To what extent can science tell us what to do about the future, especially in relation to novel phenomena like climate change. Flooding across the Southwest since early January and now across the south and some areas of the Home Counties saw a link being made last week in the mainstream media for the first time with human-influenced – or ‘anthropogenic’ climate change – ACC for short. A resident of a village on the Somerset Levels was widely quoted as saying
‘I’m used to seeing floods on the Levels, but this is just something else,” he said, noting that when the area flooded less severely last winter “we were told it was a one-in-100-year occurrence”, but that ‘the following year it happens again — only worse!”’
Several climate scientists were quoted as suggesting that we have found ourselves in an entirely novel situation. Myles Allen from Oxford University stated that:
What is clear is that just looking back at the historical record to plan flood defences or set insurance premiums is increasingly misleading.
Nigel Arnell from Reading University said that:
We have long been exposed to risk from flooding, but climate change is loading the dice
The implication of these statements is that the future will not be like the past, that the probabilities of flooding and other events will change, perhaps radically. The implications of this are important, both for understanding the role and limitations of scientific knowledge in helping us prepare for this future, and for thinking through the moral significance of the resulting uncertainty. Dr Chris Groves (Social Sciences, Cardiff University) explored these implications, and suggested that, even if scientific knowledge faces crucial limitations in what it can tell us about the future, this should not prevent us drawing firm conclusions about how we should act.
One of the key problems climate modelling faces is that it aims to help us understand a process (ACC) that, if it is indeed underway, is unprecedented in human experience. Scientists have no observational data about the phenomenon of human-caused climate change and no theoretical models that have been developed on the basis of such data, because this phenomenon has not been observed before. ACC is a hypothesis that has been put forward in order to explain an upward trend in average global temperatures over the last century or so, with a particularly dramatic spike occurring since the 1980s. When an unprecedented phenomenon is being investigated, there is no choice but to use models – in this case, General Circulation Models (GCMs) of the global climate.
Uncertainty about the future is exacerbated by the fact that, If human-caused emissions of carbon dioxide and other greenhouse gases are, via the mechanism of the greenhouse effect, causing the atmosphere to warm, then it may not be the case that this upward trend continues. The climate system is complex, composed of various subsystems, including the surface ocean, the deep ocean, the atmosphere, the biosphere (changes in vegetation cover etc), sea ice and ice sheets, and solar radiation. These are coupled together in various non-linear ways that potentially create feedback loops whose exactly direction and effect is not necessarily specifiable ahead of time. For example, melting permafrost in the arctic caused by a relatively small increase in average temperature may release methane currently trapped within it. As methane can have a more potent short term effect than even CO2, the release of these gases may exacerbate climate change greatly, increasing average temperatures by an extra couple of degrees. On the other hand, increased evaporation of water may create more clouds that may help reflect more solar energy, reducing the rate of warming significantly.
The Intergovernmental Panel on Climate Change (IPCC), responsible for integrating empirical data and GCM outputs, has produced a range of around 40 scenarios. Each scenario indicates a best estimate of temperature increase, based on variations in model parameters, and accompanied by an estimate of uncertainty, so that the best estimate is 95% likely, and other possibilities for the same scenario are less so – but not impossible. For some scenarios, the potential rise is over 6 degrees. Overall, the expected range of average global temp increase documented in last year’s 5th IPCC assessment report is between 1.5 and 4 degrees C over the next century. Added to these scenarios produced by climate models, there are other integrated assessment models (IAMs) which are used to attempt to understand the possible impacts of climate change on human societies. What will happen to agriculture, to water availability, to demographics? Will there be mass migrations from areas that experience climate change more severely?
Some sceptical objections to these attempts to model future possibilities suggest that modelling the future in this way is epistemologically suspect, and that as a result we cannot draw any firm conclusions about what will happen. Without firm conclusions, we cannot justify any particular course of action – especially not a policy of decarbonising energy production. But as already noted, modelling the future of an unprecedented problem is the only way to explore what might happen. At the same time, it’s true that GCMs and other models cannot be tested in the same way as hypotheses generally are in science.
To test the accuracy of a model, we’d need to keep levels of carbon dioxide constant and see what happens. But if the models are correct, we cannot afford to treat the world in this way as a grand experiment. We cannot wait to verify models – in fact, we do not want the models to turn out to be true! To require that models be tested against empirical data is to require that we wait and see. Do we want to do this? If we cannot wait to ‘tune’ models based on whether or not they are accurate depictions of what in fact current concentrations of CO2 may be doing to the climate, then the only reason for using them as evidence is, as the philosopher of science Stephen Haller has put it, is ‘according to their ability to give intellectually satisfying explanations of currently observable phenomena’ (Haller, Apocalypse Soon? 2002, p. 32). So we could try to reproduce the observed warming trend with them, which as I noted is what scientists have done. However, the ability of models to reproduce the extant record of temperature change is not proof of their continuing validity for predicting the future either. Historian of science Naomi Oreskes has noted [PDF] that ‘the capacity to mimic data is not evidence that you have captured the underlying causal processes, and therefore not evidence of predictive capacity’. What is needed is the ability to make predictions about the future that would not be justified without the model. But again, this is impossible, given that we want to avoid the future that would confirm the validity of the model. We have a Catch 22 in which we must act despite irremediable uncertainty, and indeed because of it.
Add in the complexities involved with modelling the social impacts of climate change, then the uncertainty increases. Nonetheless, a choice has to be made as to what to do – a wager, if you like. Usually in a game of chance you know exactly what the outcomes will be (imagine betting on a die throw). Here we do not. Because we have to rely on modelling due to the novelty of our situation, there are limits to what we can find out about what the future will look like – and therefore limitations on the ability of scientists to tell us about the potential outcomes of what we might choose.
If ‘we’ need to make a wager without knowing all the outcomes, then ‘we’ here also included anyone sceptical of the IPCC line on climate change. Not doing anything in the face of the possibility of climate change – business as usual, continuing to emit greenhouse gases at the same rate – is itself an implicit wager on the future, one which affirms that benefits will outweigh costs.
So how do we decide what to do – and importantly, make a decision in in a way that we can justify to others and perhaps convince them of? Chris suggested there are good philosophical reasons, independently of what science can tell us, for deciding on a particular course of action. Suppose you’re playing a game of craps in your local casino. There’s a one in 6 chance of rolling two dice to get sevens. You have £50. Should you bet some or all of it on the next throw? This will depend on a lot of things – not just the odds of winning on the next throw. Whether or not you can justify to yourself using some or all of the money for gambling depends on reasons outside the game itself and what counts within it – that is, in relation to the odds of winning – as a rational decision.
A relevant philosophical distinction is that between reasons internal to a practice and reasons external to that practice, as discussed by John O’Neill. The reasons for playing internal to the practice of playing craps are not enough to reach a decision about whether we continue to play – we need reasons from outside. When we are reflecting on the practices we engage in, whether cooking, welding, writing, looking after children, political activism, or whatever, when we take decisions about what to do next we have particular reasons for our decision that are internal to the practice. And these reasons relate to what we believe it means to cook, weld, write, care for the kids well. But what it is to cook well is not what it is to look after the kids well. And by extension, a good reason for doing something in one practice is generally not a good reason for doing something in another: you can’t cite the need to avoid browning garlic while frying it as a reason to not smack your children. Reasons internal to one practice do not count within another practice.
If we turn back to climate change, then science in general has its own standards of excellence, which relate to standards of evidence, experimental practice and so on. What counts as a good reason to do something as part of the practice of science – say, to accept a hypothesis, like ‘human activity contributes to climate change’, as a good explanation of something – is a given standard of reliability. If you show that your tests of this hypothesis meet standards of scientific excellence, then this gives other scientists a reason to do something, say, change their beliefs about the world. They also then have reasons to take certain other kinds of actions: they might themselves seek to test your hypothesis for example.
But the point is that what motivates scientists to do something like this are reasons that are internal to the practice of science. Most of the discussions heard in the media in recent weeks about climate change – and most discussions on this topic in general – are debates about what counts as scientific truth, the standard of what it is to do science well. But when we ask: should we take any action? And more concretely: to what extent should we try to mitigate ACC and to what extent should we try to adapt to it? then we are not taking part in the practice of science. We are talking as people engaged in a variety of other practices where things other than scientific truth matter. Above all, we are engaged in the practice of being a citizen, which if you go back to Aristotle, is all about giving ethical and political reasons for action. We are like the craps player: knowing the odds is not enough.
The implication is that, even if we did not face the catch-22 of not being able to wait for models to be verified, in order to decide what to do about climate change, we need ethical and political reasoning – reasoning about what kinds of people we are, what sort of society we want to live in. Scientific evidence about the climate system is hugely important, but the space in which the decision has to be made is an ethical and political one. There is no straightforward, mechanical path from scientific evidence to policy. We have to decide if there are things that matter to us so much that the loss of them in a climate-changed future would be absolutely unacceptable. Such things might be a way of life, an ideal of social justice and reduced inequality, or other things that might be threatened by increased flooding, increasing global hunger, or whatever. If there are such things, then we need to act to make sure they are preserved. This is an ethical, precautionary argument – not one that turns on whether any particular model turns out to be true or not. It does not rely for justification on what we know – it aims to justify itself as making sense precisely because we are uncertain about the future and to a large extent will remain so.
In discussion, several audience members affirmed that the ethical and political ‘space of reasons’ was key to understanding what to do about the possibility of climate change, from both a moral and a pragmatic point of view. To achieve changed, it was suggested, we need arguments that revolve around the better world that we hope would result from decarbonising energy production, consuming less, and so on. These oblique strategies may be expected to receive more assent and support than messages which simply rely on scientific evidence to justify tighter regulation and prohibition. The attractiveness of such future visions to individuals might be significant, others pointed out, but the obstacles to change are not perhaps individual preferences, but the concentration of power to shape how we live in the hands of corporations and other entities. The question of who the effective agents for change might be has been raised before at Philosophy Cafe, in the context of related debates around topics like sustainability. How do we change ourselves and the societies we live in – beginning with the individual, or with the collective – remains perhaps one of the most pressing questions of our time.