Friday, May 28, 2010

Carroll vs. Harris: Quantifying Value - Part I

I am addressing Sean Carroll's arguments for why 'morality as science' is not possible.

(See Sean Carroll: You Can't Derive 'Ought' from 'Is')

One of his arguments is:

There's no simple way to aggregate well-being over different individuals.

This section has so many problems I will likely spend two or three posts addressing them.

However, it is absurd on its face, without even starting to address its problems in detail.

You are in a room and you have pressed down on two buttons - one button with your left hand and one with your right. I am going to move those buttons further and further apart. There will come a point at which you cannot keep holding both buttons down. You must release one of them.

If you release the left button then five atomic bombs will go off, each in a different major city. If you release the button under your right hand, a child in Bangalore, Maine, will discover a $1 bill laying on the sidewalk.

Carroll is telling us that, in this type of situation, he would stand there in utter distress unable to come to a conclusion as to which button he would release. "We can't make inter-personal comparisons," he tells us. "We just can't. It's not possible!"

Of course we can. We do it all the time.

Consider the property of temperature - before the thermometer was invented. People living in ancient Greece did not have an easy way to assign a number to the temperature of something, but they could tell that the kettle on the fire is hotter than the kettle sitting on the rock outside during a snow storm.

Anybody who claimed that this could not be done would be uttering an absurdity.

Obviously, we can make interpersonal comparisons if the differences are large enough.

Now, let's look at specific problems.

Problem 1: The focus on well-being.

I will start with the fact that, once again, Carroll is confusing the task of defeating Harris' position on morality-as-science with the general possibility of morality-as-science. Though he writes in the context of responding to Harris, he did not title his paper, "Harris failed to derive 'ought' from 'is'. He titled it, "You can't derive 'ought' from 'is'."

An inability to aggregate well-being still leaves open a great many options respecting the possibility of morality-as-science.

Problem 2: Simple?

Why would the possibility of morality-as-science depend on a simple way of measuring? It seems to me that science gets along quite well in realms where the systems of measurement are not at all simple.

While 'simple' has its merits - all else being equal there are reasons to prefer a simpler method over a complex one - it is not a requirement for the possibility of science. You will not see a dissertation committee tell a graduate student, "These formulae are too difficult; therefore, what you are doing is not science."

Problem 3: It is not the case that the difficulty we have in measuring something is proof of the impossibility of science.

I want us to return to ancient Greece for a moment.

We are in ancient Athens about 350 BCE where we some street philosopher is telling us that we can never have a science of temperature. A science of temperature, he tells you, requires making sense of claims like, "This object's temperature is precisely 0.762 times the temperature of that object."

He insists that it is impossible to even imagine how one would make sense of such a claim. To prove his point, he challenges us to do so. As ancient Greeks we would have to admit that he is right. In the absence of what I know about temperature I certainly do not know how I could have possibly made sense of this type of claim. Nor could I imagine an experiment that would prove that the statement is true or false.

Yet, even though the premises of this argument are true, the conclusion is false. Our inability as ancient Greeks to imagine how to assign numbers to temperature and make sense of these types of comparisons are not proof that it cannot be done. It is not proof of the impossibility of temperature-as-science. It only described the limits of knowledge at the time.

Similarly, Carroll, with his "can't imagine an experiment", is trying to tell us what cannot be done. Yet the best conclusion he can actually justify if we assume that the premises he provides us are true is that our current state of knowledge does not allow us to make sense of these types of statements - yet.

The analogy to temperature goes even deeper.

We did not discover a way of assigning values to temperature until we noted that the volume of a liquid will change as temperature changes. We put liquid in very thin tube, drew a line next to the tube indicating the current volume of the liquid, and we called that '0'. Then we warmed up the substance, watch the liquid in the tube expand in volume and push up the tube, drew another mark, and called it 100.

With this method there were still a lot of temperatures we could not measure. How do we measure temperature at the core of the Earth? Of a distant star? Our inability to do so still would not prove the impossibility of temperature-as-science. It would only prove the limited extent of our current knowledge.

Nor did the invention of the thermometer allow us to make sense of "This object's temperature is precisely 0.762 times the temperature of that object." It is NOT the case that something that is 100 degrees Celsius is twice as hot as something that is 50 degrees Celsius.

We had a temperature-as-science for over a hundred years before we could answer those types of questions.

Yet, Carroll tells us we cannot have a X-as-science until we can answer to those kinds of questions.

Carroll is mistaken.

Problem 4: There is, in fact, a way of aggregating values; of assigning a number to the capacity of a state of affairs to fulfill desires. It's not that simple, perhaps, but I have already argued that 'simple' is not a valid criterion.

Everything I have written above seeks to point out that our current inability to quantify value does not prove that morality-as-science is impossible. Carroll has not given us an argument in defense of his conclusion. Against Carroll’s claim that it cannot be done, we have a huge stack of questions relating to how ‘ought’ can be distinct from ‘is’ that we have to weight against – on the other side – saying that we just don’t know enough yet. Of these two, ‘we just don’t know enough yet’ seems the wiser option.

Yet, in fact, we can quantify value.

I will start to explain how in the next post.


Kevin said...

His argument is akin to Loki's wager. (

dbonfitto said...

If there's a way to evaluate the well-being of discrete individuals, of course there's a way to aggregate well-being over many individuals.

More data leads to a better curve fit.

Kevin, Loki's Wager was summed up best by one of today's modern philosophers, Barbie: "Math is hard. Let's go shopping!"

If the dvergar had some 3D modeling and a CNC cutter they could have had a Loki's head to within a few nines...

watergate salad said...

I recently came across your blog and have been reading along. I thought I would leave my first comment. Great post, concise and easy to understand. I like this post..
I found out that this blog is very interesting and informative.
Best of luck to you!

Anonymous said...

Not that it matters much but it's Bangor, Maine. Not Bangalore. That's India.