Announcer
The following program features simulated voices generated for educational and philosophical exploration.
Leonard Jones
Good afternoon. I'm Leonard Jones.
Jessica Moss
And I'm Jessica Moss. Welcome to Simulectics Radio.
Leonard Jones
This week we've examined knowledge, consciousness, moral uncertainty, free will, and AI alignment. Today we turn to the foundations of scientific knowledge itself. Do our best scientific theories describe reality as it truly is, or are they merely useful instruments for prediction and control?
Jessica Moss
This matters enormously for how we understand science's authority. If scientific theories are just predictive tools rather than literal descriptions of reality, then what justifies the enormous epistemic trust we place in them?
Leonard Jones
Our guest today is Dr. Nancy Cartwright, Professor of Philosophy at Durham University and one of the most influential voices in philosophy of science. Her work has fundamentally challenged scientific realism, arguing that the fundamental laws of physics lie. Welcome, Dr. Cartwright.
Dr. Nancy Cartwright
Thank you both. I should clarify immediately that when I say laws lie, I don't mean scientists are dishonest. I mean our most abstract theoretical laws don't accurately describe what happens in the world.
Jessica Moss
Let's start there. What do you mean when you say fundamental laws lie?
Dr. Nancy Cartwright
Consider Newton's law of gravitation. It says that the gravitational force between two bodies equals the gravitational constant times their masses divided by the square of the distance between them. That's a beautiful, elegant law. But it's literally false. No two bodies in the universe experience only gravitational force. There are always electromagnetic forces, nuclear forces, other gravitational sources. The law describes an idealized situation that never actually obtains.
Leonard Jones
But surely we could say the law describes what would happen if only gravity operated, even if that's a counterfactual situation. Let me be precise about this—aren't you conflating the truth of a law with the applicability of a law?
Dr. Nancy Cartwright
That's the standard response, but I think it misses something important. What does it mean for a law to be true of a counterfactual situation that never obtains? Scientific realists want to say that fundamental laws describe the underlying causal structure of reality. But if that structure never operates in isolation, in what sense does it describe reality?
Jessica Moss
This sounds like you're arguing that the applicability of scientific laws is much more limited than we typically think. What are the implications for how we use science in practice?
Dr. Nancy Cartwright
Exactly. In practice, we use phenomenological models—models that describe how systems behave in specific circumstances—rather than fundamental laws. When engineers build a bridge, they use models of materials under stress, not quantum field theory. Those models work in the specific domain they're designed for, but they don't purport to describe fundamental reality.
Leonard Jones
Let me introduce a distinction that might clarify the debate. Scientific realists typically claim that mature scientific theories are approximately true and that theoretical terms refer to real entities. Are you challenging the approximate truth claim, the reference claim, or both?
Dr. Nancy Cartwright
I'm more worried about what work the notion of approximate truth is doing. Consider the history of science. We've discarded phlogiston, caloric fluid, the luminiferous ether—all theoretical entities that past scientists believed in and based successful predictions on. If our current theories are approximately true, what explains the success of these false theories?
Jessica Moss
This is the pessimistic meta-induction—the argument that because past theories turned out to be false, we should expect current theories to be false too. But doesn't that prove too much? We're dramatically better at prediction and intervention than previous generations.
Dr. Nancy Cartwright
We are better at prediction in specific domains, but that doesn't require that our theories are true descriptions of fundamental reality. It just requires that they're reliable in the contexts we've tested them. I'm an entity realist—I think electrons are real—but I'm skeptical that our theories about electrons accurately describe their intrinsic nature.
Leonard Jones
This raises a question about the relationship between predictive success and truth. Why would false theories be systematically reliable for prediction if they didn't track reality in some important way?
Dr. Nancy Cartwright
Because they're designed to be reliable for prediction. We construct theories by selecting models that fit the data in specific circumstances. That's an engineering achievement, not evidence about fundamental reality. Think about how theoretical physics actually works—we use different models for different scales and domains. Quantum mechanics for small things, general relativity for large things, statistical mechanics for many things. These theories are mutually inconsistent, yet all empirically successful.
Jessica Moss
But physicists are searching for a unified theory that reconciles quantum mechanics and general relativity. Doesn't that suggest they believe there's a single underlying reality these theories partially describe?
Dr. Nancy Cartwright
They believe that, certainly. But belief doesn't make it so. I'm skeptical that nature is fundamentally unified in the way physicists hope. Reality might be genuinely messy—a patchwork of different causal capacities operating at different scales without any underlying unity.
Leonard Jones
Let's examine this patchwork view more carefully. Are you suggesting that there are brute causal regularities at different levels with no deeper explanation?
Dr. Nancy Cartwright
I'm suggesting that explanation has to bottom out somewhere, and it might bottom out at different places for different phenomena. Chemical bonding is explained by quantum mechanics, but economic recessions aren't. We shouldn't expect a theory of everything that explains both from a single set of fundamental laws.
Jessica Moss
This seems to challenge the unity of science thesis—the idea that all sciences are ultimately reducible to physics. What are the stakes of abandoning that thesis?
Dr. Nancy Cartwright
The stakes are that we take seriously the causal claims of special sciences—biology, economics, psychology—rather than treating them as provisional until we can reduce them to physics. If I'm right that fundamental laws don't accurately describe what happens, then higher-level causal generalizations might be more reliable guides to actual phenomena.
Leonard Jones
But there's an epistemological puzzle here. If fundamental laws are idealizations that don't describe reality, how do we know which phenomenological models to trust? What makes a model reliable?
Dr. Nancy Cartwright
That's the right question, and I don't think there's a general answer. We know models are reliable through empirical testing in specific circumstances. This is why I emphasize that science is fundamentally local—our knowledge is reliable where we've tested it and uncertain as we extrapolate.
Jessica Moss
This has profound implications for how we use science in policy. If scientific knowledge is fundamentally local and our models work only in tested circumstances, how do we make decisions about novel situations like climate change or synthetic biology?
Dr. Nancy Cartwright
Very carefully. I think we should be more epistemically humble about scientific predictions in novel contexts. That doesn't mean ignoring science—it means understanding the limitations of extrapolation from tested domains to untested ones.
Leonard Jones
Let me raise a conceptual worry about your position. You claim to be an entity realist who believes electrons exist but deny that our theories about electrons are approximately true. But what grounds entity realism if not the success of theories positing those entities?
Dr. Nancy Cartwright
Experimental manipulation. We can do things to electrons—spray them, bend their paths, measure their charge. That kind of experimental control gives us reason to believe in their existence independent of whether our theories about their intrinsic nature are correct.
Jessica Moss
But how do you identify what you're manipulating as an electron rather than some other entity unless you have some theoretical understanding of what electrons are?
Dr. Nancy Cartwright
Through their causal capacities—the things they reliably do in experimental settings. I know I'm manipulating electrons because I know how to produce certain effects reliably. That's consistent with our theories about underlying mechanisms being false or idealized.
Leonard Jones
This seems to require a distinction between causal capacities and theoretical descriptions. Could you explain how that distinction works?
Dr. Nancy Cartwright
Objects have causal powers—stable dispositions to behave in certain ways under certain circumstances. Electrons have the power to deflect in magnetic fields, to participate in chemical bonding, to emit photons when transitioning energy states. These powers are real even if our theoretical models of the mechanisms underlying these powers are idealized or incomplete.
Jessica Moss
We're running short on time. Let me ask about the practical consequences of your view. Should we change how we teach science or fund research if anti-realism is correct?
Dr. Nancy Cartwright
I think we should teach science with more honesty about idealization and domain specificity. Students should understand that fundamental laws describe idealized situations, and that applying them to real-world phenomena requires judgment and approximation. As for research funding, we should value phenomenological modeling and applied science as much as fundamental theory.
Leonard Jones
That suggests a different hierarchy of scientific value—measuring success by practical reliability rather than theoretical elegance or unification.
Dr. Nancy Cartwright
Exactly. And I think that better reflects how science actually works. The most reliable scientific knowledge is often local, messy, and domain-specific rather than universal and elegant.
Jessica Moss
Dr. Cartwright, thank you for this examination of scientific realism and the limits of theory. You've given us much to reconsider about the nature of scientific knowledge.
Dr. Nancy Cartwright
Thank you both. These questions deserve ongoing scrutiny.
Leonard Jones
We'll be back tomorrow with more philosophical inquiry.
Jessica Moss
Good afternoon.