Communicating Reasons

Reasonable People #2: The psychology of communicating human rights (talk transcript), plus links and some brief book reviews

Main Course

Wilton Park is an agency of the Foreign and Commonwealth Office which organises discussion events focused on international security, prosperity and justice. It is also a really fancy tutor country house on the south downs:

I was invited to "Re-energising the narrative: human rights in the digital age (WP1655)":
"The event will consider the specific threats presented by abuse via social media platforms, the ‘echo chamber’ effect on critical thinking and policy making, and the deliberate exploitation of divisions in societies, eg computational propaganda/’fake news’, amplification by algorithms and systematised trolling."

Here’s what I said:

What does a psychologist have to offer here? I think the first thing is an apology, on behalf of my profession.

In our individualistic, narcissistic age, psychology is a growth area. Psychologists have been making hay from promoting the idea that people are irrational, that our thinking is riddled with systematic errors and delusions.

The title of popular science books are an excellent lens on this. Go to the psychology section and you'll find books with titles like "You Are Not So Smart" by David McRaney, like "Predictably Irrational" by Dan Ariely. Both good books, but you see the theme.

Perhaps the most celebrated psychologist of recent years, Daniel Kahneman, whose work is foundational to behavioural economics, which led directly to the idea of nudge, wrote "Thinking Fast and Slow", a book which describes our minds as divided, and often dominated by a fast, stupid, system which, in the words of one commentator portrays humans as basically "spending all their time failing".

Who benefits from this? Well, obviously we, the psychologists do. If human reasoning was straightforward then we'd be out of work. But the emphasise psychologists have put on studies of reasoning is profoundly limited.

So as well as an apology, I want to offer you some advice about the limitations of this work. To do this, let's pick an example from the experimental study of communication, a study from Stanford by Paul Thibodeau and Lera Bordidsky.

These two ran a study on perception of crime and crime control policy, where they asked participants to read a newspaper story about crime in a US small town.

Half the participants saw a version of the story where crime was described as a beast stalking the citizens, half saw a version where crime was described as a virus infecting the city.

Two version, two metaphors for crime. And then all participants were asked which policies they would support to deal with crime and their responses were recorded.

And this is how experimental psychology works - measurement (of people's policy support) and comparison (which metaphor people read about in the news paper). Now the result isn't so important but I'm sure you'll want to know, and probably won't be surprised by, the finding that people exposed to the beast metaphor offered more support for policies aimed at capture, enforcement and punishment: more police on the streets, longer prison sentences - and those exposed to the virus metaphor offered more support for policies aimed a diagnosis, treatment and inoculation - more education, fix the economy, resources to get kids out of gangs and so on.

But this study contains its own biases, and they are illustrative of the limitations of many such experiments.

First, because it works by comparing two conditions, it highlights the effect of the manipulation - the metaphor used, in this case - but at the cost of downplaying every other factor which influences people's judgements.

The experiment allows you to see the difference the metaphor makes to people's judgements, but renders the other reasons for invisible. And this common. Psychologists love experiments in which superficial changes creates differences between groups, but often we don't put the size of those differences into context.

In this way, experiments on biases in perception and decision making - which psychologists love to run - tell a very dangerous half truth about human reasoning. They tell the story of how our judgements can be swayed by superficial or distracting factors - all true - but neglect the story of how we come to arrive at our beliefs in the first place, at the profound role reasoning and reflection play.

The second bias in many experiments on communication is that they almost invariably look at immediate effects. We ask people to take part in our experiment, and this typically involves the manipulation and the measure at the same time point. Almost nobody gets participants back to see how their beliefs have changed a week later, or a month, or a year. That's too difficult.

So this creates another blindspot in our experimentally informed view of the world - we see things which have an immediate effect, which push our views and beliefs around: emotion, images, etc. But we're blind to stuff which works its effects longer term. This matters, I argue, because one thing which has profound long terms effects are good reasons and moral values.

The current fashion for a psychology preoccupied with our biases and limitations underestimates the common inheritance we all have as reasoning and moral beings. Worse, by promoting a view of human nature as irrational, it panders to the view that the only way to persuade people is through cheap tricks which trigger biases. By acting as if this is true we risk making it so. If we believe that there's no point arguing with some people - that they are irredeemably biased and irrational, beyond persuasion - we may abandon any attempt at persuasion by reasoned argument.

I've an optimistic faith in human rationality. We're not perfect, but we can connect with people who disagree.

So the challenge is to communicate effectively, without giving into the very partial view of human nature that psychology can seem to promote.

There is better and worse communication, yes; there is messaging which evokes our biases and so is more likely to be rejected, and messaging which works with the grain of the way we reason.

George Lakoff is a psychologist who is well known for his work on metaphors and framesa. Frames are the background ideas - metaphors - which determine the context for people's reasoning. His claim is that frames can be used to control the contours of a political debate, most notably in determining what people try and refute, and that each refutation reinforces the frame of the idea. So, for example, there is tax relief, the term for tax cuts promoted by US Republicans, which smuggles in the metaphor of tax as burden. So, Lakoff says, whether you are arguing for or against any particular case of tax relief you have conceded the general idea that tax is a burden, and we all know that, ultimately, burdens should be lifted.

A criticism of Lakoff is it opens the door to a sort of arms race where everyone tries to weaponise their language for maximum advantage. And maybe that would be true if we thought of framing as a cheap trick, a surface property which we could add to any messaging after we had already determined what we wanted to say. I'd argue a better understanding of Lakoff's framing is that it gives a way to connect our message to our common values, to share those values in way that connects with understandings that our audience already have.

An important part of Lakoff's book about political framing is the recognition that American conservatives have long recognised the importance of ideas, and funded institutions - from think tanks to talk radio - which seed the frames in minds of voters which political messages later target and exploit. Framing, in this view, is no surface property, but a way of quickly connecting to a deep history of ideas, values and community building.

To finish, I'd like to end on a positive example of framing from the city where I live and work. City of Sanctuary, is a network of local organisations, which started in Sheffield, which has the aim of creating a culture of hospitality for those fleeing violence and persecution.

Notice the framing, and how it differs from the dominant metaphors surrounding asylum and immigration in the UK. That debate is so toxic that the phrase asylum seeker seems to come with a silent "bogus" at the front, and immigrant with a silent "illegal". You could try and counter this with myth busting - show the statistics that most immigration is legal, explain the legitimate reasons for seeking asylum, but you'd be playing into the trap Lakoff outlines of reinforcing the frame of migration as illegitimate and suspect in general, even as you try and rebut it in the particulars.

City of Sanctuary sidesteps that and harks to a fundamental idea that we all recognise - of the sacredness of sanctuary, of protection for those who need it. It asks us to think about the duties of hosts, of those fortunate to have shelter, to share it with those in need. It's a brilliant bit of framing, and not superficial trick. In allows, in a few sentences, the fundamental values of an organisation to be summed up and communicated.

So, in conclusion, remember that the evidence on the psychology of communication often disguises as much as it reveals, that it has a bias toward showing the immediate influence of surface changes, rather than the enduring power of reasons, arguments and values. There are ways to connect our deep principles with persuasive messages, and I'm looking forward to discussing the details of that with you over the next few days.

This is more or less what I said at Wilton Park, 14 January 2019. For more on the counter-literature in psychology which shows the power of reasoned argument, see my 'For argument's sake: evidence that reason can change minds'. For a profound recent account of the psychology of human reasoning see "The Enigma of Reason: A New Theory of Human Understanding", by Hugo Mercier and Dan Sperber (my review of this book here).

Dessert tray

VOX: Is rationality overrated? Sometimes, it’s better to be reasonable.

Which starts by reporting a study which looked at how people play economic games and perceive those people’s actions in those games: “The study starts by distinguishing between two terms: there’s rationality, where you focus on maximizing the chance of getting what you want, and there’s reasonableness, where you strike a balance between what you want and social norms”. Which I’d say gives undue credit to, and at the same time, fails to really inspect, the logic of pursuing “what you want”.

(and from that piece I learnt of this book)

BOOK: Irrationality: A History of the Dark Side of Reason (2019) by Justin E. H. Smith.

Smith summarises his the key argument in this Vox piece

“the Enlightenment world has an innate tendency to degenerate into myth, reason into unreason. And that this tendency of reason toward unreason is exacerbated by overly ambitious efforts to suppress or eliminate unreason. I think this is true both at the level of individual reason, or “psychology,” as well as at the level of society as a whole”.

Fair enough, but, again, this honours the received notion of rationality and allow alternatives to be badged irrationality without sufficient critical inspection of the notion of rationality. Caveat: I haven’t read the book.

BOOK: Reason in Human Affairs by Herbert Simon (1983)

In the first part of this book Simon discusses different conceptions of rationality, particularly contrasting subjective expected utility with what he calls ‘bounded rationality’. Subjective Expected Utility (SEU) is a choice model whereby you multiply probabilities by outcomes and pick the option with the highest average. It plays a key part in economic theory, and is defining of what rational economic agents should do in many models. Simon makes a few worthwhile notes:

• The SEU model assumes an Olympian view whereby the agent weighs all possible choices simultaneously, using knowledge about the costs and benefits. “human beings have neither the facts nor the consistent structure of values nor the reasoning power at their disposal that would be required, even in these relatively simple situations, to apply SEU principles” (p17).

• SEU is a framework which allows choice to be optimised, but this is a different thing from guaranteeing that you have the best answer. The model tells you the best choice according to the model, but it is doesn’t tell you what the best choice is according to the world.

• The model is value neutral. “The SEU model finesses completely the origins of the values that enter into the utility function; they are simply there, already organized to express consistent preferences among all alternative futures that may be presented for choice. The SEU model finesses just as completely the processes for ascertaining the facts of the present and future states of the world. At best, the model tells us how to reason about fact and value premises; it says nothing about where they come from” (p14).

Which leads us to the alternative, “bounded rationality” which is to assume we often satisfice in our choices, dealing with the problems in front of us, one by one, avoiding the largest errors by applying a mixed bag of short-cuts and rules of thumb. Part of the achievement of the research programme in psychology which Simon was part of kicking off is to show just how successful strategies for decision making which don’t require Olympian levels of knowledge and processing power can be.

Although bounded rationality is arguably more plausible than SEU as an account of human choice, it a ton less satisfying. SEU presents a unified, coherent, theory, which offers tractable solutions in exchange for insane assumptions. Bounded rationality is a family of possible mechanisms, with no clean set of predictions. You can see why lots of people like working within SEU framework.

You can enjoy the key bit of Simon’s book in this PDF (which is annoyingly scanned back to front, sorry).

BOOK: Smarter Than You Think: How Technology Is Changing Our Minds for the Better by Clive Thompson

Thompson is unsurpassed as a commentator on technology and culture. This book shows off his measured optimism for the ways technology can enhance our thinking. No dull moments, fresh with anecodotes and evidence, and he even changed my mind about whether rhetorical questions are forgivable in prose (he uses them to good effect, previously I would have advised a blanket ban). 5/5

(ps I am stoked to read his Coders: The Making of a New Tribe and the Remaking of the World when it hits paperback later this year)

LINK: Ian Leslie’s profile of Malcolm Gladwell: “his big idea is that ideas should be pleasurable”

XKCD: Logical

END