Pluralistic Folk Psychology and the Theory-Simulation Debate: Mapping the Conceptual Landscape, Annie Mukherjee

On Wednesday, 3 June, Annie Mukherjee gave a PhD lunchtime talk titled ‘Pluralistic Folk Psychology and the Theory-Simulation Debate: Mapping the Conceptual Landscape’.

The abstract is below: 
The Theory of Mind debate in philosophy of mind examines how we understand and predict others’  mental states. Traditionally, this debate has been framed by the distinction between theory theory  (TT), which holds that we rely on an implicit theory of folk psychology, and simulation theory (ST),  which claims that we understand others by imaginatively simulating their perspectives (Goldman,  2006). Although extensively developed in philosophy and enriched by empirical research in  psychology, neuroscience, and cognitive science, the debate has remained largely conceptual or  experimentally pursued by non-philosophers. Then came the pluralistic folk psychology accounts  (Spaulding, 2018), emphasising that folk psychology is complex, context-sensitive, culturally  variable, and often narrative-driven. On this view, such diversity undermines the traditional TT-ST  framework as too narrow. However, these pluralist critiques are themselves largely theoretical and  do not decisively demonstrate that surface-level variability reflects deep conceptual disunity.  

My thesis argues that experimental philosophy (Knobe & Nichols, 2008) provides a methodological  bridge by treating the conceptual architecture of folk psychology as an empirical question. Using  systematic surveys of ordinary intuitions, it examines whether people rely on the traditional  framework at all and, if so, whether they primarily employ theory-like reasoning, simulation, or a  flexible combination of both across contexts. Rather than discarding the traditional model, this  approach tests its empirical resilience and investigates whether TT and ST operate as  complementary strategies within an adaptive conceptual structure.

About the Speaker:
Annie Mukherjee is a Senior Research Fellow at the Department of Philosophy, Jadavpur  University. Her work is situated at the intersection of Philosophy of Mind and Experimental  Philosophy, with a particular emphasis on the application of experimental methodology to evaluate  certain theories of mind. She has two forthcoming publications in this domain. She completed her  B.A., M.A. (as a Gold Medallist, securing first position in her class), and M.Phil. in Philosophy  from Jadavpur University.  
Beyond her academic pursuits, she is passionate about music and regularly performs on various  platforms.

Transcript

SPEAKER_00 0:04

Hello, welcome everybody, and thank you so much for joining us today. And welcome to Annie Muguruji, today's speaker. She is a senior research fellow at the Department of Philosophy at Jada Pur University. Her work is situated at the intersection of philosophy of mind and experimental philosophy with a particular interest on the application of experimental philosophy. Sorry, experimental methodology to evaluate certain theories of mind. She has two forthcoming publications in this domain. She completed her BA, MA as a gold medicine as a goal medalist, securing her first position in her class, an M Phil in Philosophy from Jada Fur University, beyond her academic pursuits. She's passionate about music and regularly performs on various platforms. Thank you so much for joining us, Annie. Over to you.

SPEAKER_01 0:53

Thank you, Julia, for that introduction. Good afternoon, everybody, and thank you to everybody everyone who's joining this session right now. I would like to directly go into my presentation. So just let me try sharing my screen. So the talk that I'm going to deliver is titled Pluralistic Folk Psychology and the Theory Simulation Debate, Mapping the Conceptual Landscape. So, what the term folk psychology actually refers to is something that we all do constantly in our daily lives. That is the act of attributing mental states to others for the prediction and explanation of behavior for others and often for ourselves as well. When you wonder whether your friend is upset or why the person in the Metro queue pushed you to get ahead, you're actually doing folk psychology. What I want to trust today is how philosophers have tried to theorize this capacity and whether a recent wave of pluralistic critiques requires us to abandon the traditional framework of theory and simulation. I'll be covering a fair amount, so I want to be moving at a pace now. So getting into the phenomenon directly. What is folk psychology? Folk psychology is also known as mind reading, theory of mind in the literature. It refers to our socio cognitive abilities aimed at understanding, predicting, and explaining others' minds through mental state attributions, mainly in the form of propositional attitude attributions, that is the attribution of beliefs, desires, intentions. It's analogous to something called folk physics. Just as folk physics is the commoner's understanding of physical forces and the behavior of nonsentient material things, folk psychology is similarly a commoner's understanding of mental states and behavior. Importantly, there are two senses in which we can speak of folk psychology. First, as an activity. Second, as a substantive, theoretical and conceptual structure that we deploy in individual acts of mind reading. This distinction matters because the two main traditional families of theories each latch on to one of the senses, as we will see shortly. The debate that we're discussing has a very well-defined historical arc. It begins in 1944 with Haider and Seymail's classical study involving geometric shapes. When participants observed triangles, circles moving on a screen, they spontaneously described them as intentional agents, chasing, hiding, feeling scared, bullied and getting bullied. So this pointed to something automatic and universal in our capacity for mentalistic attribution. Then in 1956, with Sellers' Empiricism and the Philosophy of Mind, Sellers initiated a more explicit philosophical discussion, proposing that terms like belief and desire are to be treated as theoretical terms, not direct reports of our inner states, but part of a theoretical framework which we deploy to explain behavior. This set off an intense debate in the philosophical arena in the 1980s and 1990s, and this is the time when the theories like theory theory and simulation theory came up, as we will see shortly. Then at the beginning of the 21st century, this debate gives rise to what is called pluralistic folk psychology, which is the subject of the second half of this talk. So, what are the standard theories? Theory theory on the left posits that mind reading capacity is something we possess, is the outcome of a theory of mind, a body of knowledge to which we refer in order to ascribe mental states to others and to explain and predict their behavior. It is taken to be a theory because it involves hypothesizing unobservables, which are mental states, from observables, which are speech, expression, behavior of different kinds, and using those unobservables to explain further observables. This shares a parallel feature with scientific theories if we think. Now think of David Lewis's formulation. Collect altitudes you can think of regarding the causal relations of mental states, sensory stimuli, and motor responses. That is just the structure of folk psychology. Now, simulation theory on the right takes a much less convoluted approach. It claims that we understand others' minds not by applying a theory, but by simulating their mental states in our own minds, mentally slipping into their shoes. Now, the proponents were God and Goldwyn Heel, and it is argued that when I predict my chess opponent's next move, I kind of swap roles mentally with her. I feed her relevant beliefs and objectives into my own decision-making system, my own process, and then project the output, the result which I think onto her. Both theories generated substantial empirical and philosophical literature, including debates about whether this capacity is innate or acquired, universal or culture-specific, conceptual or subpersonal, later came hybrid theories combining elements of both. But the focus remained on prediction and explanation of belief desire reasoning. What is now called pluralistic folk psychology marks a significant shift in the point of emphasis. The received view until then was that folk psychological abilities were primarily to be understood as prediction and explanation. Sorry, mind reading abilities analyzed in the terms of prediction and explanation of behavior. Other uses of folk psychology, the more directed acts of mind reading, were kind of largely ignored, according to these philosophers. Now, PFP challenges this. Pluralists contend that we engage in folk psychology for many purposes besides prediction and explanation. Our thoughts about minds change depending on the situation and our evolving folk psychological objectives. We rely on sociocultural situational norms, conventions, stereotypes, implicit biases, individual character traits, the full gamut of social cognition. And crucially, PFP is a big tent, it is not a singular theory. Some pluralists like Andrews and Zavitsky aim to supplant the standard theories. Others like Magyar aim to reinterpret and reframe the function of folk psychology. Others like Vestra and Spaulding, who are important in my context, work within the traditional framework aiming to improve the traditional framework. I'll cover representative examples from each of these camps. Coming to Andrews, Kristen Andrews introduces pluralistic folk psychology as a new research initiative exploring the diverse range of strategies and aims people use when trying to understand others' behavior. Her key move is to draw on social psychology to argue that rather than focusing solely on belief-desire reasoning, people regularly use alternative tools such as stereotyping, trait attribution, egocentric biases, and reliance on situational cues. These strategies work together and in dynamic ways. She also makes an important distinction. The theory theory and simulation theory, she thinks both overlook is that prediction and explanation are two different processes with different goals. Drawing on Baswan Frossen's Pragmatic View, Andrew argues that prediction aims at accuracy. While explanation mainly aims at curiosity, it tries to satisfy the why question in a way that meets the needs of the questioner. This leads to her pluralistic folk psychology, which has the following key claims. Folk psychology is a social skill or competence. It relies on a variety of cognitive mechanisms. Intentional behavior arises from a range of elements beyond beliefs and desires, like moods, emotions, personality traits, historical context, and recognizing intentional agents in the first place is a basic requirement of being a folk psychologist. Now, coming to Maggie's regulative folk psychology, Maggie's approach, in Maggie's approach, folk psychology is largely about forming and regulating one's own and others' mental states in light of a socially shared norm rather than simply interpreting or explaining them. She makes three central claims. First, mind shaping comes before mind reading. Our ability to successfully interpret others often depends on shared social practices that shape how people think and behave. A child learns to take turns and share, not because she immediately understands the beliefs and desires or the feelings of the adults, but it is because the adults kind of reinforce these behaviors, shape these behaviors. The predictability created by such shaping makes the social world an easier place to live in and it makes an interpretable world. Second, social support continues into adulthood. Folk's psychological abilities don't just develop in childhood and then operate independently. Even as adults, we rely on social norms, expectations, and feedback to guide how we understand ourselves and others. Think of how we actually behave with each other in a workplace, in a meeting, in a social gathering. Third, and perhaps most importantly, mind reading or mentalizing is a skill, not a module. It is not a built-in module to be specific. It is a learned cultural skill that reflects trained context-sensitive know-how, not the output of a built-in mental module. This has an important implication. It invites the question that whether there is a single universal folk psychology or many culturally specific folk psychologies. Now, mind shaping by Zavitsky. According to Zavitsky, what primarily distinguishes human from animal social cognition is not our capacity to read minds, but our capacity to shape them. Mind shaping, as he calls it, manifests in four main forms. And this capacity that he's talking about is complex, diverse, and flexible. Sophisticated imitation, children learn to express socially acceptable emotional expressions by imitating others, pedagogy. We teach each other not just what to do, but how to think, adopting cognitive styles, norm compliance, there are rewards and punishments for acceptable and unacceptable behavior, and narrative self-constitution. People shape their identities by the stories they tell themselves. If one is repeatedly told that they are responsible, at a certain point that person starts behaving responsible by internalizing that role. Zavitsky's central claim is that without mind shaping, advanced language, collaboration, and sophisticated mind reading would not be feasible. By placing mind shaping first, he suggests that the capacity to attribute propositional attitudes, that is classical mind reading, depends on being shaped in shared cultural, narrative, normative, pedagogical practices that make minds sufficiently similar or intelligible to each other. Propositional attitude attribution only becomes reliable after a certain shaping has already occurred. And according to him, without mind shaping, no other components of the human socio-cognitive syndrome, what he terms as human socio-cognitive syndrome, would be possible. That is, language cooperation and what he calls sophisticated mind reading. So we can see a common thread running through these theories, but these are just my examples to elaborate what PFP is basically aiming at. Now I come to the two important theories in the context of my thesis here. So Spaulding's model theory. Shannon Spaulding's model theory reconceptualizes mind reading as a form of theoretical modelling. Drawing from philosophy of science, specifically from Ron Gehr's view, that scientific theories are best understood as dynamic models rather than sets of propositions. On this view, mind reading involves constructing and applying simplified psychological profiles, which are models, to understand complex social interactions. The model takes inputs, which can be situational context, social categories, implicit associations or in-group or out-group features, personality traits. Then the processes can be a process of careful deliberation, of projection, stereotyping, self-interest distortion, and then finally the output, which can either be explanation or prediction, social regulation, social signaling, a mix or whatever, according to the context that that person is in at that moment. But another key claim that Spaulding actually brings up is that mind reading is more error-prone than traditional theories assume. In real-life situations, people often rely on heuristics and stereotypes, especially when quick decisions are needed or when maintaining social harmony matters more than accuracy. Different patterns of explanation also emerge depending on whether we succeed or fail. Failure is attributed externally, success is internalized. Think of a student who blames the difficulty of the paper when they don't do well and but credits their own intelligence when they are actually faring better. Also, a good example is our favorite football clubs, and we can all think how we externalize and internalize their successes and failures depending on how they win or lose. Now, coming to Wester's points of difference with regards to Spaulding's account. His first challenge is that Spaulding is overtly pessimistic about error. The heuristics she flags as especially error-prone may in everyday actual practice be much more reliable than she permits it to be. Think of a simple example. I meet a friend after work and I notice that she is constantly distracted and worried about something. So I guess that something at work must have gone wrong, it must be bothering her. Later, when I call her up in the evening, she confirms that is exactly what the case is. So Vestra argues that Spaulding doesn't give sufficient credit to these regular mundane successes. Second, Vestra questions the measurement of accuracy. Current empirical methods for tasting mind reading performance tend to highlight fringe or unusual scenarios where predictability is poor or bias is more likely. This may overstate the degree of error in real-life everyday mind reading. Third, and more constructively, Vestra argues that heuristics might be good enough. Egocentric projection, for instance, tends to be accurate precisely because people self-select into contexts where beliefs overlap significantly. For example, take this academic meeting right now. Think of workplaces, think of online communities. Similarly, many demographic and cultural stereotypes track actual patterns even if they are not universally applicable. He frames this as a trade-off between efficiency and precision. Heuristics are ecologically rational, just as pointing out optical illusions doesn't prove that vision is generally very unreliable. Listing biases alone doesn't tell us how often they actually lead us to errors. Now coming to his own positive proposal, which is the action prediction hierarchy. The idea is that we don't simply observe someone act and then attribute beliefs and desires from scratch. Rather, there is an organized system in which higher level, more stable features like character traits shape our expectations, at a lower level, more specific features. Like higher level features like trait attribution may narrow down our belief-desire expectations, which may help us to attribute mental states or explain and predict behavior. So at the highest level, we attribute stable character traits, for example, generosity, impulsivity, dishonesty, which helps us form general expectations about a person's behavior across different contexts. Below this, we rely on social roles and situational scripts. Further down, we attribute goals and intentions, and at the lowest level, we interpret immediate beliefs, perceptions and emotions from real-time cues like gaze, posture or facial expressions. Westra illustrates this with a vivid example. Suppose you're observing a person called Tom and whom you believe to be dishonest. A woman walks past him and accidentally drops her wallet in front of him. Because you know him to be dishonest, you assign a high probability of the hypothesis that Tom desires to steal the wallet and you predict a series of corresponding actions. The prior trait attribution serves as an overhypothesis, raising the prior probability of mental state hypothesis consistent with that trait. Now, if by chance Tom comes and returns the wallet, that creates a prediction error, which then revises the initial trait attribution. This hierarchical structure offers what Westra calls a cognitive economy. Now, three overarching problems with PFP. I am basically drawing on Gibson's 1929 2019 paper, which raises three broad problems that can be fielded against the pluralists in general. The first is the interaction problem. PFP advocates for a variety of methods for prediction and explanation. But how do all these methods actually interact? How do belief desire reasoning, role reasoning, and characteristic reasoning complement or supersede one another? PFP provides plurality, but it does not actually provide a story or a picture of how these strategies combine and come about in a given situation in real life. The second is the prediction problem. Many of the non-belief desire prediction techniques that pluralists propose, such as trait attribution and dispositional reasoning, only turn out to be successful because they implicitly presuppose a tasked attribution of propositional attitudes underneath them at a lower surface. For example, take Sunny, who is a very generous man, who makes a large donation to charity. So we infer that he is generous and later predict that when he goes to a restaurant, he will tip them very well. But this prediction only works if there is a hidden belief-desire assumption underneath. Even when we don't consciously think that Sunny believes X and desires Y, Sunny believes that tipping well in a restaurant is an appropriate or meaningful way to actually express generosity. Sunny wants to act in ways consistent with his generous disposition. Thus, trait based explanation works only when the relevant belief desire connect the trait to the action, the trait here being generosity and the action here being tipping. So, um, sorry, um the third Is the difference problem? If trait attribution frequently incorporates belief, desire, reasoning and all methods interact, then in what sense are the strategies truly distinct from one another? Are they genuinely different mechanisms or merely surface variations on a single underlying process? Now coming to the limitations of both proposals. So we saw that Vestro attempted to resolve the interaction problem by using the action prediction hierarchy model. But there are serious difficulties that come up here. On the empirical side, children are able to attribute beliefs and desires well before they can actually reliably use trait-based explanations. This suggests that propositional attitude attribution develops earlier and may be more fundamental than trait attribution, which is the reverse of what the hierarchy predicts. Studies by Barston Wellman support this. Furthermore, empirical work by Malin Holbrook shows that people make quicker and more frequent inferences about beliefs, desires, and intentions than about character traits, again undermining the idea that trait inferences have a higher level priority in the routine prediction. Model theory, meanwhile, faces a different set of difficulties. It risks over-intellectualizing mind reading by reducing folk's psychological understanding to a set of cognitive skills involving model building and application. Now, this creates a tension within the PFP's own presuppositions as a project that folk psychological strategies like trade attribution, mental state inference, are grounded in distinct cognitive mechanisms and not just general modeling abilities. Spaulding also does not quite specify what is the exact nature of this model she proposes, which means that there is currently no way of getting it tested in real-time situations. So we see, in short, that neither model theory nor Vestra's hierarchy provides a cohesive solution to the interaction problem. Now, I want to make a broader claim here. If the PFP theorists often characterize mind reading as something which is quite messy, and we find this word repeatedly if we go through the literature that they have produced. Now, what is messy? Messy is context sensitivity, variability, susceptibility to bias. But this rhetoric risks conflicting messiness with complexity. Complexity implies structured, self-governed systems that, while intricate, are still coherent and theoretically tractable, like think of language or formal logic. Calling mind reading messy suggests an absence of order or explanatory regularity, which may misrepresent the cognitive architecture of social cognition. Long before theory theory and simulation theory entered the scene, classical thinkers like Kant, Davidson, Strawson, Putnam, they all highlighted that the normative role of rationality in interpreting mental states, the rational expectation that agents act in ways intelligible in light of their beliefs and desires and goals provides the structured framework for prediction and explanation in the very first place. This structure forms the backbone of any viable account of social cognition. When PFP gives priority to cognitive biases and contextual factors to such an extent that it risks undermining the role of rational patterns, it is actually swaying off the branch that it is sitting on. The explanatory strength of any theory of mind reading lies in its ability to generalize across contexts and provide systematic predictions about behavior. What we cannot trade off theoretical precision for empirical inclusivity, because that would ultimately diminish the theory's utility. Both PFP in general and model and Spaulding's model theory in particular, in their attempt to reflect the empirical richness of human mind reading, how fascinating a phenomenon it is, risks becoming so open-ended that it might end up accommodating any behavior post hoc. Without clear constraints, PFP risks becoming a descriptive umbrella rather than a cohesive explanatory framework. So now what is the way forward? My claim is that pluralistic folk psychology provides important insights. Mind reading is indeed goal-sensitive, sometimes biased, and deeply embedded in social context. But these features do not necessitate abandoning the theory theory and the simulation theory framework in itself. Instead, what the gesture towards is the need for nuanced, empirically tested extensions of these models or these frameworks. The conceptual core of TT and ST remains quite stable. Just as grammar underlies varied speech acts or vision persists despite optical illusions, the conceptual architecture of mind reading endures beneath its messy surface. Moreover, many pluralist suggestions can be incorporated within the traditional framework itself. Heuristics and context-dependent rules can be made to work within the TT paradigm. Projecting ourselves into someone else's perspective is a fundamental aspect of simulation. We can adjust simulations according to both our own and the other person's perceived goals. So the goal-based context dependency claims comfortably sit within the ST paradigm as well. Theories can evolve while keeping the conceptual core stable. Think of how theories of gravity in physics have developed from Newton to Einstein without actually abandoning the core commitment to gravitational explanation. Pluralism in this reading becomes a research program for extending and refining TTST and not a wholesale critique of it or a wholesale abandoning of it. Now, the pluralist critique of TTST remains largely theoretical. It offers no substantive empirical evidence that disconfirms the TTST central claims as we have seen. This is where I think experimental philosophy might enter the scene and become especially valuable. By integrating methods from psychology, cognitive science, philosophy, experimental philosophy allows us to ask rather than presuppose whether people rely on generalizable explanatory principles which would support TT or whether they tend to simulate others' perspectives, which would support ST, and under what conditions one strategy is favored over the other if they're favored at all. Systematic manipulations of background knowledge, cultural norms, social context, cognitive load can reveal whether a core conceptual structure persists beneath surface variability of folk psychological explanation. In TT and ST, if TT and ST withstand these tests even in a minimal or a local form, that provides evidence for some degree of shared conceptual structure in folk psychology, even if its outward expressions appear varied, pluralistic, or narrative-bound. The presence of bias and context sensitivity in mind reading may not be taken as a flaw in the TTST framework, it is a prompt to refine it. Pluralism may draw our attention to the surface-level diversity of folk psychological explanation, but until it demonstrates this diversity reflects deep conceptual disunity at the level that TTST addresses, it remains a cautionary perspective and not a decisive alternative altogether. So I would like to end by making a few summary claims. Folk psychology is undeniably complex, but complexity is not chaos. The surface level diversity of our mind reading practices does not mean there is no underlying structure. Theory theory and simulation theory do remain vital and evolving research programs, not obsolete paradigms. We have already seen theorists like Spaulding and Bestra trying to work within these frameworks to forward their pluralist positions. Pluralism draws attention to the surface level diversity, yes, but it does not yet demonstrate that this diversity reflects deep conceptual disunity or dysfunction from the underpinnings that the traditional theories have provided or continue to provide. And finally, empirical interdisciplinary inquiry, not only theoretical pluralism, should be the next step. We should focus on testing and optimizing TT and ST rather than rejecting them on the basis of theoretical arguments alone. What I wish to say is I want to conclude by suggesting that the paradigms of explanation provided by TTST are deeply rooted in our intuitive mind reading uh practices and they cannot be just nipped off in the bud without shaking the roots. So that was my presentation. Thank you. I'm ending it here.

SPEAKER_00 31:12

Thank you so much, Annie. That was really interesting. I'm now gonna stop the recording for the QA. So if you're watching the recording, thank you for tuning in.