Saturday, 20 October 2007

Is less harm is preferable to more harm?

I have discovered that I am developing a theme, very much a work in progress, evolving in public here, that there can be such a thing as ethics - the study of codes of conduct - as a science. I did not assume this when I started here but it appears to be the case. For now I want to consolidate the last few posts into a minimal system of conduct whilst mentioning, mostly as conjectures, the plausible science that can lay behind it and validate it.

First of all there are three and, only three, key principles behind this minimal system of conduct:
  1. Less harm is preferable to more harm
  2. Act so as to minimise harm
  3. React so as to thwart harm
Lets take these three in turn. We will deal with the first principle today and then the following two in the next post.

When asked as a question as in the title of the post, it almost seems absurd. Yet this is a very important question. As a principle stated here, this is almost self evident but not quite. The fact that most people would agree with it is insufficient to justify it. There are many things that majorities have or do agree upon that have been shown to be false, so usually we cannot rely on such communal reinforcement but can we do better here?

Is harm, by definition here, anything that prevents someone and everyone flourishing, fulfilling their preferences, achieving their desires and goals? What about when achieving such goals is predicated upon or requires the harming of others? Are there clear cut boundaries to delineate harm-free versus harmful motives, action, goals and unforeseen consequences or are they mostly or even all fuzzy and open to dispute?

What this principle is pushing for is to focus on what harms there are both in general and specifically or casuistically - case by case - and how these two interrelate. It is not so much an initial assumption but rather the definition of what codes of conduct are essentially about (without being a foundationalist). It is an a priori and in this way self-evident statement that sets the framework, for both a scientific and pragmatic analysis of the problem and an evaluation of solutions.

As such whilst it is part of this specific system here, it is indeed either explicitly or implicitly part of all systems of conduct whatever other principles, assumptions and beliefs are included. So it is not something to be proved or validated rather it is always the means of validation and evaluation. However different systems of conduct will have different means for determining harms, who by, who to and how far in both space and time; with differing weights and preferences to evaluating harms; and with various ways of prioritising or ignoring incommensurate harms. Is it possible to avoid the value laden assumptions inherent in all these and somehow scientifically evaluate them?

Surely in any specific case, however tightly or broadly delineated, one can empirically appraise all the underlying factors that could be labelled 'harm'. Different systems will apply these labels differentially, selectively and dependently based on their principles. Nevertheless, however difficult to ascertain, there are facts of the matter or states of affairs that can be realistically obtained with minimised uncertainty. How do we avoid being mistaken in such allocation of labels with their related weighting and so on?

In order to minimise mistakes and eliminate error here we have to admit that there is both the empirical possibility and basis to identifying all states of affairs and facts of the matter and that there are different sets of assumptions involved in labelling these states of affairs as 'harm'. On both points it is clear that we want to make as few assumptions as possible and those that are made must be open to review, revision, replacement or rejection, since the more assumptions one makes, especially those are are unquestioned or unchallengeable, the more likely error and mistakes will occur.

Hence different systems of conduct will make differing claims as to what harms (and weights etc.) there are in any situation and we can compare and contrast them given the common empirical ground behind all of these approaches. Still how can we do this objectively without biasing one over another? There three other empirical factors that can be obtained apart from the shared common empirical ground to which they are applied.

  1. A schedule and specification of harms to be minimised
  2. The code of conduct or set of prescriptions designed to achieve this harm minimisation
  3. The means of enforcement to ensure this code of conduct is adhered to.
These can then be evaluated against the shared common empirical ground to measure how successful these systems of conduct are. There are two types evaluations
  1. We can evaluate the internal efficiency - the cost of enforcement - of each system independently and rank them accordingly.
  2. We can evaluate the external effectiveness -how well do they achieve their goals regardless of cost - of each system independently and rank them accordingly.
We can then determine a means to combine these two metrics - not necessarily an easy task. Note that in none of this has one preferred or imposed one set of assumptions of harm over another and so this is, in this sense, bias free. Is this sufficient to have well grounded science of ethics? That is an open question. What I am arguing is that this is a good starting point for such a science or at least a proto-science and nothing more. I conjecture that this alone will show that many current and past systems are deeply flawed and basically failed systems regardless of any other philosophical arguments.

Whilst continually wanting to avoid developing a moral theory with all of the philosophical conundrums I seem to be taking a related path here. So be it! I am aware of classical objections to such a science such as non-cognitivism but will deal with that and others once I have completed my tentative system here, after the next post. However this approach or something similar may have been covered in other fields such as sociology (unlikely given the non-empirical emphasis I always seem to discover when I read up on this but still possible) or somewhere else. I would certainly be interested in that if someone is familiar with such work.


Miguel Picanco said...

While I'm certainly no expert in ethics.. it does seem as though you're trying to reinvent the wheel here.

I could pick apart the aspects of this post that scare me the most.. but it might actually be easier on all of us to first direct you to a similar line of thought already well in progress and see where things go from there.

Are you familiar with Fyfe's Desire Utilitarianism? I can almost imagine that if we looked back at Fyfe's very first notes on ethics, it might very well look like this blog post.

Although there's tons to read and sift through (the archives of his daily blog posts go back to mid-2005), you can find Fyfe working through his thesis at

martino said...

If anything these ideas come from the first part of Mackie's "Inventing Right and Wrong" and a number of extended conversations with non-cognitivists over the years. In fact, reflection, these ideas are old and I am rediscovering them but, this time, sharing them.

Given the connection between Mackie and Fyfe's Desire Utilitarianism, it is not surprising that I sought someone like him out and am indeed an active reader and commenter on his blog, including today.

However I have issues with DU which I will develop in due course there, the key one which Fyfe openly acknowledges - the dependency on BDI theory, which, with a background in AI, I remain unconvinced of.

Secondly, unlike Fyfe, I am not nor intend to be a moral philosopher, which I have repeatedly emphasised here. Indeed part of my general approach is that one does not need to be a professional nor serious amateur psychologist, biologist, mathematician, statistician or philosopher to have available reasonable suitable tools to analyse and deal with the many pragmatic situations we need to today.

Finally I am just writing what I feel like writing about. I know the topics that would make this blog more popular and am not interested in pursuing that. And I do welcome people picking part my posts, please do. However there is a second part that I will deal with next before I go onto the second and third principles mentioned here

martino said...

P.S. If you had looked you would have seen he is on my blog roll :-)