The creator of the now famous TED Conference, Chris Anderson, started a podcast to ask questions that take longer than a 15-minute talk to answer and he interviewed Sam Harris who gave an early TED Talk on his 2010 book, The Moral Landscape. 

Note: this is an audio only video. 

I loved the interview and it gave me a chance to reflect on things I hadn’t though about in a while as Sam has spent the last several years interviewing guests on his podcast and his own opinions have been slightly submerged. (Except, possibly his opinions on He Who Shall Not Be Named.) I adored The Moral Landscape when it came out but I realize that my views have evolved a bit in the time since it came out.

There are two main topics covered in the podcast and I’ll probably save the second one for a later date.

The thrust of The Moral Landscape is that morality can be grounded in science and what is “right” or “wrong” can be objectively determined scientifically. Or, more simply, “Science can tell us what we ought to do.” And while I think Sam is largely on the right track, there is something missing in the discussion.

What the question of morality needs is the application of “Darwin’s strange inversion of reasoning” (a quote Daniel Dennett invokes frequently in his books and talks). This involves grounding a picture of the world from the bottom up instead of from the top down.

The way we understand the universe is that it’s built in layers out of increasingly complicated components. So if you put quarks and electrons together, you get atoms. If you understand the behaviour of atoms, you get chemistry. Extremely complicated chemistry becomes biology and extremely complicated biology becomes psychology and sociology, etc. And at each level, the details of the level below have a tendency to wash out and that’s not normally a problem. That’s why, when the doctor tells you to try and eat less salt, that conversation doesn’t reference the atomic weight of sodium atoms.

What we call moral values are agglomerations and simplifications of smaller moral units. “Murder is wrong” while we may agree it’s true, isn’t true in a fundamental sense. It’s a general principle extracted from aggregate data over decades looking at the consequences of murdering and not murdering, and the observation after the fact that reducing the amount of murder has a strong tendency to make people “better off”. And this observation is robust across different cultural understandings of “well off”.

The added benefit is it makes sense of our intuitions that sometimes murder might be permissible. Because even though in this aggregate sense, reducing murder makes people well off, you can still run moral thought experiments (the usual strategy is to invent a hypothetical situation involving someone with their finger on a detonator near some young children and cute puppies, and our intuitions will say that a well positioned bullet between this person’s eyes is probably a good idea) without getting a contradiction.