptions at play about Solomonoff induction/MML, what the bitstring means exactly, what “probability” refers to, etc. All of this means I think it’s perfectly fine to use the term “Bayesian epistemology”. I don’t think the subjective-epistemic/objective-ontological distinction is obvious or clear, nor is it clear how either concept works with “necessity”.

Again though, I don’t have a problem if you just reject necessity as BS. All I’m trying to say is that it doesn’t sit easily or obviously with a certain interpretation of Bayesian epistemology.

  (Quote)

Leon November 17, 2011 at 9:41 pm

I don’t think this is stated correctly, I think it should be (changes in bold):

Thanks, those changes look good.

So, if we were to complete 20 such problems, each with equal confidence in our answer, we should expect, on average to have one answer line up with the necessary truth and the rest to not.This means that, on average, one answer will correspond with the necessarily true conclusion and the rest will correspond with the necessarily false conclusion.However, this doesn’t mean that answers we think are false are necessarily false.The same holds for the true answers.We are not applying the attribute of necessity to our epistemology, it is applied to the ontology of the claim being considered.

Honestly, I don’t quite understand this. Could you try explaining it another way? (Feel free not to if you think the thread is dead.) As I understand it, if something is necessarily untrue, it isn’t possible to form a coherent representation of what it would be like if it were true. So claims about necessity necessarily impact on our epistemology.

Leon- Bayes is a mathematical theorem about the relationships between probabilities.It doesn’t have to be wed to a particular means of calculating probability.The key isn’t to come up with some weird definition of “possible worlds” or “bayesian epistemology,” its to get the basic math right and not confuse subjective epistemic probability with objective ontological probability.If you keep that line clean, there’s no paradoxes around at all.

I agree that Bayes is a mathematical theory about the relationships between probabilities, but it certainly does provide a means to calculate probabilities. Also, when we’re talking about “God’s bitstring”, there are a large number of assumptions at play about Solomonoff induction/MML, what the bitstring means exactly, what “probability” refers to, etc. All of this means I think it’s perfectly fine to use the term “Bayesian epistemology”. I don’t think the subjective-epistemic/objective-ontological distinction is obvious or clear, nor is it clear how either concept works with “necessity”.

Again though, I don’t have a problem if you just reject necessity as BS. All I’m trying to say is that it doesn’t sit easily or obviously with a certain interpretation of Bayesian epistemology.

  (Quote)

Leave a Comment