Monday, 11 June 2012

The Agnostic Argument - 1 (An Assertion-based Framework to Assess Theistic Claims)

[An abridged version of this post appears as a guest article on my friend Gill Eapen's blog.]

I consider myself an agnostic and not an atheist. I confess this has always embarrassed me a little, because I cannot boast of any of the certainty that theists and atheists display in their respective positions. I do try to be a rationalist though, and by this measure, I must say I have found both theists and atheists wanting. Lately, I have stumbled upon a conceptual framework that allows me to be more structured in my approach to the theist-atheist debate, and now I'm no longer embarrassed about my lack of certainty. I now know what we can be certain about and what we cannot. The lack of rationalism in the theistic position is well-known (it emphasises faith over reason, after all), but this framework for the first time allows us to see exactly where the atheist argument fails.

Let's say your five-year old tells you one night that he just saw an eight foot tall bogeyman moving about in the garden. What would you do?

You would probably break your child's report down into three parts, and deal with each of them differently:

Statement 1: There's someone in the garden.
Statement 2: That someone is the bogeyman.
Statement 3: The bogeyman is eight feet tall.

You can probably safely dismiss Statement 3 as the product of a child's imagination. No one is eight feet tall. Well, perhaps if they're wearing stilts, but on the balance of probabilities, you can probably rule that out.

You probably don't have to give any credence to Statement 2 because the term "bogeyman" is just a name that holds a special fearful meaning to a child. Any other name could be used in its place and it would add no value to the observation.

That leaves Statement 1, which may have to be taken seriously. The child may in fact have seen someone moving about in the garden, and this is something that calls for investigation.

What this tells us is that we may often hear statements that contain more than one assertion, and we need to tease apart those various assertions in order to deal with the overall statement sensibly. It would be irresponsible, in our example, to refuse to investigate whether there is someone in the garden merely because there is no such thing as the bogeyman or because no one can be eight feet tall.

Religions routinely make such compound claims. As a rationalist, I would not take them as literally true or as the word of God. Religious texts were authored by human beings. They make various claims, and these claims are artificially fused into what the religion in question would consider "core beliefs". An adherent must believe in all the claims to be considered a believer. Selective unbelief is generally not tolerated (although in today's world, most educated theists do exhibit only selective belief in the core scriptures of their religion).

My experience as an IT professional working in the area of Identity Management has exposed me to structured ways of analysing compound claims. What people hold to be truths can be considered "assertions", the validity of which needs to be proven. The technology standard called Security Assertion Markup Language (SAML) talks about three types of assertions that can be made about a subject: Identity Assertions (i.e., that a subject exists and can be uniquely identified as such-and-such), Attribute Assertions (i.e., that a subject has certain properties) and Entitlement Assertions (i.e., that a subject may be permitted to do certain things).

I believe the theist-atheist debate can also be meaningfully structured using this assertion-based framework. The various statements made about God need to be deconstructed into a set of "assertions" before we can start to prove or disprove any of them. I believe this is sorely needed because I have seen atheists make statements to the effect that since Evolution adequately explains how life evolved on earth, it proves that God does not exist. To my mind, this is analogous to responding to your child saying that since no one can be eight feet tall, there cannot be anyone in the garden. To be truly rational, we need to understand the scope of what we are proving or disproving and not to conflate a limited proof to one with a wider scope.

Here's my attempt at this:

Existence Assertions (EA):
  • God exists
Attribute (or Capability) Assertions (AA):
  • God is omniscient (all-knowing)
  • God is omnipotent (all-powerful)
  • God is omnipresent
  • God is all-merciful
Deed (or Event) Assertions (DA):
  • God created the Universe
  • God created all life on Earth
Intent Assertions (IA):
  • God has a plan for all of us
  • God does everything for the good
[There could be additional assertions too, for example an Attribute Assertion that God is male. However, this is not likely to be universally accepted by theists. The Vedantins, for instance, claim that there exists a Supreme Consciousness but make no assertions about the gender of this being. Worshippers of the Mother Goddess see divinity as feminine.]

[Not all theists accept all of the above assertions. I personally know of one person who implicitly rejects the Intent Assertions above because of their paranoid belief that God is malevolent and vindictive, and who prays only to humour him.]

Now, all these assertions are independent (of course, the Existence Assertion is an obvious precondition for all of the others). Importantly, we should recognise that one or more of these assertions may be proven or disproven without impact upon the others. We should also recognise that some of these assertions may be partially proven or disproven (e.g., that God may not be all-knowing, but knows more than all humans put together).

While theism takes all the above assertions as valid without proof, atheism doesn't always do a convincing job of disproving them either. Part of the reason is that atheistic arguments take the compound claims of religion at face value and do not tease apart the various assertions within them. This approach may be good for winning an argument, but not for establishing the truth. As a rationalist, I am more interested in the truth than in proving a prior position.

For example, even a definitive proof that the Universe was created by a Big Bang and not by God would not automatically disprove the existence of God, since the Existence Assertion does not depend upon the Deed Assertion. (At best, it would spell the end of "God as we know it".)

The same goes for the Theory of Evolution. The Theory of Evolution only disproves a single Deed Assertion, not the Existence Assertion or even any other Deed Assertion. [Incidentally, the use of the word "theory" for Evolution does not mean it is an unsubstantiated idea. In science, an idea has to have fairly strong evidence in its favour before it can be elevated to the status of a Theory. If it was just an idea without evidence, it would probably be called the Evolution Hypothesis or Darwin's Conjecture. Evolution is as much a theory as Gravitation, and a pretty strong refutation to the Deed Assertion that God created all life on Earth.]

But of course, even such a convincing refutation of the Deed Assertion does not automatically disprove the Existence Assertion. Many atheistic arguments make the irrational leap from one to the other, and this is where they can be seen to fail. By accepting without question the theists' bundling of assertions instead of challenging such bundling, atheists are in fact abandoning rationality and critical thought.

The assertion-based framework I introduce here allows us to play with ideas of "limited theism" by tweaking the degree to which each of the above assertions is valid. One could, for example, postulate the existence of a networked human consciousness that evolved along with the human species, that knows as much as all humans put together and has some limited ability to influence the actions of people. This hypothesis then accepts the Existence Assertion but only partially accepts the Attribute Assertions of classical theism.  Obviously, this consciousness did not create the Universe nor was it responsible for creation, since it was itself the product of Evolution. Hence, this limited theism rejects the Deed Assertions. In intent, the consciousness could be seen as benign, since by definition, a by-product of evolution probably is aligned to human survival and well-being. The hypothesis therefore accepts the Intent Assertions to a great extent. Such a hypothesis might be useful in exploring why prayer seems to work or in researching certain kinds of parapsychological phenomena without stigma. Without such a framework, a discussion of phenomena like the efficacy of prayer would call for taking sides between absolute atheism (which refuses to entertain any notion beyond the purely material) and outright mumbo-jumbo, where all rational thought is suspended.

We should be able to have a richer and more nuanced discussion about theology and similar subjects relating to the nature of our existence without being trapped within absolute, black-and-white positions. Hopefully, a framework such as this can provide a useful starting point to structure the debate.


Unknown said...

Having same (not just similar) views in almost everything as you, I’ve been waiting for a chance to disagree with you! Now I got one. Will chat with you some time when free. In summary, my point is without defining any attribute (capability and attributes such as volume are too distinct to club for your single AA), the EA is meaningless. Define some attributes and then a probability of 0 to 1 can be assigned to EA. But by not defining any attributes, it is like retroactively changing goal post after a goal and so EA probability is always 1.

In other words, I say that there is some thing called XYZ which will be in some form and has done some thing and can do some thing and will do some thing. With such definition, ofcourse XYZ will exist for ever

prasadgc said...


I take your point. Mere existence without ANY attributes is meaningless. However, unless we eliminate ALL the claimed assertions (AA, DA and IA), we cannot disprove the existence assertion.

(Even Richard Dawkins only claims that God "probably" doesn't exist.)

prasadgc said...

Hi Gladston,

It appears particle physicists agree with me. See my other blog post on identity and the Higgs boson:

i.e., the Standard Model of particle physics says it is possible to dissociate even a seemingly inherent attribute like mass from a particle. In similar fashion, it is conceptually admissible to think of an 'God' entity without any attributes and analyse all assertions independently on their merits. Perhaps it has some properties that are of no relevance or visibility to us, but conceptually, we can think of an entity that exists but has none of the properties that we listed.