Mariannetalbotdotcom.files.wordpress.com



Lecture OneCausation is hugely important to human beings: it is by appeal to causation that we:explain events: b happened because a happened (because a causes b)predict events: if a happens b will happen (because a causes b)manipulate events (if I bring about an a, then I will bring about a b (because a causes b) Certainly causation is of importance to philosophy. We have causal theories of reference, perception, knowledge, content and numerous other things. If it were to turn out that causation doesn’t exist, we would be in serious trouble!Causation is so important that it has been said that: “With regard to our total conceptual apparatus, causation is the centre of the centre”, and called ‘the cement of the universe’. How Do Philosophers Think About Causation?Philosophers are interested in the nature of causation, and in its existence. They want to know what causation is and whether it exists. The first question is, of necessity, prior to the second. We are not in a position even to ask whether something exists unless we know what it is whose existence we are questioning. So as a first step in our investigation of the nature of causation philosophers engage in ‘conceptual analysis’. They analyse our concept of causation, asking how it works in our reasoning, and what this implies about causation itself.Let’s try a bit of this. Here is a thought experiment: If the Queen had watered my tomato plants they would not have died. If my tomato plants are dead through lack of water it is true, isn’t it, that had the Queen watered them they would not be dead? Is the Queen’s not watering my tomato plants a cause of their death? If not why not?The intuitions triggered by such thought experiments are important to philosophers – it is by analysing them (particularly when they conflict) that we acquire an understanding of the role our concept of causation plays in our reasoning.Intuitions are the starting point of philosophical discussion. They are not the end point (or should not be). The job of a philosopher, having identified the intuitions stimulated by a thought experiment, is rationally to pin them down, and formulate a coherent theory on the basis of them. This might lead to the jettisoning of some intuitions. But of course philosophers are not concerned only with our concept of causation. Ultimately we are concerned with causation itself: with what we are thinking about when we use our concept of causation. So in asking about causation philosophers are concerned with semantics and conceptual analysis, but also with metaphysics and ontology. We want to know whether our concept of causation applies to something that exists independently of us (or if not independently of us, in what way it depends on us). And if so what this something is. Causation is central to science. But it tends to be studied only by scientists interested in thinking. So psychology and neurophysiology are concerned with the empirical study of how our concept of causation works. Applied robotics is also concerned with thinking about causation – if we are to build robots sensitive to causal relations we must give them a concept of causation, so we need to know what this concept of causation is.Philosophy and science complement each other (or should do). Without the conceptual analysis with which any enquiry should start, science is working blindly. Without the empirical results generated by science, philosophy risks becoming a loose cog.So in thinking philosophically (or indeed scientifically) about causation we must start by analysing our concept of causation. Ideally we will end up successfully doing one of the following:reducing causation to non-causal relations and matters of fact;eliminating causation;admitting causation as primitive.Our Starting Point: Hume’s Regularity Theory of Causation:Philosophers have been thinking about causation since at least 2500 BCE. But modern discussions started with the Scottish Philosopher David Hume (1711 – 1776).From 1739 when Hume published A Treatise of Human Nature, or perhaps more accurately since the publication, in 1748, of the Enquiry Concerning Human Understanding, philosophers discussing causation have done so under Hume’s shadow. The study of causation is now emerging from Hume’s shadow, but to understand causation we have to start with Hume. In this lecture I shall introduce the traditional interpretation of Hume, two key problems it faces, and the solutions to them. The Traditional Interpretation of HumeAccording to the traditional interpretation of Hume, he embraced what is known as a Regularity Theory of Causation (RTC). According to the traditionalists this is Hume’s argument for RTC:Premise one: Our idea of causation seems to be the idea of necessary connection – i.e. that given the cause the effect had to happen.Premise two: As empiricists we should accept that all our ideas come from impressions (i.e. from experience and reflection on experience).Premise three: We do not (and cannot have) any experience of necessary connection (we can experience an event’s happening, but not its having to happen).Conclusion: our idea of causation is not an idea of necessary connection.So far this gives us Hume’s negative thesis – that our idea of causation is not an idea of necessary connection. This leaves us with two questions:If our idea of causation is not an idea of necessary connection what is it an idea of?Why are we so certain our idea of causation is an idea of necessary connection? Hume’s answer to question one:Experience, according to Hume, tells us that a cause is temporally prior and spatially contiguous to its effect, and that events similar to the cause are constantly conjoined to events similar to the effect. Our idea of causation is therefore that it is a relation involving only (a) temporal priority, (b) spatial contiguity, and (c) constant conjunction. The RTC, on this interpretation, is a reductive theory – it reduces causation to non-causal facts and relations. Hume’s answer to question two:Our experience of constant conjunction between events similar to the cause and events similar to the effect forms in us a habit of mind: when we observe an event similar to the cause we expect an event similar to the effect. This expectation is projected by us onto the world as the idea of a necessary connection between cause and effect. The RTC, therefore, on this interpretation, eliminates the idea of necessary connection as something that exists in the world independently of our expectations. On the traditional interpretation, therefore, the RTC is a theory that (a) reduces causation to non-causal facts, (b) eliminates necessary connection, and (c) sees causal regularities as epistemically prior to singular causal relations. Problems for the Regularity Theory:There are many problems for RTC on the traditional interpretation:is empiricism the correct theory?Is there really no impression from which we might get our idea of necessary connection?Correlations do not have a direction but causation does. How can a relation that depends on similarity be objective?But two key problems are:Surely regularity is not sufficient for causation? Surely regularity is not necessary for causation? Let’s discuss these problems:Regularities are not sufficient for causationImagine we discover that every male in this audience is a second son. I bet you’d reject the idea that this regularity is causal? I don’t think you’ll think there is something about being a male in this audience that makes someone a second son? Or something that prevents men other than second sons being in this audience? This is an accidental regularity.Now think of the barometer’s falling just before the storm starts. Does the falling barometer cause the storm? No, both are effects of the lowering of atmospheric pressure. This regularity is not accidental. But nor is it causal. Both situations are counterexamples to RTC given that RTC says there is no more to causation than regularity. If we are to save RTC in the face of this objection we must add something to exclude accidental generalisation, and other non-causal regularities. Regularities are not necessary for causationWe say that smoking causes cancer. Yet we recognise that some smokers do not get cancer. Exceptionless regularities, therefore, do not seem to be necessary for causation.We probably believe that the existence of the universe was brought about by the Big Bang. But none of us observed this, and we certainly didn’t observe it more than once. Must we agree with Hume, furthermore, that we do not observe causation in the individual case? When we slice a loaf are we not conscious of the efficacy of the pressure we are putting on the knife as it cuts into the bread? If we can observe causation in the individual case, why should we accept that regularity is required for causation? Again RTC is threatened and must be modified if it is to resist such objections.Solution to Problem One: The Laws of NatureThe solution to the first problem rests on the introduction of laws of nature. These enable us to distinguish between regularities that are not causal and those that are. So whilst you will probably deny the existence of a law of nature ensuring that any male in this audience is a second son, you will probably accept that there is a law of nature underpinning the fact that every child born with Down’s Syndrome has trisomy 21. And whilst you would accept that there are laws of nature underpinning relations between a fall in atmospheric pressure and a falling barometer, and between the former and the onset of a storm, you’re likely to insist that the relation between the falling barometer and the storm is a function of these laws, not of a separate law governing the regularity between the fall in the barometer and the start of the storm.RTC can claim it is only regularities underpinned by a law of nature that are causal. This saves it from the objection that regularity is not sufficient for causation. Solution to Problem Two:John Stuart Mill and Donald Davidson both insist that regularities are necessary for causation. It only appears to be the case that there can be causation without regularity. Mill explains this appearance in terms of the fact that whatever we identify as the cause of a given effect will in fact only ever be part of the cause. So we might say that the short circuit caused the fire, but the short circuit only caused the fire because of the presence of flammable material and the absence of a sprinkler. There is no exceptionless regularity relating short-circuits and fires. But Mill believes that were we to identify the whole cause of the fire we would see that any occurrence of this cause, the total cause, would suffice for the occurrence of a fire; i.e. that there is a law of nature underpinning the causal relation between this short-circuit and this fire. Davidson explains the appearance in terms of the fact that, pace Hume, we have no need to observe a constant conjunction in order to identify the relation of cause and effect. When we observe a case of causation, however, Davidson thinks that we will always believe that there is a law subsuming cause and effect. What we won’t know is what that law is. So we see that this short-circuit caused this fire. This tells us there is a law of nature involved. But it doesn’t tell us what this law of nature is. Both Mill and Davidson believe, with Hume, that where there is causation there is regularity (i.e. a law of nature). But both deny Hume’s claim that we must observe a constant conjunction between similar causes and similar effects to know we have a case of causation. Each philosopher allows that we see causation in a given case (or in a few cases) and later, through observation and experiment, we discover a regularity. Causation and Individual CausesRTC insists that regularities (in the form of laws of nature) are necessary and sufficient for causation. But reflection on the claims of Mill and Davidson make it clear that in everyday parlance individual causes need be neither necessary nor sufficient for their effects. The philosopher John Mackie argued that individual causes must be at least INUS conditions for their effects. A cause is an INUS condition of an effect iff the former is an Insufficient, but Necessary part of an Unnecessary but Sufficient condition of the latter. So the short circuit is not itself sufficient for the fire, nor necessary, but it was a necessary part of a condition that was sufficient (albeit not necessary) for the fire. In other words: there is a law of nature such that the short circuit at t, together with the other conditions present at t, sufficed for the outbreak of fire at t+1 (and no other set of conditions sufficient for the outbreak of the fire was present at t).The Canonical Statement of RTCThese (and other) modifications of RTC, made in response to other objections, have resulted in what might be thought of as the canonical statement of RTC:C is a cause of E iff for some time earlier than E, C belongs to a set of events occurring at t that non-redundantly suffices for E. Next week we shall look at another account of causation that was for many years considered a rival to RTC.Marianne TalbotDepartment for Continuing EducationUniversity of OxfordMichaelmas 2015 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download