Philosophy of science deals largely with two general topics: Metaphysics and Epistemology. These are general topics of philosophy, and in the philosophy of science they deal only with the metaphysics and epistemology of science. So there are no overarching debates about how you can tell if you’re dreaming, or whether we are all brains in a matrix-style vat. But there are local issues, as it were, that reflect these general concerns of philosophers.
Metaphysics covers many things
Metaphysics is a hard field to define. It is named after the book of Aristotle, which followed on from the discussion of the physics of his day and formulation. It has a number of traditional topics that get discussed in various degrees of abstraction. Metaphysics in science covers a number of issues that affect the nature of scientific explanations. For example
Causation: What is it to say that one thing causes another? Are there only ever one cause for each event, or many? Is causation due to the operation of laws of nature, or it it just regularity in events? Is a cause a necessary process? Are there uncaused events, and so on (for example, quantum foam, or mutations)?
Causation is tied closely to question of what good explanations are, which gets covered under the epistemology heading.
Realism: Another widely debated question is whether or not the entities that play a role in theories of science are real. The classical example is electrons – one philosophy has famously said, if you can spray them they are real, but is this the case? More relevant to biology are questions about the reality of things like species, ecosystems, communities, and even organisms.
Those who are scientific realists think that science, or at any rate an idealised final science that we may or may not approach with our best theories, describes the way the world is, while others think that all we have are our theories, and that if they work, that is the best we can hope for. These folk are often referred to as “internal realists”. Some claim we can never get outside our theories of the world, so the question whether they are right doesn’t even enter into it. The most extreme, often called social constructionists, believe that science is “just” a tale we tell ourselves, and some even think it isn’t all that important and that other tales will do just as well. Recently, one of these social constructionists defended creationism, in the Dover trial in America. Those who think science is a way of gathering knowledge that outshines all others have little time for this sort of approach, but given that science really is a human activity, and that we are fallible and have no magic method for ascertaining the state of the world, we must at least consider why it is we think science is the best. And that leads us to…
Ontology: this is the study of what is, what things are. One influential view is that the ontology of our best theories, the furniture of the world or domain the theories explain, gives us our ontology, while others ask whether or not our “folk ontologies” might have some relevance to science. For instance, a famous example is the “two table” metaphor of James Jeans early in the 20thC. Jeans claimed there are two kinds of table – the one we see, eat off, and bump into, and the one that physics teaches us exists, which is largely empty space, constituted by particles that are waves, constantly in dynamic flux. Science, he said, undercuts our intuitions, and it is not “organised common sense” as the prior traditions had thought.
In biology, we might ask if the folk psychology of every day life is really accounted for by neurology or evolutionary and physiological research into the emotions, or if organisms are single entities or metacommunities of bacteria, parasites and commensuals as well as the obvious organisms. A book written in the 1980s by Leo Buss pointed out that it is very hard to identify what the individual organism is when you are dealing with colonial organisms like slime molds and jellyfish, or with vegetatively propagating stands of trees and fungi.
Many terms used in biology are either hard to define or are used in various ways that are not entirely consistent, like “species”. Given that often these terms refer to what are supposed to be the “units” of the theory or field of study, it is bothersome to have this ambiguity. Or is it? Perhaps this ambiguity is useful in allowing scientists to play with their concepts and develop more exact ideas without being trapped by a single clear definition. Or it might be that the world really is messy. But often ambiguities like this can lead to false inferences, because we use incommensurable kinds in our comparisons. Is it really the case that a species of microbe living in a patch of soil is equivalent in terms of biodiversity to a large bodied mammal? Can we compare the rate of extinction of species in the paleontological record with rates of extinction today? A study by Sepkoski in the 1970s estimated past extinction pulses in terms of families – are families real things, or are they artifacts of classification, in which case aggregating all the extinction rates is to mix apples and oranges, or worse?
One of the main tasks of the philosophy of science is to consider these questions. They aren’t empirically determinable, often. They are conceptual questions.
Epistemology (how to know): The other general topic in philosophy of science is epistemology. This is nothing to do with the over-consumption of alcohol at the Friday night get-together. From the Greek word “episteme” (to believe), it is the study of how we come by our best beliefs about things, or in other words, how we know what we know, if anything. In the Sepkoski case, we might ask if we even can come up with a reliable estimate of extinction pulses in the past, and whether they can be compared to current extinction rates. This is the area in which we ask what science is; whether there is a scientific method, or many scientific methods, and so on. It is particularly important when trying to isolate science from the many pseudo-sciences that are out there and competing for attention from the media, governments, and educators. I’d like to briefly cover a few of the major issues of scientific epistemology.
An obvious concern is the nature of data. We sometimes tend to treat data in the way the original sense of the word, as “givens”, but every scientist knows that you have to work hard to get good data. And this raises the spectre of data being “constructed” on the basis of our theoretical interests, or even our political and economic interests. Consider how the data that supported smoking were gathered, or how pharmaceutical data are sometimes accrued and used. One influential view is that our theories determine what we measure and so determine what the data are that the theories are to explain.
Then there is the question of accuracy. Any statistics course will cover how one can justifiably eliminate outliers that are clearly (or not) artifacts of the experiment. Why are these justified? And if you get a too-clean data set, like Mendel did, are you faking your data, or are you just lucky? How to spot fraud is an ever-present problem in science, but there is a deeper question whether your best measurements are accurate.
Sometimes data is used just because it’s the best thing we can get at present. We may have an assay that works, or a technique that is authoritative because of the weight of tradition. An example might be the way microbial species are delimited – the “gold standard” for a long time was 70% DNA hybridisation, but recently we have moved on to 98% or better similarity in a few chosen genes – either cytochrome C or 16s sRNA. How well do they represent the things themselves? Perhaps they are used simply because nothing better is about, and everybody else uses them too. The recent “species barcoding” proposal is hotly debated because there is no guarantee that variation in the genes used match actual diversity of taxa.
But the big issue in philosophy of science is explanation. What sort of process is explanation? If you have a model that happens to fit the data, have you explained the data? How? Some think that to explain a phenomenon you need to give the laws that govern the phenomena, while others hold instead that it is enough to give a causal sequence or story. This goes strongly to the matter of what a good theory is, which I’ll get back to in a minute. Another major topic is whether or not to give a good explanation we ought to reduce one domain (say, ecology) to another (physics) using the laws of physics, or whether biology has its own laws and objects of study that are not just physical. We often hear, for instance, of “emergent properties” in computational and other biology. This was first proposed by philosophers to attend to evolved properties like mind or growth. If the properties of the molecules that make up living systems do not have, for instance, the informational character of genes, but rather only genes in a living cell have them, that is an emergent property. Should biologists remain content with describing these properties, or must they uncover the ways the physical properties cause them? Scientists seem to have voted with their feet on this question, through the molecular biology research program, and in other areas by the use of thermodynamics in ecological systems, but some argue that biology has its own laws. If so, what are they? Laws in physics are supposed to be without exception, while the Harvard Law – under carefully controlled conditions organisms do what they damned well please – suggests that biology is not like physics in this regard.
I won’t go through all the other major topics – this is getting like a shopping list as it stands. But one thing ought to be noted – what makes a good theory, and is there a discovery method for making them? We are all taught at some point that there is a Scientific Method, with the capital letters, and when we try to defend science against, say homeopathy or astrology or creationism we appeal to this method and the lack of it in those disciplines. But extensive study of science by philosophers and historians has indicated that in fact there is no such beast. There are local methods used, in one discipline, but it is all too easy to find good science using a different set in another, or at another time.
What makes science science? Is it just, as one cynical biochemist friend tells me, what scientists do (in which case I think beer plays a major role in the scientific method)? This lack of definitional criteria seems to undercut the privilege science enjoys in our epistemology. Is it, as the sociological critics argue, that science is just another political and social game, on a par with ritual magic or religion? What is special about science? To answer this, I will present the views of David Hull, a philosopher of biology who I think has given a good, if incomplete, account of what science does and why it works.