Friday, March 16, 2007

Patricia Churchland: The Biology of Morality

Since I started this series on the presentations of Beyond Belief 2006, I have been aching to get to this section – the section on morality. In the section on morality, I have been aching to get to Patricia Churchland’s presentation.

This is because Churchland’s position represents a massive speeding train which is heading entirely down a wrong track. Like logical positivism in philosophy, behaviorism in psychology, and communism in economics, Churchland takes a road that is fundamentally flawed.

Churchland's Relationship between Morality and Biology

Churchland is the chair of the Philosophy Department at the University of California, San Diego. In her presentation, she attempts to describe a relationship between morality and biology.

She begins with this pair of claims:

Evolution sets the brain’s style of drives and emotions. Experience in a culture shapes the style into specific habits and preferences using the reward system.

This I happen to agree with. Our aversion to pain, desire for sex, preference for certain types of food, our drive to seek out a comfortable environment avoiding excessive cold or heat, are all under evolutionary influences. We evolved as beings that will tend to have desires that have, in the past, perpetuated the species.

(Note: It is tempting to say that we desire things that will perpetuate the species. However, this is a mistake. There is no law of nature that says that the desires that have perpetuated the species in the past will continue to do so as the environment changes. Those desires that will perpetuate the species will survive into the future – that much is axiomatic. However, there is no guarantee that all or even most of us have those desires.)

From there, she went on to tell the story of two species of voles (fuzzy little mole-like mammals). From studies of the voles she concludes, “In male prairie voles, you can predict the degree of monogamy . . . as a function of the density of vasopressin receptors in certain very specific parts of the brain.” With females, behavior is related to oxytocin levels.

I found this to be just an astonishing result. Because here is something that many people have thought is really a cultural concept, that is whether you should be and monogamous whether you should be promiscuous, and it turns out that it’s importantly related to your specific biochemistry.

Whoa! Stop!


'Is' and 'Should'

The research that Churchland has described says nothing about 'should'. What this research shows is that whether a vole is monogamous or promiscuous is importantly related to specific biochemistry. They have not discovered anything about should.

Here’s the rub. Unless you believe in some type of magical power, all behavior is going to be linked in some way to brain chemistry or brain structure. What else is there that controls behavior?

One of these days, we may well discover a specific brain structure associated with rape, murder, theft, whatever. When we do so, this brain structure will indeed be importantly related to whether a person does or does not commit rape. Yet, it will not at all be related to whether the person should or should not commit rape.

Churchland provides her own counter-example. In the discussion that followed her session, she spoke about a tendency among chimpanzee males, when all is comfortable within the group, to wander to the edge of their territory, find a lone chimpanzee from a different tribe, and to kill it in a horribly brutal way. She wants to know the brain chemistry of this type of behavior so that it can be stopped. Yet, if we accept her analysis above, then this brain chemistry determines if the chimpanzees should go seek out a chimpanzee from a neighboring tribe and kill it.

Churchland's Answer to 'Is'/'Should'

Churchland, of course, is aware of the claim that one cannot derive an 'ought' (or 'should') from an 'is'.

She acknowledges that the inference from ‘is’ to ‘ought’ is formally invalid. However, she says, there are a number of ways in which people draw inferences other than through formal logic. She mentions how we take in a set of observations and draw an “inference to the best explanation.” She uses as an example a mother knowing that her daughter has been playing outside in the brush on a warm spring day coming home with a rash on her arms. She infers “my daughter has been in poison ivy,” without this being a formal deduction from true premises.

She claims that we do the same thing with morality. You are hiking. You see a bear and you retreat back up the trail from which you came. There, you see other hikers. It is not a deductive inference from ‘is’ to ‘ought’ but an informal inductive argument that leads you to the conclusion that you ‘ought’ to tell the hikers that there is a bear up ahead.

While in the process of illustrating how this type of inference works, Churchland in fact provides the proof against her original claims. She puts up a slide of an optical illusion where two squares that are in fact the same color (or shade) appear to be different – because of their context. The mind looks at the image and draws an inference of a difference where, in fact, no difference exists. Now, she should ask, "What is to keep the mind from looking at a situation, drawing an inference of an ‘ought’, where in fact no ‘ought’ exists?"

Worse, she does not address the possibility that the mind might infer an ‘ought’ where there is, in fact, an ‘ought not’ or vica versa.

In the case of the child with the poison ivy and the two squares on the optical illusion, we look at the evidence and draw a hypothesis. We know what it means to say that the daughter has been in poison ivy or that the two squares are of different color, from this we make further inferences, we can then test those inferences to verify or falsify our earlier hypothesis.

But when we infer that one should be monogamous, or should tell the hikers about the bear, or should go to the edge of our territory and brutally kill any lone individual from the next tribe that we might find there, how do we test that hypothesis? How do we determine whether our ‘inference to the best explanation’ is an inference to the right explanation as opposed to an inference into error?

Another problem with Churchland’s view is that it makes morality extremely easy. How do I determine whether I should be monogamous? Churchland seems to suggest that I go to my doctor and have my oxytocin and oxytocin receptors checked. One result would tell me that I should be monogamous, another result that I should be promiscuous. This, as I said, is an extremely convenient morality because the first result also says that I do not wish to have an affair and the second result will suggest that I do wish to have an affair. So, conveniently, morality will tell me that I should do exactly what I wish to do.

As I said, Churchland is not consistent on this. Though she seems to speak about deriving the should of monogamy or promiscuity directly from brain science, she does not speak about deriving the should of ripping apart a lone individual from another tribe in this way.

An Alternative to Churchland

Now, I am going to offer a solution to some of these problems. To do that, let me introduce some hypotheticals. Please note, I am analyzing concepts here, not the actual moral value of monogamy and promiscuity. I am going to suggest how this can be determined. It will be different than Churchland’s suggestion, but more consistent with her overall remarks.

Let us assume for the sake of argument that a disposition towards monogamy is related to oxytocin levels. This is a simplifying assumption. However, as I said, I am analyzing concepts here.

Let us further assume that evolution sets a base level (or range) of oxytocin levels, thus setting the base disposition towards monogamy.

So far, monogamy has the same moral status as height and weight. We have not said anything that suggests whether an individual should be monogamous.

However, let us further assume that social forces can affect oxytocin levels. Specifically, let us assume that social praise of monogamous individuals tend to increase oxytocin levels in individuals who witness that praise. Churchland herself discussed research that shows that oxytocin levels can be influenced by a simple massage. So, perhaps the prospect of an ego massage for monogamous individuals can cause higher oxytocin levels, which increases the strength of the disposition towards monogamy. On the other hand, condemnation of monogamous individuals will decrease oxytocin levels and, thereby, promote adultery and promiscuity.

Now, we have a question to ask: Should we be using praise to increase or decrease oxytocin levels? In other words, do we have reason to promote higher oxytocin levels in individuals (through praise of monogamy), or to decrease it (through condemnation of monogamy) – or does it really matter?

Let us further assume that higher oxytocin levels make individuals kinder and more considerate of others. Certainly, we have reason to make others in our community kinder and more considerate of others. This would make them kinder and more considerate of us. It will also make them kinder and more considerate of those we care about – our children, other family members, friends.

We also must consider the fact that others have reason to increase our oxytocin levels. To the degree that they do so, we have even more and stronger reasons to make people kinder and more considerate of our children, other family members, and friends. Because, now, we care more about their well-being.

Now, I have introduced a ‘should’ into the equation. What is the nature of this ‘should’? Is it subject to the same problems as Churchland’s ‘should’?

This ‘should’ simply means ‘has more reason to than reason not to’. The ‘should’ of increasing oxytocin levels in others is merely asking whether I ‘have more reason to than reason not to’ increase oxytocin levels in others. This is not a mysterious form of ‘should’ that cannot be derived from ‘is’. This is an ordinary form of ‘should’ that translates into ‘is the case that I have more reason to than reason not to’.


When we apply this to the examples we discussed so far, we get the following results:

Is it the case that we generally have reason to promote the desire to kill lone members of a neighboring tribe, or to inhibit such a desire? Let us say that we have the option of talking to members of the other tribe and, through dialogue, create an inhibition in them from attacking our lone members. At the same time, they create in us an inhibition to attack their lone members. Under these conditions, our members will have more liberty to travel alone in safety – and so would theirs. The members of our two tribes have a great many reasons to inhibit this desire to attack lone members of other tribes.

These ‘reasons to inhibit a desire’ are the reasons that Churchland quite naturally (but unconsciously) called upon when she said that she wanted to know the biochemistry of this behavior so that we can put an end to it.

On the issue of monogamy versus promiscuity, Churchland said that oxytocin levels determined whether one should be monogamous – a claim she does not make about killing lone individuals from a neighboring tribe.

Assessing Churchland's Claims

I suspect that Churchland began with the assumption that monogamy or promiscuity were permissible – or, at least, their morality were ambiguous. If she had truly approached the question with the idea that promiscuity were wrong, the way rape and killing a lone individual from another tribe is wrong, she would have seen this discovery as a way to eradicate a disposition towards acts – the way she would view a similar discovery about the biochemistry of rape or killing a lone individual from a neighboring tribe.

In other words, she assumed, unconsciously that there are no reasons to act so as to promote or inhibit desires relating to marital fidelity. As such, individuals were permitted to act in whatever way their natural oxytocin levels dictated. This absence of reasons for promoting or inhibiting desires explains and predicts why the monogamy case was handled differently from the killing case.

Even though Churchland explicitly says that she is looking to brain states to determine if something is right or wrong, her judgments suggest that she is, in fact, looking to something else outside of the brain state to judge the brain state good or bad. I would further argue that this ‘something else’ is whether the brain state itself tends to fulfill or thwart other desires.

Please note, I have not refuted or even doubted any of the brain science. I am not saying the brain science is wrong. Nor am I bringing in any bizarre properties – no supernatural (or non-natural) ‘oughts’ to clutter our ontology. My claim is not that biology fails to tell us anything useful about morality. My claim is that biology plays a different role than Churchland (sometimes – though not consistently) says that it plays.


Anonymous said...

i agree with your assessment of Churchland and the mess that ensues from sloppily mixing biology with ethics. Sober and Wilson, have a much better approach (of course, that's a philosopher and a biologist working together).

I do have qualms about your 'Alternative' to Churchland. You are justifying a certain 'ought' with propositions that give "more reason[s] to than reason[s] not to". However, the reasons you give, making people "kinder and more considerate of others" are themselves moral reasons, just less controversial ones than the 'both sexual practices go' reason that Churchland implicitly gave.

A yet better argument against Churchland would follow Hilary Putnam in explaining that facts of biology and values of ethics are fundamentally intertwined; thus Churchland's definitions of 'promiscuity' and 'monogomy' are already morally-influenced ideas that cannot thus claim to provide an unshakeable, scientific foundation for further ethical reasoning.

Alonzo Fyfe said...


The "reasons that I give" were not "making people kinder and more considerate of others."

The reasons I give are "the reasons we have for making people kinder and more considerate to others."

Some of those reasons are basic biological desires - aversion to pain, desire for food, desire for a comfortable place to live, desire for sex, desire for the well-being of one's children (which is a basic biological desire whose strength may be altered through social factors but that exists even without social factors).

Some of those reasons are acquired reasons (acquired desires) such as a desire to study astronomy, a desire to play football, a desire to write a novel.

All of these reasons give us reasons to be surrounded by people who are kinder and more considerate to others.

It is the reasons for being surrounded by people who are kinder and more considerate to others that I am referring to, not the kindness and consideration itself.

Note: I hold that nothing has intrinsic value. All value consists in relationships between states of affairs and desires. Even the value of a desire is determined by its relationship to (other) desires.

That is the part that Churchland missed out on - questioning the value (in terms of the fulfillment of other desires) of having particular oxytocin levels, to the degree that we can act in ways that influence the oxytocin levels in others.