‘The philosophy of science is as useful to scientists as ornithology is to birds.’ – Richard Feynman
So why should you study the philosophy of science, then? In particular, if you’re a scientist then what is to gain from all this metatheory? How does it impact upon your day to day practice of research and theorising and so-forth?
Well, I can’t speak for you but I can speak a little bit about why I have found it helpful. Mind-altering, even.
So normally, most people think of science as trying to make valid causal inferences: the search for cause and effect. This is thought to occur via something called The Scientific Method whereby the scientist proposes a hypothesis and runs an experiment to see if the hypothesis ‘comes true’. If it doesn’t come true, the hypothesis is falsified and discarded. If the hypothesis seems really robust, we can start to call it a theory or even a law of nature. Even if one’s attitude isn’t quite so narrow as to what constitutes ‘science’, most still view prediction by induction to be the raison d’etre of scientific enquiry, where the subject of that enquiry are a reality that we can approximate in our theories and which exists in more or less the same form regardless of our ideas about it (eg when the tree falls in the forest and nobody is around, it sill makes a sound).
I am a social scientist. This means that I study what people do as part of society, and what sort of things characterise and constitute that society. It follows from the above definition of science that I should be looking for the causes and effects of social activity, and for general patterns of the type ‘whenever A happens, B follows‘. But people are really, really complex, and the methodologies available for studying them – those rules of analysis and problems or questions of interest that together define a research agenda – vary widely between fields and within them. Should I be trying to falsify hypotheses? Should I be using economic models? Should I be, I don’t know, just asking people to tell me why they do the things they do?
Let me give you an example of one of the puzzles that social scientists face; something that gives us cause to wonder if many of our common explanations involve some deeply paradoxical or counterintuitive assumptions. Consider the notion of social structures: if social structures affect our actions, but social structures only exist because of the actions of individual people, then doesn’t that mean that we’re both cause and effect? And if structures are more than the sum of their parts – us – how is that metaphysically possible? Isn’t that like saying 1+1=3? And if structures are nothing more than us, and can’t influence us, then what do they do? I’m not actually going to try to answer these puzzles – though I do have some ideas about them – but I do think they give you just a small glimpse into some of the conceptual difficulties that social scientists must face.
There is a critical immediacy to discussions on how to conceptualise notions like truth, causality, observation, explanation, and scientific progress when it comes to social science that don’t seem as pressing in the so-called ‘natural’ or ‘experimental’ sciences. That doesn’t mean they don’t matter deep down, but there doesn’t seem to be as much choice between radically different methods and assumptions for studying the same general thing. Some philosophers of science have suggested that the mark of a ‘mature’ science is this sort of methodological uniformity, but if that were true, then it would be grim news for social scientists, because I have the distinct sense that we’ll never get there. Or at least, we’ll get there only by radically changing the way we talk about people, in potentially impractical ways.
I think that a basic understanding of the philosophy of science – which I’ve already said is particularly important in social science – should consist of an understanding of the various debates on the following (interconnected!) issues:
- What are facts and what is a true statement?
- How do we find facts or determine whether a statement is true?
- What is the link between observation, theory, evidence, and truth? What gives us warrant to assert a claim?
- What is an explanation?
- How do we develop more ‘knowledge’ or make progress on our understanding and explaining?
- What does it mean to ‘do science’?
These are hard questions. Very hard questions. Of course, we will need to take up certain substantive positions on them if we are going to support our particular choice of methodologies, and of course I have my own opinions on them, but the only way to take a reasonable position is to know the basics of the various options, and to have some idea of where those options take you.
I won’t go into many specifics of who to read and why – at least not extensively – but some excellent starting choices are Alexander Rosenberg‘s concise and cogent introductions to science and social science, Larry Laudan’s pragmatic theory of scientific progress, Laudan’s ‘Confutation of Convergent Realism‘ (for the slightly more advanced), and Peter Winch‘s short but powerful ‘The Idea of a Social Science‘ (note: link is to an ebook and its only 160 pages!). I would highly recommend to anyone else in International Relations and political science to at the very least read Patrick Thaddeus Jackson’s book on how philosophy of science pertains to the field, though Jackson’s book would be a good start even for sociologists, anthropologists, critical theorists, etc simply for its breadth. Or just go through the readings in an introductory syllabus and see where they take you.
If you’re looking for some key insights right away, let me share a few with you:
- There are more ways to do science than to make hypotheses and test them by seeing if their predictions come true. While experiments of this nature can give us powerful ways of controlling the world, they may not give us the right kind of answer to other sorts of questions. For example, instead of determining which independent variables are stronger ’causes’ than others for ‘dependent variable’ effects, an event or outcome may be best understood as occurring when a whole set of factors come together in just the right way as to produce an effect. In other words, rather than a set of possible causes to find the real ones, the scientist studies one situation that they know was the real cause, and looks to see how it created the effect – in short, a search for causal mechanisms.
- The above alternative to the hypothetico-deductive method – ie experiments and hypothesis testing – is particularly pertinent to looking at very complex things, such as society or certain kinds of biology. That is because there are many causal factors, and it is very hard to know just precisely how they came together. Some of them might only need to be present, but others might do different things depending on their quantities and ratios. In a sense, some causal structures simply make something possible while others are part of a sequence in the process that leads to it happening in the way that it did.
- In the social sciences, quite often the sort of ‘explanation’ we are looking for is an answer to ‘why did X find it reasonable or appropriate to do what they did?’ This kind of explanation requires us to try to understand X’s situation in the same way X did, and look for the rules of behaviour or thinking that influenced X’s judgement. Often we can only really get an answer by deep, narrative study of X and X’s relation to their environment.
- There are powerful challenges to the notion that we could ever describe a kind of objective reality, not just in terms of cognitive biases but in principle. These challenges come from the nature of language (ebook), of consciousness, and of causality and metaphysics itself. They are worth studying in detail, because deciding how to deal with them is necessary to produce a coherent methodology. One way out is to be entirely instrumentalist, but this escape comes at great cost.
Even though my main area of study is International Relations and political sociology, I have found that studying the philosophy of (social) science has helped me think about how to view things like ‘states’ or ‘politics’ or ‘rhetoric’ or ‘power’ in hugely relevant ways, and to figure out what I can contribute in studying them.