Friday, December 05, 2014

What's the point of social science?

By chance, our visit to France coincided with this conference:

International Conference : Confidence, Credibility, and Authority in Climate Sciences and Politics

It looked interesting, so along we went...

The talks from scientists were generally straightforward, but the social science talks inevitably left us waiting for the punchline. They would get to the end and stop, before reaching any real conclusion. This has been a common impression we have both got from a number of similar events. The speakers tend to be long on historical description and retrospective analysis, and short on anything amounting to overall vision, substantive advice or predictive claims. (There have been notable exceptions to this general impression, but they are rare.)

A lunchtime discussion provided more insight than a day and a half of talks. We were sitting next to a social scientist, let's call him Bob (because his real name was Ian). jules asked him pointedly what the purpose of social science was, as it never seemed to say anything of substance to us, i.e. concrete advice. IanBob admitted rather frankly - proudly, even - that it wasn't supposed to have a point. Does it have to have utilitarian value to be worthwhile, he countered. I readily admitted that art and poetry had some value without making falsifiable predictions. But I had been hoping for more from the, um, scientists, involved in social science. The entire system of science can possibly be summed up as the making of falsifiable predictions and this is what most clearly separates it from religion (probably a bit over-simplistic, I don't claim any great authority on the topic). So asking for some testable theories didn't really seem too unrealistic to me.

The first talk was actually a history of the establishment of confidence in climate modelling, and though it was basically a valid review of the literature, it lacked a little (in my view) in describing confidence and consensus as something that seemed to emerge by default over time, and failing to recognise the emergence of consensus as primarily an indication of the limits of credible disagreement. IMO, this is the most fundamental aspect of consensus-forming and scientific progress (as we argued in this piece), but of course the failure to generate credible alternative theories is not really obvious from the literature. For a currently topical case, consider Tim Palmer's call for new high resolution climate modelling centres. Tim has some ideas for improvements to climate models and climate modelling, which may be wrong or right (his new article is at least an improvement on previous versions of his argument, IMO), but at least they are plausible and concrete. In contrast, Judith Curry waves her hands and asks for "fundamentally new model structural forms" but without any actual ideas as to how these fundamentally new models might be created nor what they could bring to the table, it's just hot air and hand waving. While I'm on the topic, if it's not the job of people like Curry to actually create such new models, then who exactly does she think should be doing it, and how? But I digress.

Anyway, back to the story, Ian argued that the main point of social science was to provide stories - his word - that described how human society worked. And these stories were to be judged primarily on how plausible or convincing they sounded. The concept of "truth" as a scientist would interpret it didn't come into the matter - truth was basically determined as whatever ideas were currently popular, nothing more. Of course scientific "truth" is actually a bit of a slippery concept. For example, Newtonian gravity is not actually true, but it's near enough for very accurate predictions over a wide range of applications. We don't think we are really describing truth, but we are at least attempting to approximate it and the demonstration of this is that the theories reproduce and predict the world, rather than merely being attractive to an audience. Note again the importance of useful predictions in this. Moreover, the stories were not expected to be generalisable to other situations. They were just what happened in that particular case. No over-arching theories, or even any consideration that this could - in principle - be one of the eventual goals.

Of course Ian's argument was somewhat undermined by the number of speakers wailing that "things need to change" (in order to make progress in the policy debate, which went almost without saying as the underlying purpose of the conference). This sounds almost like a predictive claim, i.e. that a change in behaviour might lead to some observable result, but stopped some way short, in that they didn't actually describe what the required changes were, nor what results would likely be observed. Next time I hear a social scientist going on along similar lines, I will simply sigh and try to treat it as a Just So Story, only not as good.


Steve Crook said...

For all your social science and ethnology requirements please visit BBC R4 "Thinking Allowed". Laurie Taylor and his guests talk for half an hour and, as John Ebdon used to say, "come to no serious conclusion"

Anonymous said...

There's more to social science than you might have seen at this conference. Much of it really is data-driven and cumulative, with testable hypotheses and replicable results. Also a clearer sense of purpose.

CapitalistImperialistPig said...

I think you might be looking for social sciences in the wrong places. The central question of social science is what is the nature of this animal we call "man"? The answer appears to be pretty complicated, but it's absolutely fundamental to understand some aspects if you hope to persuade society of the necessity of any difficult course of action - like, say, limiting carbon emissions.

Let me recommend, among others, the book Moral Origins, by Christopher Boehm.

James Annan said...

Perhaps my question could be more precisely expressed as, what does social science offer us in climate science? Cos there are all these projects and conferences where they poke their noses in. Is there any anticipated benefit to anyone? From where I'm sitting it's hard to see any meaningful outcomes.

I did allude in my original post to some notable exceptions, but they seem rare. Work on perception of risk, for example, which demonstrates that some presentations are more easily understood than others. That seems to be a valuable result.

Anonymous said...

I'm not sure this (just published) directly addresses your need for practical advice, but I'll argue that it is a climate-relevant instance of social *science*.
"A four-party view of US environmental concern"
Chris Mooney in the Washington Post wrote it up, reproducing the original graphics, here:

The paper itself, not paywalled, is here:

There is much more social-science work in this vein, addressing science communication and why things are like they are.

Anonymous said...

Re: "Note again the importance of useful predictions in this. "

In your opinion, what are a few short-term (e.g. 5 year) "useful predictions" from climate science?

John Cook said...

Time for a social scientist to defend the honour of his field :-)

I was at a workshop a few months ago chatting with several climate scientists (both of whom I respect a great deal). They were both decrying the field of social science based on a single dodgy social science study. I was struck by the irony that they were using the same type of anecdotal reasoning that social science is criticised for. Reading this blog post, I'm having a deja vu experience.

Much of social science is data-driven and purposeful, with practical application. I straddle two worlds - climate communication and cognitive psychology. The lessons learnt from my psychological research are directly and practically applied in my real-world climate communication. I have written several papers on adapting the lessons from social science into practical communication that improve climate literacy:

Cook, J., & Jacobs, P. (2014). Scientists are from Mars, Laypeople are from Venus: An Evidence-Based Rationale for Communicating the Consensus on Climate. Reports of the National Center for Science Education, 34(6).

Cook, J., Bedford, D., & Mandia, S. (2014). Raising climate literacy through addressing misinformation: Case studies in agnotology-based learning. Journal of Geoscience Education, 62(3), 296-306.

Not only does social science provide practical, tangible results, I would argue that its imperative that scientists adopt a data-driven, evidence-based approach to science communication, using the research from social science.

Coincidentally, I'm presenting two talks at the AGU Fall Meeting in a few weeks based on these two papers.

Sou said...

I've read elsewhere people who have attitudes like Bob/Ian's. But those attitudes aren't universal.

Some social science might be simply telling a Just So story, but a lot of it is useful in monitoring behavioural change and bringing it about. People's conversion to recycling or quitting smoking or immunisation or family planning didn't happen by accident. It was because of public awareness programs informed by social sciences and science in related fields.

There's a lot of good research being done in the science communication arena and the cognitive science fields too, which, while maybe not falling under the umbrella of social science aren't that far removed. They are helping with getting the message out about climate change and the importance of climate action, so that it will be heard and acted upon.

Sou said...

My comment overlapped with John Cook's. I had his work in mind, among others, when writing it. Good to see.

Sou said...

Here is an account of another program I had in mind when writing my comment. Although it may be difficult to pin down just exactly what was responsible for a change in behaviour, it would seem reasonable to conclude, given the timing, that at least some of the change could be attributed to the behavioural change program - which was based on social science research.

James Annan said...

Fair enough :-) But interesting to hear acknowledgment that BobIan's position is a recognised one within the field.

Mondomensch, I think you are reading too much into what I'm talking about here. I don't think the 5y climate change predictions are currently much use (I've blogged about that a few times), but such predictions are a definite goal of some, and climate science has made and continues to make many other predictions.

Anonymous said...

I think one does have to be careful of falling into the "physicists are very clever" trap (said with acknowledged irony :-) )

I tend to have a great deal of respect for many social scientists, partly because I know that I don't understand many of the topics particularly well.

My personal issue is with social scientists who try to put themselves at the science policy/society interface. In my discussions with some of them (a smallish sample, to be fair) many appear to have little understanding of how the physical sciences work. They seem unwilling to accept that there really are physical laws and that it is really difficult for something that is wrong to remain accepted for a lengthy period of time.

Of course, there have been ideas that have been wrong that have been accepted for some period of time, but once enough people have illustrated the issues, these ideas are rejected (by most, at least).

So, my main issue with the social sciences (in the context of climate science at least) is the role that some seem to be trying to play without actually trying to develop any understanding of how the physical sciences actually work. This is, of course, not true for everyone, but I have seen quite a number who I think fit this description.

jules said...

I'd originally thought that social scientists had been encouraged into the climate world in order to help us work out what to do.

What I learned from Bob is that it's all a bit more parasitical - they are involved because they see something interesting going on, and they can get grants to make up stories about why it is the way it is. They talk about the "failure" of the climate scientists to get real policy changes, and tell us that history shows that there was never a chance of it working anyway, while offering no advice as to a practical way forward.

I'm not convinced there has been a failure. I think the world listened to the science, understood it, and decided to stick with existing economics, adapt rather than mitigate, while gently encouraging some easy to do green things, like recycling, and putting a bit of money into alternative energy and efficient technologies.

Yes - communications science is the bit that is potentially useful, and it is in that area that we have seen the really interesting and potentially useful stuff, but, sadly, makes up quite a small percentage of the social science morass that we have seen talking about climate science.

Anonymous said...

I'd originally thought that social scientists had been encouraged into the climate world in order to help us work out what to do.

Yes, I think that was my confusion too. I had assumed that those social scientists who saw themselves at the science policy/society interface were aiming to help with the communication of science. Instead, what seems to be the case (which is similar to what I think you've concluded from talking with Bob) is that they see an opportunity and, if anything, are trying to suggest that scientists stick to doing science, leaving them to be the ones who provide the link to policy makers and society. Given that, from what I've seen, they have little understanding of the physical sciences or how physical science works, this would seem to be a very poor way forward.

@ReinerGrundmann said...

I think there are various red herrings in the above.

Science does not only explain and predict, it also describes and explores. Not all good science has to falsify hypotheses.

Being data-driven is certainly true of my colleagues who relish detailed historical case studies. Based on these studies they do not, and cannot, predict anything.

My own take on the relevance of Sociology for climate change debate can be found here (just one example):

Some social scientists do contribute to the policy debate, myself included. Again see one example here:

Apart from science communication studies, comparative research is important. Some weak generalisations can be drawn from them, including some hypotheses. This is an area of research which has great relevance for me and others.

Anonymous said...

Since you've made an appearance, I'll put you on the spot. An example of something related to the science/policy interface that I find very unimpressive is from this Roger Pielke Jr post written by you and about your book.

In the post, it says There is an eerie similarity between race science and climate science in that both see their services as essential for solving pressing social problems.

If I understand the argument you're making, it is that race science was motivated by a desire to influence policy and so is climate science. Now, I'm not formally a climate scientist, but I'm amazed more of them aren't openly appalled by such a suggestion. At this point you can of course clarify what you meant and indicate that my understanding is wrong.

As a physical scientist, however, I also find the argument (assuming I understand it) very odd. Climate science is largely a physical science and we have understood for a very long time the basic impact of increasing atmospheric greenhouse gas concentrations. It's certainly my view that climate science has become policy relevant because of what it has indicated might happen if we continue to increase our emissions. It hasn't, as far as I can tell, become policy relevant simply because it is seen as a mechanism to influence policy. I find such a suggestion remarkable and - as I said above - amazingly insulting to climate scientists.

Of course, I'm more than happy for you to elaborate on what you were suggesting so as to either convince me that even though I do understand your argument, it has merit, or to convince me that my understanding of your argument is wrong.

EliRabett said...

Ok, think Gavin Schmidt, you and Jules and Tim Ball. Ball was, to the extent he was a climate scientist, a number gatherer, he had not made the transition to cladistics, looking for common characteristics within the data set, such as you and Jules. Gavin is a modeler who tries to relate basic principles to the ordered data sets.

Peter said...

Political science seems particularly relevant to studying climate change politics. It answers why support among voters for climate change legislation is so low. Explores what attributes of political systems explains differences in support for climate change legislation across countries.
Public choice theory explains climate inaction in terms of interest group politics and concentrated costs/benefits and disbursed benefits/costs. Climate change legislation has immediate costs for a concentrated group of energy companies, but disbursed benefits for society a long way in the future, so energy companies will lobby harder and we'll get less action on climate change.
Economics would seem to be the only science (social or physical) that would predict the costs and benefits of climate change, and consequences of climate change policies.
Basically, there seems to be plenty of useful social science research regarding climate policy, yoou just gotta know where to look.

JohnMashey said...

Social science, as per Wikipedia.

Like anything else, there is a distribution of usefulness within each of the disciplines.

In terms of usefulness to climate science, social scientists are akin to physicists or statisticians in the following sense: the best ones bring their skills and learn enough to be very useful ...
at the other end, they can be worse than useless.

A social scientist need not be a regular attendee at AGU to be useful, but some do come or have spent years making sure they understood the science well enough ... for others, not.

Anonymous said...

There are cultural variations, too, within social science and also between its modal European and US manifestations. The latter (with many exceptions both ways) tends to be more quantitative, for one thing. Data analysis can provide a common language for interdisciplinary collaboration, and sociologists publishing data-centered papers in natural-science journals such as GRL or IJOC, or presenting at AGU, as John Cook mentions above.

Anonymous said...

Under auspices of the American Sociological Association, Riley Dunlap and Robert Brulle have been editing a book called Sociological Perspectives on Global Climate Change. Coauthors Sandra Marquart-Pyatt, Andrew Jorgenson and I wrote a chapter for this about "Methodological approaches for sociological research on climate change." Our chapter includes 14 pages of references ranging from Allison (2009) Fixed Effects Regression Models to Zaval et al. (2014) "How warm days increase belief in global warming," which I hope could give a broader perspective on what sociologists are doing, or trying to contribute.

Or, drop by the talks at AGU by John Cook or me, to hear what some of us are up to and why it might be of interest to many folks at that venue.

JohnMashey said...

I've long thought the right sorts of social scientists were invaluable, dating back to the late 1970s at Bell Labs, when i built a group that included cognitive psychologists with computer software folks.

Researchers in some area almost always know who in their specialties knows what they're talking about. It is a specific skillset to assess the credibility of those in different fields. Not everyone gets much training or experience in that skillset, but it is rather valuable for high-tech managers, troubleshooters and those who perform due diligence on tech investments.

Anonymous said...

I get your frustration with some social science contributions to issues like climate change, but I think some of this comes from deeper epistemological and ontological differences in how we see the world. If you maintain a positivist view of social inquiry then yes you are likely to be frustrated and dismissive of a large body of research in the humanities and social sciences which use a more interpretivist and social constructionist perspective. However, I would suggest positivism is a fairly limited perspective in trying to understand human society.

My own research into climate change and corporate capitalism is explicitly social constructionist but I find this both a far more informative and interesting approach in seeking to answer what I consider to be the really big and important questions in this space; specifically, why despite blindingly evident scientific evidence of catastrophic anthropogenic climate change – human society has decided to ignore the threat? The answer, I argue, is that the ideology and political power of economic elites such as major business corporations makes it (to quote Fredric Jameson) “easier to imagine the end of the world than to imagine the end of capitalism”.

The only way to really get at these sort of issues (questions of economic power, the construction of hegemonic discourses of ‘economic growth’ and ‘free markets’, the creation and maintenance of a capitalist imaginary of endless growth, the political strategies employed by corporate elites) is via a qualitative, interpretivist and social constructionist lens.

After all, our knowledge of anthropogenic global warming stretches back over two centuries and we have had at least four decades of political hand-wringing about climate change and yet humanity’s combined greenhouse gas emissions continue to grow and are accelerating. How to make sense of this insanity? This is where critical social inquiry is essential.

You can read more on this on my blog:

I also have book coming out next August which outlines this research called "Creative Self-destruction: Corporations, Climate Change and Capitalism" (Cambridge Uni Press).

jules said...

"why despite blindingly evident scientific evidence of catastrophic anthropogenic climate change – human society has decided to ignore the threat?"

The answer is that you are asking the wrong question. There is no blindingly evident scientific evidence of catastrophic anthropogenic climate change. That's why people are not responding as if there is. The problem word is "catastrophic".

For example, if you go by the Stern Report, the cost of climate change to Japan is (iirc) the same order as the 2011 tsunami. It was/is certainly catastrophic to some people and some places, but, as a whole, the Japanese nation lives on yet quite comfortably.

Without an actual imminent catastrophe, no one is going to vote for the massive reduction in standard of living that is presently required to bring about the mitigation-lifestyle.

Anonymous said...

Well I agree with you in one sense re engaging the world's citizenry around some mind-boggling catastrophic event. However we have had a nice series of extreme climate events over the last 20 years to get people's attentions!:
- 2003 European heatwave, which resulted in tens of thousands of fatalities
- 2005 Hurricane Katrina decimates the US city of New Orleans.
- 2010 Pakistan experiences its worst floods in living memory. An estimated 20 million people were directly affected.
- 2010 Russia endured its worst-ever heatwave and drought. Around 56,000 people died as a result.
- 2011 the US was plunged into the most devastating drought in its history. At the same time the Mississippi suffered massive floods that matched the ‘great floods’ of 1927 and 1933.
- 2012 Arctic summer sea ice melted to an all-time low. The decline was so great that scientists now project the Arctic Ocean could be ice-free in only a few decades.
- In the same year New York was hit by Hurricane Sandy leading to the flooding of the world's financial capital.
- 2012/3 also saw devastating bushfires across Australia during the country’s hottest-ever summer. The heat was so intense that new colours had to be found to depict its severity on weather charts.
- 2013 Super-typhoon Haiyan hits the Philippines
- 2014 scientists announced the collapse of the West Antarctic ice sheet. This process is expected to result in a sea-level rise of as much as five metres and has been described by glaciologists as ‘unstoppable’.

So the extreme weather is there and mounting. The point is there is a powerful discursive and political battle going on to convince people that there is no problem, no climate crisis and 'business as usual' can continue.

It is in studying that process - of creating a 'fossil fuels forever' imaginary - that critical social theory is particularly useful.


It's a an older problem than generally realized.

Steve Bloom said...

Cause for hope on this front? Or not.