Admissions Straight Talk

I was recently contacted by a fan of the show asking for advice in the choice of their research topic. Oddly, the best advice I could give them pertained to philosophy of science. In this episode, I expand on what I told them, to explain the most important ideas in the philosophy of science that I think are worth knowing about. My ultimate target is Imre Lakatos. If you can understand Lakatos' idea of research programs then you have all you need. However, in order to properly understand Lakatos, you need to know about Thomas Kuhn's concept of paradigms, "normal" science, and scientific revolutions; and to understand Kuhn, it helps to have some grounding in what Karl Popper had to say about falsificationism before him. I add in Francis Bacon as the founder of science and William of Occam just to mention his timeless razor. The "feel" of this episode is to say that science is done by people, and people are imperfect. Philosophers of science have long discussed the ways in which science should be (or is) done, and as we get to Kuhn and Lakatos it becomes more a case of explaining the often infuriating way that people don't seem to rationally update their worldviews. Paul Feyerabend makes an appearance between Kuhn and Lakatos in this episode to show off some of the more anarchistic views of science, and to help demonstrate through the apparent "cheating" of Galileo and the failings of the Copernican system how sometimes a little stubborn imagination in the face of rational refutation has been good for scientific progress. I am also very happy to bring in my favourite scientific paper of all time, The Myth of Language Universals: Language Diversity and its Importance for Cognitive Science by Evans and Levinson. I use it as an example of how a longstanding theory seems to be alive and kicking despite what seem to be decades of continual refutation, and how this exemplifies Lakatos' idea of research programs, while also seizing the opportunity to talk a little bit about the wonders of linguistic diversity. Overall, I hope that I can convince some people of the relevance of ideas in the philosophy of science to the much more practical issues of research, including why people who are wrong are so hard to convince. Enjoy the episode.

When we ask the question of whether something is "nature or nurture", we are implicitly suggesting a dichotomy, or excluded middle - it is either nature, or nurture, or a mix of both, but not a mix of both plus something else. In The Hidden Half, Michael Blastland takes us on a journey of skepticism which somehow magically reveals the presence not only of a "third factor", but shows, startlingly, that such a factor has been known to account for as much as half (!) of the variation in some traits. References to dark matter immediately spring to mind. After reflecting on a paradigm-shifting species of parthenogenic crayfish, the author discusses, among other things, how you only have a 50% chance of developing the same mental disorder as your identical twin you were raised with is already suffering from (shared genes and environment, remember - shouldn't this be close to 100%?); the mysterious inability to transfer infant mortality reduction measures to new regions; and the still inexplicable sudden drop in teenage pregnancies in the UK around 2008. One could see Blastland as occupying allied but opposite territory to the human irrationality parade, spearheaded by the likes of Daniel Kahneman and Dan Ariely. Blastland contends that yes, we do misunderstand the world a lot, but rather than focusing on the human weaknesses of irrationality, he draws our attention to the fact that the world is just really, really, intractably hard to understand. In that sense, you could see him loosely affiliated with Nassim Taleb. Another way of seeing Blastland is as a voice (however inadvertantly) in the incredibly important but widely ignored literature on transfer of learning. I covered Robert Haskell's book on the topic recently, and one of the central messages was that people don't seem to be very good at transfer of learning. Blastland's position would be that the world is not very amenable to transfer of learning. Again, it's not people's fault, the world just is that way. This book came as a powerful message to me on the limits of human knowledge (relevant to the enterprise of this podcast) and an alternative view on human irrationality and transfer of learning (relevant to its content). I would like to get a picture of a marmokrebs crayfish and put it in a frame on my desk, so that I never forget the book's central message. Enjoy the episode. *** RELATED EPISODES 110. Transfer of Learning by Robert Haskell 108. Expert Political Judgement by Phillip Tetlock 11. Thinking, Fast and Slow by Daniel Kahneman

So far, one of my most downloaded episodes has been number 42, on Sir Ken Robinson's talk Do Schools Kill Creativity? Numerous members of the audience have told me that they appreciated my critical eye on the matter. But at that time I had not read any of Ken Robinson's several bestsellers. "Don't you think you should? How can you be critical of him when you haven't even read him?" It was goading from someone else, asking me to rise to my own intellectual standards, that made me finally give in. I decided to read a book of his. Suffice to say, it was not a very good use of my time. Out of our Minds comes with the subtitle Learning to be Creative, and yet it gives very little concrete advice on that point. A jumbled book with no clear order, it mixes worn-out platitudes with click-baity list articles ("Nine qualities of a creative leader" - all that was missing was "Number six will shock you!!"). He weaves in talk of a "schism" between arts and sciences without properly defining what he means by this, as well as the occasional anecdote about children and potential that he is so well known for. Probably the most startling of his theses is that the current school system (by which he means children sat at desks learning knowledge from books and a teacher) is a product of the Enlightenment. Besides the questionable relevance of this to the book's pruported subject matter, the idea itself immediately makes no sense to anybody with the faintest inkling of the history of education. How, I wonder, does he think school looked before the Enlightenment? Did children in ancient Babylonia engage in "active learning" according to their kinaesthetic learning styles? Did Jesuit schools organise project-based learning and focus the curriculum around dance? Were Chinese boys studying for the Imperial examinations letting their curiosity go wild through the freedom of discovery learning? The irony that the intellectual tradition to which Robinson himself belongs owes much to Jean-Jacques Roussea's 1763 book Emile, and so is itself a product of the Enlightenment, is completely lost on him. But the problem I want to focus on is not the structure or even the content of the book, but simply the methodology. Robinson is revered around the world, and yet his arguments can hardly be called cogent, thorough, or research-backed. Sprinkling in clever anecdotes and aphorisms might work for a casual chat, but it doesn't do much when you're trying to seriously prove a point. I suppose this book is not meant to convince unbelievers, but for consumption by those already sympathetic to his message. I would like to remind you that I was at first a fan of Robinson, all those years ago. It was a combination of his charisma and my values that made me like him, and I was intrigued by his arguments (such as they were). Had I been better armed with knowledge of how learning really works I could have realised what was wrong from the outset. At least now, all those years later, I can share my criticism of him and hopefully make more people realise the flaws in his ideas - and his methodology. Ultimately, the questions that the work of Robinson raises are worth discussing, but properly. Let's not simply swallow his words whole without thinking about their validity. Let's think about what he is suggesting, and investigate it with proper scientific scrutiny and broad humanistic erudition, as far as we can. Enjoy the episode. *** RELATED EPISODES 42. Do Schools Kill Creativity? by Ken Robinson

Let's set things straight - intelligence isn't really *all* that matters. The editor seems to have forced a provocative title that even the author doesn't agree with. But intelligence does really matter, and the evidence on this point is overwhelming. Early on in my study of education, I was enamoured with Carol Dweck's Mindset research, and in all of my growth mindset zeal I couldn't bear to even consider that people might differ in some apparently "fixed" way. However, with time I have had the courage to face this issue; or, more realistically, it has beseiged me enough that I have had to give in. Intelligence is real, it varies from person to person, and it has a large heritable component. Research on intelligence has continued for well over 100 years, and it has several findings which are very well supported by evidence. The most important finding is the positive manifold, which states that all mental capacities - from vocabulary size to social intelligence to mental rotation to reaction time - are positively correlated with one another. The standard way of explaining this is that they all share a reliance on some general intelligence, or g-factor, in common. IQ tests aim to measure the g-factor. Another finding is that almost all of the things that one could wish for in a life well-lived also correlate with IQ: health, grades, earnings, and a successful marriage, to name a few. IQ also predicts patience, willingness to cooperate, win-win thinking, and generally being a good citizen. (For many of these, causation has also been shown.) IQ does not merely concern how well you can do a paper-and-pencil test, as many would argue. Due to the positive manifold, it basically concerns everything, which is an only slightly hyperbolic way of putting it. Intelligence: All That Matters serves as an introduction to intelligence and IQ research, a subject I would like to get into more on the podcast with time. I hope the general topic of IQ research will open your mind as much as it has mine. Enjoy the episode.

This is the second part of the episode about Robert Haskell's book Transfer of Learning. In this part, we go in detail into the importance of rich declarative knowledge for transfer. Topics include: The difference between surface structure and deep structure of problems, and how experts can see through the former to get at the latter, allowing transfer to happen How pure scientific discoveries with no application resulted in groundbreaking technological breakthroughs decades later Prolific inventors' advice on what it takes to be a great inventor (hint: "irrelevant" knowledge) How even specialists in procedural knowledge ("skills") openly admit that rich, well-structured declarative knowledge is the cornerstone of transfer Enjoy the episode.

One central question which I find very difficult to answer is "What is education for?". There seem to be many parallel purposes, most of which are subjective, ill-defined, and hard to measure. It is difficult to navigate between the Scylla of unrealistic expectations far from the core activities of school (e.g. developing well-adjusted entrepreneurial job-ready good citizens) and the Charybdis of uninspiring flat-footed apparent irrelevancies (e.g. hoping that they at least remember Pythagoras' Theorem). However, there is one aim that seems to me to pass both of these criteria: the ability to use what one has learned in novel situations. It seems absolutely necessary to make for a justifiable education - after all, if you are unable to apply your knowledge anywhere outside of the classroom, what's the use of learning it in the first place? On the other hand, it also seems eminently achievable and related to real classroom content. Application of what one has learned to new scenarios is known in the psychology literature as transfer of learning (or knowledge transfer), and it is the subject of Robert Haskell's book. Haskell makes the compelling point that gaining knowledge and skills plus the ability to adapt them to a range of unfamiliar situations is what can properly be called education; if you can only rigidly perform a set of tasks in their original circumstances, this should be considered merely training. After spilling considerable ink explaining what the term might really mean (scoring points with me for taking defining one's terms so seriously), the author tells us how, in over a century of research, psychologists have repeatedly shown transfer of learning to be nearly impossible to achieve. After walloping us with that shocker, he rallies us with the promise of his Eleven Points, a sort of manifesto which, he claims, can make transfer of learning possible after all. He then spends the rest of the book detailing what these points mean, and showing evidence (some convincing, some circumstantial) of the importance of each of the points in his program. Transfer of learning is an incredibly important idea, and it is shocking that it is so little talked about in education circles. Instead, we get lectures about "21st century skills", with very little basis in actual science, as far as I have been able to figure out. After critiquing "critical thinking" in the previous episode, this one serves to draw attention to a more promising-looking educational aim with around 120 years of research behind it, but which nobody seems to be talking about. Enjoy the episode.

"Critical thinking" is an idea commonly discussed in education. Most people who talk about it say we need more of it. Almost nobody seems willing or able to define it. I have trouble believing in it. With anything that I believe, I keep an open mind and even force myself against my cognitive biases to hear out those whose opinions I disagree with. This has been very useful to me in the past, as there have been a number of education-related ideas that I have had to eschew on further investigation. In line with this attitude, I was happy to give this book a chance, particularly since the author claimed to be a fan of the show, and therefore would presumably have some sense of my predisposition to this issue. The book opened with promise. The author writes that he is aware of the many criticisms levied at critical thinking - that "it cannot be defined, ... or that it takes away from the content of the course, or that it is different in every discipline, or that it depends on knowledge..." (all of which would be exactly my criticisms). "You name the excuse; we heard it." Clearly, I thought with a mixture of trepidation and excitement, the author will show in this book why those criticisms don't stand up to scrutiny - I will have to part with a long-held belief of mine again to get closer to the truth. He goes on to mention the success that he has had in teaching critical thinking: "Together [with collaborator Dave Carillo], we were charged with raising critical thinking outcomes across campus. And we did." (Emphasis in original.) So not only was I wrong in theory, the author wanted to show me in practice how effective he could make such interventions! Alas, I found no more mention of the details of his interventions, no description of what he did and evidence showing that it worked. Neither did he ever again mention the criticisms that I had, saying that he'd heard them all before, but declining to explain what is wrong with them. To add insult to injury, right at the outset he deliberately omits discussing what critical thinking actually is, claiming "that's a very complex question" which would require "a whole other book in itself", and overall it would be a "tediously lengthy affair". I would rather a lengthy affair than an undefined concept lacking evidence and avoiding facing up to criticism. Among the many failings of the book, there is one place where I picked up something new and useful - it seems that the research on project-based learning (PBL) is less negative in its conclusions than I had understood thus far. I have taken note to look into this more in future. To be clear, I still hold that critical thinking might be a thing, it's just that (a) the cognitive science I've learned so far seems to have no need for the concept, (b) I only ever hear people talk about it in a stand-on-your-soapbox high-sounding sermon without reference to any actual evidence or even theory, and (c) this book does a terrible job of supporting the idea, as do so many others who love to talk about it. Perhaps one day I will encounter work that explains what it really is, shows that it really exists, and demonstrates how important it is using proper evidence rather than rhetoric. Or maybe all there is on the topic is impassioned oratory rather than solid science. In case you are interested, Steven Pearlman co-hosts two podcasts: The Critical Thinking Initiative and Smarterer!. Enjoy the episode.

Phillip Tetlock is an expert on expertise, but of a different kind to the late K. Anders Ericsson. While Ericsson's work focused on experts within "kind" domains (as defined by Range author David Epstein) such as music and chess, where feedback is near-immediate and clear and the rules are known to all and stated at the outset, Tetlock is interested in those who specialise in "wicked" domains, such as economics and politics. These are fields in which we can't run experiments or train for specific, recurring situations; where the rules are unknown; and where the situation at hand is not bounded, but can be influenced by a myriad of unpredictable forces. The author's most important finding is that cognitive style plays a major role in deciding who is good or bad at predicting world events. He reaches for Isaiah Berlin's concept of the Hedgehog and the Fox: "the fox knows many things, but the hedgehog knows one big thing." Hedgehogs tend to view the world through their particular favourite lens, basking in the power and glory of their pet theory. Foxes, on the other hand, are much patchier in their choice of models, and continuously second-guess themselves as they piece together a tentative view of the many ways things could unfold. While it seems that hedgehogs would make good political leaders (and are much more charismatic), foxes outperform them on prediction tasks to a stunning degree - or perhaps we should say, hedgehogs perform so terribly that foxes have an easy win. Tetlock has published another book on the topic of expert predictions in the political and economic arena called Superforcasting: The Art and Science of Prediction, which is aimed at a general audience. Expert Political Judgement is an earlier work, written for academics, and therefore deeply concerned with methodology and epistemology, and more willing to discuss probability theory and mathematical modelling. The full details of this are not amenable to sharing via audio, but the approach does provide a reassuring amound of skepticism to his own conclusions. Enjoy the episode. *** RELATED EPISODES Expertise (in "kind" environments): 18. Bounce by Matthew Syed; 20. Genius Explained by Michael Howe; 22. The Talent Code by Daniel Coyle; 24. Outliers by Malcolm Gladwell; 49. The Art of Learning by Josh Waitzkin Breadth of learning (characteristic of "foxes"): 97. The Polymath by Waqas Ahmed; 98. Range by David Epstein Moral hazard of people who talk too much (generally scathing about economists): 84. Skin in the Game by Nassim Nicholas Taleb

Jaron Lanier is a Silicon Valley veteran and a pioneer of virtual reality technology. In Ten Arguments for Deleting Your Social Media Accounts Right Now, as you might expect, he outlines his dissatisfaction with the status quo of the Internet. He admits that his generation of Silicon Valley engineers and entrepreneurs made a grave error in the early days of the Internet by paradoxically believing in the power of everything being "free", while at the same time hero-worshipping successful billionaire capitalists such as Steve Jobs. Lanier is a very witty writer, and the book is peppered with eminently quotable phrases. I structure the recording around these. Here are some of my favourites: "Social media is biased, not to the Right or Left, but downward."; "Your character is like your health, more valuable than anything you can buy. Don't throw it away."; "Hypnosis might be therapeutic so long as you trust your hypnotist, but would you trust a hypnotist who is working for unknown third parties?". Having just recorded the epic nine-part episode of The Anthropology of Childhood, I wanted to go for something a bit lighter here. The topic is serious, but the approach is laid back. Enjoy the episode. *** RELATED EPISODES 104. Addiction by Design by Natasha Dow Schüll

This is the last part of the multi-part episode on David Lancy's The Anthropology of Childhood. I share the main points of what I've learned, what I think of the book, and the way that Professor Lancy summarises the main points of the text. Among the things we've learned are the differential ways that WEIRD (western, educated, industrialised, rich, and democratic) societies and most societies throughout history approach topics such as: Reproduction; Family structure; Parenthood; The social status of children; The value of children, economic and sentimental; The stages of childhood (baby, toddler, child with "sense", adolescent); Children's relationships with adults; Play; Learning and teaching; School; Discipline; Children's work and responsibilities; and How young people change society. I hope you learned a lot - I certainly have! Enjoy the episode.