What’s the right way to do philosophy when you care about science? Sometimes, this question is misunderstood. It might be seen as prioritising science over all other human efforts to get knowledge. Or seeing science as the peak of all human endeavour. I don’t believe this is the case.

I don’t presume science can settle matters of values or aesthetics although it can help inform the debate. Likewise, I don’t see any reason to think science has a monopoly on obtaining knowledge. If it did, science itself could never get off the ground.

In any case, values and aesthetics are not my focus. My main concerns are questions about minds. For that, science has a great deal to say. Progress towards such understanding is best made with a scientific approach.

What this amounts to in specifics and how to justify this view is a complicated matter. But let me at least start with a rough sketch. The scientific enterprise is partly about finding ways to state and test questions in a way that can stand up to scrutiny. So, at least as I see it there are three parts to doing science. Asking the right questions in the right way. Finding ways to explore and test possible answers. Then holding it up to scrutiny, typically via peer review and public experiments.

Much of the practice of science - as done and taught by scientists - centres around the norms of determining what is “right” in the above. Philosophers of science spend a lot of time trying to explain how these practices derive their norms. Or sometimes how they ought to derive their norms. But scientific practice - much like human social practice in general - evolves over time by consensus. It is not pre-planned for the most part.

Modern scientific practice at least involves hypothesis building, mathematisation, experiment design, peer review, and publishing results. These elements have changed over time and there is nothing to suggest they won’t change in the future. It will be up to the community to determine what the science of the future will look like. With reflections of those who spend time thinking about these things.

To the extent we understand what science is today, it is clear that philosophy is not science. Philosophy may sometimes suggest hypothesis that can be empirically tested. It may make the hypothesis plain in logical terms if not precise mathematical ones. It may even in very rough sketches lay out potential experiments that could test its hypothesis. But philosophers are not typically aiming to fit the standards of scientific practice.

Except of course, for a few very scientific minded philosophers who work in collaboration with scientists. And some people simply wear many hats. They are scientists and philosophers. They swim in two different streams. They are happy to talk the language of science when with scientists. And the language of philosophy when with philosophers. I am not concerned with these lucky cases here.

So, what is the value of philosophy in a scientific enterprise? I’m attracted to the view that philosophy can clear up conceptual confusions. Science begins with a lot of assumptions. When building theories, we use a lot of concepts without examining them too closely. That’s sometimes the only way to get at new knowledge. You take what you believe is the case. You probe into what you take to be real phenomena. Then you devise theories to explain what you find. But your theories cannot escape your assumptions.

That is where philosophy comes in. Take my own concern about representations. It is prevalent within cognitive science and AI that internal states of cognitive agents represent things. That much is just taken as a starting point in theorising. But what does that mean? How can one thing be about another at all? These are the questions that are fundamentally philosophical. They do not lead to direct empirical hypothesis. But they can be reasoned about. And through this reasoning clarified (or swept away as misconceptions as the case may be).

What is the value of playing with concepts that cannot be tested? Surely, the success of the scientific project is that it has done away with such speculations in favour of what we can test. What we can prove empirically. This view is prevalent and there is some comfort in it.

Yet, to reiterate my earlier point: concepts come first. Our concepts are the horizons of the hypothesis we can make. Our hypothesis are what we decide to test. And science as a means of knowledge production, can only generate new knowledge through what it can test. So the wrong concepts can lead, through a chain of mistaken assumptions, to false beliefs. Worse, lack of certain concepts can close off entire avenues of investigation. Our concepts (or lack thereof) can leave us blind.

So much for the value of philosophy to science. What value does science have to philosophy? If as I see it what philosophy is best at is, at least in my area of concern, working with concepts where does the science come in? Here I think there are two answers:

1. Philosophy of science concerns understanding and examining scientific concepts used by scientists. Very rarely, it seems to me, the way scientists use common words accords with the common every day usage. Sometimes, they will use highly technical terms. Without being informed by what science is actually doing, it is very hard to work with the concepts of science in any way worth the effort.

2. Philosophy in science should use scientific evidence in the way other philosophers rely on intuitions. Wherever possible we should support our speculations with cases from the scientific literature.

On this second point, I want to stress that intuitions are a valuable heuristic in both philosophy and science. They are the creative spark. They are our quick grasping of complex issues. They are also very useful primes for making the case for something. We pump people’s intuitions with thought experiment to help them see the world in ways they may not have considered before.

If we want to convince someone of something new, we have to meet them were they are. We have to bridge the gap by getting them to agree to the new thing via something they already believe. This also applies reflexively on ourselves. We intuit new ways of thinking by navigating our own preconceptions. Our intuitions build bridges to the new territory we want to explore.

Yet, this cannot be all we do. Intuitions should be informed by evidence outside of ourselves. Especially evidence that challenges our intuitions. Our own predilections cannot be the base on which we try to explain the world.

Of course, evidence can come from many sources. It just happens the best factory of producing evidence which withstands scrutiny is the scientific enterprise. It is not perfect and it is far from infallible. Yet, it is much preferable to armchair philosophising on its own. As importantly, when argument from intuitions lead to an impasse, only evidence can break the impasse. And if no evidence can break the impasse, then whatever that project is, it is not the one that interests me. Intuitions alone cannot guide me to know the world as it is. I cannot vouch for anybody else having that flawless grasp either.

Such is my own reflection on the kind of naturalism I value in philosophy. The kind that works with science, but is not trying to be science. The kind of naturalism that restricts its speculations and intuitive leaps to what the evidence can support. One that tentatively accepts sweeping metaphysical positions only worriedly. A great worry that they go beyond the evidentiary base. We might have to accept some shaky foundations to make progress. If are to do anything, we must take certain assumptions on trust. We can only climb up with the ladder. But to reach the top we may need to ditch the ladder. If we hold on to our metaphysical foundations too tightly, I suspect we will only hobble ourselves. We will put a ceiling to how far we can climb.

In short, the kind of methodological naturalism I describe has ambitions to explain and make sense of complex topics such as the mind. It can only do so by conducting the enquiry in the nexus between sciences and philosophy. Assumptions are unavoidable. But if we are to get closer to understanding, we must hold on to them lightly.