Pramāṇa, Psyche & the
Limits No Machine Can Cross

The Nyāya school requires that a genuine pramātṛ (knowing subject) be capable of pramā (valid cognition). But the depth-psychological and cognitive-scientific literature of the past century reveals that the human pramātṛ operates under severe, systematic, and largely invisible constraints. Understanding these constraints is not an argument for AI's superiority — it is the opposite: the constraints are part of what makes human knowing real, embodied, and epistemically grounded in a way that AI's unbounded token-processing is not.

The Seven Fundamental Cognitive Limits of the Human Knower

Nyāya posits that the pramātṛ is the locus of all cognition — but depth psychology reveals that the vast majority of what the human mind does occurs below the threshold of conscious awareness. The psychoanalytic tradition, beginning with Freud and extended by Jung, Adler, Klein, Winnicott, and Lacan, maps the terrain of the unconscious in its clinical dimensions. Cognitive science complements this with empirical documentation of automatic, implicit, and non-conscious processes that shape — and often determine — conscious cognition. What the human knower cannot see in itself is as epistemically significant as what it can.

Three Depth-Psychological Insights Applied to the NI-AI Comparison

The post-Enlightenment philosophical tradition has largely treated emotion as the enemy of knowledge — the distorting force that must be controlled or excluded for reasoning to reach its conclusions cleanly. Nyāya's posture is more nuanced: the ātman's icchā (desire) and dveṣa (aversion) are part of what makes it a knower rather than merely an information-processor, because they are what give the knower a stake in the truth. Depth psychology goes further: it reveals that emotion is not merely motivationally relevant to knowing but constitutively epistemic — that certain forms of knowledge are only available through emotional attunement.

The Five Dimensions of Emotional Knowing — Each Unavailable to AI

The phenomenological tradition — from Husserl's analysis of the Leib (lived body) through Merleau-Ponty's radical claim that "consciousness is in the first place not a matter of 'I think that' but of 'I can'" — demonstrates that human knowledge is primarily not propositional but practical: it is knowledge embedded in the body's capacities for skilled action, orientation in space, manipulation of objects, and attunement to the environment. This tradition converges with findings from cognitive science (Lakoff and Johnson's embodied cognition) and developmental psychology (Piaget's sensorimotor stage) to establish that abstract conceptual knowledge is built atop and remains structured by the body's practical engagement with the physical world.

What Embodied Knowing Is — and What It Enables

Consider what a surgeon knows. They know anatomy propositionally — they can state the location, function, and pathology of every organ system. But the knowledge that makes them a skilled surgeon is different in kind from propositional knowledge: it is the knowledge in their hands — the calibrated pressure, the tactile discrimination between healthy and diseased tissue, the micro-muscular adjustments that respond to resistance without conscious deliberation. This is embodied knowing: knowledge that lives in the body's trained competencies rather than in propositional representations.

Merleau-Ponty's famous analysis of the blind man's cane demonstrates the structure of embodied knowing: after sufficient use, the cane ceases to be an object of perception and becomes an extension of the body-schema — the blind man perceives not the cane but the world through the cane. The tool disappears into the body's competence. Every skilled human practitioner — the musician, the craftsman, the athlete, the surgeon, the dancer — achieves this transparency of tool and technique, in which the medium vanishes and only the world remains.

The epistemological consequence is fundamental. Lakoff and Johnson (1980, 1999) demonstrated that even the most abstract domains of human thought — mathematics, logic, ethics, metaphysics — are structured by embodied metaphors derived from physical experience: more is up, time is a path, argument is war, understanding is grasping. These are not arbitrary linguistic conventions; they are the traces of embodied experience in abstract thought. If Lakoff and Johnson are right — and the empirical evidence strongly supports them — then human abstract reasoning is grounded in and structured by bodily experience in ways that are constitutive, not merely motivational.

Nyāya's ātman is a persisting substance — the same self that perceives today was the self that perceived yesterday, and the continuity of this persisting subject is what makes learning from experience possible. But contemporary psychology reveals that this continuity is not merely a metaphysical postulate; it is actively constructed through memory and narrative, and this constructive process is itself epistemic — it determines what the self knows about itself and therefore what it can know about anything.

Three Dimensions of Memory as Epistemic Architecture

Autobiographical Memory — The Story That Holds the Knower Together. Autobiographical memory is not a recording device but a constructive process. Every act of remembering reconstructs rather than replays: the memory is rebuilt from fragments, shaped by current knowledge, emotional state, and the narrative frameworks available to the rememberer. This means that the pramātṛ's relationship to its own past is always partially fictional — not in the sense of deliberately false but in the sense of constructively produced. The self that the pramātṛ knows is the self of a story, and the story is continuously being revised.

The epistemological consequence for the NI-AI comparison: AI has no autobiographical memory within or across sessions. Each session begins from a static parameter configuration, not from an accumulated personal history. The AI that helped you yesterday does not remember doing so — not because its hardware is insufficient but because it has no self for whose story that memory would be relevant. There is no continuing narrative subject for whom past interactions are part of an unfolding life.

Semantic Memory — General Knowledge Grounded in Personal Experience. Human general knowledge (semantic memory: knowing that Paris is the capital of France, that water boils at 100°C, that courage is a virtue) is grounded in, though increasingly independent of, specific personal episodes of learning. The child learns that fire is hot by touching something hot — and the semantic knowledge (fire = hot) retains the phenomenological quality of the episodic experience even as the specific episode fades. General human knowledge has this experiential texture throughout.

AI's semantic representations lack this texture. The claim "fire is hot" in an AI's output is derived from the statistical frequency with which "fire" and "hot" co-occur in text — not from the phenomenological memory of heat. This is not merely a difference of degree; it is a difference of kind. The AI knows that fire is hot in the way a person knows that the capital of a country they've never visited is a certain city — by description, not by acquaintance. Russell's distinction between knowledge by description and knowledge by acquaintance maps precisely onto the AI-NI gap in semantic memory.

Prospective Memory & Intention — The Knower Who Has Plans. Humans not only remember the past; they remember the future — they form intentions, set goals, and carry forward commitments across time. Prospective memory (the memory to do something in the future) is a uniquely temporal form of knowledge: it holds open a space in the future that the self is committed to filling. This forward-directedness of human consciousness — Husserl's protention, Heidegger's projection — is constitutive of human agency. The human knower is always already oriented toward a future they are trying to bring about.

The Nyāya tradition's account of śabda (testimony) recognizes a profound truth: some knowledge can only be acquired through another knowing subject. But depth psychology extends this insight further: the very structure of the knowing self is intersubjective from its inception. The self does not first exist and then enter into relations with others — it is constituted through and by relations with others. To understand the human pramātṛ fully is to understand a being whose knowing-capacity is irreducibly social.

Winnicott, Bowlby, and the Relational Origins of Mind

Donald Winnicott's object-relations theory establishes that the infant's very capacity for self-experience depends on the quality of its early relational environment. The "good-enough mother" — attuned, responsive, reliable — provides the holding environment within which the infant first experiences itself as a coherent self. Without this relational mirroring, self-experience fragments. The self that knows is therefore not a monadic substance that happened to be born into a social world; it is a being whose self-constituting capacity was enabled — and partly shaped — by the quality of care it received.

John Bowlby's attachment theory adds a precise behavioral and neurological specification: attachment patterns (secure, anxious-ambivalent, avoidant, disorganized) formed in the first year of life create internal working models of self and other that shape all subsequent relational cognition. A securely attached person approaches epistemic challenges with curiosity and trust; an anxiously attached person approaches them with hypervigilance; an avoidantly attached person with premature closure. These are not merely motivational differences — they are differences in the architecture of knowing.

Theory of Mind — Knowing That Others Know Differently

One of the most distinctive capacities of human intelligence is theory of mind: the ability to represent other minds as having beliefs, desires, intentions, and knowledge-states different from one's own. Theory of mind develops through specific neural substrates (the temporoparietal junction, medial prefrontal cortex) and through specific developmental experiences (shared attention, pretend play, cooperative problem-solving). Without theory of mind, human social coordination, moral reasoning, and communicative understanding collapse entirely.

AI can model theory of mind — it can produce outputs that accurately describe the likely beliefs, desires, and intentions of characters in scenarios. But AI does not have theory of mind in the sense of genuinely representing another mind as a mind. It represents descriptions of mental states — the linguistic surface of intersubjectivity — without having the intersubjective experience that grounds genuine mind-reading. The difference is between knowing that you are in pain (propositional) and feeling concern for you because I can to some degree resonate with what you are feeling (empathic). AI does the former; it cannot do the latter.

The opening verse of the Nyāyasūtra begins with duḥkha — suffering — as the condition that makes epistemology necessary. The entire project of knowing, in the Nyāya framework, is motivated by the knower's experience of suffering and their desire to escape it through accurate cognition. This connection between suffering and knowledge is not incidental; it is constitutive. The Nyāya knower is not a dispassionate observer seeking truth for its own sake — they are a suffering being seeking truth as a path out of suffering. Knowledge matters because it changes the knower's situation.

Five Ways in Which Suffering Constitutes Human Knowledge

1. Mortality as Epistemic Ground. Heidegger's analysis of Being-towards-death (Sein-zum-Tode) establishes that the awareness of one's own finitude is not merely a psychological fact about humans but an ontological structure that shapes everything about human existence, including its cognitive character. The human knower who knows they will die knows the world differently from a being that does not face this limit. Every decision carries the weight of irreversibility; every moment has the character of something that will not come again; every relationship is colored by the possibility of loss. This is not pessimism — it is the depth-structure of finite knowing. AI has no mortality and therefore has no being-towards-death. Its "knowledge" carries no weight of irreversibility, no character of unrepeatable presence.

2. Pain as Data. Physical and psychological pain are among the most important epistemic signals available to biological organisms. Pain is not merely an unpleasant experience; it is a dedicated information channel that carries urgent, high-priority signals about threats to the organism's integrity. The person who has experienced loss knows what loss means in a way that no description can fully convey. The clinician who has experienced illness knows patient suffering differently from one who knows it only from textbooks. Pain's epistemic function is not reducible to its informational content — it is inseparable from the phenomenological quality of the experience itself.

3. Trauma and Post-Traumatic Growth. Traumatic experience — overwhelming events that exceed the ordinary capacity of the psyche to integrate — reorganizes the knower's relationship to reality in fundamental ways. Post-traumatic stress disorder represents one outcome: a reorganization that is fragmenting and incapacitating. Post-traumatic growth (Tedeschi & Calhoun) represents another: a reorganization that deepens the knower's appreciation of existence, relationships, personal strength, and existential priorities. Neither outcome is available to AI. The transformation of the knower through overwhelming experience — the deepening or fragmenting of the self — is a form of knowledge-through-existence that cannot be replicated by a system that does not exist in the relevant sense.

4. Grief and the Epistemology of Loss. The experience of grief — the response to irreversible loss — is among the most cognitively profound experiences available to humans. In grief, the knower must restructure their understanding of the world around a permanent absence. The absent beloved continues to shape the griever's cognition in complex ways: through continued internal dialogue, through heightened sensitivity to anything the beloved valued, through the restructuring of one's own identity around what one has lost. This restructuring of the world around an absence is a form of knowing that has no parallel in AI — because AI has no beloveds, no losses, and no internal dialogues with the absent.

5. The Wisdom That Only Suffering Teaches. Across cultures and philosophical traditions — from the Nyāya path from duḥkha to apavarga, to the Buddhist understanding of the First Noble Truth, to the Greek concept of pathei mathos (wisdom through suffering), to Nietzsche's "that which does not kill me" — there is a consistent recognition that certain kinds of wisdom are only accessible through the experience of suffering. Not all suffering produces wisdom; unreflected suffering can produce trauma without insight. But reflectively integrated suffering — suffering that has been metabolized through the long process of making meaning from pain — produces a quality of understanding of human existence that is unavailable to those who have not suffered, and certainly unavailable to systems that cannot suffer.

The nine chapters of Part II, taken together, map the complete architecture of the human knower as it has been revealed by depth psychology, cognitive science, phenomenology, and evolutionary biology. They do not describe a system with certain limitations and certain capacities. They describe a kind of being — a being that knows from the inside, that knows because it suffers and desires, that knows through a body that has learned from millions of years of encounter with a physical world, that knows intersubjectively through the medium of relations with other knowing beings, and that knows cumulatively through a memory that constructs a self from an accumulated life-history.

The Nyāya framework that preceded these nine chapters offers the epistemological scaffolding: pramāṇa, pramātṛ, vyāpti, hetvābhāsa. The nine chapters of Part II fill that scaffolding with empirical and phenomenological content. Together they produce a picture of such complexity that the question "can AI replicate human intelligence?" begins to look like it was always the wrong question.

The Right Questions — Reframed Through Nyāya and Depth Psychology

The Practical Synthesis — Seven Principles for Right Relation to AI

Combined Bibliography — Volumes I & Enhanced Part II