Science and Religion series -- "What is science?"
In this second essay in our Science and Religion series, Fr Jonathan Jong offers a view of science, not just as a collection of facts or theories, but as a way of observing and interpreting the world. This way, he considers how science is well-equipped to answer some questions, but not others, including questions about crucial concepts in science itself.
See also the first essay in this series here.
In the first post in this series, I tried to say something about God and what it means for God to be the creator of the world, and why it’s a mistake to think that the doctrine of creation could be contradicted by some scientific theory or the other. I would like now to try to say something about the scientific enterprise, and how it is related to what philosophers and theologians do.
[NB: There is going to be very, very little theology in this post.]
Judging by how science is often taught in schools, it would seem that we think of science as a body of knowledge: in classrooms and textbooks, we learn facts about the natural world and about the laws that govern them. Judging by media coverage of science, it is also quite common to think of science as a source of authority: the phrases “According to science…” and “Science says…” are ubiquitous. What science is meant to be an authority on is, of course, the aforementioned facts about the natural world. So these two common ways of thinking about science are not unrelated.
These are not altogether terrible ways to think about science. It’s true that science deals in facts, though what this means is a slightly complicated affair; more on which later.. And it’s true that science is an appropriate authority to adjudicate upon questions about all sorts of things. But these approaches also provide a misleadingly static and bounded picture of science. Misleadingly bounded, because there is no actual list of Things-that-Science-has-Discovered for us to consult. Misleadingly static, because it leaves out the process of discovery that arguably sets science apart from other efforts to know the world.
Science: a history of observation
Science has a history. As it turns out, it’s a history it shares with philosophy and theology, which we now consider to be totally different disciplines, done by different people, in different university faculties. From the time of Aristotle until the 19th century, what we now called science was called “natural philosophy”: it was the branch of philosophy that was concerned with the natural world. Long before the Ernest Rutherford published his paper on the Geiger-Marsden experiments, Democritus in the 7th century BCE argued that everything was made of much smaller things, which he called “atoms”. Long before Darwin published the Origin of Species, Anaximander in the 6th century BCE argued that fish-like life initially appeared in the water, with humans coming later when dry land eventually emerged.
Democritus’s atomism isn’t quantum physics; indeed, Rutherword's work was the beginning of the end of his idea that atoms were indivisible. Nor is Anaximander’s story about the origin of species the modern synthesis of evolutionary theory; he thought fish gave birth to humans. Our ideas about the world have changed, even if some salient features of current thinking can be detected elsewhere in the history of philosophy. But we miss something if we focus too narrowly on the history of ideas. The great genius of modern science is not really that it enhances our creative ability to speculate about the world: a cursory reading of Plato’s Timaeus should suffice to disabuse us of that prejudice. For my money, the most interesting thing about science—modern science, as it has developed since the 15th and 16th centuries—lies in its powers of observation.
When the ancient Greeks were arguing over the anatomical location of the intellect—the two candidates being the brain and the heart—they already had at their disposal the ability and tendency to dissect animals. Alcmaeon of Croton, on the brain side, referred to poroi, which he found connected sensory organs to the brain. Aristotle, on the heart side, believed that thoughts were passed through the blood, and observed that the brain was relatively bloodless. Galen, in the first century CE, puts the debate to rest in part by vivisecting pigs: cruel, but effective. Now, we can implant electrodes directly onto the brains of living humans to measure and even manipulate neuronal activity: creepy, but illuminating.
Let’s take a less gruesome example. Aristotle believed that the Earth was a very large sphere. He believed this because he observed that, during a lunar eclipse, the Earth casts a curved shadow on the Moon; furthermore, as one travels between North and South, the stars change their locations in the sky. Nearly 2,000 years later, astronomy was still done via observations by the naked eye: Tycho Brahe’s celebrated tracking of the celestial movements was done without the benefit of a telescope, though he did have other instruments like quadrants and sextants. A decade after Brahe’s death, Galileo points his spyglass to the heavens for the first time, and observes things that no one was able to see before: the irregular surface of the Moon, the moons of Jupiter, the composition of the Milky Way. Even then, Galileo’s telescope was a crude thing: only 10% of the devices he constructed worked well enough to make these observations, and even then they were blurry and distorted. In 1990, NASA sent the Hubble Space Telescope into orbit, which has helped us see things over 10 billion lightyears away, to the beginning of the Universe. (Incidentally, in its early days, Hubble also produced distorted images because its main mirror was faulty, but this was soon fixed.) Besides optical telescopes, we also have radio telescopes that do not really have lenses or mirrors at all: they look like massive satellite dishes pointing to the sky, not only magnifying far away objects, but also detecting all manner of invisible things that even the Hubble Telescope cannot see.
Theory as the handmaid of observation
Now, let’s bring theories back in. Our ability to observe the world around us is not just a matter of constructing fancy gadgets. Neither Alcmaeon nor Aristotle had the concept of nerves, and so even if they could see them, they would not have understood what they were. When Robert Wilson and Arno Penzias first picked up signals of the big bang from their radio telescope, they thought that they were the result of bird droppings messing up their antenna. Our ideas about the world and our observations of the world interact intimately. So, the narrative of science as a source of information about the world—whether in terms of facts or theories—is itself made richer by this shift on emphasis on the role of observation in science.
Theories affect what we see in another way too. I mentioned earlier that Galileo’s telescope produced distorted and blurry images. Why, then, should his contemporaries have trusted these instruments? Indeed, there were those who didn’t (see Note 1, for an irrelevant aside). One obvious route to justifying trust in telescopes is to check that they can magnify other objects that we can see via other means: we know telescopes work well enough to see faraway ships, so it’s just an extension of this knowledge to infer that it can help us see faraway planets. The other way to argue for the veridicality of the telescope is by positing a theory for how a telescope works: this would be a theory about optics, and in particular in the case of Galileo’s telescope, a theory about the refraction of light. Similarly, to return to our neuroscientific example, we might want to argue for the validity of fMRI technology, and if so, we would have to appeal to theories about the physics of electromagnetism and the oxygenation of blood.
All of this is to say that scientific facts are themselves the products of observations and theories, including theories about observations. Particular sciences—astrophysics, neuroscience—have to go beyond themselves to justify their methods of observation and to interpret the deliverances of those methods. But it is not always to other particular sciences that they turn. Scientists also rely on theories that are not themselves based on empirical evidence at all: they rely on mathematics, theories of which are not verified or falsified on the basis of observations as such, but on logical grounds. Newton may have invented the calculus to serve his physical research, but proving the fundamental theorems of the calculus does not involve going out into the world and collecting data. Similar things can be said of the statistical techniques applied to the biological and psychological sciences.
On laws of nature, causes, and truth
Technological, theoretical, and mathematical instruments are not the only tools a scientist requires. There are also conceptual tools, which are just about as abstract as mathematical ones. Consider, for example, the concept of a law of nature. What is a law of nature? Some people think of laws of nature as causal entities: which is why we say, for example, that the apple falls to the tree because of the law of gravity. On this view, a law of nature is a kind of thing, albeit an abstract thing. (Speaking of mathematics, some people also think that numbers are abstract entities: that is, they are objects that exist, even through they are not physical objects.) Other people deny that laws of nature are things in the world: instead, they are convenient fictions that exist in our minds only. We say that the apple falls because of the law of gravity, but this is just a manner of speaking, perhaps to express the idea that apples always fall from trees.
Speaking of causes, this too is a concept often deployed in science. But what is a cause? Some people say that A is the cause of B if and only if B would not have happened if A hadn’t happened. Others think that this is too strict, so they come up with more probabilistic accounts of causation. Some people say that for A to be the cause of B, A has to precede B in time. Others deny that causality requires temporality. Arguments over theories of causation typically proceed by thinking about whether the account in question is logically coherent, and also whether there are uncontroversial counterexamples to the account: that is, whether it leaves out cases that really do seem to involve causation and/or it includes cases that really don’t. These debates are on-going: there is no clear answer to the question of which, if any, of the existing theories is true.
Speaking of truth, what—to quote Pontius Pilate—is truth? Most scientists are really very interested in whether their theories are true. But what does it mean for Theory A to be “more true” or “closer to the truth” than Theory B? One plausible idea—which we owe to Karl Popper—is that Theory A is more true than Theory B if and only if it entails more true and fewer false propositions than does Theory B. (Or, put another way, if Theory A makes more true predictions and fewer false predictions than Theory B.) This view also allows for both theories to be true (i.e., to entail no false propositions) and yet differ in their truthlikeness if one theory entails more true propositions than the other.
But things take a worrying turn when we try to consider two false theories, and ask if one can be closer to the truth than the other. This is very important in science because the history of scientific theories is basically the history of false theories (i.e., they entail at least one false proposition) that, we hope, are nevertheless steps toward the truth. So let’s consider two theories, A and B, both of which are false; and let’s say that they both do entail some true propositions, but A entails more of them than B does. This means that there will be some true propositions that are entailed by A but not B: let’s call them ta. We have already said that both theories are false, and so even A must entails some false propositions: let’s call any false proposition f if it is entailed by A. Now, since A entails both ta and f, then the complex proposition “ta and f” is also entailed by A (see Note 2), and “ta and f” is obviously false. This means that for every uniquely true proposition entailed by A, false propositions are also entailed by the conjunction of each ta and each f. Furthermore, “ta and f” is not entailed by B, even if B does entail f: the implication of this is that A cannot actually entail more true propositions without also entailing more false ones than B. Which is just to say that there cannot be two false theories, one of which entails more true propositions and fewer false propositions than the other. Which, in turn, is to say that our plausible definition of “closeness to the truth” turns out to be defective on purely logical grounds (see Note 3).
Don’t worry too much if you glazed over the last few paragraphs. My point is just that, besides machines and mathematics, science also relies on concepts—e.g., laws of nature, causality, truth—that are not themselves verifiable or falsifiable on the basis of observations. The answer to questions like the ones we have just raised are not to be obtained by running more experiments. It is the powers of our intellect and imagination that are required, not our powers of observation. Now, some scientists dismiss these questions as foolish or pointless. But given that these are questions about central concepts in science, this seems like a belligerently anti-intellectual move. What is the alternative, after all? Should we just pretend that we know what a law of nature is, what causation is, what truth is? Or should we keep talking about these things even though we literally have no idea what we are talking about when we do so?
On metaphysics (and theology)
Finally, then, what does all this have to do with philosophy and even theology? If there is something special about science, it is that scientists have developed very sophisticated and powerful methods of observing things. But not all of the important questions that human beings have are questions that can be adjudicated by observation. As we have seen, not even all the important questions to science are themselves “scientific questions” or, better, empirical questions; and we did not even cover normative questions like “Should we do science at all?”. What would an answer to that question look like? A social scientist might say, “Yes, because—according to the data—science makes people healthier and happier”. But why are health and happiness good things, such that we want to increase them? That does not seem like an empirical question at all, amenable to scientific investigation. And yet, it would simply be prejudicial to dismiss it as meaningless.
The branch of philosophy that deals with the natural world was once called natural philosophy, and is now called science. Questions like those about what laws of nature, causation, and truth are belongs to metaphysics, from the name given to one of Aristotle’s major works. Questions about God also belong to metaphysics, though they do not necessarily begin as questions about God per se. Rather, they might begin as questions about the world itself, chiefly: why does it exist? Now we are in familiar territory, from the first essay in this series. The answer to this question—or, at least, the answer to this question that Christians have ended up with, and which we also share with Jews and Muslims—involves a necessary being that is radically unlike any created thing. God is not like a brain or the moons of Jupiter: God cannot be vivisected or seem through a telescope, no matter how powerful. And if that’s how science knows things, then so much the worse for science.
There is, of course, more to Christian theology than the metaphysics of theism. Just as methodological questions arise in science, they also arise in theology. We might know why telescopes should be trusted. By why should biblical texts be trusted? And just as epistemological questions arise in science, they do too for theology. As we have seen, it is difficult to work out what it means for a scientific theory to be true: what does it mean for a theological theory—of the atonement, say; or of the eucharist—to be true? And so forth. Theology and science—and, for that matter, philosophy—turn out not to be so different, really. They are each attempts to say as many true things and as few false things as possible about the world: or, better, we are constantly doing science and philosophy and theology when we probe the world with our powers of observation and reason and imagination, asking and answering questions about Jupiter's moons or the morality of meat-eating or the goodness of God. These are, I'm sure we are aware, not the same kinds of questions that require the same tools. I suppose this essay is just a long-winded way of saying that.
Note 1: It is often said that some people refused to look into Galileo's telescope. There are only two named culprits, both professors of Aristotelian Philosophy. Cesare Cremonini may actually have had experience with the telescope (“to observe through those glasses gives me a headache”, he said), but it’s unclear if he ever tried to replicate Galileo’s findings. Guilio Libri probably did turn Galileo’s offer down: our only source for this is Galileo’s letter to Kepler, in which he mocks Libri who had recently died (just months after Galileo’s findings were first published) saying “perhaps he’ll see [the moons of Jupiter] on the way to heaven”. In contrast, Kepler leapt to Galileo's support without ever having peered into a telescope himself. Christopher Clavius’s—arguably the Church’s chief scientific authority at the time—was sceptical at first, because Galileo's findings contradicted the prevailing theory of the day, but changed his mind after he and other scientists at the Roman College confirmed Galileo’s findings for themselves. On this note, the Jesuits’ most senior astronomer was a better scientist than Kepler, and indeed, many of us today.
Note 2: Consider a weather app that predicts both precipitation and temperature. Say it predicts that tomorrow it will rain and also that it will be 30º C. And say that tomorrow, it does indeed rain, but the temperature turns out to be −5º C. If so, then it would have correctly predicted that “It will rain tomorrow” and wrong that “It will be 30º C”. But it will also be wrong that “It will rain and it will be 30º C tomorrow”.
Note 3: This argument was originally proved by two different philosophers, Pavel Tichý and David Miller in 1974. There is actually another side to it that begins with the premise that B entails more false propositions than A, and shows that this cannot be so without B also entailing more true propositions than A. I have left out this side of the argument, partly for brevity, but partly also because it relies on the counterintuitive logical rule that conditionals with false antecedents are always true.