Olivia Guest

We commend Sanbonmatsu et al. (2025) for centring metatheory and metamethod as remedies to “difficult research problems”. However, we wish to depart from their conceptualisation of what models are and what role they play in psychological theorising. Under computationalism, models are not a container for observations through being fit to neurobehavioural data and should not be held to the standard of providing us with numerical predictions. Computational cognitive models can only play their role of mediating from theory to data and back, if and only if we buttress their special status against confusions with other model types, such as with inferential statistical models of the data. This is a realistic goal for our fields. Ultimately, theory crisis or no, blurring the boundary between model, theory, and data in our metatheoretical calculus, results in many of the problems all authors have pointed out.

Contemporary artificial intelligence models owe much of their success and discontents to connectionism, a framework in cognitive science that has been (and continues to be) highly influential. Herein, we analyze artificial neural networks: (a) when used as scientific instruments of study and (b) when functioning as emergent arbiters of the zeitgeist in the cognitive, computational, and neural sciences. Building on our previous work with respect to analogizing between artificial neural networks and cognition, brains, or behavior (Guest & Martin, 2023), we use metatheoretical analysis techniques (Guest, 2024), including formal logic, to characterize two distinct tendencies within connectionism that we dub classical and modern, with divergent properties, for example, goals, mechanisms, and scientific questions. We also demonstrate how we, as a field, often fail to follow important lines of argument to their end—this results in a paradoxical praxis. By engaging more deeply with (meta)theory surrounding artificial neural networks, our field can obviate the cycle of artificial intelligence winters and summers, which need not be inevitable.

Artificial intelligence (AI) companies and their rhetoric infringe on academia in harmful ways, mirroring past uncritical acceptance of industry logics, such as those of tobacco and petroleum. In this position piece, we tease apart and explain why phrases like 'generative AI' impede scholarly discussion because by design these expressions are used to dazzle and sidestep scrutiny. Furthermore, we contend with the AI industry's logics to enable rejecting frames such as: that we must embrace the future, that this hype cycle is unique, that anthropomorphism and circular reasoning hold water when discussing AI systems, and that students are now all cheating or all need to use AI. To these ends, we expound on why universities must take their role seriously to a) counter the AI industry's marketing, hype, and harm; and to b) safeguard higher education, critical thinking, expertise, academic freedom, and scientific integrity. For each point we raise, we include pointers to relevant work to further inform and convince our colleagues.

While it seems sensible that human-centred artificial intelligence (AI) means centring "human behaviour and experience," it cannot be any other way. AI, I argue, is usefully seen as a relationship between technology and humans where it appears that artifacts can perform, to a greater or lesser extent, human cognitive labour. This is evinced using examples that juxtapose technology with cognition, inter alia: abacus versus mental arithmetic; alarm clock versus knocker-upper; camera versus vision; and sweatshop versus tailor. Using novel definitions and analyses, sociotechnical relationships can be analysed into varying types of: displacement (harmful), enhancement (beneficial), and/or replacement (neutral) of human cognitive labour. Ultimately, all AI implicates human cognition; no matter what. Obfuscation of cognition in the AI context — from clocks to artificial neural networks — results in distortion, in slowing critical engagement, perverting cognitive science, and indeed in limiting our ability to truly centre humans and humanity in the engineering of AI systems. To even begin to de-fetishise AI, we must look the human-in-the-loop in the eyes.

The current AI hype cycle combined with Psychology's various crises make for a perfect storm. Psychology, on the one hand, has a history of weak theoretical foundations, a neglect for computational and formal skills, and a hyperempiricist privileging of experimental tasks and testing for effects. Artificial Intelligence, on the other hand, has a history of conflating artifacts for theories of cognition, or even minds themselves, and its engineering offspring likes to move fast and break things. Many of our contemporaries now want to combine the worst of these two worlds. What could possibly go wrong? Quite a lot. Does this mean that Psychology and Artificial Intelligence can best part ways? Not at all. There are very fruitful ways in which the two disciplines can interact and theoretically inform the interdisciplinary study of cognition. But to reap the fruits one needs to understand how to steer clear of potential traps.

The cognitive sciences, especially at the intersections with computer science, artificial intelligence, and neuroscience, propose 'reverse engineering' the mind or brain as a viable methodology. We show three important issues with this stance: 1) Reverse engineering proper is not a single method and follows a different path when uncovering an engineered substance versus a computer. 2) These two forms of reverse engineering are incompatible. We cannot safely reason from attempts to reverse engineer a substance to attempts to reverse engineer a computational system, and vice versa. Such flawed reasoning rears its head, for instance, when neurocognitive scientists reason about what artificial neural networks and brains have in common using correlations or structural similarity. 3) While neither type of reverse engineering can make sense of non-engineered entities, both are applied in incompatible and mix-and-matched ways in cognitive scientists' thinking about computational models of cognition. This results in treating mind as a substance; a methodological manoeuvre that is, in fact, incompatible with computationalism. We formalise how neurocognitive scientists reason (metatheoretical calculus) and show how this leads to serious errors. Finally, we discuss what this means for those who ascribe to computationalism, and those who do not.

Huettig and Christiansen in an earlier issue argue that large language models (LLMs) are beneficial to address declining cognitive skills, such as literacy, through combating imbalances in educational equity. However, we warn that this technosolutionism may be the wrong frame. LLMs are labor intensive, are economically infeasible, and pollute the environment, and these properties may outweigh any proposed benefits. For example, poor quality air directly harms human cognition, and thus has compounding effects on educators' and pupils' ability to teach and learn. We urge extreme caution in facilitating the use of LLMs, which like much of modern academia run on private technology sector infrastructure, in classrooms lest we further normalize: pupils losing their right to privacy and security, reducing human contact between learner and educator, deskilling teachers, and polluting the environment. Cognitive scientists instead can learn from past mistakes with the petrochemical and tobacco industries and consider the harms to cognition from LLMs.

Contemporary AI models owe much of their success and discontents to connectionism, a framework in cognitive science that has been (and continues to be) highly influential. Herein, we analyze artificial neural networks (ANNs): a) when used as scientific instruments of study; and b) when functioning as emergent arbiters of the zeitgeist in the cognitive, computational, and neural sciences. Building on our previous work with respect to analogizing between ANNs and cognition, brains, or behaviour (Guest & Martin, 2023), we use metatheoretical analysis techniques (Guest, 2024), including formal logic, to characterise two distinct tendencies within connectionism that we dub classical and modern, with divergent properties, e.g. goals, mechanisms, scientific questions. We also demonstrate how we, as a field, often fail to follow important lines of argument to their end --- this results in a paradoxical praxis. By engaging more deeply with (meta)theory surrounding ANNs, our field can obviate the cycle of AI winters and summers, which need not be inevitable.

We use the myth of Pygmalion as a lens to investigate and frame the relationship between women and artificial intelligence (AI). Pygmalion was a legendary ancient king of Cyprus and sculptor. Having been repulsed by women, he used his skills to create a statue, which was imbued with life by the goddess Aphrodite. This can be seen as one of the primordial AI-like myths, wherein humanity creates intelligent life-like self-images to reproduce or replace ourselves. In addition, the myth prefigures historical and present gendered dynamics within the field of AI and between AI and society at large. Throughout history, the theme of women being replaced by inanimate objects (e.g. automata, algorithms) has been repeated, and continues to repeat in contemporary AI technologies. However, this socially detrimental pattern in technology — what we dub Pygmalion displacement — is often overlooked, whether due to naive excitement about new developments, or due to an unacknowledged sexist history of the field itself. As we demonstrate herein, Pygmalion displacement prefigures heavily, but in an unacknowledged way, in the original Turing test, the imitation game: a central thought experiment, foundational to AI. With women, and the feminine generally, being both dislocated and erased from and by technology, AI is and has been (presented as) created mainly by privileged men, subserving capitalist patriarchal ends. This poses serious dangers to women and other marginalised people. By tracing the historical and ongoing entwinement of femininity (from a patriarchal perspective) and AI, we aim to understand, make visible, and start a dialogue on the ways in which AI harms women.

In the cognitive, computational, and neuro-sciences, practitioners often reason about what computational models represent or learn, as well as what algorithm is instantiated. The putative goal of such reasoning is to generalize claims about the model in question, to claims about the mind and brain, and the neurocognitive capacities of those systems. Such inference is often based on a model’s performance on a task, and whether that performance approximates human behavior or brain activity. Here we demonstrate how such argumentation problematizes the relationship between models and their targets; we place emphasis on artificial neural networks (ANNs), though any theory-brain relationship that falls into the same schema of reasoning is at risk. In this paper, we model inferences from ANNs to brains and back within a formal framework — metatheoretical calculus — in order to initiate a dialogue on both how models are broadly understood and used, and on how to best formally characterize them and their functions. To these ends, we express claims from the published record about models’ successes and failures in first-order logic. Our proposed formalization describes the decision-making processes enacted by scientists to adjudicate over theories. We demonstrate that formalizing the argumentation in the literature can uncover potential deep issues about how theory is related to phenomena. We discuss what this means broadly for research in cognitive science, neuroscience, and psychology; what it means for models when they lose the ability to mediate between theory and data in a meaningful way; and what this means for the metatheoretical calculus our fields deploy when performing high-level scientific inference.

I present an ontology of criteria for evaluating theory to answer the titular question from the perspective of a scientist practitioner. Set inside a formal account of our adjudication over theories, a metatheoretical calculus, this ontology comprises the following: (a) metaphysical commitment, the need to highlight what parts of theory are not under investigation, but are assumed, asserted, or essential; (b) discursive survival, the ability to be understood by interested non-bad actors, to withstand scrutiny within the intended (sub)field(s), and to negotiate the dialectical landscape thereof; (c) empirical interface, the potential to explicate the relationship between theory and observation, i.e., how observations relate to, and affect, theory and vice versa; (d) minimising harm, the reckoning with how theory is forged in a fire of historical, if not ongoing, abuses—from past crimes against humanity, to current exploitation, turbocharged or hyped by machine learning, to historical and present internal academic marginalisation. This work hopes to serve as a possible beginning for scientists who want to examine the properties and characteristics of theories, to propose additional virtues and vices, and to engage in further dialogue. Finally, I appeal to practitioners to iterate frequently over such criteria, by building and sharing the metatheoretical calculi used to adjudicate over theories.

The idea that human cognition is, or can be understood as, a form of computation is a useful conceptual tool for cognitive science. It was a foundational assumption during the birth of cognitive science as a multidisciplinary field, with Artificial Intelligence (AI) as one of its contributing fields. One conception of AI in this context is as a provider of computational tools (frameworks, concepts, formalisms, models, proofs, simulations, etc.) that support theory building in cognitive science. The contemporary field of AI, however, has taken the theoretical possibility of explaining human cognition as a form of computation to imply the practical feasibility of realising human(-like or -level) cognition in factual computational systems, and the field frames this realisation as a short-term inevitability. Yet, as we formally prove herein, creating systems with human(-like or -level) cognition is intrinsically computationally intractable. This means that any factual AI systems created in the short-run are at best decoys. When we think these systems capture something deep about ourselves and our thinking, we induce distorted and impoverished images of ourselves and our cognition. In other words, AI in current practice is deteriorating our theoretical understanding of cognition rather than advancing and enhancing it. The situation could be remediated by releasing the grip of the currently dominant view on AI and by returning to the idea of AI as a theoretical tool for cognitive science. In reclaiming this older idea of AI, however, it is important not to repeat conceptual mistakes of the past (and present) that brought us to where we are today.

A key strength of connectionist modelling is its ability to simulate both intact cognition and the behavioural effects of neural damage. We survey the literature, showing that models have been damaged in a variety of ways, e.g. by removing connections, by adding noise to connection weights, by scaling weights, by removing units and by adding noise to unit activations. While these different implementations of damage have often been assumed to be behaviourally equivalent, some theorists have made aetiological claims that rest on nonequivalence. They suggest that related deficits with different aetiologies might be accounted for by different forms of damage within a single model. We present two case studies that explore the effects of different forms of damage in two influential connectionist models, each of which has been applied to explain neuropsychological deficits. Our results indicate that the effect of simulated damage can indeed be sensitive to the way in which damage is implemented, particularly when the environment comprises subsets of items that differ in their statistical properties, but such effects are sensitive to relatively subtle aspects of the model’s training environment. We argue that, as a consequence, substantial methodological care is required if aetiological claims about simulated neural damage are to be justified, and conclude more generally that implementation assumptions, including those concerning simulated damage, must be fully explored when evaluating models of neurological deficits, both to avoid over-extending the explanatory power of specific implementations and to ensure that reported results are replicable.

Contemporary methods of computational cognitive modeling have recently been criticized by Addyman and French (2012) on the grounds that they have not kept up with developments in computer technology and human–computer interaction. They present a manifesto for change according to which, it is argued, modelers should devote more effort to making their models accessible, both to non-modelers (with an appropriate easy-to-use user interface) and modelers alike. We agree that models, like data, should be freely available according to the normal standards of science, but caution against confusing implementations with specifications. Models may embody theories, but they generally also include implementation assumptions. Cognitive modeling methodology needs to be sensitive to this. We argue that specification, replication and experimentation are methodological approaches that can address this issue.

In order to understand cognition, we often recruit analogies as building blocks of theories to aid us in this quest. One such attempt, originating in folklore and alchemy, is the homunculus: a miniature human who resides in the skull and performs cognition. Perhaps surprisingly, this appears indistinguishable from the implicit proposal of many neurocognitive theories, including that of the 'cognitive map,' which proposes a representational substrate for episodic memories and navigational capacities. In such 'small cakes' cases, neurocognitive representations are assumed to be meaningful and about the world, though it is wholly unclear who is reading them, how they are interpreted, and how they come to mean what they do. We analyze the 'small cakes' problem in neurocognitive theories (including, but not limited to, the cognitive map) and find that such an approach a) causes infinite regress in the explanatory chain, requiring a human-in-the-loop to resolve, and b) results in a computationally inert account of representation, providing neither a function nor a mechanism. We caution against a 'small cakes' theoretical practice across computational cognitive modelling, neuroscience, and artificial intelligence, wherein the scientist inserts their (or other humans') cognition into models because otherwise the models neither perform as advertised, nor mean what they are purported to, without said 'cake insertion.' We argue that the solution is to tease apart explanandum and explanans for a given scientific investigation, with an eye towards avoiding van Rooij's (formal) or Ryle's (informal) infinite regresses.

Psychology endeavors to develop theories of human capacities and behaviors on the basis of a variety of methodologies and dependent measures. We argue that one of the most divisive factors in psychological science is whether researchers choose to use computational modeling of theories (over and above data) during the scientific-inference process. Modeling is undervalued yet holds promise for advancing psychological science. The inherent demands of computational modeling guide us toward better science by forcing us to conceptually analyze, specify, and formalize intuitions that otherwise remain unexamined—what we dub open theory. Constraining our inference process through modeling enables us to build explanatory and predictive theories. Here, we present scientific inference in psychology as a path function in which each step shapes the next. Computational modeling can constrain these steps, thus advancing scientific inference over and above the stewardship of experimental practice (e.g., preregistration). If psychology continues to eschew computational modeling, we predict more replicability crises and persistent failure at coherent theory building. This is because without formal modeling we lack open and transparent theorizing. We also explain how to formalize, specify, and implement a computational model, emphasizing that the advantages of modeling can be achieved by anyone with benefit to all.