The Great Psychology Delusion with Marek McGann
"Human beings are basically the least ergodic things in the known universe. We all individually change over time. We do things differently as individuals, and differently from context to context — let alone different people doing the same thing in different conditions."
“You can be pluralistic and still be doing good science. You can be disciplined and scientific without thinking there’s a single objective truth you have to be aiming for. Being pluralistic is very, very different from the idea that anything goes. It is simply a recognition of multiplicity.”
Read The Great Psychology Delusion here.
This is an academic psychology-focused episode with lecturer Marek McGann, whose work spans enactive cognitive science, embodiment, politics, feminist philosophy, and Science and Technology Studies. Andrea and Marek discuss his co-authored book The Great Psychology Delusion with Craig Speelman. McGann explains why “delusion” fits psychology’s persistence in treating long-critiqued assumptions as valid, especially the aggregation delusion: averaging group data and applying it to individuals despite human non-interchangeability and change over time, linked to the ergodic assumption and ergodic theorem conditions rarely met in human behavior. They discuss how averaging can create misleading “laws” (e.g., power law of learning), the research–practice gap in clinical work, psychology’s history and method-driven identity, and the need for disciplined, pluralistic, scale-aware science that better integrates perspectives and practitioner expertise.
00:00 Show Intro And Intro Marek
01:23 Book Thesis And Stakes
02:24 Aggregation Delusion Explained
03:54 Research Practice Gap
04:49 More Detailed Book Summary
07:47 Averaging Artifacts And Ergodicity
09:29 Careful Critique Not Anti Psychology
11:06 Warm Reorientation Sendoff
11:51 Conversation Begins
15:17 Why Call It Delusion
20:11 How Psychology Became Method Led
31:08 Aggregation Delusion Deep Dive
33:35 Ergodic Fallacy in Humans
35:21 Scale Slippage and Delusion
37:59 Research Practice Gap Explained
41:01 Clinician Code Switching
42:46 Many Scales of Mind
43:57 MRI Averaging Pitfalls
48:32 Method Silos and Identities
52:43 Care, Careers, and Canalization
55:27 GPS Model for Pluralism
01:00:33 Pluralism Not Relativism
01:02:58 Why Marek Cares
01:06:06 Psychology’s Moment of Change
01:06:56 Closing Thanks and Wrap
Marek McGann has been a lecturer in the Department of Psychology since 2005. His principal research is theoretical work on the enactive approach to cognitive science, which examines the mind more as something we do rather than something we have. This is also related to ecological approaches to psychology, which explore how behaviour and mental life can be examined by looking at what your head is in, rather than what is in your head. He also has a related interest in critical considerations of theory and scientific practice in psychology more broadly.
Marek co-convenes the ENSO Seminars, a series of online seminars with researchers from enactive and ecological cognitive science.
The paper Andrea mentions: Facing Life
"Human beings are basically the least ergodic things in the known universe. We all individually change over time. We do things differently as individuals, and differently from context to context — let alone different people doing the same thing in different conditions."
"We do one thing, think we're doing something else, and don't have the capacity to reflect and realize the sliding between scales. Assuming that research with samples and averages has given you insight into individuals and their individual behaviors — that's a delusional thing."
"You can be pluralistic and still be doing good science. You can be disciplined and scientific without thinking there's a single objective truth you have to be aiming for. Being pluralistic is very, very different from the idea that anything goes. It is simply a recognition of multiplicity."
— Marek McGann
Transcript
Andrea Hiott: Hi Marek. Thanks for being on Love and Philosophy.
Marek McGann: Pleasure to be here. Thanks for inviting me on.
Andrea: I met you in Austria about six months ago. I realized we had a lot in common, and then you mentioned a new book coming out — which is now out. I'm constantly thinking about research methods in cognitive neuroscience, so I'm really glad to find it. It's called The Great Psychology Delusion. Do you want to introduce it?
Marek: Sure. This is work I've been doing with my colleague Craig Spielman at Edith Cowan University in Perth, Australia. We've been working together for some time, and the book brings together different strands around theory and practice in psychological research. Our first publication together, back around 2013, looked at the use of averages and the mean in psychological research — the limitations, the underlying assumptions. Nothing terribly new to most psychologists, and yet despite that, everyone still continues on as though everything's fine, when we think it really isn't. This book is an attempt to bring those threads together: the often-implicit assumptions about human beings that underlie psychological research, and the implications for how we understand, interpret, and apply the findings. We try to elucidate the underlying issues in a fairly clear, and sometimes polemical, way. We note in the book that almost none of the issues we raise are new — they have decades of history behind them. But for whatever reason, they've never been adequately addressed. In the current fervor for methodological reform within psychology, it's worth articulating them again.
Andrea: The word "delusion" is really strong. Why did you choose it? It feels particularly important right now, across so many areas of life, but especially in academic work and psychology. As I read the book, I started to think about the contradiction almost embedded in that word in a positive sense — as if noticing what's going on could itself open a path into a more pluralistic, optimistic place. But I want to hear why you chose "delusion."
Marek: It is perhaps a little hyperbolic and polemical, but it articulates something we both find particularly frustrating in the dialogue around these issues within the profession. We open with the Oxford dictionary definition: "an idiosyncratic belief or impression maintained despite being contradicted by reality or rational argument." For so many of the points we address — the problematic use of averages, the weak and undisciplined way theoretical work is done in psychology — these have decades of criticism behind them. And yet mainstream psychological research just plows on as though all is well, or as though we can dig ourselves out of the hole by just digging faster and deeper. Or, in the current vigor of methodological reform, the idea is that if we just do our stats harder, the underlying in-principle problems will be massaged away by discipline and good intent. We really want to call that out. The community, or a large chunk of it, seems to blindly follow established practice without a clear enough understanding of why those practices exist or what they actually produce. If we can be a little more explicit about what we're interested in knowing, we might start to break this delusion. But we have to see it and name it as such first.
Andrea: That's kind of what I felt by the end. It seems like you're trying to make us realize how general and disparate and diverse all of this is, so that we can then be more specific. Does that make sense?
Marek: That's a very nice, succinct way of putting it. Yes, absolutely.
Andrea: And it's interesting because psychology is probably one of the most popular subjects people read about, study, or want to be part of. And yet you're showing that the fundamental ways we do it academically are deeply flawed. How do you see that contradiction? Is it because it's so important to us that we want to believe we're making progress?
Marek: There are a number of forces in play. One thing that underpins the whole book is that we draw on work in the history and philosophy of science — going back through science and technology studies and sociology of science, about forty years of work — which makes it very clear that science is a deeply human endeavor. It brings with it the characteristics of all human endeavors: the messiness of coordinating actions and fitting behaviors to standardized practice within a given community. Once you apply that lens to psychology as a discipline, you can track its development as a set of courses through universities and applied settings, and see how many of its threads are actually quite separate and never properly drawn together. Where they are combined, it's often for political or pragmatic ends rather than because of a coherence of scientific worldview.
There's a contemporary analysis, from the late eighties onward, of psychology as a scientific discipline being defined not by the set of phenomena it's interested in, but by the set of methods it uses. That is a characteristic not of scientific practice per se, but of a community of professionals. The narrative is: I'm interested in people, I want to understand the mind, I'm interested in behavior and experience. But to become a psychologist is to learn to use certain research methods, and the phenomena almost get pushed to one side. We have a tendency to force our phenomena into certain ways of asking questions rather than adapting the research methods to suit the phenomena. Psychology's founding narrative is that it was established as a laboratory discipline — around 1876, with the setting up of laboratories. Every other science starts with naturalism, with systematic observation of what's in the world. Psychology didn't really go through that phase. As a result, our phenomena are very much defined by that which can be produced in a laboratory, rather than that which tends to happen naturally in various contexts.
Andrea: I see the same pattern so often: what we care about most, we try to get rigorous about — and we end up going to the methods. That's understandable, because the methods really do take enormous skill and study. And because of that specialization, it can seem like whatever you find with those methods must be right. Do you see a relation between science having to defend its methods to remain credible, and the way what we actually care about gets lost in the process?
Marek: Yes, that's exactly the kind of logic we're looking at. We construct our phenomena in a way that is unusual for scientists. As a discipline, we've decided that certain ways of constructing phenomena are the only valid ones — and everything else gets relegated to the messy outside, the domain of practitioners, not hardcore scientific work. The general logic is that when physicists bring things into the lab, it's to stabilize an otherwise too-dynamic phenomenon from the real world. Psychology applied the same logic, except we didn't bring something in from the real world — we created it. There was nothing systematic about our observations of the real world that then said, let's stabilize this in the lab. The work really comes from philosophical traditions. In cognitive psychology particularly, our notions of reasoning, decision-making, and problem-solving are drawn from philosophical analyses of these things — how to go about them in a normative manner — rather than from going out and watching what people actually do day in, day out. And it's just taking us a while to really come to grips with that.
Andrea: The methods are also developing in different ways. Maybe we could talk about the aggregation delusion specifically. We've set up an experimental habit where one thing looks at another thing and applies methods to it. Part of what I'm pushing toward is: how do we break that into something more like a constellation, a multiplicity? It's so habitual — the stimulus-response framing, the group averages applied to individuals — that it can be really hard to even notice it. Could you walk through the individual-versus-group confusion?
Marek: This really was the foundational issue where Craig's and my work began. It's simply the underlying assumption that human beings are sufficiently interchangeable — that when you want to study the mind, you can gather a large number of people, produce data according to some protocol, and average across all of it, which "boils off the noise" so you're looking at something like a true measurement of whatever psychological variable you're interested in. But to do that, you have to make the assumption that human beings are interchangeable — that we're all essentially doing the same thing when we perform a given cognitive or behavioral task. This is sometimes referred to as the ergodic assumption: that the average of the group is sufficiently similar to the average of the individual, so that analyzing the behavior of the group gives you insight into what an individual does. That's fine for hydrogen atoms, or various aspects of chemistry. But human beings are basically the least ergodic things in the known universe. We all individually change over time. We do things differently as individuals, and differently from context to context — let alone different people doing the same thing in different conditions.
The notion of aggregating scores to approximate a truer score, which underlies almost all tests of mean difference, which are in turn fundamental to the vast majority of laboratory psychological work — that really is an untenable assumption that goes almost entirely unquestioned. There is a literature on it, as I said, and there's very little in the book that is genuinely novel to Craig's and my work — we're drawing on Peter Molenaar, Ellen Hamaker, and a whole host of others. But that idea of aggregating across human scores is diagnostic of an assumption about what human beings are and how they work that just doesn't stand up to scrutiny. It misrepresents phenomena at one scale — the behavior of a sample — as though they were phenomena at another scale — the behavior of an individual. Psychologists will say they're interested in understanding the mind, which tends to occur one per person. But in a large amount of our research, we don't look at individual human beings at all. We only look at samples and their averages. Not being able to see that mismatch is the delusion. Doing one thing, thinking you're doing something else, and not having the capacity to reflect and realize the sliding between scales — and then assuming that at the end of research with samples and averages you've got insight into individuals — that's a delusional thing. And it's indicative of the conceptual slippage that is desperately common in too much of psychological science.
Andrea: Right — it's this generality-specificity thing you bring up often. And something like over 80% of papers make general claims at the group level while being applied at the individual level. And as you say, whatever we mean by mind — this embodied experience of being in the world — is going to be different for every individual in very significant ways. But you also say somewhere in the book that we do get a lot from studying at the group level. What would it look like to open that up so we're getting more specific about what we're actually studying?
Marek: I think an awful lot is being covered by individual professional sensitivity. Particularly when you talk about the application of psychology in clinical or counseling or educational settings. There's what's called the research-practice gap, and depending on who you ask, it means different things. For researchers, the research-practice gap is practitioners not doing what they're supposed to be doing — not listening to researchers and applying the theories in a disciplined way. For clinicians, the research-practice gap is the extent to which the research just doesn't fit their needs. Because the research is conducted with samples — carefully selected, clean samples — in ways that individual cases almost never are. There is therefore a mismatch between the kind of knowledge produced in research settings and the kind needed in practitioner settings. That mismatch, that failure to grasp that there's an underlying shift in scale between a sample average and an individual case, can't be bridged by clever statistics alone.
Insofar as a practitioner is able to take insights from research and apply them effectively in day-to-day work with individuals, the sensitivity and individual responsiveness of that clinician to their individual clients is doing a huge amount of the heavy lifting. It's not that research provides no orientation — but the extent to which it actually supports practice with an individual client is because it's being digested, transformed, and implemented by the often implicit expertise of the practitioner. There's a lovely observation by Sanneke de Haan in Enactive Psychiatry, where she was a philosopher embedded in a working psychiatric hospital and also attended psychiatric conferences. She noted — without calling it code-switching explicitly, but we might describe it that way — that the psychiatrists tended to switch their ways of speaking in different contexts. At conferences, they talked about the brain all the time, with the assumption that we'll find underlying neurological function that explains psychiatric experience. But in the hospital, working directly with clients, they spoke in the language of clients: experience, behavior, concrete realities. The references to the brain were attenuated. And the clinicians seemed almost unaware of that code-switching. So even where a person has their feet in both worlds, their practice is doing something quite different in each. If we can just recognize that these movements are happening, we can reflect on them and orient ourselves more effectively. The issue is not that one scale is wrong and one is right, but that you should know when you're moving between them and what changes when you do. And in fact there are almost certainly more than just two scales in play. There's sociology and anthropology at the level of group behavior. There's the individual. There's neuroscience, cellular structure, and so on. These are all different scales at which phenomena occur, and they don't simply relate to each other in a simple reductive manner. When you switch between scales, you must be sensitive to how your perspective on the phenomenon is changing.
If I use an MRI machine to study neural activation while a person presses a button, then do that with lots of people and average the results, and call all of it "decision making" — the pressing of the button, the neural activation, the behavioral task, the average across participants — I've bludgeoned a whole set of crucial distinctions in the hope of gaining insight, and actually blinded myself. Whereas if we can shift between perspectives deliberately, knowing what we're doing when we do it, we'll end up with a much higher resolution of knowledge. And I think we will eventually learn to move between these different scales more effectively, which will start to help close the research-practice gap and force us to become much clearer and sharper in our theories and concepts.
Andrea: Yes — I sometimes talk about holding paradox and it can sound easy, but I think it's actually one of the hardest things. Opening to multiplicities is genuinely difficult. You talk about the mechanical and the clinical, the research and the practice, and we're still talking in either/or terms, which I understand, but it can be so overwhelming if you're in any of these fields. Especially to hold all of it in mind — that there's something very real, there's perspectival realism in a way, and therefore the group scales are giving us something important, and yet we also have to take every individual into account. And all of that we've just discussed is at play simultaneously. There are these multiple, fractal levels. A lot of people just can't — or don't want to — hold that. It's really hard. And I wonder if part of it is about just being aware of it, as you're saying, so that we can help each other find ways to hold it together. Because it's really hard to deal with all of this in the same way that what we care about the most, we need to make scientific — we need grants, we need to publish — and so we end up putting our time and energy into either very specific details or very abstract things, depending on the discipline we choose. But what we're getting to is: we need all of it. Does this make any sense in terms of the feedback loop problem you were describing?
Marek: I think it really does. And I think we're seeing — or will see — paradigm shifts arising because of it. The need for a pluralistic view is real. Even many of the psychologists I would include in the criticisms I've been making would probably feel hard done by and say, of course, I don't mean this to apply everywhere, of course I'm not saying this is universal. But when Craig and some of his students looked at how psychologists actually write about their results, we do tend to write in generalizing ways that make this organic switch between scales. And we don't have a consistent, ongoing practice of allowing for pluralism. Instead, what we tend to do is balkanize ourselves — set up an identity and fixate on it. There's a whole group of psychologists who might listen to this and think: well, I don't do that, I'm a qualitative researcher. Which is a delightful notion — qualitative as opposed to what, exactly? Why are we defining ourselves by our methods rather than by the phenomena we're interested in? And how often do the qualitative people actually talk to the quantitative people? How often is there a real effort to integrate and draw insight across them? The mixed-methods people are not being invited to the qualitative conferences. None of them are invited to the quantitative conferences in the same numbers. There's a comfort with separation — a kind of methodological apartheid of: these are the people I'm comfortable talking to, and there's plenty for us to say to each other without making it messier by talking to that other group. And there is, unfortunately, plenty enough — enough to sustain individual careers. But at the scale of the profession and the scientific community, it prevents insight and theoretical development. We've got to break those silos. We've got to get people away from identities based on methods. And that has to happen as a community, not just as individual researchers.
So it really is just that kind of forcing ourselves to step back and take note. And that's one of the things Craig and I are trying to do in the book — cover a number of those angles, because they're interlinked. The aggregation problem, the poor discipline around theory and conceptual development, the research-practice gap — these are not separate phenomena. It's trying to show the ways in which they're interlinked in a way that demonstrates how important the sociological interventions are going to be, not just the statistical and methodological ones. Maybe we need to not fully unpick the tangle — as you say, we just have to be willing to shift perspective and be a little more mobile than we've been.
Andrea: Yes. We have to hold the tension. It's uncomfortable, but we could help each other do it — we could still specialize, if we developed a way to come together and discuss it as multiplicity rather than either/or. My method is better than your method. Because I feel like a lot of it developed through competitive dynamics. And that's not necessarily a bad thing. But instead of competition, we could talk about care. Most of us come into these disciplines because we really care about these issues. And then it gets really hard — it's hard to make your career, hard to get through all the schooling, hard to just be alive and in relationships in these environments. And a lot of people really do come into it from a place of care. But then you get burned in certain ways, you find a group where you feel safe, you get uncomfortable because you don't know something everyone else in a different field knows, or people treat you a certain way because you're not in their group. All of this feels like part of the scales you're talking about too. And when you bring up the multiplicity and federation of psychology — ways we can open to that in our research — doesn't it also happen at these other scales, in our lives, in the way we can return to a motivating force that isn't hurting us?
Marek: Yes. And it's not just about getting closed off or burned — we also get channelized, or canalized, as it's sometimes described in neurological terms. I've become this type of researcher. I've adopted this set of methods. I can build a career with this. If I step outside of it, I'm actually weakening my career. And I can achieve a lot of success by just continuing to do this one thing. There's no bridge back. But I think it's all motivated by care. Most psychologists get into this because there's some aspect of human experience that is very much close to home — often personal and intimate. And I think if we start paying more attention to practitioners and including them in research discussions with a bit more agency, there's a real opportunity to overcome our fear of pluralism. Because that's where you see pluralistic work being done.
An example I always come back to is from Hasok Chang's work — I think it first appears in his book on water, H2O. It's an example that anyone who knows me is sick of hearing, but I think it's brilliant: the GPS system. That's an engineering achievement, but it depends on satellites launched into orbit using Newtonian physics. Those satellites carry atomic clocks that can only work because of our understanding of quantum mechanics. Those clocks use Einsteinian relativistic mathematics to identify a coordinate on the globe, which is then represented on a flat-earth map to allow for navigation by a human being moving on a street. All of those are mutually exclusive descriptions of the universe. You can't be Einsteinian and Newtonian at the same time. You can't be Newtonian and quantum at the same time. You can't be flat-earth and round-earth at the same time. And yet the GPS system works brilliantly. It is all of those things simultaneously, embedded in a large collection of human practices that makes it all make sense and doesn't worry about the ways in which things don't match. That kind of pluralism can be immensely productive.
Now, a big part of our complaint in the book is that psychology is messy and incoherent in the way it theorizes. But the messiness isn't the problem. It's the way we conflate that mess with reductionism — assuming that in any given case, this one concept is the same thing and should be treated as the same. Plural definitions of concepts are okay, so long as we're clear what we mean, how concepts relate to other things, and where they don't relate, so we don't commit what James Flynn calls "conceptual imperialism" — applying a concept developed at one scale uncritically to another. Being comfortable in the tension of pluralism is a practice that we as a discipline really need to start figuring out. Incorporating practitioners and therefore facing the experience of individual human beings as something crucial and un-averageable is a crucial first step. And just building into undergraduate teaching and professional practice some deliberate little kicks — go and listen to the qualitative people for a while, go and listen to the social people, go and listen to the cognitive people — to build in that silo-breaking as a standard practice, to help each other hold the tension. It's a strong disciplinary task, but they've done it elsewhere, and there's no reason to believe we can't do it too.
Andrea: I love that. And I want to add: it doesn't mean anything goes. That's also part of the tension — realizing all of this matters, it's real, it's connected, it's inseparable, it's influencing each other. We tend to either say "I'm the scientist looking at the thing from outside" or we start fighting about consciousness or matter. The GPS example is wonderful because you don't need to reconcile those perspectives. They're just different ways of trying to understand something inseparable and shared, and for that very reason they can't be fully reconciled. It's not an anything-goes idea, which is maybe why it's so tense — but also why that tension and care can be just as motivating as competition. Is there anything you want to add to that?
Marek: No, I fully endorse it. That is the fear for so many scientists — that the value of science lies in there being a single objective truth to which we can aspire. That's the thing that sets us apart and allows for progress. And while that might be a crucial motivating narrative that we'll have to address sensitively, it is clearly not true. Scientists are very capable of holding multiple truths in mind at the same time and being fine with it. The idea that this can only be managed if every one of those scientists also believes that ultimately there's a single objective truth they're all approaching from different angles — that they just haven't got there yet — that's clearly not essential, because we've been doing this for hundreds of years and we're still not there. And every time someone predicts the end of science in the next fifty years, everyone else knows they'll be wrong. You can be pluralistic and still be doing good science. You can be disciplined and scientific and not think that there's a single objective truth you have to be aiming for. Being pluralistic is very, very different from the idea that anything goes. It's very different from simple relativism. It is simply a recognition of multiplicity. And looking at and appreciating the constraints we're working under — that's valuable as individuals and certainly as a community.
Andrea: That speaks to what I was saying about how delusion becomes the tension we have to hold — that we are going to be a little bit delusional in a way, but noticing it changes it. We're constantly perpetuating the thing we're studying, and so there will always be a little bit of delusion. But noticing it changes it. I know we need to go, but just to close: why do you care about this, Marek?
Marek: I came to care about this — I guess I've always had a philosophical bent. My undergraduate degree is in psychology, but my early postgraduate work was in philosophical psychology. I then expanded into cognitive science, did a master's, and then a doctorate at the University of Sussex, with the group there that included the likes of Hanne De Jaegher and Ezequiel Di Paolo and Rob Clowes — a whole host of enactive thinkers. That's where I was introduced to enactive thinking. Throughout all of that I was also teaching an undergraduate module on critical issues in psychology, which suited my philosophical leaning, and gradually blossomed into one arm of my career. As I dug deeper into that material, it became increasingly difficult to teach certain things without being able to leaven them with the kind of critical perspective that those more conceptual and historical aspects of psychology require. So those two strands — the research around philosophy and cognitive science and enactive thinking, and the teaching of conceptual and historical issues in psychological science — gradually coalesced over the past decade and a half into a number of individual papers, pretty much all of them with Craig, around scientific practice in psychology. And the book is one culmination of that, where all of these things arrived together.
Andrea: You really do care about psychology as a discipline, then.
Marek: I really do. For a lot of psychologists, I'm not a psychologist — I don't do enough empirical work for their tastes, I get dismissed as a philosopher. But I consider myself a psychologist. I'm paid by a department of psychology. And psychology is going through a period of change. These discussions have gone back decades, but there's a vigor and passion to them at the moment that's unusual in the history of the discipline. And I think there is certainly reason to take that seriously and to work seriously with it. There's a real possibility for transformation and for an ever-increasingly successful psychological science on the back of it, if we can just figure out a few of these supposedly philosophical questions.
Andrea: Well, you've been holding that tension in exactly the way we were talking about, and that's a real service. Both you and your co-author demonstrate it very well in the book. We'll have to talk again. Thanks for doing this, for holding the tension, and for the book.
Marek: Thanks so much for having me. Great to see you.