Process-Based Non-Cognitivism

by Francois Tremblay

James Lazarus’ excellent article on ‘The Argument from Non-Cognitivism’ discusses in detail what I consider to be the most powerful line of evidence for strong-atheism, the meaninglessness of religious language and specifically the term “god”.

This article is an extension of his argument, examining what he calls Obtained God-Belief (OGB), as opposed to Maintained God-Belief (MGB). MGB is the position that a god’s definition and existence must be accepted a priori, while OGB is the position that the definition and existence of a god is justified by various facts of reality. These are two complementary approaches that have been the subject of hot debate in theological circles for almost a century now. Regardless of this debate, we can examien both approaches and determine whether they are valid.

Contents:

1. What Is An OGB Approach? The Example of Neptune
2. Formal arguments
3. Is There A Proper Observation? Analyzing the Classical Arguments
4. Conclusion

I. What Is An OGB Approach? The Example of Neptune

OGB does not put forward a proposition but rather uses propositions as a springboard: it is the domain of the hypothesis. We can compare this to the process of scientific discovery. Sometimes, such as in the cases of black holes, electrons and some distant planets in our solar system, we discover something many years before we observe it.

This process is of the same type as OGB, finding evidence of the relation between an unknown quantity and things we do observe. The meaning of such an unknown quantity will therefore be proportional to whatever effects we do observe, until the quantity in question can be observed, or its effects prove its existence beyond rational doubt.

The example of Neptune may shed some light in what I mean. In 1845, John Couch Adams, a Cambridge mathematician, applied Newton’s law of gravitation to Uranus’s erratic orbit to discover that another planet was creating this disturbance. He presented his findings to the Astronomer Royal, who dismissed his hypothesis as trivial. It is only after another mathematician came to the same conclusion and got Berlin Observatory to turn their telescopes to the task, that Neptune was actually observed and acknowledged as a planet in July 1846. It had never been recognized as a planet of our solar system before, let alone the cosmic body that led to apparent disturbances in the orbit of Uranus.

Before we continue, it’s important to highlight the fact that a hypothesis is not a “wild guess”, as it may sometimes be used in the vernacular. A hypothesis is a tentative explanation for an observed phenomena. If we have no phenomena to explain, we have no hypothesis. If Newton’s Law of Gravitation already explained Uranus’ orbit, it would not have been considered erratic, in need of explanation, and Adams would not have looked for an explanation. If he had posited a new planet in the solar system on that sole basis, he would have uttered a meaningless proposition. Without the erratic nature of the orbit being an observation that can be explained by this new planet, there is no observation that we can use to say “this planet is X and not Y”. With the observed deviations from Newton’s predicted orbit, we can predict the mass and the orbit of that new planet. Without it, we have nothing.

So there are three factors to consider in our example.

• The quantity introduced by the hypothesis: a new planet.
• The observations supporting this hypothesis: Uranus’ orbit.
• The process used to deduce one from the other: the fact that Uranus’ orbit differs from what we would expect using Newton’s Law of Gravitation, supported by the subsequent calculations.

And then they come together in the notion of specificity. I do not need to reinvent the wheel on this question, since James has already defined the notion of specificity:

By specificity, I mean the proportion by which a meaning identifies the referents of a term. A meaning which specifies the referents of a term and nothing else is completely specific, and is thus sufficiently meaningful. A meaning which does not permit us to identify the referents of a term, however, lacks specificity, and is therefore inadequate—or lacking of sufficient meaning.

In the specific case of OGB-type approaches, the specificity of a hypothesis is proportional to the scope of the process of deduction. The specificity of the concept of this “new planet” is therefore proportional to the results Adams obtained when he made his calculations. In this case, he could know at least the mass and orbit of this new planet, making is a pretty specific hypothesis, sufficient to find its referent a mere year later using the technology available at the time. Granted, he could know little else, but the scope of his deduction was sufficient for the hypothesis in question.

II. Formal arguments

Here is James’s formal argument as it relates to MGB:

Posit that we attempt to define “god” by MGB.

1. There are three attributes of existents which concern us particularly, these being:
   1. Primary Attributes
   2. Secondary Attributes
   3. Relational Attributes
2. B as well as C are dependent upon and must be related to an existent’s A in order to be considered meaningful. The term “God” lacks a positively identified A.
3. Because of this, the term “God” holds no justified A, B, or C. (From 2)
4. However, an attribute-less term (a term lacking A, B, and C) is meaningless.
5. Therefore, the term “God” is meaningless. [From 3 and 4]
6. Therefore, the god-concept is invalid.

I define process-based noncognitivism (my term for the refutation of OGB) as such:

Posit that we attempt to define “god” by OGB.

1. To be considered a valid OGB-type hypothesis, a concept must be a viable explanation for a given observation or set of observations.
2. There is no observation that the god-concept can viably explain.
3. The god-concept cannot be considered a valid OGB-type hypothesis. [from 1 and 2]
4. Therefore, the term “god” is meaningless.
5. Therefore, the god-concept is invalid.

The theologian may attempt to define “god” by MGB, OGB, or both. In either case, the two arguments above demonstrate that his attempts are futile. Depending on the type of defining that the theologian uses, one should use one or the other argument to defend noncognitivism. We can therefore express a general form of these two types of noncognitivism, which we can call a General Argument for Theological Noncognitivism. It would look something like this:

1. The god-concept can be either defined by assertion (MGB) or as a hypothesis (OGB).
2. Posit that we attempt to define “god” by MGB. There are three attributes of existents which concern us particularly, these being:
   1. Primary Attributes
   2. Secondary Attributes
   3. Relational Attributes.
3. B as well as C are dependent upon and must be related to an existent’s A in order to be considered meaningful. The term “God” lacks a positively identified A.
4. Because of this, the term “God” holds no justified A, B, or C. (From 2)
5. However, an attribute-less term (a term lacking A, B, and C) is meaningless.
6. The term “god” cannot be defined my MGB. [from 4 and 5]
7. To be considered a valid OGB-type hypothesis, a concept must be a viable explanation for a given observation or set of observations.
8. There is no observation that the god-concept can viably explain.
9. The god-concept cannot be considered a valid OGB-type hypothesis. [from 7 and 8]
10. Therefore, the term “god” is meaningless. [From 1, 6, and 9]
11. Therefore, the god-concept is invalid.

III. Is There A Proper Observation? Analyzing the Classical Arguments

There are only two premises in the process-based argument, premises 1 and 2. We have already explained premise 1, and only premise 2 remains. Is it indeed the case that “there is no observation that the god-concept can viably explain”?

Well, it is not our burden of proof to demonstrate that there are. Fortunately for us, trying to point out such observations has been the preoccupation of theologians for centuries. There is a whole class of theological arguments that aim to demonstrate that specific facts of nature prove the existence of a god: we call them “classical arguments”. The main categories of such arguments are the Cosmological (First Cause) Arguments, Ontological Arguments, Teleological (Design) Arguments, and different arguments from mind-entities or processes which I divide between Arguments from the Intellect and Arguments from Emotionalism.

Now, it is easy to see that, if these arguments are valid, then they serve as sufficient placeholders for the needed observations. For instance, if a Cosmological argument is valid, then we can say that “the fact that there are caused things” is our observed evidence for the hypothesis of the god-concept. Three conditions must be fulfilled by such arguments:

1. The deductions from the observations to the hypothesis must be logical and valid.
2. The observations must not be already explained, otherwise the god-concept is not a viable explanation for them.

To express this more clearly:

1. If the argument is invalid then we have no valid observations, since there is no way to link them to the god-concept.
2. If the observations are already explained, then there is no need for a hypothesis to explain them at all, and the god-concept is irrelevant.

We have already seen both of these points in the Neptune example. So now we must ask the question: are any classical arguments valid and refer to observations which are not explained? Due to the number of such arguments, this is a rather tedious question. In this article, I will only address the categories of arguments that I have listed above. Other sources refute the individual arguments eloquently, including my very own Handbook of Atheistic Apologetics, and many other books and web sites available in our References section.

Before I begin, I have to point out that virtually all classical arguments lack specificity. That is to say, their logical conclusion is not that there is a god, but rather a supernatural process or impersonal Creator. In some cases, the logical conclusion is even less specific. I will note this lack of specificity in my discussion of each category. One notable exception to this rule is William Craig’s version of the Kalam Argument, which I refute in my article “Dr. Craig’s Unsupported Premise”. Suffice it to say that Craig’s argument is a laudable effort but, like any such endeavour, lacking in logic.

Cosmological Arguments:

1. Observations: The fact that everything we observe is in movement/in a state of change/contingent.
2. Deduction: Everything we observe in the universe is in movement/in a state of change/contingent.- The universe must have a First Mover/First Changer/First Cause – This being is God.
3. Logical problems: Virtually all Cosmological Arguments commit the fallacy of composition and special pleading. They commit the fallacy of composition by transposing the property of being in movement/in a state of change/contingent to the entire universe without justification. They commit special pleading by stating that anything in movement/in a state of change/contingent requires a cause, but excluding God from that deduction.
4. Lack of specificity: The deductions do prove the existence of a supernatural being (if they were true), but do not prove its nature. They do not rule out a supernatural process or an impersonal Creator.

Ontological Arguments:

1. Observations: None.
2. Deduction: God is defined as perfect. – Existence is better than non-existence – God must exist.

Since this category of arguments does not bring to bear any observation, it is pointless to discuss it in the context of noncognitivism. But I may add that the ontological arguments suffer from many flaws, one of them being its total lack of specificity, as we can define anything as “perfect”. Furthermore, the term “perfect” is absolutely meaningless in this context.

Teleological Arguments:

1. Observations: The universe (or parts thereof) is complex/intelligible/an interacting whole.
2. Deduction: The universe (or parts thereof) is complex/intelligible/an interacting whole. – Either this is achieved by chance or by design. – Not chance. – The universe (or parts thereof) was achieved by divine design.
3. Logical problems: All such arguments commit false dichotomy and special pleading. They commit false dichotomy by positing that the only alternative to divine intervention is pure chance. But we know that natural law is the most reasonable explanation for these observations – whether in biology or cosmology – and not chance. They also commit special pleading by refusing to acknowledge that a god would also have to be complex/intelligible/an interacting whole (at least by induction).
4. Lack of specificity: While these arguments are more specific, all that they prove at best is the existence of a supernatural designer, not a god specifically.

Arguments from the Intellect and Arguments from Emotionalism:

1. Observations: The existence of intelligibility, eternal truths, moral obligation, desire for God, conscience, the idea of God, common consent about God, religious experience.
2. Deduction: X exists. – Only God can explain X’s existence. – God exists.
3. Logical problems: The main problem in all these arguments is to prove that natural processes are insufficient to bring about the existence of these X. It would be lengthy to examine all these arguments in turn, but none of them are very good. The Argument from Consciousness, for example, commits the same fallacies than teleological arguments. The Argument from Truth presumes that only an eternal bring could grasp eternal truths, when there is no such correlation. And the Moral Argument presumes that no atheistic position can derive moral obligation, when this is patently false (Objectivism and Humanism being two counter-examples). Emotionalist arguments also generally suffer from the flaw of presuming that desires or belief are somehow relevant to reality.
4. Lack of specificity: These arguments generally prove that some being has succeeded in implanting various things in our minds, but it needs not be a god, or even supernatural.

We can express our general argument in this section as the following:

1. For the god-concept to be considered a valid OGB-type hypothesis, it must be a viable explanation for a given observation or set of observations.
2. For the god-concept to be considered a valid OGB-type hypothesis, the theist has the burden of proof to point out these observations and why the god-concept is a viable explanation for them.
3. The theistic attempts to prove (2) are called the classical arguments.
4. Since all the classical arguments fail, the burden of proof to point out these observations and why the god-concept is a viable explanation for them has never been met.
5. There is no observation that the god-concept can viably explain. The god-concept is not a valid OGB-type hypothesis. [from 2 and 4]

The theist here has the burden of proof, since he is the one making the positive claim (that the term “god” is meaningful”). He must present evidence for that claim. In the absence of observations that we can use to make the god-concept a hypothesis, we must conclude, at least for now, that the requirement of premise 2 in process-based noncognitivism has not been properly met. We could also add an inductive argument on the low possibility of such observation ever being found.

IV. Conclusion

The meaning of “god” here is tied to the classical arguments. If a classical argument was found to be valid, then it not only would prove the existence of a god, but it would also gives us some meaning, as well as a way to find further meaning, for “god”. But since no such argument is valid, we are left with still no meaning of “god”.

There is one point that I have omitted. As I detail in my article “The Impossibility of Divine Intervention”, it seems to be the case that we can never attribute any natural event or entity to a god, regardless of the evidence. If this is the case, then process-based noncognitivism can be said to be true regardless of the success of classical arguments. I have still examined the classical arguments to make the case more forceful, but one can make the case that premise 2 is always true because of the facts I discuss in that article.

On the vast scope of this argument and noncognitivism in general, I once again invite you to read ‘The Argument from Non-Cognitivism’. I have only given here the specific argument as it relates to OGB. This argument, just as the argument against MGB, proves that the term “god” is meaningless, that its existence is an impossibility, and that any use of religious language can only be meaningful in the strong-atheistic context (or, depending on your position on the compatibility of noncognitivism with weak-atheism, the atheistic context in general).

Last updated: October 9, 2004