Question For A Non-Theist: How Does Morality Fit In?

Posted on by Ryan Downie

An Axiomatic Approach

The purpose of this blog is not to speak on behalf of all non-theists. Non-theism isn’t a unified system of thought. Instead, I’m going to use this space to lay out my personal view of morality as an answer to the proposed question aimed at me. Along the way I will offer suggestive reasons in support of my model, but the primary aim here is to simply formulate a sufficient outline as a launch pad for further consideration. It should be understood that this is an initial sketch and I fully acknowledge the existence of many “kinks” to smooth out.

The primary concern of morality (at least as I see it) is the classification of actions. More specifically, morality can be described as a special kind of partition on the set of all actions. For reference, label the set of all actions by \textbf{Act}. Any moral system essentially amounts to partitioning \textbf{Act} into three subsets: G, B, and N. The labels are suggestive in that G is to be the set of all ‘good’ actions, B the set of all ‘bad’ actions, and N the set of all neutral actions. Intuitively, actions within G are preferred in some sense. They are actions to be done, while those in B are to be avoided. These classifications are intended to be a guide to action.

Different moral theories end up with different partitions based on how they establish their reference (or axioms), but there seems to be significant overlap between all theories that have any significant influence and following. I suggest that this is properly explained by my approach, which will be discussed below. Morality is more than creating arbitrary partitions (which can be done in infinitely many ways). We are definitely referring to something and the way we partition \textbf{Act} is an attempt to get at this something, which is fundamental to the actions themselves relative to agents.

Agents Are Necessary

Talk of morality really on makes sense with respect to worlds with agents. In this context, define an agent as follows:

An agent A is a being capable of selecting from a given set of actions at any given time based on reasons, which supply the explanation for carrying out the action.

When agents operate in a world, they are involved in what I will call situations. Situations contain agents along with relations amongst them. Furthermore, for any situation S containing agents A_1, A_2,..., A_n, each agent in S has an associated set of possible actions. Call this Act_{A_{n}\in S}.  A set of actions will be classified as axiological if its intersection with either G or B is non-empty.

Agents Act According to A Nature

Any agent possesses a nature \mathcal{N} consisting of attributes that act as a framework for values. This, in turn, significantly influences (if not determines on some level) the sorts of actions an agent will engage in. For example, an agent possessing a nature that leads the agent to only value itself will not (with high probability) choose actions that are self-sacrificial.

Agents Act According To Beliefs/Reasons

Agents also don’t act randomly. Each agent operates from an epistemic/doxastic state or status that induces a ranking on sets of available actions.

Well-Being Is The Way To Go

First let me offer the following distinction: actions themselves (as abstracta) are either ‘good’ or ‘bad’. It is the instantiation or execution of a good or bad action by an agent that is moral or immoral. With that said, I suggest (as a first pass) that an agent P acts morally in situation S when P is (1) morally motivated, and (2) chooses an action from Act_{P\in S}\cap G. It should be stressed that this view does not require there to be a single “correct” moral action.

At this point I need to say what makes an action good or bad. In other words, I need to divulge how my partition gets generated. As mentioned above, creating a partition is very easy. All one needs to do is divide up \textbf{Act} into any three disjoint and exhaustive subsets and then attach labels. But this isn’t meaningful. We want something that naturally captures the sense we have of goodness and badness. For the theist, G and B (and therefore N) are determined by referencing what is believed to be the choice or nature of a divine agent. I believe there are several serious flaws in this approach that I will expound upon in another post. Until then, I will simply offer an alternative that I believe shows far more promise in explaining and understanding morality.

Where I agree with the theist is that the determination of G and B must be grounded in a nature. However, I suggest that the only nature that can produce a meaningful morality is the nature of the agents for whom morality is to be a guide. The nature to which I am referring is very general and should not be confused with individual idiosyncrasies.

The most basic, irreducible, or primitive value shared by any agent is well-being (or perhaps fulfillment). This is not a trivial claim, but for now I will simply issue it as a premise. I will also purposely leave “well-being” vague, since it can amount to different things depending on the nature of the agents in question. As humans, we have a good enough intuition as to what that means for us.

Next, morality emerges from how actions affect the well-being of agents. After all, what need could there be for any guide to action in worlds where actions taken literally have zero impact on other agents? This leads me to believe that good actions consists in those actions that promote well-being and bad actions are those that obstruct or discourage well-being. Of course this is overly simplistic, but it’s in the right direction. Agents therefore act morally insofar as they seek to choose actions from G that are available in any given situation S according to their ability. This can be difficult, since it may not always be obvious what actions belong to G or if there are “more correct” options than others in a given circumstance. One key principle for helping agents decide what actions belong to G and which of those might be most appropriate is the following symmetry principle:

Principle of Symmetry: This is an empathy principle whereby all agents are treated as value indistinguishable. There is no preferred local point of view. Thus the evaluation or classification of an action depends (in part) on operating from behind a veil of ignorance (Rawls) whereby one cannot assume which part he or she will play in the interaction (the one acting or the one being acted upon).

Some Final Remarks

Morality is a deep and difficult subject. What I have presented above is woefully incomplete. There are many areas that need elaboration, exploration, unpacking, tweaking, etc. Nevertheless, I hope it provides a sufficient sketch that can act as a catalyst for further inquiry and discussion.

Category: Questions For A Non-Theist