Yes, in one form or another, I think that's the great wall that separates sinful humans from God.

The abductive reasoning part occurs before the relevant observations are made and tested and definitely before the new phenomena are predicted. Jibninjas put forth "abductive reasoning" as a potential missing link in:

Aim: To test my theism
Method: ?
Results: God exists

Abductive reasoning can be used (even has to be used) to decide what experiments to do. It doesnt lend any support to the claim on its own.

It so happens, I can abductively reason that all number plates end with an odd number (from looking at the seven in the carpark I can see - the chance of them randomly all being odd is pretty slender). Not much of a compelling hypothesis if I stop there though, right?

But when creationists "interpret" this, they do not end up in a place that logically follows from the evidence. Let's say it goes something like this:

Evolutionary biologist:

"The discovery that chromosome 2 is actually two chromosomes fused together tells us that not only is our DNA 98% similar to apes, but that we , at one, time, had the same number of chromosomes as well. This information, along with various other pieces, leads us to believe that Apes and humans evolved from a common ancestor.

Creationist:

(assuming they actually accept that chromosome two is two chromosome fused together)

The discovery that chromosome 2 is actually two chromosomes fused together tells us that not only is our DNA 98% similar to apes, but that we , at one, time, had the same number of chromosomes as well. This information, along with various other pieces, leads us to believe that God made all of his creatures with certain similarities.

Now my arguments may not exactly be correct to the letter but I think I have it basically right.

The evolutionary biologist can say that these are example of common descent because that follows logically from the evidence. In the case of the creationist, they are simply positing an unknown entity to explain the evidence... How is this unknown entity a better explanation of facts than the one the scientist posits?

As NR said, abductive reasoning doesn't seem to be reasoning in itself, as reasoning is required in order to justify what you have gleaned from abduction. In this case, do you disagree that the evolutionary biologists explanation is superior to the creationists in terms of what is more likely?

I have a different perspective, but it would get boring if we agreed about everything, wouldnt it?

Forgive me if this is the incorrect link. I am at work so I can't preview it but I am pretty sure it is correct.

Plantinga speaks about belief in God as a basic belief/axiom.

http://www.youtube.com/watch?v=oL5rykiekBs

edit: after hearing so much about Plantinga I was somewhat underwhelmed by his arguments. He still makes some good points though.

This is at odds with my understanding of abduction and inference to best explanation:

At around about the time I became an atheist again, I was playing around with fleshing out an argument for theism based on this. I generally think it's hard to dispute a theist's right to declare the existence of God self-evident (or for them to say they have performed some "abductive reasoning" and found the God hypothesis to be the best explanation for the data we have so far observed). That doesnt mean those assumptions are correct, of course.

I had wondered though if an argument could be made that, since as a general rule the addition of any proposition to a set of axioms will usually result in an inconsistent set of axioms* - it therefore followed that, if the addition of an axiom results in no observed inconsistency, we have evidence that the new axiom is indeed true (or at least can be taken to be so). I hadnt convinced myself, but it seemed to me there might be an argument to be made.

* I dont know if this is, in fact true (or even meaningful) but it seemed reasonable to me (irony noted).

I suspect we're talking past each other since I agree with that list of principles we should use when formulating hypotheses (did my number plate example mean anything to you?)

We do reference observation and current theories when utilising abductive reasoning to come up with new hypotheses. These arent the relevant observations though, when it comes time to test the theory.

As I hoped to illustrate with the number plates - having formed a 'best explanation' of available data, we then construct experiments to test our theory. There's no point looking at the data we used to come up with the theory because (unless we're particularly idiotic) it will always do a great job explaining those data.

Having formed my odd-number theory I now test it but looking at some different cars (and find it lacking). Or try and make predictions (about what number plate will come off the assembly line after XYZ 921, for example).

A theist is behaving sensibly to look at the eye and declare that it is 'most plausibly' the product of design. If he wishes to claim empirical support for his theory, he cant then go looking at the eye as evidence for design. He needs to then proceed on the basis of "If my design theory is true, what else would I expect to see or not to see?"

I'm not sure it's talking past each other as that I think you are knocking down a straw man. You are claiming that those explanatory virtues are not part of a (good) abductive argument but only come after it. And it doesn't help that, while I have provided links to over half a dozen papers which agree with my take on what inference to best explanation involves, you are so far only invoking Jibninjas in order to demonstrate your take on it.

For another source that confirms my view that the explanatory virtues are part of abductive reasoning see the SEP entry:

So, it's not that there are no problems with abductive reasoning, but your characterizing it as nothing more than just coming up with any old hypothesis that fits the data and declaring "makes sense to me!" completely ignores the essential aspect of comparing multiple hypotheses and (trying to) objectively choose between them using the explanatory virtues.

No, I'm not.
Well, I would claim that those explanatory virtues we choose are those which "makes sense to us". Nonetheless, to speak directly to the point - I'm not ignoring that there are principles to help guide us to 'good' hypotheses, I acknowledge that and think it's an important part of the abductive process.

I obviously havent been clear in my criticism (or perhaps overstated it - early on I did concede that my knowledge of this was rusty and that I have a very low opinion of it, meaning I'm not very interested in dusting off the textbooks to dismiss it all over again).

To be clear - it's not that I think those things are absent in our formulation of hypothesis. It's that, once we have utilised those to select the hypothesis we consider to be 'the best explanation' we are no more closer to have establishing its truth/utility/persuasiveness - pick your chosen criteria. The first step doesnt add any evidence - it tells us where to look for it. (Barring the success of an argument along the ones I sketched a few posts up). Having come up with whichever hypothesis we consider to be the most promising, we then have to perform further tests if we expect our hypothesis to be considered supported by evidence. (Which is where this tangent started).

This is not my characterisation. "Makes sense to me" means something.

I had wondered though if an argument could be made that, since as a general rule the addition of any proposition to a set of axioms will usually result in an inconsistent set of axioms* - it therefore followed that, if the addition of an axiom results in no observed inconsistency, we have evidence that the new axiom is indeed true (or at least can be taken to be so). I hadnt convinced myself, but it seemed to me there might be an argument to be made.

* I dont know if this is, in fact true (or even meaningful) but it seemed reasonable to me (irony noted).

Bunny,

I don't follow what you are saying in the bolded above.

What you said in post 109 makes sense re the human eye and reasoning.

Maybe. I'm happy to be shown the errors of my ways as I am something of a philosophy noob, but whoever does so will kinda need to be dusting off textbooks to do so, or at least provide some clickable links.


Did you edit this bit? Regardless, I'm afraid I disagree (no really, I'm afraid, as I always am when disagreeing with someone smarter than me ) as I consider the search for the best explanation to be 'what science is'.

I think this is very good example of what I am talking about, so let's look at this closer.

First, I think you need to replace the word "logically" with something else. There is nothing in the evidence that says that evolutionist conclusion follows logically, as in it is logically necessary. What I believe you mean is it follows reasonably or may rationally. Both conclusions could be true as neither follows necessarily. I just don't want to step on each other with using different definitions of words.

So in your example there is nothing rigorous that forces us to choose the evolutionists conclusion over the creations. What does urge us to choose the evolutionists conclusion over the creationists is more so that it is a more reasonable conclusion (amongst other things as this is complex subject).

I agree that the explanation from the evolutionary biologists is superior, but not for any reason that escapes the accusations of bunny/NR about abductive reasoning. This is just as much of an example of "seems to make more sense" as being thrown at the existence of God. In other words, there is nothing here that applies that could not also apply to my original statement about using abductive reasoning about the existence of God.

So a question for you, what form of rigorous logic does the above example use other than "it seems to make more sense"?

One fundamental point is that the "theory of a common ancestor" was proposed before DNA was discovered. Once DNA came on the scene the hypothesis of a common ancestory made predictions about that previously unknown data.

firstly, you're right, "logically" wasn't quite what I meant to say there, lets go with "reasonably."

To answer your question: none.

What separates the two is that one has empirical evidence to support a conclusion, while the other doesn't.

It's not simply that "it seems to make more sense" but that "it seems to make more sense based on the evidence."

Granted I am knew to "abduction," but t seems to me to be a very different form of reasoning (leaving aside the previous objections by bunny and NR that it isnt actually reasoning.) In a deductive argument, if the premises follow to the conclusion, then that's the end of it.

Here, it seems that more work needs to be done in order to justify our conclusion...

This is going to sound like a copout, but it genuinely isnt...

I'm just about to go on holiday for a week. My objections to abductive reasoning can be made more carefully and with more referencing, but not by me in the near future I'm afraid.

Happy to argue about it on the 17th, if you're still interested.
Sure - but we do that by forming hypotheses (using abductive reasoning, applying those virtues you'd previously cited as objectively as possible) then we go and test them. My crucial critique (of abductive reasoning as empirical test in itself, which was the original claim) is that stopping at the 'best explanation' step does not actually constitute evidence.

Again I refer you to my number plate example. I gaze at half a dozen cars all with odd number plates. It's quite reasonable (assuming no prior knowledge) to form the hypothesis that all number plates end in an odd number. However, at this stage I havent actually done anything other than describe the world (as I've seen it) succinctly.

To test this hypothesis, I need to go look at some new cars - not the data I used to come up with the hypothesis. This step is the empiricism bit. If a Godless universe seems reasonable to someone, that's all fine and dandy - but that opinion doesnt constitute an empirically tested claim (no matter which virtues they used to come up with it) until they go out and actually do some experiments.

As I say - I feel a bit wimpish, but real life beckons. I can read and post, but I cant actually go and look stuff up - sorry. (I bet the SEP has some decent critique of abductive reasoning though).

I might just be misinterpreting you, but it seems to me that you are misrepresenting the proposed role of abduction in theory choice (at least, you are certainly de-emphasizing its importance in a way I disagree with). It doesn't add new evidence (but then, neither does deduction). Nor is it only relevant for coming up with hypotheses. What it does is make more explicit the criteria we use to decide between theories. Say we have two competing hypotheses: Intelligent Design versus the modern evolutionary synthesis. How do we decide between them?

We don't use deduction: both theories are internally consistent. We can't really use induction--famously, Intelligent Design has almost no empirical predictions. So what is our basis for preferring the modern evolutionary synthesis?

I would say it is abduction. That is, we have developed certain criteria that we think are characteristic of good explanations--things such as simplicity, scope, testability, and the other things zumby mentioned. When we compare these two theories to see which of them exhibit more of the virtues of a good explanation, the evolutionary theory is a clear winner. And thus, we accept it as the better theory.

Now, this isn't a proof (in the deductive sense) that Evolution is true and Intelligent Design is false. It could certainly be the case that what seems to us to be the better explanation is in fact false (an example of this would be the geocentric theory in ancient times). But it does give us a good reason to accept one theory over another.

Now, it is true that abduction can also act as a spur on to more empirical work. We might think that a hypothesis is simple and elegant, but has little empirical support. If it were competing with another hypothesis that is less simple and elegant, but with greater empirical support, we have a dilemma. One way to resolve the dilemma is to seek for more empirical support for the more elegant hypothesis.

Incidentally, I think the SEP article on the underdetermination of theory by evidence I linked to earlier is relevant to this discussion. Part of the reason why we need to use abductive reasoning in theory choice is because (in my view), it is impossible to falsify a scientific theory by experimentation alone.

This is where we disagree. The universe is full of empirical evidence to support the existence of God in the same way that the biological world is full of empirical evidence to support the theory of common decent.

And this is exactly what I would say about the existence of God.

agreed

Ok, I mean you've sort of shifted bit here, I thought we were still speaking in the context of creationism vs evolution, but that's fine.


In either case you aren't really using abduction to reason, as far as I can tell, you are simply using the available evidence, which as I stated before, is ambiguous enough that it can be championed by either side.

Oh, sorry I misunderstood that.


Again I disagree. I am using abductive reasoning for the existence of God in the same manner that it is being used for the theory of common decent.

In the example of common descent vs creationism, the empirical evidence cannot be used to deduce that creationism is the best explanation. ( In fact, I would say that creationists are aware of this, which is why they often misrepresent or ignore large chucks of it)

When we talk about the empirical evidence for god, what about can be used to support a universe with a creator which is a slam dunk for it being the best explanation, in your opinion?

Edit: I would argue that we aren't really using abduction for the theory of common descent either, we are just following the evidence in the place where it leads.

See my post above re vacations - taking logic texts on vacation with me is a great way to annoy my wife . I'll have it out with you when I get back, but fwiw (admittedly speaking from 15 year old recollection) I think it's obvious, not-terribly-important stuff dressed up in complicated language to make it sound cool.

I dont have any confidence the bolded is as objective as your post implies - "we consider theories with these properties to be superior" is what I meant by "seems reasonable to me". A probably true statement with very little import - I think the spur to empirical or rational testing is where the true value lies.

I was mindful that there may have been a connection to that previously linked article, cheers.

What do you mean "cannot be used to deduce that creationism..."? This seems muddled to me.

It is not any one individual piece of empirical data it is when you look at all of the data as a whole. Just like there is no slam dunk piece of data for common decent. If you want to look at one piece of data I would point to the laws of physics.

What does it mean to "follow the evidence in the place where it leads"? How does one get there?