How safe is safe enough?

November 20, 2009


We don’t do very well at dealing with risk and uncertainty. Maybe because it’s so darn scary!  Risk means danger, and uncertainty only adds to our fear, even if the risk, as a quantitative value, is very small!  Here we have an example of small risks attacked with big solutions that cost lots of money!

The marvelous water supply system of New York City brings some of the best tasting and safest drinking water in the world to nine million people, mostly in the city’s five boroughs.  It directs nearly 1.5 billion gallons per day (bgd) from a reservoirs system in upstate New York,  90% of which from the Catskill-Delaware systems about 90 miles from Manhattan.  The City is now spending approximately $3 billion to build an ultraviolet disinfection plant for the water supply, and to build a cover over the Hillview Reservoir, one of the last holding points for the supply.  That’s a lot of money, even in NYC!  What are we getting for it?

Except for the 10% from the East-of-Hudson reservoirs just north of the city, the water is unfiltered.  It is of such high quality, and spends so much time in enormous reservoirs, that it does not require cleaning.  Cities that draw their water from the ground or from rivers, gaaaggg!, must carefully filter the water.  The water is disinfected with chlorine to kill harmful bugs (pathogens), like the ones that used to cause cholera and typhus epidemics.  The water is safe!  Why the UV plant?

With the advance of public health science, new “disease vectors” have been identified.  In water supply, the latest are cryptosporidium and giardia, two very tiny critters that can cause intestinal disorders in humans, and if the victims have compromised immune systems, possibly lead to death.  These bugs are not killed by chlorine, but people can protect themselves by drinking boiled water.  They are very rare in NYC water.  There has never been a documented outbreak of any public health risk in NYC due to these bugs.  They can be serious risk in many small and improperly run water suppliers, especially those in agricultural areas, where farm animals produce lots of manure with the bugs that may get washed into water supply areas.  UV sterilizes the tiny bugs, preventing them from reproducing, which is as good as killing them.  Nobody has found a good way to kill them, other than boiling them, which is obviously impractical for 1.5 bgd.

So, we are spending $1.4 billion on a UV plant to eliminate a bug that is rare and impossible to monitor, which has never caused a disease outbreak in NYC, and from which the few at higher risk can protect themselves by drinking boiled water?  There was a serious outbreak about ten years ago in Milwaukee, but that system had a malfunctioning filter (which would normally capture the bugs) and happened during an extreme weather event that would not have a similar effect on NYC’s huge system.  In addition, NYC has a strict watershed protection program in place, which is why the US EPA does not require it to filter most of its water. 

Well, if you were at risk, you would certainly want to have that UV plant online!  But then, looking at it from the public health perspective, $1.4 billion would buy an awful lot of work in preventing TB, AIDS, veneral disease, and other sourges that are killing people now.  What’s the cost-benefit?

The story with the cover is much the same.  Birds pooped in the reservoir, the presence of E coli bacteria spiked, the EPA noticed it in the report and ordered a cover.  The problem was pretty much eliminated with other programs to frighten away and discourage birds, change the way water was withdrawn, etc, but the ruling was kept in place.  No exceptions.  $1.6 billion to build a cover for a reservoir that will not appreciable improve the lives of anyone but engineers and contractors working on it. 

Ahhh…but we can all breathe so much easier, knowing that at least the risk has been reduced to nearly zero!

Free Will, and All That …

February 10, 2005

“Who is Number 1?”
“You are Number 6.”
“I am not a number. I am a free man!”


Is Ahab, Ahab? Is it I, God, or who, that lifts this arm? But if the great sun move not of himself; but is as an errand-boy in heaven; nor one single star can revolve, but by some invisible power; how then can this one small heart beat; this one small brain think thoughts; unless God does that beating, does that thinking, does that living, and not I …Where do murderers go, man! Who’s to doom, when the judge himself is dragged to the bar?

Two views of freedom for us poor pismires crawling about the surface of the globe. The Prisoner knows he is a free man, while Melville’s Ahab sees his actions as the movements of a puppet on a string, a string pulled by God, or something, perhaps the blind material universe. His insane murderous behavior is not his fault, it was caused, determined, preordained outside of his powers of volition.  I’ve been thinking about this free will jazz for a while, about thirty years. It occurred to me that something I said in my post about the so-called Intelligent Design theory might be said of free will, to wit, when you stop assuming it exists, you don’t see it anymore. That’s what some folks would argue about free will anyway, that it’s an illusion, that we are all, in fact, some sort of thinking automata that wonder if we have will and volition. Noooo. I don’t think so.

When a person becomes convinced of the soundness of evolutionary theory, he or she stops seeing design in nature, it’s true, but the person will also stop talking about design. Yes, they may say, “nature designed this in such a way..,” but that’s just shorthand for long phrases such as, “the random variation of types was winnowed by natural selection over a long period of time to yield…” Darwinians don’t believe in design in nature. People who claim to have demonstrated the illusory nature of free will still, however, use terms like choice, we, I, think, and so on. “When we make choices, we think we are free, but all our actions are as determined as the plunk and click of billiard balls moving about on a table.”  How can you speak of choice without free will?  Do thermostats make a choice when they go on as the temperature rises?  No, their actions are determined by physical law.  So, unless you think that people and their minds are machines like those we see about us, and unless you think that we never choose, you can’t claim not to believe in free will.

The issue of free will shouldn’t even be discussed – let’s ban it!  It’s a question badly framed at too high a level of abstraction.  First we have to decide what consciousness is, or mentality, because that’s what we think distinguishes us from purely deterministic machines.  Free will is just that question in another form.  And beyond that, there is the deeper question of what is determinism itself, what is causality, and what is time?  Free will is trivial if you can deal with those.

I like to say that the notion of free will requires determinism.  We act for reasons, acts are ‘determined’ by reasons, otherwise they would be random.  And random acts are not what we usually think of as the products of conscious volition.  So without strict determinism, there is no free will because there is no will.  Will is a determinate principle, whatever that means.  David Hume made sort of this argument when he said that the dispute was purely semantic and that our acts were free and determined.  Free in that that were not constrained, determined in that they had reasons.  I happen to agree with this for the most part but I don’t think it goes far enough in explaining why the anti-free will position just makes no sense.  I’m going to start by digressing onto a topic that I usually avoid, quantum physics.

I’ve just been reading an absolutely marvelous little book by Richard Feynman called The Character of Physical Law.  Feynman is brilliant, profound, and funny all at once, and if you think you’re not the type to read treatises by quantum physicists, this is the book for you.  Really – it was actually written for you!  The reason I avoid discussing quantum physics is because I’m sick of hearing people who don’t appear to know any more about about physics than I do say things like, “Of course, this applies to everything except the quantum level,” or “Of course, that only applies at the quantum level.”  I don’t know that much about physics, but it always seems to me that these arguments are kind of like deus ex machina.  And then, I wonder, how valid can an argument about everyday life be if it has to rely on quantum physics for a justification?  Something seems out of whack.

Now thinking again about that beefeater, David Hume, I’ve come up with a new thought: People get all hot and bothered about Heisenberg’s Uncertainty Principle because they feel it contradicts what they know to be true about life.  What if it were just a confirmation of what we knew to be true?  What if life is not predictable, not determinate, but is sort of random?  And how does that square with what I said earlier?

Hume examined the notion of cause in detail and found that he couldn’t really pin it down.  A causes B because event A always precedes event B.  More than that, there is nothing to say about what cause is, anymore than we say what gravity is.  (Or what is is?)  We arrive at the notion of causality through induction, i.e., by watching and observing the same sequence happen over and over, and then drawing the conclusion that A causes B, but we cannot prove that someday A won’t cause B, or that B will happen without A happening first.  Notice that causality seems to work in one direction, forward in time (whatever that means.)  Hume is rather skeptical about this notion of cause, although he recognizes its utility, and he does not believe that it can be justified in any genuine way other than that it seems to be quite useful.  So, what if cause is an illusion?  Then the notion that we have no free will would be an illusion, wouldn’t it?  And wouldn’t free will be impossible since I claimed that it requires determinism?  No again, causality and determinism are not the same thing.

The conflating of predestination, determinism, and causality is a big problem with all of these discussions, but right now I want to point out that while causes appear, to us, to work forward in time, determinism always seems to be after the fact – something figured out in hindsight.  “Aha,” we say, “Now I see why that happened…because this, then that, then those, then finally this!”  This is exactly how it is with free will:  A person makes a choice; we ask why?  The person gives reasons; the anti-free will philosophers say, “Your choice was determinedby those reasons. You have no free will. It was all determined.”  But discerning determining circumstances is not the same as demonstrating causality.  If we didn’t have reasons for acting, we wouldn’t have a will, but having reasons doesn’t mean our choices were unfree.

Into this rambling discussion I now must vent my wrath against the sizable community of counter-factualists, those people who worry and wonder about what would have happened if Hitler hadn’t overslept on D-Day, if Richard’s horse hadn’t lost a nail, if John Wilkes Booth had missed his target, and so on.  You can wonder, but you can’t make much sense if you try to predict the future of the past as if the past had been different.  I have to fall back on my ironclad law of historical causation: Things happened as they did because that’s how they happened.  It wasn’t predetermined, it wasn’t predictable, it wasn’t predestined, and it only seems deterministic in hindsight.  But it was all caused, one piece at a time, and each free choice was determined by reasons one at a time. We really have to keep these concepts separate since they may not even make all that much sense taken one at a time.

If Hitler hadn’t overslept, perhaps NOTHING would have changed.  And if Hitler had died in that bomb plot, perhaps NOTHING would have changed, except he would have died.  We can’t even begin to say something sensible about this, except to carefully speculate on what the possibilities were.  (It was not possible that Martians would lead the Nazis to victory!)  And no matter how much data we have on what the situation was, and what the possible choices were (what is an impossible choice?) we can’t make any sensible statement (determination) about what counter-factual thing might have happened.  Call it Lichanos’ Historical Uncertainty Principle if you like.

So finally, I get back to Richard Feynman, who was explaining in his little book the well known puzzler about the electrons shooting through the sheet of metal with two holes in it, and interference patterns and all that, and it comes down to the simple inexplicable fact that you just can’t tell which hole an electron is going to come through as they pop out toward the sheet.  Just no way to predict, no matter what you know about the total situation.  Might be this one, might be that one.  “Nature herself does not know,” is how Feynman quotes one physicist.  Now this reminds me of another situation, much more familiar to readers of magazines and newspapers:

Two boys growing up next door to one another in a very poor, crime ridden neighborhood.  Same age, same family status, same education, same…everything!  One goes on to become a hood, and ends up badly, in jail for murder.  No surprise, it was a foregone conclusion, social determinism, environmental determinism, etc. etc.  The other one goes on to become a doctor and spends his life setting up free clinics for children all over the world, has a beautiful, brilliant doctor wife, and five lovely children.

Why one, and not the other?  Now, I’m not saying the environment doesn’t count, because it obviously changes and can constrict the choices available to one, but one still has choices.  Even with a gun to your head you can choose to obey your torturer or to die – it’s a bad set of choices, but you have a choice.  But maybe life is not deterministic as we usually take that word, and this is just one more example of it.  And maybe the world is not deterministic in the way that we think our notions of causation make us see it, and Feynman’s puzzler is just one little example of that.  Maybe…the electron…and the poor boys…are behaving in much the same manner.  (I don’t mean the electrons choose, but that both are not deterministic in the simple, can-be-predicted way we think of it usually.)  You see, we covet choice in the world, and want to get rid of it too.  Choice for people, causality and determinism for thermostats.  Maybe causality and determinism aren’t quite what we think for all and everyone.  Freedom for all!

[A simple thought-experiment: A man faces two doors; which will he pick?  He picks the red one: Aha, it was determined because of this, that, and the other thing, including his childhood experiences with … But if he had picked the blue door, we would have made the same comments mutatis mutandi, we would have found a string of determining circumstances, so what does this tell us about anything?]

And what about randomness?  Huh, what about it?  What the hell is that?  Does anyone even try to explain why if I take 10,000 ball bearings and drop them in the top of a device with a grid of pegs over which they bounce as gravity pulls them down they will fall out the bottom and always make a bell-shaped distribution as they pile up?  Even if I drop each one in exactly the same way?  Is randomness some sort of effect of hidden causes, or, could it be, like those electrons? Just some irreducible fact of nature?  It seems that with ball bearings, as with poor boys, the same determining factors can lead to different outcomes.  (That’s not choice, but surely it leaves room for choice.  But what makes the choice?  What is the will after all?  Arrrrgghh!!)  Which would mean that our notion of cause should probably be more like David Hume’s than the rock-solid axiom we think it is.  Which means that free will isn’t so odd in the context of a ‘deterministic’ universe, and quantum effects are not so odd either.

The best book I ever read on this topic, before Feynman, was by Erwin Schrodinger, who certainly knew this stuff.  A wonderful little book, in which he said that when your ideas lead you to conclude that matter acts simultaneously as a particle and a wave, and that this causes contradictions, then it’s time to come up with a different notion of matter.

Sometimes I wonder if all these questions aren’t just a problem of scale.  As the scale of things changes, some things disappear.  As we walk around, we are not aware of quantum effects at the sub-atomic level; we aren’t even aware of molecules, not to mention that we are oblivious to the fact that matter is mostly empty space.  What if the same sort of effects relate to time – what would that do to our notion of causality and determinism?  As we ‘zoom’ our time-scale out to the enormous, everything would appear to be happening at more or less the same time, wouldn’t it?  Events that seem of long duration now become an eye-blink. Stretches of intervening time between events become eye-blinks. (Remember that eye blinking as Dave makes his descent to Jupiter in 2001?)  And finally, everything is happening simultaneously. What came before what? If you don’t know, you can’t talk about causes. And if an event is always happening, or has always happened, e.g. the universe has always been here, then it was never caused at all, which is how some medieval philosophers demonstrated that there could be an event without a cause, which might or might not have been God…

I don’t pretend to know the meaning of what I just wrote there, but I do think our notions of time are very crude, and have only changed slightly in the last 100 years since Einstein gave them a jolt. Things may change a lot more in very many ways. Meanwhile,

It’s my life and I’ll do what I want
It’s my mind and I’ll think what I want
Show me I’m wrong…