On Outreach and Education for the Foundations of Physics

On Saturday 25th April, I took part in a discussion panel sponsored by FQXi at the “New Directions in the Foundations of Physics” conference in Washington DC.  My co-panellists were Sabine Hossenfelder and Dagomir Kaszlikowski.  Sabine has already posted her comments on her blog, and I largely agree with what she had to say.  However, since I wrote out my comments before the discussion, I might as well post them here, as it is just a cut-and-paste job for me.

The discussion was very interesting, and it evoked more passion from the audience than I thought it would.  Given the limited time, I did not give the best responses to all of the comments from the audience, so I have added a few thoughts on the discussion, particularly the points raised by Mile Gu and David Wallace.

Without further ado, here is my intro.

How do we convey foundational physics concepts to non-physics audiences?  It is obviously hard to do so in a way that is both accurate and accessible.  As this conference shows, it is difficult to do so even amongst ourselves.  However, this is not the main problem we should worry about.  To explain why, requires a diversion on the broader aims of outreach.

I think outreach has three main goals: INSPIRATION, EDUCATION, and ACTIVATION.

Inspiration is making physics seem cool and interesting, so that, for example, a high school student might decide to study physics at university.

Education is the obvious, we want people to understand more physics after the outreach than they did before.

Activation is perhaps less obvious, but it means that we want people to actually DO something after the outreach.  This might be voting for a politician who supports evidence based policy and science funding, or it might mean persuading people not to employ the services of a new age “quantum healer” who claims to resolve health issues holistically using quantum entanglement.

To my knowledge, there is very little research into the effectiveness of outreach for these various goals.  That’s worrying because it feels good to win an FQXi essay (as well as being good for the bank account), to get immediate feedback on a blog post, to give a public lecture to a large audience, and, despite the fact that I have no personal experience of this, I imagine it feels good to have a bestselling popular science book or appear on TV in flashy a documentary.  In absence of hard data on how best to spend our limited time and resources, we will continue to do the things that feel good, regardless of whether they are the most effective.

Nonetheless, I think it is fair to say that the likes of Neil de Grasse Tyson and Brian Cox are doing a pretty good job on the inspiration front, so the rest of us should not devote too much time towards that.  Regarding education, there is now voluminous evidence from Phs Ed research on what are the best methods of teaching physics to high school students and lower level undergraduates.  This research is still ignored by the vast majority of institutions, and given that we know that these methods work, I think we would be better off putting our efforts into implementing research validataed cirricula in schools and universities rather than trying to do it with outreach, the effectiveness of which is largely unknown.  Incidentally, one of the things we do know about good physics pedagogy is that it is largely uncorrelated from the personality of the instructor, which leads me to be suspicious of the personality-driven nature of much scientific outreach.

That leaves activation, and I think we could be doing a much better job here.  Not everyone is going to take action in the name of science, but that does not matter, so long as those who do it do so loudly.  We are past the age of mass media, so we need no longer always cater only to the GENERAL public, instead going for smaller niche audiences who are currently underserved.  In particular, I am thinking of the science fanboys and girls, such as the community of skeptics who like to debunk pseudo-science.  They may be a relatively small community, but they are also the ones most likely to act in the name of science.  Most of them can give you a coherent acount of evolution and why it is true, but ask them about quantum theory and you’ll likely get some vague mumblings about waves, particles and the uncertainty principle.  They’d like to understand things more deeply, but we haven’t given them the tools to do so.  I think this is because we have been far too focussed on making our popular accounts accessible to everyone, e.g. publishers always advise against
including any equations in pop physics books.  This advice is appropriate for the mass audience, but not if we are targeting niche audiences, who are probably bored of hearing the same vague and inaccurate descriptions in fifty different popsci books.

So, turning back to the question of how we should convey foundationsal physics concepts to non-physics audiences, it is almost impossible to do so accurately for the mass audience, and it is probably best to go for inspiration in that case.  However, we can, and should, target more accurate explanations, with more math and more subtle details, to those smaller communities who are already passionate about physics, and who are more likely to act on the knowledge when they have it.

Following these remarks, there are two points from the discussion that I want to address.  Firstly, Mile Gu raised the point that we want to direct outreach to the broadest audience possible, as we need popular support to change government policies on science funding, or at least to keep it at a reasonable level.  To this, I responded that only a tiny minority of people are going to change their vote based on science policy, compared to the big issues like the economy, education, and healthcare, so we are better off focussing on that minority.  I know think that this is wrong.  If there is a general consensus within society then this can influence the policy of all major political parties, regardless of whether it is a vote-changing issue.  An example of this is the issue of gay marriage.  Very few people in the UK would have changed their vote based on this issue alone, but because there was a general feeling in the population that allowing people to marry whoever they choose is a good thing, there was a political consensus that pushed the issue forward.  Similarly, if there were known to be a general consensus in the population that science funding for basic research without ties to immediate applications is a good idea, then there would be political consensus on that too.  For this, I think we need to go beyond inspiring the general public into thinking that science is cool, by also emphasizing that the process of science and technology development as a whole does not work without the freedom to think freely, without knowing in advance what, if any, applications there may be.  We also need to emphasize that science is not just a machine for generating economic growth, but also a key part of human culture, comparable to the arts and humanities, all of which we should fund for their own sake because they enrich the human experience.

Secondly, David Wallace cautioned against my advice to verify the effectiveness of outreach via empirical research, suggesting that to emphasize research too much might make us too bogged down to actually do much outreach.  Instead, he suggested a “let a thousand flowers bloom” approach.  Let people go ahead and do the outreach they want to do, and presumably there will be enough different approaches that we’ll eventually have the desired effect.

I think I answered this badly on the day, effectively conceding David’s point.  However, David’s approach is only valid if we think there is not such a thing as bad outreach, i.e. activities that actually harm the goals we are trying to achieve.  This is especially true if there are not selection mechanisms in place that automatically weed out the bad outreach in favour of the good.

There is a compelling analogy here with physics education.  Professors have been left to their own devices to teach in whatever way they want for decades, and they almost universally choose methods that are pedagogically sub optimal, such as just lecturing from the front for an hour.  These methods can actually harm people’s perception of physics, reinforcing the idea that the subject is too hard for them.  Personally, I think it would be better if all the future medical doctors undergoing their required physics courses came out with a positive impression of the subject, and a good understanding of it, rather than regarding it as an alien subject that is irrelevant for their careers.  It is only through rigorous research that we have developed better pedagogy that is gradually being accepted in physics departments, although we still have a long way to go.

My position on outreach is that, although we shouldn’t encumber every attempt at outreach with a rigorous research investigation, if we think there are widely employed methodologies that are actually harmful to the aims of outreach then we should verify this empirically, try to figure out what works better, and encourage change.

If there are harmful aspects in current outreach, I suspect they are mostly in things like TV documentaries and popular science books, which are driven by popularity and sales.  A literary agent giving advice on how to write a popular science book is not giving advice on how to best convey the science, but rather on how to best sell it to a publisher, who is in turn concerned with how many people will buy the book.  So the usual advice to avoid any equations and to emphasize personal stories over the science, might not be good advice for communicating the science, even if they increase popularity and sales.

I think the focus on popularity leads to many popsci tropes, which might turn out to be actively harmful.  For example, there is the focus on stories of “great men struggling with grand ideas”, which may accidentally reinforce the impression that science is too hard for most people and so they should not engage with it, and discourage under-represented minorities from entering the subject.  Similarly, there is an excessive focus on speculative wild-sounding ideas, as opposed to the basics, which may inadvertently give the impression that “anything goes” in physics, and make people question why they should believe scientists over and above politicians and/or their local preacher.

One experiment I would suggest to address this would be to give a bunch of people a popular science book containing a lot of speculative ideas, and a couple of weeks after finishing the book ask them to classify how speculative the various ideas presented in the book are.  A good choice would be Max Tegmark’s “Mathematical Universe” because he goes to great pains at the beginning to classify how speculative his various multiverses are, even including a table.  My hypothesis is that most readers won’t remember how speculative the ideas are, and that ideas from standard model cosmology would be conflated with those of various multiverses in terms of the degree to which they are established.  I expect people will mostly recall the ideas that sound cool, rather than those that are most supported by evidence.  I also suspect that it won’t matter how careful the author is to distinguish speculation from established science, which could be checked by comparing results from Tegmark’s book with any randomly chosen Michio Kaku book.

If my hypothesis is confirmed, then perhaps we could persuade authors to hold back on the speculation a bit, in favour of established science, particularly in a society where the general level of science literacy is quite low.  If they do include speculation, perhaps it would be better to do so with a more skeptical treatment, including a detailed criticism of the ideas.  Perhaps a book written by a small group of experts with conflicting opinions on the speculative ideas is a better way to do this than the traditional single-author popsci books.  Whatever you think about this, these are ideas that we could clearly benefit from investigating empirically.