The Limits of Science

Saturday March 20, 2021

Consciousness and Subjective Experience

I believe that it is fundamentally impossible to conclusively answer questions about the nature of consciousness and subjective experienceI’m using “subjective experience” to describe what IIT seems to call “intrinsic existence” — the name “subjective” could be somewhat misleading, because while it is subjective, its existence is in a way, the only thing that one can ever be certain of. using science. There is a fundamental bootstrapping problem here: definitionally, all thoughts exist within the bounds of subjective experience, which prevents us from asking questions about things that exist outside of those bounds. We can’t say for sure whether any given system is “conscious” in that it has a subjective experience, because we are only able to exist in our own subjective experience.

Notably, it’s possible to step around this by defining “consciousness” as something unrelated to subjective experience, or by taking conceptions about the way that the system of all subjective experiences “should” look as axioms,This is the trick that IIT pulls — bundling a huge set of assumptions in their axioms, and then building giant mathematical castles in the sky on top of their unprovable axioms. but that limits the questions you can ask quite severely.

Ethics and Morality

Science can answer questions about what the world is and how it behaves, but that does not tell us how we should behave. While many ethical and moral systems choose to draw from science, science itself is unequipped to make moral statements.

People often choose to sell their moral ideas as “scientific”, but this is always a trick — the “science” is typically window dressing to distract from questions about the fundamental moral principles the ideas are based on. This is perhaps most visible with eugenics — people trying to push eugenics will center the conversation on the “scientific facts” about things like racial differences in order to distract from the part of their ideology that makes a link between whatever “scientific” metric they’re looking at and people’s worth and decency. Unless you’re careful, it’s easy to get sucked into arguing about how accurate skull calipers are, without asking what exactly people are proposing to do with the measurements, which is the far more important question.

We also see this with coronavirus regulations: people say to “follow the science”, but really we’re using science to inform our behaviour in relation to the moral principle that we should avoid causing others to come to harm. How important people feel avoiding causing harm to others is varies a huge amount, and science does not help us answer questions about how important avoiding causing harm is.

Any convincing argument for why people should do something should have two parts: A scientific argument about what concrete impact a change will have, and a moral argument about why that impact is good.If you’re going to skip one of these parts, skip the science — at least virtue ethics doesn’t misrepresent itself. It’s OK to believe that something is inherently morally good, the trouble comes when you convince yourself that that moral good is objective.

Selecting Questions

Say we understand the limits of science for morality and consciousness — there’s still a problem: the scientific method relies on asking questions and answering them, but there are an infinite number of possible questions we can ask. How do we select our questions?

In order for science to be useful, it must be a creative pursuit: creativity is the fundamental process by which we choose what questions to ask, and choosing to ask important questions is the way we get anything useful from science.

There are many parallels between science and mathematics research here. At first, mathematics appears to be a strictly logical discipline, but the skill of writing novel proofs is almost entirely about using intuition to choose which path to take out of an infinite set of options. What’s worst of all is that mathematical publications take the result of this process, and only describe the paths that were taken, leaving out the dead ends, but the dead ends are incredibly important for developing an intuition for what questions to ask and what paths to take. The scientific community is moving more towards preregistration and publishing negative results, which could conceivably help people develop this intuition, but I’d love to see more discussion of how people decide what to study, both individually and collectively.


Science is an incredibly important tool, but it is just a tool, and it is important to understand the limits of the tools we use. Don’t let yourself be tricked by moral arguments masquerading as “scientific” ones.

changelog
: Initial post
: Remove intro