Why Faith Can’t – and Shouldn’t – Disentangle Science and Politics

I recently appeared as a panelist on the BBC debate show The Big Questions, debating the topic ‘should we have more faith in science?’. So here are some more extended thoughts on that topic. The programme ranged from climate change to homeopathy, but I’ll stick to the former.

It might sound odd, but I’m not sure many scientists would want us to have ‘faith’ in science.  It somewhat goes against the ideal image of science as built on claims which have survived rigorous experimental tests.  However, in reality doing science does require some minimum amount of faith.  For instance, the act of ending an experiment, of saying ‘well we might have missed some hidden factor or source of error, and there’s still a 0.0001% chance that this result is just random background noise, but dammit I’m getting bored of this I think we can say we’ve got it’, is a leap of faith.  The trick – and this is where science might arguably be distinguished from other pursuits – is making the leap as small as possible, of having the absolute fewest possible points at which you have to say ‘well I’m not 100% sure, but I still think that…’.

However, there’s a crucial difference between scientists having faith in their own work and other people’s faith in scientists.  Again in an ideal science-world everyone, from policymakers to the general public, scrutinises the evidence behind every claim we ever hear.  However, not only would this make general conversation intensely annoying, but in the 21st century it’s also a very idealised image.  Even top scientists get rapidly less capable of critically scrutinising scientific evidence once you move outside their particular field.   All the science going on nowadays is pretty darn complicated and pretty darn specialised.  I’d love to be able to double-check the Higgs Boson discovery using my own backyard hadron collider, but it ain’t going to happen.  I’ll stick to the somewhat easier option of believing the newspapers and their quotations from scientists.  Is that a loss to the world of high-energy physics, or to anyone except me? Most likely not.

But it’s a different matter for sciences that have broader implications.  It was claimed in the opening to our debate that six in ten people don’t believe humans are causing climate change (although interestingly, given that the topic was about faith, none of us on the panel questioned that figure – and I’ve since struggled to find where it’s from1).  If that’s true, and if they’re wrong, then the implications for everyone are pretty severe.  The reasons behind (dis)belief in climate change become important topics for discussion.  But I don’t think such discussions should be about having more faith in science.  That’s rarely where the actual conflict lies: just as in homeopathy vs. conventional medicine debates, both climate change proponents and sceptics claim they have faith in science.  They’re just alternative, competing ideas of science.  And arguing what is and isn’t ‘science’, or even ‘good science’, isn’t a particularly productive way forward – that’s a personal worldview which people use as a basis to assess information, so it’s not something which can be changed by information.

Instead, I’d suggest that it’s more important to think about how people are getting and assessing their information.  Despite the differences in political stakes, I strongly suspect that the situation is actually quite similar to stories like the Higgs Boson or gravitational waves: who is saying things, and how they’re being heard, is at least as important as the data.  We may not all be able to assess climate change data, but if scientists are revealed to be funded by certain businesses, or if emails are leaked showing supposed manipulating of results… well, there’s some data many more people feel comfortable drawing inferences from.2  This where I disagree with fellow panellist Prof. Tim Palmer’s suggestion that we should try and disentangle the science and the politics of climate change debates.  Even if scientists did somehowmanage to insulate themselves from political and social concerns, they’d still need to convince everyone else that they had done so (and I doubt many people would take such a claim on faith alone).  Indeed, it’s exactly at the intersection between science and the political that many of the doubts and debates lie – from the moderate ‘I don’t think the evidence is strong enough to justify policy action’ to ‘I think your results are biased by the neo-Marxist cabalistic brainwashing of your climate science community’3.

So instead of just using occasional massive reports to accumulate and transmit information, bodies like the IPCC need to use their expertise (both scientific and political) in much more transparent and open discussions – whether face-to-face, through mass media, or even the internet.4  This idea of dialogue rather than transmission is a popular one in the field of ‘public engagement with science’, and is being increasingly employed by policymakers for national issues – but I have yet to see any large-scale similar efforts to help laypeople navigate climate change debates, or raise their own questions and viewpoints.5  But I think it would be better than what anything debaters are using currently.

I’m imagining something like this: So, climate sceptic, you think my data’s flawed?  Here is all my data, feel free to analyse it.  Ah, you think any data I present you with will be flawed because of the influence of some political bias?  Well let’s examine the political culture within my lab, even something as simple as voting patterns or accounts of funding sources might be instructive.  But you’re still unconvinced because we keep talking about ‘uncertainties’ in our report? Well, let’s talk about the fact that this is how science is done, it’s not a particular flaw of climate change.  You still think the uncertainties are too large to justify policy action? Well, when would policy action be justified – bearing in mind that uncertainties will never be zero, and that policy action gets less effective the longer you delay?

This image is an ideal one, and flawed in many ways.  But if we’re targeting ‘swing voters’ of climate change, I think it points the way towards better cultures of communication.  Conversations need to be reacting to actual doubts, not hoping that throwing more data at doubters will do the trick (even just a bit more widespread discussion of surveys on public attitudes to climate change which go beyond ‘six in ten people are sceptical’ could be helpful).  Currently, we’re in an old ‘here you go, now be off with you’ culture of communication.  Studies1 show that the public get scientific information from the mass media – short news items where opposed sides talk past each other for three minutes, with longer programmes and experienced communicators generally reserved for galaxies or DNA while avoiding politics.  Or they get it from the internet – a confusing and tribalistic deluge where no-one knows what to believe.  There need to be more open and less overwhelming alternatives.  Most importantly, talking about science and politics in the same discussions might change how many debaters see climate change – not as a scientific problem to be conclusively ‘solved’ before we act, but instead seeing it as a Pascal’s wager.  There’s danger in every decision, and there’s always going to have to be a leap of faith.  Everyone wants to look before they leap; we need to find out what it is they’re looking for.



(An interesting recent article, with a similar premise but somewhat opposed argument, can be found here: http://www.rsablogs.org.uk/2014/socialbrain/scientists-strike/)



1 = It’s particularly interesting given that two recent polls – the Wellcome Trust Monitor and the Public Attitudes Survey – both suggest that general trust in science and scientists remains high.

2 = An important thing to note is that scandals like these, particularly Climategate, often rely on a wide misapprehension of how day-to-day science actually works – for instance, manipulating data on a graph to get rid of known errors is a standard, and justifiable, practice in science.  The same is often true of (often quite uninformed) criticisms of the use of computer models, as if this somehow made climate change ‘less scientific’ than other sciences.

3 = The latter being a claim I have seen raised in one particularly fraught climate change discussion thread – but I can’t remember where, and please don’t make me try and find it again. I’m not ready to go back to that place.

4 = The internet is a tricky one – trust me on that one, it’s my current area of research – in that it tends to polarize opinions rather than produce any agreement (to put it mildly).  It only takes a mercifully brief glance at the comments thread in any climate-change related article to see that they’re rarely the space for anything productive.  However there are ways to construct online forums which minimise this, such as forcing people to provide evidence of their claimed expertise in the subject, having a strict (but transparent) set of moderation criteria, or dividing up discussion topics into very specific areas.

5 = I’ll admit I might be missing something – climate change debates are difficult to stay comprehensively informed on (which, in essence, is the problem I’m discussing here) – but the IPCC’s current media and outreach website doesn’t give a lot to go on.

This entry was posted in Uncategorized. Bookmark the permalink.

3 Responses to Why Faith Can’t – and Shouldn’t – Disentangle Science and Politics

  1. Jim Grozier says:

    Apologies that this mainly concerns a topic in your intro which is not part of the main argument … but I don’t think the “leap of faith” you mention is ever actually made, at least not in experimental physics. Results of experiments are always announced in terms of the value of some parameter together with its uncertainty; if comparison with a theory is relevant, this is only done in the most vague, qualitative way, if at all. Even the Higgs result was announced in terms of the probability of getting the same result if the Higgs did not exist – the standard way of comparing experiment with theory. If there is a “leap of faith”, it is made by others, possibly (effectively) by the funding bodies in deciding there’s no point in funding the experiment any more. But most of the time an experiment reaches the limit of available precision, where the background is irreducible – I don’t think anyone – certainly not the experimenters – ever says “I’m getting bored with this and I think we can say we’ve got it” or words to that effect. Because – as you hint elsewhere in your post – we can never really say we’ve got it.

    When I was doing my PhD on the neutron EDM experiment, we were always being told that our experiment, together with its predecessors, which together spanned half a century, had disproved more theories than any other. But try as I might, I could not find any paper which said categorically that “… therefore theory X is wrong”. All you ever get is a parameter plus uncertainty.

    Finally onto something more relevant – it amuses me that climate change deniers concentrate on trying to show that the warming we’ve experienced to date is not anthropogenic. Assuming for the time being that they’re right, surely that means that – given that the only completely unassailable scientific fact is that atmospheric CO2 captures some of the earth’s infrared and beams it straight back at us – we ought then to be even more careful about emissions than otherwise, since there will be anthropogenic warming above and beyond the “natural” warming. But I guess that isn’t quite where they’re coming from …

    • So the “getting bored” bit was facetious, but otherwise I think (?) we’re in agreement about the leap of faith – as you say there comes a limit of available precision, and in the absence of tests beyond that there just has to be faith. Another side of the ‘we can never say we’ve really got it’ idea. The Higgs Boson is actually a problematic example in this respect in that it is still being expressed in terms of parameters + uncertainty, but think of positrons – the chances of them not existing, and of there being a more correct explanation for all the phenomena that we currently ascribe to ‘positrons’, are even more vanishingly small than the Higgs Boson (though think of phlogiston, the aether…). So vanishingly small that we base numerous other assertions on them (e.g. ‘all particles have an associated antiparticle) without ascribing uncertainties. I could go even further and say ‘giraffes don’t exist, they’re a sensory illusion caused by demons’, or ‘the gravitational constant is not constant, it’s going to change on the first Tuesday of 2917’. These philosophical doubts must be dismissed if scientists are to do anything at all – but that dismissal can only be based on faith, not evidence.

  2. Jim Grozier says:

    Well, yes, I agree that we broadly agree … but … let’s look at that positron example. If the statement “positrons exist” is to mean anything, we need to define what we mean by a positron. If we choose to define it as just “the thing that made the track in Carl Anderson’s cloud chamber”, then its existence is hardly in doubt. But if we want a more useful definition – “a small particle with a charge opposite to that of the electron” then we need to define “small”, since this could equally well cover a proton. Now, if we include in the definition “with a mass of 9.1 x 10^-31 Kg”, we can’t hope to ever find out whether they exist, since we will never measure the mass precisely enough. We might add “+ or – X” where X is the uncertainty in the measurement, but remember that the uncertainty is only a parameter and not a strict cut-off.

    We might settle for “a positron is a particle with a mass which is definitely smaller than one hundredth of that of a proton”, which gives you plenty of margin of error if the only competition is the proton, since it is nearly 2000 more massive. But once you get pions and muons coming into the picture you need to be a bit more careful; and in any case, even with a huge margin we can never say definitely that they exist or don’t exist, since whatever the theoretical mass, the probability of measuring the mass at any given value does not go to zero, it only becomes very small – that’s all the while we accept the idea that measurements are normally distributed.

    Of course we could also add other stipulations such as ” a positron does not feel the strong nuclear force” … but how would you experimentally verify that? You’d end up doing what an awful lot of current experiments are doing, namely looking for something (the interaction between a positron and a strongly-interacting particle) that is possibly zero but possibly just very very small, smaller than you can measure. Or you could say it annihilates an electron and produces two gammas, but that definition would probably run into similar trouble at some point …

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s