Posted by: dacalu | 4 March 2012

Utility Functions

At some point, I feel a need to move from analysis to synthesis, from talking about how things are to trying to make a useful system.  With luck, this post will make that switch.  Having noted a common experience of agents and preferences, and having pointed out why science is fundamentally incapable of dealing with them in a meaningful way, it seems prudent to attempt some non-scientific, yet still rational system.  I’ve spoken earlier of my system of atoms and my system of agents;  this is the latter.

We start with the premise (axiomatic, but also practically necessary) that some entities exist with agency – the ability to have chosen otherwise.  I consider my-self to be an agent based on experience and I project that agency onto other human individuals and possibly other non-human individuals (cats, for instance).

I have noted that individuals lacking limbs continue to have the appearance of agency.  Likewise, individuals with mental challenges (up to, but not including those in a coma or brain dead or just plain dead) continue to have the appearance of agency.  As agency is a binary operator – you have it or you don’t – I’m left in something of a bind with regard to the relationship between agency and the body.

First, I have never been a fan of “free will” in the strict sense.  Just because I believe we have some freedom to choose, does not mean that we have the freedom to choose anything.  I cannot choose to not obey the law of gravity.  Similarly, I believe my will to be constrained by evolutionary, social, and psychological conditioning.  Some choices I will not ever imagine, others I can imagine, but cannot imagine being open to me personally.  As my own will is constrained, I will posit that others’ wills are likewise constrained, some more and some less.

I suspect that physical and mental integrity (the health of body and mind) impact the extent to which the will is constrained.  Research suggests that a lack of sleep will impair decision making.  It may also limit the ability to think creatively about problems.  Torture involves some assumption that the will can (at least in some cases) be circumvented, or at least redirected, by environmental factors.  So agency could be a function of the body.

On the other hand, I would note that individuals like Stephen Hawking have severe physical limitations without an apparent diminishing of intellect or will.  Some people seem partially immune to the effects of sleeplessness and torture.  Correlations between physical state and agency can be drawn for populations, but there also seem to be exceptions.  Thus I’m extremely hesitant to strictly link the two.  Unless and until some mechanism can be found, I want to recognize the correlations, without reducing one to the other.  As science cannot measure agency, the discovery of a mechanism may never occur.

I attribute agency broadly, for reasons of compassion.  That is to say, I generally associate agency with preferences and I seek to account for and respect (without necessarily agreeing with or honoring) the preferences of all others.  I will err, then, on the side of attributing agency to non-agents.  I will consider the likely preferences of those in comas, even if I have no evidence of actual preferences.  [Again, don’t jump ahead of me – I have not claimed equal value for the preferences of all.]

The language got a little heavy, so let me say it more simply.  I weigh my own preferences against those of others.  I try to think of the preferences of everyone involved in an action, before I make a final decision.

In order to do this sort of reasoning, I model the preferences of agents.  I model my own preferences, those of my neighbors, and yes, those of God.  [If you don’t believe in God, skip that part.  It’s not essential to the train of thought.]  As I desire to have good models of preferences, I use Ochkam’s razor, constructing the smallest set of variables to explain the broadest range of observed preferences.  Perhaps Johnnie always chooses the activity that involves the least work.  Perhaps Susie regularly avoids conflict. Very rarely does a single variable suffice – people are complex – but I ‘m a scientist by training, and so I look for elegant explanations.  When talking about the suite of variables that explains preferences, I call it “utility function.”  I could also call it a “value system,” but that both begs the question of moral realism (do values exist independent of valuers) and fails to highlight my criterion of elegance.

I want to have a better understanding of my own utility function as well as the functions of others.  To this end, I feel a need for some form of evaluating preferences.  Call it “decision theory,” “psychology,” or “moral reasoning,” it still falls into the category of trying to figure out why we choose as we choose.  I have no doubt that evolutionary history has constrained my preferences.  I like fats and sweets.  I have no doubt that social forces have constrained my preferences.  Analytic philosophy makes more sense to me than Rabbinic and Taoist philosophy.  Nonetheless, I have reason to believe that my final decisions reflect more than just my conditioning.  I think I can choose differently than my conditioning suggests.  I think I can train myself to eat better and appreciate other types of reasoning.  And, yes, I recognize that that reflects an adaptive slow thinking process (evolution) and years of college (society), but at some point, I also want to take responsibility and say that active, rational engagement in decision making is important. [And know (empirically) that belief in agency makes it easier to make prosocial decisions.  grin.]

So I would encourage you, whatever your metaphysical predisposition, to model your own decision making and preferences, so that you can consciously change them – should you so desire.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

%d bloggers like this: