Jerry Fodor – 1935 2017

Biography

Jerry Alan Fodor (1935–2017) stands as one of the most intellectually formidable, fiercely independent, and stylistically distinct figures in the history of analytic philosophy and the foundations of cognitive science. Emerging onto the philosophical scene during the twilight of Oxford-style ordinary language philosophy and the ascendance of Skinnerian behaviorism, Fodor accomplished nothing short of a revolution in the scientific and philosophical conception of the human mind. Across six decades of relentless intellectual output, he mounted an uncompromising, rigorously constructed defense of commonsense mentalism. Where behaviorists sought to reduce mental states to stimulus-response pairings, and where reductive materialists sought to dissolve the richness of subjective belief into neurochemical exchanges, Fodor insisted on the literal truth of folk psychology: humans possess internal, causally efficacious mental states—beliefs, desires, intentions, and hopes—that genuinely explain what they do.

Fodor’s monumental contribution was to wed this unapologetic intentional realism to the formal machinery of the computer age. Drawing deeply from Alan Turing’s mechanical formalisms and Noam Chomsky’s generative linguistics, Fodor formulated the Computational Theory of Mind (CTM) and its indispensable corollary, the Language of Thought (LOT) hypothesis, often referred to as Mentalese. In doing so, he provided an ontological and mechanistic bridge across the ancient Cartesian chasm between mind and body. Mental states, for Fodor, are relations to internal, physically realized symbolic representations whose syntactic properties mirror their semantic contents. Thought proceeds via formal, mechanical operations executed over structured mental representations, allowing the physical brain to function as a syntax-driven engine that respects semantic rationality. In this single, audacious move, Fodor granted cognitive psychology its scientific autonomy, showing how psychology could study the formal algorithms of the mind without collapsing into the physiological taxonomy of the underlying neurobiology.

Yet Fodor was neither an uncritical cheerleader for artificial intelligence nor an orthodox system-builder. He was, by temperamental disposition and philosophical instinct, an iconoclast. His career was marked by deep-seated skepticism toward the dogmas of his contemporaries. He famously argued for the strict modularity of perceptual input systems while delivering a devastating, pessimistic assessment of our ability to computationally model central, non-modular reasoning. He launched blistering assaults against connectionist artificial intelligence, semantic holism, prototype theories of concepts, inferential role semantics, and even neo-Darwinian adaptationist accounts of cognitive architecture. Written with extraordinary wit, razor-sharp polemical zeal, and a profound philosophical commitment to common sense, Fodor’s oeuvre permanently altered the contours of the philosophy of mind, language, and cognitive psychology. To study Fodor is to confront the foundational questions of what it means to possess a mind, to think a thought, and to know an external world.

1. Biographical Trajectory and Intellectual Formation (1935–2017)

1.1 Early Life, Academic Education, and Formative Mentors

Jerry Alan Fodor was born in New York City on April 22, 1935, into a Jewish family whose intellectual curiosity and cultural sensibilities shaped his formative years. Reared in the vibrant, contentious intellectual climate of mid-century New York, Fodor entered Columbia University in the early 1950s. At Columbia, his intellectual trajectory was decisively steered by the idiosyncratic, brilliantly dialectical philosopher Sydney Morgenbesser. Morgenbesser, legendary for his lightning-quick Socratic wit, streetwise pragmatic rigor, and refusal to let philosophical abstraction drift untethered from common sense, instilled in the young Fodor a profound allergy to philosophical pretension and intellectual fashion. Under Morgenbesser’s guidance, Fodor learned that philosophical seriousness was entirely compatible with irrepressible humor, and that the ultimate benchmark of any philosophical theory of mind was its ability to account for the real, unvarnished phenomena of human cognitive life.

After graduating from Columbia in 1956 with an A.B. summa cum laude, Fodor moved to Princeton University to pursue doctoral studies in philosophy. Princeton was at that moment becoming an epicenter for the emerging post-positivist revival in metaphysics and the philosophy of science. Crucially, Fodor’s doctoral research was supervised by Hilary Putnam, who was then in the process of formulating machine-state functionalism—the revolutionary idea that mental states are defined not by their physical constitution, but by their functional and causal roles within a system. Putnam’s radical insight, which utilized the mathematical formalisms of Alan Turing’s work to dismantle both Cartesian substance dualism and type-identity physicalism, provided the fertile soil from which Fodor’s own philosophies would germinate. Fodor completed his Ph.D. in 1960 with a dissertation titled Conditions for Meaning Rules, a work that engaged closely with Ludwig Wittgenstein and ordinary language philosophy while already betraying a restive desire to transcend their linguistic insularity.

The early phase of Fodor’s development marked a deliberate, calculated pivot away from the ordinary language paradigms of Gilbert Ryle and J. L. Austin, which dominated Anglo-American philosophy at the time. Fodor recognized that the linguistic maneuvers of ordinary language philosophy, which treated talk of mental states as mere behavioral dispositions or grammatical customs, were radically ill-equipped to engage with the nascent cognitive revolution. Instead, Fodor sought an intersection between formal logic, the rigorous empirical commitments of early psychology, and psychological realism. He refused to treat mental concepts as mere linguistic artifacts to be dissolved through conceptual clarification; rather, he viewed mental states as real, internal, unobservable mechanisms whose causal interactions demanded theoretical postulation, empirical investigation, and formal modeling.

1.2 Institutional Appointments and Academic Milestones

Upon completing his doctorate, Fodor joined the faculty of the Massachusetts Institute of Technology (MIT) in 1961 as an assistant professor in the Department of Humanities, eventually transitioning into the newly founded Department of Linguistics and Philosophy. Fodor’s twenty-five-year tenure at MIT placed him at the historic geographic epicenter of the cognitive revolution. Working in immediate daily proximity to Noam Chomsky, Morris Halle, and early pioneer cognitive scientists like Merrill Garrett and Zenon Pylyshyn, Fodor immersed himself in the revolutionary paradigm of generative grammar. Chomsky’s rigorous formalization of linguistic syntax, coupled with his ferocious critique of B. F. Skinner’s verbal behaviorism, provided Fodor with the empirical and theoretical paradigm he needed: human mental capacities were governed by rule-governed, generative internal computational structures that were biologically innate rather than conditioned from environmental stimuli.

During his decades at MIT, Fodor’s intellectual output was explosive. In 1975, he published his seminal monograph, The Language of Thought, which crystallized the computational-representational paradigm for an entire generation of philosophers and psychologists. This was followed in 1983 by The Modularity of Mind, an architectural tour de force that remains one of the most widely cited texts in modern cognitive science. In 1986, Fodor transitioned from Cambridge to the Graduate Center of the City University of New York (CUNY), where he served as a Distinguished Professor of Philosophy. His presence at CUNY, alongside figures like Stephen Neale and Jerrold Katz, turned New York into an intellectual powerhouse for the philosophy of language and mind, drawing graduate students and international scholars into his orbit.

In 1988, Fodor accepted an appointment as the State of New Jersey Professor of Philosophy at Rutgers University, where he remained until his retirement in 2016. At Rutgers, Fodor was instrumental in building what became universally recognized as one of the world’s preeminent graduate programs in analytic philosophy, particularly in the philosophy of mind, cognitive science, and metaphysics. Alongside his close friend and frequent collaborator Ernie Lepore, Fodor transformed the Rutgers philosophy department into an intellectual salon characterized by fierce debate, uncompromising intellectual standards, and an utter disdain for academic pomposity. Throughout his career, Fodor delivered prestigious lectures across the globe, including the John Locke Lectures at the University of Oxford in 1993 (published as The Elm and the Expert), leaving an indelible institutional and disciplinary imprint on the cognitive sciences.

1.3 Polemical Style, Wit, and Philosophic Sensibility

Beyond his foundational technical contributions, Jerry Fodor was celebrated—and occasionally exasperated his contemporaries—for possessing one of the most distinctive, electrifying, and humorous prose styles in twentieth-century analytic philosophy. Rejecting the dry, sanitized, pseudo-scientific tone common to academic treatises, Fodor wrote with visceral energy, biting irony, and an inimitable conversational voice. His papers and books were populated by a cast of eccentric recurring archetypes: “Granny” (the embodiment of unsophisticated, common-sense folk wisdom), “Auntie” (the conservative, dogmatic defender of philosophical orthodoxy), and baffling hypothetical scenarios involving the consumption of toxic shoe polish, telepathic martians, or the idiosyncratic culinary choices of his beloved basset hounds.

This legendary wit was not merely ornamental; it was an indispensable weapon in Fodor’s philosophical arsenal. Fodor used hyperbole, satirical reductios, and vivid thought experiments to puncture academic pieties and expose what he viewed as the profound absurdities underlying reigning intellectual fashions. Whether he was demolishing behaviorism, skewering connectionist neural networks, mocking the excessive claims of evolutionary psychologists, or dismantling the baroque machinations of semantic holism, Fodor deployed humor as an instrument of rigorous conceptual hygiene. He possessed a rare capacity to distill dense, highly abstract philosophical debates into stark, inescapable dilemmas, forcing his opponents to confront the untenable costs of their theoretical commitments.

Underneath his polemical bravado lay a philosophic sensibility marked by deep, unabashed conservatism regarding the reality of the everyday human mind. Fodor regarded the widespread twentieth-century philosophical inclination to eliminate, instrumentalize, or neurobiologically reduce ordinary mental phenomena—beliefs, desires, intentions—as a form of intellectual pathology. His entire intellectual project was fundamentally defensive: to construct an impenetrable metaphysical and scientific fortress around common sense. This singular synthesis of unapologetic realism, razor-sharp technical mastery, rationalist deduction, and rollicking humor exerted an immense influence on transatlantic philosophy, establishing Fodor not merely as a brilliant philosopher, but as an unforgettable intellectual presence whose papers were read as much for the sheer pleasure of their rhetoric as for their profound philosophical content.

2. The Computational Theory of Mind (CTM)

2.1 Turing Machines and the Mechanization of Thought

The philosophical centerpiece of Fodor’s life’s work is the Computational Theory of Mind (CTM). For centuries, philosophy was trapped within a seemingly unresolvable Cartesian dilemma: if mental phenomena (such as thoughts, reasons, and meanings) are non-physical, how can they exert causal influence over physical matter (such as the brain and body)? Conversely, if the brain is merely an intricate biological machine governed solely by the mechanistic, non-rational laws of physics and chemistry, how can its operations genuinely respect the laws of logic, rationality, and semantic meaning? Rene Descartes had opted for substance dualism, leaving the mechanism of mind-brain interaction entirely mysterious. Materialists had historically opted for reductive physicalism, threatening to reduce intentional, reason-guided thought to the blind collision of atoms.

Fodor argued that the conceptual breakthrough needed to dismantle this Cartesian impasse was provided by Alan Turing’s mathematical formalization of mechanical computation. Turing demonstrated that it is possible to design a physical mechanism—a Turing machine—whose state transitions are determined entirely and exclusively by the syntactic (formal, physical, shape-based) properties of the symbols written upon its tape, while simultaneously preserving an interpretation under which those operations strictly respect the semantic relations (such as truth, reference, and entailment) holding between those symbols. As Fodor famously and aphoristically summarized this triumph: Turing showed that you can build a device where the syntax mirrors the semantics, and the physics drives the syntax. Therefore, if the brain is understood as a computational engine, meaning can be physically instantiated without invoking immaterial substances, supernatural homunculi, or magic.

Within Fodor’s CTM, mental operations are strictly conceptualized as syntactically driven formal symbol manipulations executed over internal mental vehicles. The physical brain functions as a syntax engine: its causal state-transitions are governed purely by the formal shapes, configurations, and physical distributions of internal representations. However, because those syntactic configurations are structured to instantiate semantic rules, the causal trajectory of the machine tracks the rational trajectories of reason and logic. A person who believes that if it rains, the streets will be wet, and who subsequently comes to believe that it is raining, is physically transitioned by their neural machinery into the belief that the streets are wet. The physical mechanism operates exclusively on the formal structure of the symbolic tokens, yet the outcome preserves truth and obeys the rational principles of modus ponens.

2.2 The Synthesis of Representational Realism and Computationalism

Fodor was adamant that computation cannot exist in an ontological vacuum: there can be no computation without a medium of representation. Consequently, Fodor insisted that the Computational Theory of Mind logically presupposes and necessitates the Representational Theory of Mind (RTM). Under RTM, to be in a propositional attitude state—such as believing that p, desiring that q, or fearing that r—is literally to stand in a specific functional, computational relation to an internal mental representation that means that p, q, or r. Mental states are not amorphous holistic orientations, nor are they mere behavioral dispositions; they are genuine, token physical states possessing formal, structural vehicle properties inscribed inside the central nervous system.

This formulation marked Fodor’s total and unyielding rejection of instrumentalism, operationalism, and behaviorism. Philosophers such as Daniel Dennett famously proposed an “intentional stance,” suggesting that attributing beliefs and desires to an agent is simply an interpretive heuristic—a convenient, predictive stance adopted by an observer, rather than an assertion that there are actual, physical things called “beliefs” residing inside the agent’s skull. Fodor viewed such instrumentalism as an act of intellectual cowardice. If common-sense psychological explanations are true—and Fodor argued their predictive triumph is the most spectacular empirical fact in the history of science—they must be true in virtue of physical facts. There must be real, physical, token representations within the organism whose structural configurations causally generate behavior.

By synthesizing representational realism with computationalism, Fodor forged an elegant middle path between eliminative reductionism on the one side and intentional mysterianism on the other. RTM accounts for the intentionality, aboutness, and semantic content of mental states, while CTM accounts for their causal efficacy within the natural world. Mental representations act as physical vehicles carrying semantic freight. Because these vehicles possess physical properties (spatial location, mass, neuroelectric signatures), they interact mechanically via ordinary physical laws. Because they simultaneously possess syntactic properties that encode semantic propositions, their physical interactions preserve rational coherence. In this way, Fodor successfully reconciled semantic truth conditions with mechanical neurobiological causation.

2.3 Limitations of Purely Syntactic Engines

Despite his status as the primary architect of CTM, Fodor was acutely aware of the intrinsic limitations of purely syntactic engines. This philosophical self-awareness set him sharply apart from enthusiastic functionalists and artificial intelligence researchers who believed that syntax was entirely sufficient for mind. Fodor engaged deeply with John Searle’s celebrated Chinese Room argument. Searle argued that an agent trapped in a room, manipulating Chinese characters purely according to a set of formal, syntactically defined rulebooks, could produce fluent Chinese output without understanding a single word of Chinese. Searle concluded that syntax is neither identical with nor sufficient for semantics, and that computationalism was fundamentally bankrupt as an explanation of understanding.

Fodor’s response to Searle was characteristically subtle. He conceded Searle’s core negative premise: syntax, considered purely in isolation from an organism’s causal interactions with its environment, can never yield genuine semantic content. Merely running a syntactic program inside a closed box does not automatically generate intentionality, meaning, or reference. Formal symbol manipulations do not self-generate their own interpretation; symbols require “grounding.” However, where Searle concluded that intentionality must therefore be an exclusive biological byproduct of the brain’s mysterious neurochemical causal powers, Fodor argued that the computational theory simply needed to be wedded to an externalist, causal-informational semantics. The internal syntactic symbols acquire genuine meaning not from internal functional loops, but through their external, causal-historical relations to the world they represent.

Furthermore, toward the end of his career, Fodor delivered a sobering critique of the scope of computationalism itself, culminating in his provocative book The Mind Doesn’t Work That Way (2000). While maintaining that computation provides the only viable scientific model for local, modular perceptual processes, Fodor conceded that purely syntactic machines struggle profoundly to explain central, holistic, non-modular thought. In high-level reasoning, scientific theory confirmation, and creative abduction, the relevance of any single belief depends upon the totality of an agent’s background knowledge. Because formal syntax is local—operating strictly on the intrinsic geometric properties of individual symbols—it cannot naturally accommodate the non-local, holistic nature of general rational reflection. Thus, Fodor ended his life warning cognitive scientists against overextending the very computational paradigm he had spent his life establishing.

3. The Language of Thought Hypothesis (Mentalese)

3.1 Productivity, Systematicity, and Compositionality

If the Computational Theory of Mind claims that thinking is the manipulation of representations, what kind of structure must those representations possess? In The Language of Thought (1975), Fodor unleashed the foundational thesis that thought requires an internal, linguistic representational medium: the Language of Thought (LOT), universally known as Mentalese. Fodor’s central argument was that human cognitive capacities display three indispensable structural characteristics that can only be explained if the medium of thought possesses a combinatorial syntax and a compositional semantics: productivity, systematicity, and compositionality.

The argument from productivity mirrors Chomsky’s famous arguments regarding natural language: human beings possess a finite cognitive mechanism (a brain of finite volume and finite neural resources) capable of producing and understanding an unbounded, potentially infinite number of distinct, structurally novel thoughts. An individual who has never heard the proposition “a purple rhinoceros is reciting Shakespeare on a skateboard in downtown Minneapolis” can instantly comprehend and evaluate it. Because a finite mind cannot contain an infinite lookup table of pre-stored holistic ideas, the only known mechanism capable of generating infinite semantic variety from finite resources is a combinatorial, generative syntax. Complex thoughts must be built out of a finite stock of reusable, atomic symbolic primitives according to recursive structural rules.

The argument from systematicity is even more decisive. Cognition is inherently systematic: the ability to think certain thoughts is intrinsically, non-contingently tied to the ability to think other, structurally related thoughts. If an agent has the conceptual resources to think the thought John loves Mary, it is an empirical, psychological law that the agent also has the conceptual resources to think the thought Mary loves John. There is no such thing as an adult human who can think the first thought but is constitutionally incapable of comprehending the second. Fodor pointed out that this systematicity would be an utter, miraculous coincidence if thoughts were holistic pictures or unstructured associationist nodes. It is explained cleanly and trivially if and only if “John,” “loves,” and “Mary” are discrete, syntactically articulated components of a linguistic code whose meanings are determined by their roles within a formal compositional architecture.

Underpinning both productivity and systematicity is the principle of compositionality: the semantic content of a complex representation is an invariant function of the semantic contents of its atomic syntactic parts, together with the syntactic rules that combine them. Compositionality guarantees that an atomic concept—such as “brown” or “dog”—retains its core semantic identity across an infinite variety of distinct combinations (“brown dog,” “spotted dog,” “the dog barked at the brown car”). Without compositionality, learning, inference, and semantic coherence would collapse into utter chaos. Fodor thus concluded that whatever the physical implementation of the mind may be, its representational vehicles must be physical tokens of an internal, combinatorial language.

3.2 Mentalese as an Innate Medium of Cognition

One of the most radical and widely contested implications of Fodor’s Language of Thought hypothesis was his insistence that Mentalese must be an innate medium of cognition. Fodor did not merely argue that humans possess a biological capacity to acquire symbolic systems; he argued that the fundamental symbolic primitives of the internal code are themselves innate, biologically hardwired features of the human cognitive apparatus. To establish this staggering conclusion, Fodor formulated his famous and widely vexing “paradox of concept learning.”

Fodor observed that every prevailing scientific theory of concept acquisition in cognitive psychology—from Jerome Bruner to Jean Piaget—relied on a single explanatory framework: hypothesis formulation and testing. To learn a concept, such as the natural language word “dog” or the category “triangle,” an organism must form a hypothesis about what the target term means (e.g., “perhaps ‘triangle’ means closed three-sided plane figure”), project that hypothesis onto incoming sensory data, evaluate whether the data confirms or refutes the hypothesis, and revise its internal state accordingly. But Fodor pointed out a devastating logical consequence of this model: in order to formulate a hypothesis about what “dog” or “triangle” means, the organism must already possess a representational medium in which that hypothesis is framed!

One cannot formulate the inductive hypothesis that the word ‘dog’ applies to entities that are furry, quadrupedal, and bark unless one already commands the concepts [furry], [quadrupedal], and [bark]. If learning a language (like English, Mandarin, or Spanish) consists in learning the truth conditions and semantic mappings of its lexical items, then learning a natural language logically presupposes the possession of an internal, pre-existing language in which those truth conditions can be represented. You cannot learn your first language; you can only map it onto an internal language you already possess. Consequently, Mentalese cannot be identical to any spoken natural language. It must be an innate, universal, phylogenetically hardwired biological code. Fodor drew extensive empirical support for this conclusion from developmental psychology—demonstrating rich, pre-linguistic cognitive architectures in infants—and from animal cognition, pointing out that pre-verbal human infants and non-human animals solve complex logical, perceptual, and spatial problems without possessing a single word of a spoken natural tongue.

3.3 Objections to the Language of Thought

The Language of Thought hypothesis was greeted with immediate, visceral resistance from multiple philosophical quarters. One of the most common early criticisms was the regress problem: if understanding a natural language like English requires an internal language like Mentalese, why doesn’t understanding Mentalese require an internal “Mentalese-prime,” and so on into an infinite tower of internal interpreters? Critics charged that Fodor had simply pushed the mystery of understanding one level inward, solving the problem of linguistic comprehension by inventing a homunculus inside the brain who reads an internal teleprinter.

Fodor brushed this objection aside with characteristic ease, arguing that it fundamentally miscomprehended the mechanical nature of the Computational Theory of Mind. The regress halts immediately at the level of basic physical execution. Unlike English words, which are conventional, arbitrary cultural artifacts that require conscious interpretation, the symbols of Mentalese do not need an internal reader. They operate mechanically. The symbols of Mentalese are physical states of the nervous system whose execution is governed directly by physical, neurobiological laws. A compiler in a modern computer does not require an infinite sequence of internal programs to execute machine code; machine code acts directly on the physical logic gates of the central processing unit. Similarly, Mentalese symbols are physically instantiated syntactic tokens whose causal powers derive directly from their physical architecture. Regress avoided.

A second major challenge emerged from philosophers influenced by Ludwig Wittgenstein’s private language argument and rule-following paradoxes. Wittgensteinian critics argued that meaning is inherently social, normative, and public; the idea of a private, biologically innate symbolic code operating prior to public linguistic interaction was, they asserted, logically incoherent. Simultaneously, Daniel Dennett maintained that propositional attitudes are best treated as interpretive ascriptions (the intentional stance) rather than concrete, sentence-like inscriptions inside the brain. In response, Fodor maintained an unapologetic commitment to scientific naturalism. Social conventions and normative language games, he argued, are secondary, derivative phenomena that completely depend upon the prior, biological existence of representational cognitive architecture. To deny the Language of Thought because it offended Wittgensteinian anti-psychologism was, in Fodor’s eyes, to elevate linguistic dogmatism over empirical science.

4. The Modularity of Mind: Architecture and Constraints

4.1 Characteristics of Modular Cognitive Systems

In 1983, Fodor published his landmark monograph, The Modularity of Mind, introducing a rigorous functional taxonomy of cognitive architecture that permanently reconfigured neuropsychology, perceptual theory, and the philosophy of psychology. In this work, Fodor sought to resolve the historic conflict between faculty psychology (which viewed the mind as composed of distinct, autonomous organs) and horizontal, holistic mental theories (which viewed the mind as a domain-general, unconstrained intellectual engine). Fodor proposed that the human mind is functionally bifurcated into specialized, low-level modular input systems and a holistic, domain-general central system.

Fodor meticulously outlined the core operational characteristics that define an authentic, biological cognitive module. The defining, non-negotiable hallmark of a module is informational encapsulation. A system is informationally encapsulated if its internal processing cannot be penetrated, overridden, or influenced by information held in the rest of the mind—most notably, by an agent’s conscious beliefs, desires, and general knowledge. Fodor demonstrated this via the famous Müller-Lyer optical illusion: even after an observer measures the two lines with a ruler and knows with absolute scientific certainty that they are identical in length, the lines continue to appear stubbornly unequal. The visual input analyzer cannot access the central belief store; it is sealed off, executing its proprietary algorithms in epistemic isolation.

Beyond informational encapsulation, Fodor identified several other interrelated characteristics of modules:

• Domain Specificity: Modules perform computational operations exclusively on a strictly delimited, idiosyncratic class of environmental inputs (e.g., parsing phonemes, computing retinal disparities, detecting geometric edges).
• Mandatory Operation: Modular processing is automatic, reflexive, and cognitively involuntary. An agent cannot choose not to hear an utterance in their native tongue as meaningful speech, nor can they choose to see a color stimulus as monochrome gray; the input system executes instantaneously upon receipt of the proximal stimulus.
• Fast Processing Speed: Because modules do not need to query the entire web of belief or compute central probabilities, they operate at lightning-fast speeds, essential for survival.
• Shallow Outputs: The information outputted by a module is relatively unelaborated and semantically basic (e.g., identifying a surface or a basic category object before central systems infer its historical significance).
• Fixed Neural Architecture and Characteristic Breakdowns: Modules correspond to hardwired, localized neuroanatomical pathways, displaying predictable, localized ontogenetic trajectories and specific clinical breakdowns (such as prosopagnosia or specific aphasias).

4.2 Input Systems Versus Central Systems

The profound architectural divide in Fodor’s model lies between these encapsulated input systems (the perceptual senses and linguistic parsing mechanisms) and the unencapsulated central systems responsible for higher-level cognitive operations: belief fixation, scientific reasoning, conscious deliberation, abduction, and general problem solving. While input systems deliver fast, local, shallow perceptual representations of the proximal environment, the central system’s task is to integrate these representations, weigh evidence, resolve contradictions, and determine what the organism should actually believe and do.

Fodor characterized central systems as possessing two defining, interrelated epistemic properties: they are Quinean and isotropic. A system is Quinean (named after W. V. O. Quine) because the degree of confirmation or rational warrant assigned to any given belief depends upon the holistic properties of the entire belief system. Beliefs are not confirmed in isolation; they hang together in an interconnected web where a revision in one node can send seismic reverberations throughout the entire network. A system is isotropic because, in principle, any fact or belief known to an agent can turn out to be relevant to the confirmation of any other belief. A scientist attempting to confirm a hypothesis in astrophysics might find that a discovery in quantum mechanics, a data point in crystallography, or an insight from abstract topology suddenly becomes critically relevant.

This stark dichotomy led Fodor to formulate his darkly humorous and deeply pessimistic “First Law of the Nonexistence of Cognitive Science”: the more isotropic a cognitive process is, the less anybody understands it. Fodor argued that classical computational models and cognitive science had succeeded exclusively in the domain of perceptual input systems precisely because those systems are modular, encapsulated, and computationally local. Local processes can be modeled as formal, syntactic algorithms operating over proprietary inputs. But central, isotropic, Quinean processes resist localized algorithmic modeling. Because cognitive science possessed no viable model of non-local, global information processing, Fodor contended that high-level human rational thought remained, from a scientific standpoint, fundamentally opaque.

4.3 Critique of Massive Modularity

In the late 1980s and 1990s, a prominent school of evolutionary psychology, championed by Leda Cosmides, John Tooby, and later popularized by Steven Pinker, proposed the radical thesis of massive modularity. Drawing inspiration from Fodor’s 1983 book while discarding his crucial architectural dichotomy, these thinkers argued that the human mind is entirely modular from top to bottom—a “Swiss Army knife” composed of hundreds or thousands of specialized, genetically selected Darwinian modules designed to solve specific ancestral problems, from mate selection and predator avoidance to cheat detection and tool use. Under this view, there is no domain-general central processor; human cognition is modules all the way down.

Fodor responded to massive modularity with ferocious intellectual hostility, most comprehensively articulated in his devastating 2000 polemic, The Mind Doesn’t Work That Way. Fodor exposed what he called the insurmountable “input problem” facing massive modularity. If the entire cognitive architecture consists solely of specialized, domain-specific modules, how does the mind determine which module should process a given environmental stimulus? If a stimulus—say, an approaching human holding an ambiguous object—could be relevant to the predator-detection module, the mate-selection module, the kin-recognition module, or the trade-negotiation module, there must be a prior system that parses the stimulus and routes it to the appropriate specialized module.

However, this prior routing mechanism cannot itself be domain-specific; it must be domain-general, capable of surveying the incoming information and evaluating its broad contextual significance across multiple potential cognitive domains. The attempt to eliminate central, domain-general cognition inevitably smuggles it back into the architecture under the guise of an unacknowledged executive routing system. Fodor maintained that human cognition is characterized precisely by our ability to transcend biological encapsulation: we integrate metaphors, synthesize radically disparate intellectual paradigms, write poetry that bridges biology and mechanics, and deduce scientific laws that apply across cosmic scales. To reduce this sweeping, unconstrained discursive reflection to a fractured collection of Pleistocene reflexes was, for Fodor, a catastrophic theoretical regression.

5. Information-Theoretic Semantics and Asymmetric Dependence

5.1 The Problem of Intentionality and Semantic Naturalization

Having established that thought occurs within an internal symbolic language, Fodor was confronted by the ultimate metaphysical hurdle of the philosophy of mind: the naturalization of intentionality. Franz Brentano had famously declared that intentionality—the property of mental states being “about” or directed toward things, states of affairs, and properties—is the infallible mark of the mental, an irreducibly non-physical phenomenon that places minds outside the explanatory scope of the natural sciences. As an uncompromising physicalist, Fodor could not tolerate this mysterianism. In Psychosemantics (1987), he laid down his naturalistic manifesto: if meaning and aboutness are real, they must be exhaustively explicable in the non-intentional, non-semantic vocabulary of physical science. If intentionality cannot be naturalized, then intentional realism is false, and common-sense psychology is an illusion.

Early attempts at naturalizing semantics had relied on causal-informational theories pioneered by Fred Dretske and Dennis Stampe. The fundamental intuition was simple: smoke means fire because smoke covariously indicates fire. Similarly, a mental representation—say, a token of Mentalese reading [COW]—means cow because tokens of [COW] are causally provoked by, and lawfully correlate with, the presence of actual cows in the organism’s environment. Meaning is grounded in nomic (law-like) mind-world covariation.

However, Fodor quickly exposed the fatal vulnerability of simple causal-informational theories: the notorious disjunction problem. Mental representations can, and frequently do, misrepresent reality. An exhausted hiker in the dim twilight spots a distant, lumbering black horse and mistakenly thinks: “There’s a cow.” Under a simple causal covariation model, the token [COW] in the hiker’s head was causally provoked by a horse. If the meaning of a symbol is simply the class of things that causally cause its tokening, then the symbol does not mean cow; it means the disjunctive category cow OR horse-in-the-mist. But if the symbol means cow OR horse-in-the-mist, then the hiker’s thought was not a misrepresentation at all—it was completely true, since a horse-in-the-mist was indeed present! Simple covariation makes genuine cognitive error theoretically impossible. A viable naturalistic semantics had to explain how a representation can lawfully mean X, even when it is mistakenly triggered by Y.

5.2 The Asymmetric Dependence Condition

To resolve the disjunction problem and rescue the possibility of objective misrepresentation within a naturalized physics, Fodor introduced his famous and exquisitely ingenious Asymmetric Dependence condition. Fodor argued that the disjunction problem arises only if all causal mind-world links are treated as having equal metaphysical weight. In reality, the causal links between the world and mental tokens are structurally asymmetric.

Consider the two distinct causal links: the link between real cows and mental [COW] tokens, and the non-standard link between horses-in-the-mist and mental [COW] tokens. Fodor pointed out that there is an irreversible, counterfactual dependency holding between these two causal pathways:

• The causal link between horses-in-the-mist and [COW] tokens asymmetrically depends upon the causal link between cows and [COW] tokens.
• If horses on dark nights cause you to think [COW], that is only because you already possess an established, lawful connection between actual cows and the token [COW]. The horse-caused tokening is an epistemic parasite—an accidental, parasitic byproduct of the primary, standard causal relation.
• Conversely, if we imagine a counterfactual world where horses-on-dark-nights no longer cause you to token [COW] (perhaps you acquire better night-vision glasses), the causal link between cows and [COW] tokens remains perfectly intact.
• However, if you break the primary link—if cows cease to cause [COW] tokens—the parasitic link between horses and [COW] tokens collapses entirely. Horses would no longer trigger [COW], because the representational category itself would lose its foundation.

Thus, Fodor formulated the truth condition: a mental representation [X] means X (and not X or Y) if and only if the causal connection between Y and [X] is asymmetrically counterfactually dependent upon the causal connection between X and [X]. This elegant, purely nomological criterion solved the disjunction problem without smuggling in intentional notions like “purpose,” “function,” or “correct interpretation.” Misrepresentation is naturalized as counterfactual asymmetric dependence between causal laws.

5.3 Evaluation and Controversies of Fodorian Semantics

Fodor’s informational semantics and the asymmetric dependence condition ignited intense, prolonged philosophical debate. His primary rivals were the champions of teleosemantics, led by Ruth Garrett Millikan and David Papineau. Teleosemanticists argued that intentionality could only be naturalized by appealing to evolutionary biology and Darwinian natural selection: a mental token [COW] means cow because that representation was biologically selected to perform the evolutionary function of indicating cows to the organism. The teleosemantic approach claimed to solve the disjunction problem through normative historical teleology: a horse-caused [COW] is a malfunction of an evolutionarily designed tracking mechanism.

Fodor waged an unrelenting war against teleosemantics, rejecting it as fundamentally unviable. First, he pointed out that evolutionary historical explanations are diachronic, whereas mental causation is strictly synchronic: the causal powers of a thought in your head right now cannot depend upon what happened to ancestral organisms millions of years ago. To make current mental content depend upon ancestral history was, Fodor argued, a bizarre form of action-at-a-temporal-distance. Second, Fodor mounted the “Swampman” thought experiment: if lightning strikes a swamp and spontaneously synthesizes an exact physical duplicate of Jerry Fodor atom-for-atom, that duplicate possesses no evolutionary history, yet it surely possesses beliefs, desires, and the ability to think about cows. Teleosemantics was forced to declare the Swampman semantically empty—an outcome Fodor considered absurd.

Nevertheless, Fodorian semantics faced formidable challenges of its own. Critics questioned whether the counterfactual asymmetric dependencies required by Fodor’s model actually hold up under rigorous modal scrutiny. In complex cases involving abstract mathematical concepts, non-existent entities (like unicorns or phlogiston), or twin-earth scenarios where natural kinds diverge, establishing the precise direction of nomological dependency proved remarkably elusive. Furthermore, empirical philosophers questioned how scientists could practically verify the fragile counterfactual assertions underpinning Fodor’s semantics. Despite these controversies, Fodor’s asymmetric dependence remains one of the most sophisticated, purely physicalist naturalization programs ever devised.

6. Intentional Realism and the Defense of Folk Psychology

6.1 The Indispensability of Belief-Desire Psychology

Throughout his career, Fodor maintained an unwavering allegiance to what he affectionately called “Granny’s view of the mind”: folk psychology. Folk psychology is the commonsense conceptual scheme that ordinary human beings use every single day to navigate social reality. Under this scheme, human actions are explained and predicted by citing an agent’s propositional attitudes: their beliefs, desires, preferences, and intentions. If you wish to know why your colleague walked to the campus bookstore at noon, the commonsense answer is obvious: she desired to purchase a book, and she believed that the book was available at the bookstore.

Fodor argued that the predictive success of this commonsense psychological framework is not merely impressive; it is utterly unprecedented in the history of science. If an individual phones an airline, reserves a flight to Chicago, and states that they will arrive at O’Hare Airport at 3:15 PM on a Tuesday, an observer can predict with nearly ninety-five percent accuracy that a biological organism will materialize at a precise geographic coordinate thousands of miles away at that exact minute. No theory in quantum physics, molecular biology, or meteorology can make predictive projections across space, time, and chaotic physical environments with such breathtaking precision based on so few variables. And what variables are used? Pure folk-psychological propositional attitudes.

Because folk psychology is so extraordinarily predictive, Fodor insisted that it must be an empirically true, descriptive science. Propositional attitudes cannot be treated as linguistic fictions, literary metaphors, or mere instruments. They must be recognized as genuine, causally efficacious internal psychological states physically instantiated in the human nervous system. Furthermore, Fodor recognized the monumental moral and existential stakes of this position. If folk psychology were false—if beliefs, desires, and reasons were shown to be non-existent chimeras—the entire architecture of human civilization would instantly dissolve. Concepts of moral responsibility, legal guilt, contracts, love, rational deliberation, and human agency depend absolutely upon the premise that our conscious beliefs and desires cause our bodily behavior. To surrender intentional realism was, for Fodor, to surrender human dignity and common sense.

6.2 The Assault on Eliminative Materialism

The primary existential threat to intentional realism during Fodor’s prime came from eliminative materialism, championed aggressively by Paul and Patricia Churchland. The Churchlands argued that folk psychology is not an established empirical science, but an ancient, pre-scientific, highly stagnant theory—comparable to the medieval humoral theory of medicine or the phlogiston theory of combustion. Pointing to the persistent failure of folk psychology to explain memory consolidation, sleep, mental illness, or learning mechanisms, the Churchlands predicted that as modern, mature neuroscience advances, folk-psychological concepts like “belief” and “desire” will not be smoothy reduced to brain states; they will be eliminated entirely from our scientific ontology, just as witches, caloric fluid, and the luminiferous ether were banished by mature physics.

Fodor launched a blistering counter-offensive against eliminativism, treating it with alternating philosophical devastation and satirical contempt. First, he demolished the claim that folk psychology is a stagnant research program. Folk psychology, Fodor pointed out, was never designed to be an exhaustive theory of the brain’s physiology; it is a theory of intentional action. Within its proper domain—explaining why rational agents do what they do—it continues to enjoy unblemished, peerless empirical predictive triumph. Comparing folk psychology to phlogiston was, Fodor argued, a category mistake of the most staggering proportions.

Second, Fodor exposed the profound pragmatic self-refutation lurking at the heart of eliminative materialism. When an eliminativist philosopher mounts an argument, writes a paper, and delivers a lecture asserting that “beliefs do not exist,” what are they doing? They are communicating what they believe to be true, with the explicit intention and desire to alter the beliefs of their audience through the presentation of reasons! The very act of asserting, comprehending, and rationally evaluating the eliminativist hypothesis presupposes the reality of the propositional attitudes that the hypothesis denies. An eliminativist who claims there are no beliefs is in the position of someone shouting at the top of their lungs that they are completely mute. The eliminativist position is cognitively self-defeating and completely unlivable.

6.3 Special Sciences and the Autonomy of Psychology

In 1974, Fodor published what is widely considered one of the most influential papers in twentieth-century philosophy of science: “Special Sciences (or: The Disunity of Science as a Working Hypothesis)”. At the time, the dominant metaphysical orthodoxy was the “Unity of Science” thesis, rooted in logical positivism and championed by physicalists who insisted that true physicalism required bridge-law reductionism. Under this reductionist model, every legitimate law in the “special sciences” (geology, biology, economics, psychology) must eventually be translated into and deduced from the fundamental, physical laws of nature via bridge laws connecting special-science predicates directly to physical-science predicates.

Fodor decisively shattered this reductionist consensus by introducing the argument from multiple realizability. Consider a predicate from economics, such as “is money.” What physical predicate corresponds to the economic property of being money? In the actual world, money can be instantiated as gold coins, silver ingots, green paper notes, plastic cards, entries in a digital banking ledger, stringed cowrie shells, or packs of cigarettes in a prisoner-of-war camp. In an alien civilization, money might be instantiated as complex molecular gas signatures. There is no single, unified physical predicate—no neurochemical, chemical, or subatomic configuration—that all tokens of money have in common. What makes a token “money” is its functional, relational role within an economic system, completely independent of its physical constitution.

Because special-science kinds are wildly multiply realizable at the physical level, bridge laws cannot be formulated without resorting to endless, open-ended disjunctions of physical states. But an open-ended disjunction of physical states (e.g., “is gold OR paper OR cowrie shells OR digital blips”) does not constitute a legitimate physical natural kind. Therefore, reductionism fails. Crucially, Fodor demonstrated that one can be an absolute, unyielding token physicalist (every concrete, individual mental or economic event in space-time is identical to a physical event) without being a type physicalist (mental types or economic properties are identical to physical types). Psychology, economics, and biology are fully autonomous sciences with their own proprietary taxonomies, nomological regularities, and explanatory laws, completely insulated from being swallowed up or eliminated by fundamental physics.

7. Critique of Connectionism and Associationist Models

7.1 The 1988 Fodor-Pylyshyn Manifesto

In the mid-1980s, cognitive science was rocked by the meteoric resurgence of artificial neural networks, parallel distributed processing (PDP), and connectionism, championed by figures like David Rumelhart, James McClelland, and Paul Smolensky. Connectionists boldly proclaimed that classical, symbolic artificial intelligence—the very paradigm Fodor had championed via CTM and LOT—was obsolete. Instead of serial symbol manipulation operating on explicit sentence-like representations, connectionism modeled cognition as patterns of activation propagating through dense networks of interconnected, sub-symbolic nodes via weighted, adjustable connections. The connectionists claimed that their distributed representations were biologically realistic, brain-like, and capable of learning from statistical patterns without explicit rules.

In 1988, Jerry Fodor and his close colleague, the cognitive scientist Zenon Pylyshyn, dropped an intellectual bomb on the cognitive science community with the publication of their monumental treatise, “Connectionism and Cognitive Architecture: A Critical Analysis”. The paper remains one of the most celebrated and fiercely debated critiques in the history of artificial intelligence. Fodor and Pylyshyn argued that beneath its glittering veneer of modern computational terminology and linear algebra, connectionism was nothing more than a high-tech revival of old-fashioned, thoroughly discredited British Empiricist Associationism—the failed mental physics of David Hume and John Locke, upgraded with statistical vector spaces.

Fodor and Pylyshyn demonstrated that connectionist architectures fundamentally lacked the representational resources necessary to model the core architectural features of the human mind: productivity, systematicity, and compositionality. A connectionist network stores information not as discrete, structured symbols, but as high-dimensional statistical vectors spread across hundreds or thousands of nodes. In such a network, representations do not possess syntactically articulated constituents. A node or pattern representing “brown dog” is not physically composed of the node for “brown” and the node for “dog”; it is simply an opaque, unanalyzable vector point in state-space. As a result, connectionist networks cannot support genuine compositional syntax.

7.2 The Dilemma for Connectionist Systems

To crystallize their critique, Fodor and Pylyshyn confronted connectionism with an agonizing, inescapable architectural dilemma that became known as the Fodor-Pylyshyn Dilemma:

Either connectionist systems implement classical, symbolic architectures, or they are empirically false as models of human cognitive architecture.

The logic of the dilemma proceeded with ruthless analytical precision:

• Horn 1: Connectionism is an implementation model. If connectionists wire their neural networks so that they successfully exhibit the systematicity, compositionality, and structural sensitivity characteristic of human thought, they will have accomplished this solely by creating distributed subnetworks that function as physical tokens of symbols and syntactic rules. In this scenario, connectionism is not a revolutionary, alternative theory of cognitive architecture; it is merely an interesting lower-level story about how classical, physical-symbol systems happen to be implemented in biological neural hardware. Connectionism becomes to cognitive psychology what silicon transistor physics is to software engineering.
• Horn 2: Connectionism is an alternative architectural model. If connectionists reject the classical paradigm and refuse to implement symbolic representations with compositional constituent structures, their networks will remain intrinsically incapable of explaining the universal systematicity of thought. A non-symbolic network can be trained via backpropagation to associate the vector for [JOHN LOVES MARY], while completely failing to generalize to [MARY LOVES JOHN], because the network has no internal constituent that represents “John” or “loves” across both contexts. Since human cognition is systematically structured, any model that treats systematicity as an accidental, statistically contingent byproduct is empirically refuted by the psychological facts.

Fodor and Pylyshyn emphasized the vital philosophical distinction between mere functional approximation and authentic architectural explanation. A connectionist network might be tuned, tweaked, and trained to approximate certain outward cognitive behaviors, but that does not mean its internal mechanisms mirror the actual algorithms of the mind. Cognitive science requires an explanatory architecture where structural sensitivity and deductive inferential transitions follow naturally and necessarily from the formal properties of internal representations.

7.3 Legacy of the Classical vs. Connectionist Debate

The Fodor-Pylyshyn manifesto permanently defined the battle lines of computational cognitive science for the next four decades. Connectionist theorists fought back vigorously. Paul Smolensky proposed “sub-symbolic computation,” arguing that connectionism operated at a finer-grained level of description that could achieve systematicity via tensor products and complex mathematical vector transformations without classical concatenation. Others argued that human thought is far less systematic than Fodor idealistically claimed, pointing to widespread human irrationality, cognitive biases, and associative slips as evidence that the mind is fundamentally statistical rather than formal and logical.

Fodor dismissed these defenses as missing the conceptual mark. Cognitive errors and associative slips are performance variables, not competence variables. The underlying competence of a rational thinker requires compositional structure. In recent years, with the staggering explosion of modern artificial intelligence, deep learning, and Large Language Models (LLMs) such as Transformers, the classical-versus-connectionist debate has returned to the absolute center of contemporary philosophy and computer science with unprecedented urgency.

Modern transformer-based neural networks have achieved extraordinary empirical breakthroughs precisely by ingesting vast, planetary-scale corpuses of statistical text. Yet, despite their astonishing conversational fluidity, they continue to suffer from the very foundational structural pathologies Fodor diagnosed in 1988: semantic hallucinations, failure of reliable compositionality under out-of-distribution prompts, an inability to guarantee logical entailments, and profound opacity. As a consequence, leading figures in contemporary artificial intelligence are increasingly advocating for “neuro-symbolic AI”—hybrid architectures that combine the pattern-recognition powers of deep neural nets with the explainable, compositional, rule-governed symbolic engines that Jerry Fodor championed half a century ago.

8. Methodological Solipsism and the Internalism vs. Externalism Debate

8.1 Methodological Solipsism as a Research Strategy

In 1980, Fodor published a provocatively titled, foundational paper in theoretical psychology: “Methodological Solipsism Considered as a Research Strategy for Cognitive Psychology”. Borrowing the radical Cartesian terminology of solipsism, Fodor was not advocating the metaphysical absurdity that only the self exists. Rather, he was proposing a precise, methodological constraint on how scientific cognitive psychology must taxonomize mental states if those states are to feature in causal explanations of behavior.

The thesis of methodological solipsism asserts that psychological states must be individated and classified entirely from an internalist perspective, without presupposing or referencing the existence of an external world. Fodor based this thesis on the physical constraints of computationalism. A computational device—whether a desktop computer or a biological central nervous system—operates strictly on the formal, syntactic properties of the physical representations tokened within its boundary. The computational engine has no direct, unmediated epistemic or causal access to distal objects in the world. The brain is locked inside the dark vault of the cranium; its causal transitions are governed solely by local neural excitations and syntactic inscriptions.

Consequently, Fodor argued that for the purposes of scientific psychology, the taxonomical boundary of the mind halts at the skin-and-skull boundary. If two organisms are exact physical, computational duplicates—such that their internal syntactic configurations are atom-for-atom identical—scientific psychology must treat them as being in identical psychological states, regardless of differences in their respective external environments. Behavior must be explained by the local causal mechanisms operating inside the organism. If an agent believes there is a glass of water on the table and reaches out their arm, the causal explanation of the reaching motion involves the internal representation [WATER ON TABLE], not the external physical liquid itself. Cognitive psychology, Fodor argued, is inherently a science of internal computational states.

8.2 Engagement with Semantic Externalism and Twin Earth

Methodological solipsism immediately placed Fodor in direct, violent conflict with the ascendant paradigm of semantic externalism, pioneered by his doctoral supervisor Hilary Putnam and philosopher Tyler Burge. In his epochal 1975 thought experiment, Putnam introduced the concept of Twin Earth: an exact physical duplicate of Earth in every single detail, with one solitary exception—the clear, odorless, potable liquid flowing in rivers and coming out of faucets on Twin Earth is not H2O, but a totally different, complex chemical compound abbreviated as XYZ.

Putnam asked: when an Earthling named Oscar and his exact molecular, computational duplicate on Twin Earth named Twin-Oscar use the word “water” (or token the internal thought [WATER]), do they mean the same thing? Putnam answered with a resounding “No.” Oscar is referring to H2O; Twin-Oscar is referring to XYZ. Their thoughts have radically different truth conditions and references. Because Oscar and Twin-Oscar are internal physical duplicates, Putnam famously concluded: “Cut the pie any way you like, ‘meanings’ just ain’t in the head!” Tyler Burge expanded this externalism to social environments, showing that an individual’s linguistic meanings depend on the normative practices of their broader linguistic community.

For Fodor, Putnam’s conclusion threatened to tear the Computational Theory of Mind apart. If meanings are external, distal relations, and computational operations are purely internal, local syntactical processes, then how can computation ever be an explanation of meaningful thought? Fodor initially attempted to resolve this tension by adopting a dual-aspect theory of content, distinguishing between narrow content and broad content. Narrow content was internal, non-referential, and computational—the internal cognitive role shared identically by Oscar and Twin-Oscar, which causally drives their identical behavior. Broad content was the externalist, referential truth conditions that differed between them. However, as the 1980s progressed, Fodor became increasingly dissatisfied with the baroque metaphysical complications of narrow content, eventually making a startling philosophical reversal: he completely abandoned narrow content, surrendered to externalism regarding semantic content, and declared himself an unapologetic semantic atomist.

8.3 Reconciling Causal Relevance with Representational Content

Fodor’s surrender to semantic externalism landed him directly in front of the dread specter of epiphenomenalism: if semantic content is broad and external (constituted by causal relations to distal objects in the outside world), how can an external semantic property make any causal difference whatsoever to a physical, computational brain operating strictly within the skull? This is the core problem of mental causation. If the broad semantic fact that my thought is about H2O is external to my brain, then my brain’s physical motor cortex must be executing its bodily commands purely on the basis of local neurobiology and syntax, leaving the semantic meaning completely inert—a mere epiphenomenal passenger along for the ride.

Fodor tackled this existential dilemma in Psychosemantics (1987) and A Theory of Content and Other Essays (1990). His brilliant solution was to show that while broad semantic properties do not act as direct, local physical causes, they are nonetheless causally relevant because they are nomologically mirrored by local syntactic properties. Consider an analogy: a Coca-Cola vending machine accepts physical coins. The machine’s mechanical sorter operates exclusively on the formal, physical dimensions of the coin—its thickness, diameter, weight, and electrical conductivity. The machine has no access to the legal, social fact that the coin is legal tender issued by the United States Treasury.

Yet, the machine’s operations are not epiphenomenal with respect to financial value! Why? Because the mint designs coins such that their physical dimensions precisely track their monetary value. The physical syntax mirrors the financial semantics. In exactly the same way, evolutionary history and causal grounding ensure that the local syntactic properties of Mentalese symbols lawfully correspond to their broad semantic truth conditions. Mental causation is completely genuine because computational syntax serves as the physical vehicle through which broad semantic properties exert real-world causal control over human actions.

9. The Critique of Evolutionary Psychology and Adaptationism

9.1 The Controversy of ‘What Darwin Got Wrong’

Throughout his career, Fodor was renowned for his fearlessness in challenging dominant scientific orthodoxies, but no debate unleashed a more ferocious firestorm of intellectual condemnation than his assault on the core logic of neo-Darwinian evolutionary theory. In 2010, Fodor published What Darwin Got Wrong, co-authored with the distinguished molecular biologist and cognitive scientist Massimo Piattelli-Palmarini. The book was greeted by evolutionary biologists, scientific institutions, and establishment philosophers with an explosive mix of outrage, indignation, and absolute bewilderment.

Fodor and Piattelli-Palmarini did not question the fact of biological evolution, nor did they advocate creationism or intelligent design; on the contrary, they were uncompromising naturalists. Their radical, provocative thesis was that the classical Darwinian mechanism of natural selection—conceived as an adaptationist filter operating via external environmental sorting—is conceptually incoherent, logically flawed, and incapable of doing the explanatory work assigned to it. Fodor claimed that natural selection cannot, even in principle, solve what he termed the “selection-for” problem.

The argument rested on the empirical reality of phenotypic correlation: whenever an organism undergoes natural selection, multiple traits are invariably linked together, whether through pleiotropic genetic expression, physical geometry, or historical co-occurrence. Consider the classic textbook example of the heart: the heart pumps blood, and the heart makes rhythmic pumping sounds. In ancestral environments, every organism whose heart pumped blood also had a heart that made pumping sounds. Both traits were strictly coextensive. Humans intuitively want to say: nature selected for blood-pumping, while the thumping sound was merely an accidental, correlated free-rider. Fodor’s devastating question was: what non-intentional, physical mechanism in nature distinguishes between blood-pumping and sound-making?

9.2 Intentionality, Intensionality, and Natural Selection

Fodor exposed a profound metaphysical confusion at the heart of evolutionary adaptationism: the illicit conflation of extensional physical processes with intensional (with an ‘s’) explanatory contexts. In human artificial breeding, when a farmer selectively breeds sheep, there is a conscious, intentional mind deciding to select for wool thickness rather than sheep weight. Because the farmer’s mind possesses intentionality, the farmer can represent the distinction between the two linked properties, even if they are physically coextensive. Artificial selection can differentiate between correlated traits because the breeder’s mind creates an intensional context where truth values are sensitive to how the traits are described.

However, Fodor pointed out that nature possesses no mind, no consciousness, no representations, and no intentions. Mother Nature is completely mindless, purely physical, and strictly extensional. But purely extensional physical processes cannot distinguish between physically coextensive properties! If trait A (pumping blood) and trait B (making a thumping sound) are nomologically tied in an ancestral population, nature acts on the organism as a whole package. To claim that nature “selected for A but not for B” requires evaluating counterfactual assertions about what would have happened if trait A had existed without trait B. But in mindless nature, there are no independent historical facts that ground counterfactual claims about uninstantiated variants.

The backlash from the biological and philosophical communities was immediate and brutal. Prominent philosophers including Daniel Dennett, Philip Kitcher, and Ned Block, along with leading evolutionary biologists, published scathing rebuttals. Dennett accused Fodor of committing basic conceptual errors regarding population genetics and the nature of counterfactual reasoning in the special sciences. Critics argued that natural selection does not require an intentional agent; the counterfactual truth that pumping blood was the causal driver of survival is established by the biomechanical physics of circulatory dynamics, not by psychological descriptions. Fodor fiercely defended his position in subsequent exchanges, insisting that his opponents were persistently anthropomorphizing natural selection by covertly treating it as a literal, teleological selector.

9.3 Internalist Constraints on Morphology and Cognition

If natural selection was fundamentally limited, how did Fodor and Piattelli-Palmarini propose to explain the astonishing diversity and complexity of biological form and cognitive architecture? Their alternative was to shift the primary explanatory weight from external, adaptationist environmental pressures to endogenous internal constraints: physical, geometric, structural, and molecular laws governing biological organisms.

Fodor drew heavily upon the revolutionary insights of evolutionary developmental biology (evo-devo) and the classic critiques of adaptationism formulated by Stephen Jay Gould and Richard Lewontin in their celebrated 1979 paper, “The Spandrels of San Marco and the Panglossian Paradigm.” Gould and Lewontin demonstrated that many prominent architectural features of living organisms are not bespoke adaptations engineered by natural selection, but “spandrels”—unavoidable structural byproducts of primary geometric constraints. Fodor argued that the morphology of animals and the structure of human cognition are severely constrained by deep, internal genetic cascades, physical dynamics, and biochemical channeling that dictate what forms are biologically permissible long before external selection ever encounters an organism.

This biological internalism was in flawless, seamless harmony with Fodor’s lifelong philosophical commitments. In linguistics, he had spent decades arguing alongside Chomsky that human language is not an external cultural adaptation conditioned by communicative utility, but an internal, genetically pre-specified, domain-specific biological organ governed by universal grammar. In the philosophy of mind, he had argued that cognitive concepts are innate and internally structured rather than learned from external sensory exposure. His critique of Darwinian adaptationism was the ultimate, logical culmination of his structuralist, rationalist worldview: the human mind, like the human body, is fundamentally shaped by internal biological architecture, not by the external tailoring of natural selection.

10. Concepts, Nativism, and Semantic Atomism

10.1 The Rejection of Conceptual Analysis and Definitions

In the late 1990s, Fodor turned his formidable analytical weapons toward the theory of concepts, culminating in his incendiary 1998 book, Concepts: Where Cognitive Science Went Wrong. For centuries, philosophical tradition and cognitive psychology had operated under the nearly unquestioned assumption that lexical concepts (such as [BACHELOR], [CHAIR], or [BIRD]) are complex, structured mental representations that can be decomposed into definitional constituents, prototypes, or inferential networks. Fodor systematically demolished every single one of these paradigms.

Fodor began by documenting the empirical, historical failure of the classical definitional theory. For over two millennia, philosophers and linguists had attempted to provide necessary and sufficient conditions for common lexical concepts, and they had failed in virtually every single case. Even seemingly trivial concepts like [GAME], as Wittgenstein famously noted, resist necessary and sufficient definitions. Decades of intensive lexicographical and psychological research had produced practically zero universally accepted definitional analyses for common terms. Concepts, Fodor concluded, simply do not have definitional structure.

Next, Fodor attacked the dominant alternative in cognitive psychology: prototype theory (pioneered by Eleanor Rosch) and exemplar models. Prototype theory claimed that possessing a concept is possessing a statistical cluster of typical properties: to have the concept [BIRD] is to represent things that fly, sing, lay eggs, and build nests. Fodor dismantled prototype theory with a single, devastating argument: prototypes do not compose. The cornerstone of the Computational Theory of Mind is compositionality: the meaning of a complex expression must be an invariant function of the meaning of its parts. But consider the complex concept [PET FISH]. The prototype of a [PET] is furry, warm, affectionate, and lives in a house (like a dog or cat). The prototype of a [FISH] is gray, cold, medium-sized, and swims in the open ocean (like a trout or cod). But what is the prototype of a [PET FISH]? It is a tiny, bright orange goldfish swimming in a glass bowl!

The prototypical properties of the complex concept [PET FISH] cannot be computed or derived from the prototypes of its constituent concepts. Therefore, concepts cannot be prototypes. If prototypes cannot compose, and compositionality is required to explain productivity and systematicity, then prototype theory is completely dead as an account of human concepts. Fodor extended this exact same demolition to inferential role semantics (the view that a concept’s identity is constituted by its inferential transitions within a holistic network), proving that inferential roles also fail the compositionality test and inevitably collapse into self-defeating semantic holism.

10.2 Radical Concept Nativism

Having dismantled definitions, prototypes, and inferential roles, Fodor arrived at an astonishing, inescapable conclusion that sent shockwaves through both philosophy and psychology: most lexical concepts are primitive, unanalyzable semantic atoms. A concept like [DOORKNOB], [CARBURETOR], [RED], or [DOG] possesses no internal compositional structure. It is a simple, indivisible symbolic primitive.

Now, pair this semantic atomism with Fodor’s earlier “paradox of concept learning”: concept acquisition cannot occur via hypothesis testing, because hypothesis testing requires already commanding the concepts in which the hypothesis is framed. If a concept like [DOORKNOB] is primitive and cannot be decomposed into simpler sensory parts (like “round metal thing used to open doors”), then it cannot be learned by assembling simpler concepts. The inescapable, logically deductive consequence was Fodor’s most notorious and widely ridiculed thesis: Radical Concept Nativism. Concepts such as [DOORKNOB], [CARBURETOR], [ELECTRON], and [AIRPLANE] are fundamentally innate!

To many of his contemporaries, this claim seemed completely mad: how could human beings possess an innate concept for “carburetor” or “doorknob” millions of years before internal combustion engines or doors were invented? In Concepts: Where Cognitive Science Went Wrong, Fodor clarified his position, reframing the mechanism of nativism. Fodor was not claiming that Neanderthals possessed a detailed mental image of a twentieth-century automobile carburetor. Rather, he argued for an informational locking model. What is biologically innate is the specialized neurocomputational disposition to be “locked onto” a specific external property upon minimal perceptual triggering.

Learning a concept is not a rational process of cognitive induction, hypothesis testing, or definition assembly; it is a brute, biological process of triggering and nomic locking. Just as an encounter with ducks triggers an innate imprinting mechanism in a newborn duckling, an infant’s sensory encounters with doorknobs or dogs triggers an innate biological disposition that locks an internal atomic Mentalese token onto the distal property of being a doorknob or a dog. Radical concept nativism was not a mystical biological prophecy; it was the relentless, unyielding application of rationalist logic to the failures of empiricist theories of learning.

10.3 Semantic Atomism and Holism

Fodor’s defense of semantic atomism was cemented by his ferocious, career-long opposition to meaning holism, an assault waged in close partnership with his friend and collaborator Ernie Lepore, culminating in their 1992 masterpiece, Holism: A Shopper’s Guide. Meaning holism, widely endorsed by giants of twentieth-century philosophy including W. V. O. Quine, Donald Davidson, and Ned Block, is the thesis that the semantic content of any single mental state or linguistic term depends constitutively upon its systemic relations to all the other beliefs, inferences, and concepts within the entire cognitive system.

Fodor and Lepore demonstrated that meaning holism is an intellectual catastrophe for philosophy and cognitive psychology. If holism is true, then no two individuals can ever share the exact same belief or mean the same thing by the same word. Consider: you believe that Shakespeare was a great playwright, and you also believe that coffee tastes better with oat milk. I believe Shakespeare was a great playwright, but I despise oat milk. Because our broader webs of belief diverge, and because under holism the meaning of any single concept is constituted by its total inferential relations across the entire web, your concept [SHAKESPEARE] is semantically distinct from my concept [SHAKESPEARE]. We are literally speaking past one another, unable to express the same proposition!

Furthermore, under holism, an individual can never rationally change their mind. If you abandon a single belief, the entire holistic network shifts, meaning every concept you possessed has now changed its identity; you have not revised a belief, you have simply swapped your entire mind for a different one. Communication, mutual agreement, scientific disagreement, and psychological generalization instantly collapse into solipsism. Against this holistic nihilism, Fodor’s informational atomism stood as an unshakeable bulwark: the identity of a concept depends purely on a one-to-one, external nomic relation between an internal symbol and an external property, completely independent of an agent’s other beliefs, associations, or inferential networks. Communication is stable, disagreement is real, and the commonality of human thought is preserved.

11. Key Philosophical Debates, Critiques, and Rebuttals

11.1 The Epistemic Conflict with Empiricism and Pragmatism

Viewed in its sweeping historical context, Jerry Fodor’s intellectual corpus represents the most formidable, sustained resurrection of Cartesian rationalism mounted in twentieth-century analytic philosophy. He positioned his work in absolute, uncompromising opposition to the Anglo-American empiricist tradition that stretched from John Locke, George Berkeley, and David Hume down through John Dewey, logical empiricism, and modern neo-pragmatism. Fodor viewed empiricism’s fundamental dogma—that the human mind begins as a tabula rasa (blank slate), drawing all its concepts, structural categories, and inferential machinery from sensory experience—as an intellectual disaster that had crippled psychology and philosophy for three centuries.

Fodor engaged in fierce philosophical polemics with contemporary neo-pragmatists and inferentialists, notably Richard Rorty, Robert Brandom, and the broader Pittsburgh school of philosophy influenced by Wilfrid Sellars. These thinkers argued for an epistemic, pragmatist theory of mind and language: meaning is constituted by social norms, communal justification practices, and inferential scorekeeping in the “space of reasons.” Brandom’s inferentialism asserted that to possess a concept is simply to master the social, normative proprieties of inferential assertion and endorsement.

Fodor rejected this pragmatist turn with ferocious contempt, viewing it as a thinly disguised revival of behaviorism and anti-realist idealism. For Fodor, truth is correspondence to objective reality, completely independent of human epistemic practices, social justification, or communal consensus. Semantic content is determined by causal-nomic relations between internal symbols and the natural world, not by social language games. To prioritize epistemology or pragmatics over metaphysics—to treat meaning as dependent upon human verification or assertion rules—was, Fodor declared, to put the cart before the horse. Rationalism, biological innateness, and metaphysical realism were the only legitimate foundations for a scientific philosophy of mind.

11.2 The Frame Problem and the Limits of Computational Cognition

One of Fodor’s most celebrated contributions to the foundational philosophy of artificial intelligence was his relentless philosophical diagnosis of the computational Frame Problem. Originally formulated in 1969 by John McCarthy and Patrick J. Hayes within formal robotics, the frame problem concerned the technical challenge of writing axioms that describe how an action changes the world without having to write an astronomical number of non-change axioms stating what remains unchanged (e.g., painting a wall red changes the wall’s color, but does not change the weight of the toaster in the kitchen).

Fodor recognized that the frame problem was not merely a narrow coding glitch in symbolic AI; it was the engineering manifestation of the deepest, most intractable problem in epistemology: the problem of induction and relevance. How can an artificial system determine, in real time and without exhaustive search, which background facts are relevant to a given situation, and which facts are completely irrelevant? When you decide to pack an umbrella because the sky is dark, your central processor automatically ignores your knowledge of the telephone number of the prime minister of Australia, the atomic weight of platinum, and the score of the 1986 World Series. How does the mind pull off this miraculous, instantaneous filtering?

Fodor pointed out that because relevance is isotropic and Quinean, there is no formal, local syntactic rule that can mark a belief as “inherently irrelevant.” A fact that is irrelevant in ninety-nine contexts might suddenly become critically relevant in the hundredth context. A computer program operating via local, syntactic algorithms must evaluate every single proposition in its memory bank to confirm that it is irrelevant—a computational explosion that paralyzes the machine in an infinite computational loop. This led Fodor to a profound, unyielding pessimism regarding the prospects for Artificial General Intelligence (AGI). He maintained that classical computational cognitive science had reached an impenetrable wall: it had solved encapsulated perceptual modules, but as long as it lacked an algorithmic solution to global relevance and the frame problem, building a truly rational, domain-general computational mind was a pipe dream.

11.3 Collaborations and Scholarly Engagements

The development of Fodor’s philosophies was marked by deep, career-defining intellectual partnerships and legendary dialectical rivalries. Chief among his collaborations was his long, extraordinarily productive partnership with Ernie Lepore. Together, Fodor and Lepore functioned as a philosophical demolition squad, co-authoring a vast array of papers and books, including Holism: A Shopper’s Guide (1992) and The Compositionality Papers (2002). Their collaborative work was marked by razor-sharp dialectical rigor, relentless humor, and an absolute refusal to compromise with semantic holism, prototype theories, or conceptual role semantics.

Equally profound was Fodor’s lifelong engagement with Noam Chomsky. While Fodor and Chomsky stood shoulder-to-shoulder in their defense of rationalism, nativism, and internal mental representations against behaviorism, they parted ways on crucial architectural issues. Fodor championed the strict modularity of perceptual systems and fought against evolutionary explanations, whereas Chomsky resisted Fodorian informational encapsulation for language processing and focused on the minimalist program and biolinguistics. Their deep mutual respect provided the bedrock upon which the institutional infrastructure of modern cognitive science was constructed.

Finally, Fodor engaged in epic philosophical duels with his former supervisor Hilary Putnam over externalism, functionalism, and reductionism, and with British philosophers like John McDowell and Christopher Peacocke over the nature of perceptual content. McDowell argued that perceptual experience is thoroughly conceptual and normatively rich from the ground up, while Peacocke developed complex accounts of non-conceptual content. Fodor, fiercely defending the informational encapsulation of modular perception, insisted that low-level perceptual input analyzers deliver dumb, shallow, non-conceptual computational outputs long before higher-level conceptualization occurs. Across all these engagements, Fodor was universally recognized as an intellectual adversary of the highest order—fair, fearless, uncompromising, and impossible to ignore.

12. Jerry Fodor’s Enduring Legacy and Impact on Contemporary Philosophy of Mind

12.1 Transformation of the Cognitive Science Landscape

When Jerry Fodor passed away on November 29, 2017, at the age of eighty-two, the philosophical and scientific worlds mourned the loss of a true intellectual titan. Looking back across the landscape of modern philosophy, it is difficult to identify any single thinker who did more to shape, define, and legitimize the conceptual foundations of cognitive science. Prior to Fodor, psychology was an embattled discipline, torn between the physical reductionism of hard neuroscience, the operationalist anti-mentalism of behaviorism, and the therapeutic linguistic dissolves of ordinary language philosophy.

Fodor gave psychology its scientific soul. By formulating the Computational Theory of Mind, inventing the Language of Thought hypothesis, establishing the autonomy of the special sciences, and constructing an impregnable defense of intentional realism, Fodor demonstrated that the mind could be studied as a legitimate, autonomous computational-representational system. Today, the core vocabulary of cognitive science—mental representations, domain-specific modules, information processing, cognitive architecture, combinatorial syntax—is the very vocabulary that Jerry Fodor forged, clarified, and defended across five decades of peerless scholarship.

His impact on empirical developmental psychology and animal cognition remains monumental. The modern empirical discovery of “core knowledge” systems in human infants—pioneered by Elizabeth Spelke, Susan Carey, and Renée Baillargeon—which demonstrates that human infants possess innate, domain-specific computational systems for naive physics, numerical tracking, and psychological agency, is the direct, triumphant vindication of Fodorian modularity and concept nativism. Fodor demonstrated that common-sense folk psychology was not a relic of pre-scientific superstition to be cast onto the ash heap of history, but the profoundest scientific truth ever discovered about the nature of human beings.

12.2 Contemporary Relevance in the Era of Modern Artificial Intelligence

As the twenty-first century enters the epoch of advanced artificial intelligence, generative large language models, and autonomous systems, Jerry Fodor’s philosophical warnings and structural insights have not merely endured—they have become prophetic. Today’s machine learning landscape is completely dominated by massive, statistical, distributed neural networks: modern implementations of the very connectionist paradigm that Fodor and Pylyshyn dissected in 1988.

The contemporary struggles of modern artificial intelligence are precisely the structural pathologies Fodor predicted with mathematical clarity decades ago:

• The Failure of Compositionality: While modern LLMs produce dazzling linguistic imitations, they frequently fail at basic compositional logic when confronted with counterfactual, out-of-distribution reasoning, proving Fodor’s point that statistical correlation is not compositional syntax.
• The Problem of Hallucination: Modern neural networks struggle with truth and reference precisely because they lack an externalist, causal-informational semantics; they are syntactic prediction engines operating in epistemic isolation from distal grounding, perfectly exemplifying the limitations of ungrounded syntax.
• The Persistence of the Frame Problem: The inability of autonomous AI systems to operate reliably in dynamic, unpredictable real-world environments without massive human scaffolding stems directly from the unsolved, isotropic frame problem that Fodor diagnosed.

Far from being obsolete relics of twentieth-century debates, Fodor’s compositionality criteria, his critique of massive modularity, and his analysis of central abductive reasoning are now the core theoretical benchmarks used by computer scientists and philosophers seeking to build the next generation of robust, trustworthy, neuro-symbolic artificial intelligence. Fodor’s cautionary arguments remind us that brute statistical correlation can never replace the formal, compositional architecture of reason.

12.3 Final Assessment of a Philosophical Iconoclast

Jerry Fodor was that rarest of intellectual phenomena: an uncompromising rationalist who was simultaneously the greatest defender of common sense; an analytical logician who wrote with the dazzling wit of a satirist; an architect of cognitive science who spent his final decades warning the discipline against its own hubris. His philosophical architecture was breathtaking in its systematicity and coherence: CTM, LOT, modularity, asymmetric dependence, intentional realism, and concept nativism were not isolated ideas, but mutually supporting pillars of an integrated, grand philosophical cathedral.

Fodor took the ordinary, unvarnished human experience—that we have thoughts, that our thoughts are about the world, that our reasons cause our actions, and that we communicate these thoughts through structured language—and showed that this everyday reality is fully compatible with the physical universe. He refused to give an inch to reductionists, behaviorists, pragmatists, or eliminativists who demanded that human beings surrender their beliefs and desires in the name of a misguided, scientistic materialism. He stood before the scientific world and demonstrated that the mind is a real, physical, and profoundly rational computational wonder.

To engage with the writings of Jerry Fodor is to experience analytic philosophy at its absolute zenith: dazzlingly clever, brutally rigorous, unapologetically honest, and deeply, uproariously human. As long as human beings puzzle over the mysteries of thought, representation, language, and meaning, the extraordinary, unforgettable voice of Jerry Fodor will echo down the halls of philosophy, challenging our dogmas, puncturing our pretensions, making us laugh, and forcing us to think.

Conclusion

The intellectual odyssey of Jerry Alan Fodor remains one of the most compelling narratives in the history of ideas. From his early days in the classrooms of Sydney Morgenbesser and Hilary Putnam to his decades at the vanguard of the cognitive revolution at MIT and Rutgers, Fodor consistently chose the path of intellectual independence. In an academic culture frequently swayed by technical fads and conceptual bandwagons, he was the ultimate dialectical anchor, relentlessly subjecting contemporary orthodoxies to devastating, principled critiques. His defense of folk psychology, his formulation of the Language of Thought, and his profound architectural delineation of the modular and central mind revolutionized our understanding of cognition.

Ultimately, Jerry Fodor’s enduring triumph was to preserve the dignity of the human mind within the cold, mechanistic framework of modern science. He showed that we do not have to choose between a physicalist ontology and the rich, meaningful reality of mental life. Through the magnificent scaffolding of the Computational Theory of Mind, Fodor proved that the brain can be a physical machine without turning the mind into a biological illusion. Thoughts are real; meanings are real; reasons genuinely matter. In uniting the rigorous formalisms of Alan Turing with the rationalist insights of René Descartes and the intuitive wisdom of Granny, Jerry Fodor constructed a monument of twentieth-century thought that will continue to inspire, provoke, and guide the philosophy of mind for generations to come.

References

• Chomsky, N. (1957). Syntactic Structures. Mouton & Co.
• Chomsky, N. (1959). A review of B. F. Skinner’s Verbal Behavior. Language, 35(1), 26–58. https://doi.org/10.2307/411334
• Churchland, P. M. (1981). Eliminative materialism and the propositional attitudes. The Journal of Philosophy, 78(2), 67–90. https://doi.org/10.2307/2025900
• Dennett, D. C. (1987). The Intentional Stance. MIT Press.
• Fodor, J. A. (1960). Conditions for Meaning Rules [Doctoral dissertation, Princeton University].
• Fodor, J. A. (1974). Special sciences (or: The disunity of science as a working hypothesis). Synthese, 28(2), 97–115. https://doi.org/10.1007/BF00485230
• Fodor, J. A. (1975). The Language of Thought. Harvard University Press.
• Fodor, J. A. (1980). Methodological solipsism considered as a research strategy for cognitive psychology. Behavioral and Brain Sciences, 3(1), 63–73. https://doi.org/10.1017/S0140525X00001771
• Fodor, J. A. (1983). The Modularity of Mind: An Essay on Faculty Psychology. MIT Press. https://doi.org/10.7551/mitpress/4737.001.0001
• Fodor, J. A. (1987). Psychosemantics: The Problem of Meaning in the Philosophy of Mind. MIT Press.
• Fodor, J. A. (1990). A Theory of Content and Other Essays. MIT Press.
• Fodor, J. A. (1994). The Elm and the Expert: Mentalese and Its Semantics. MIT Press.
• Fodor, J. A. (1998). Concepts: Where Cognitive Science Went Wrong. Oxford University Press. https://doi.org/10.1093/0198236360.001.0001
• Fodor, J. A. (2000). The Mind Doesn’t Work That Way: The Scope and Limits of Computational Psychology. MIT Press. https://doi.org/10.7551/mitpress/4627.001.0001
• Fodor, J. A. (2008). LOT 2: The Language of Thought Revisited. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199548774.001.0001
• Fodor, J. A., & Lepore, E. (1992). Holism: A Shopper’s Guide. Blackwell.
• Fodor, J. A., & Lepore, E. (2002). The Compositionality Papers. Oxford University Press. https://doi.org/10.1093/0199252160.001.0001
• Fodor, J. A., & Piattelli-Palmarini, M. (2010). What Darwin Got Wrong. Farrar, Straus and Giroux.
• Fodor, J. A., & Pylyshyn, Z. W. (1988). Connectionism and cognitive architecture: A critical analysis. Cognition, 28(1–2), 3–71. https://doi.org/10.1016/0010-0277(88)90031-5
• Gould, S. J., & Lewontin, R. C. (1979). The spandrels of San Marco and the Panglossian paradigm: A critique of the adaptationist programme. Proceedings of the Royal Society of London. Series B. Biological Sciences, 205(1161), 581–598. https://doi.org/10.1098/rspb.1979.0086
• Millikan, R. G. (1984). Language, Thought, and Other Biological Categories: New Foundations for Realism. MIT Press.
• Putnam, H. (1967). Psychological predicates. In W. H. Capitan & D. D. Merrill (Eds.), Art, Mind, and Religion (pp. 37–48). University of Pittsburgh Press.
• Putnam, H. (1975). The meaning of ‘meaning’. Minnesota Studies in the Philosophy of Science, 7, 131–193.
• Pylyshyn, Z. W. (1984). Computation and Cognition: Toward a Foundation for Cognitive Science. MIT Press.
• Searle, J. R. (1980). Minds, brains, and programs. Behavioral and Brain Sciences, 3(3), 417–424. https://doi.org/10.1017/S0140525X00005756
• Smolensky, P. (1988). On the proper treatment of connectionism. Behavioral and Brain Sciences, 11(1), 1–23. https://doi.org/10.1017/S0140525X00052432
• Turing, A. M. (1950). Computing machinery and intelligence. Mind, 59(236), 433–460. https://doi.org/10.1093/mind/LIX.236.433