Abstract. The comparative study of religion has lacked a stable unit of comparison since its inception. Belief, ritual, myth, doctrine, symbol, institution, ontology, and cosmology have each been proposed as the common object that religious traditions share, and each has failed for the same underlying reason: each privileges the historical and conceptual profile of a particular tradition and then measures every other tradition by how well it fits that profile. Two failures of this kind have been especially costly. The first is the elevation of belief , private, propositional assent , into the basic unit of religious life, a category whose apparent universality conceals its origin in a specifically Protestant settlement of what religion fundamentally is. The second, less often named but equally consequential, is the field’s persistent search for a clean distinction between personal gods and impersonal forces, a search that has never succeeded because the distinction it is looking for is not the one that actually organises religious difference. This article argues that the missing tertium comparationis , the third term against which otherwise incommensurable traditions can be genuinely compared , is not an object at all. It is a task. Every religious, magical, and scientific tradition must discern the recursive organisation of reality: determine which entities are nonrecursive, selfrecursive, or interrecursive; identify where recursive agency lies and how much of it is compellable; recognise the load-bearing recursive processes that govern flourishing and its failure; and establish, on that basis, how living beings should coordinate with the world they have discerned. Science, magic, and religion turn out to be three different answers to this shared task rather than three points on an evolutionary ladder. The article develops this framework, shows that the highest-stakes disturbances of life are what force this discernment into explicit view, argues that traditions typically supply orientation within an unresolved discernment rather than resolution of it, and proposes that the traditions themselves have often already said as much, most visibly in the Hindu and Buddhist elaboration of dharma and in every tradition’s account of revelation, teaching, or the disclosed will of the divine. It reclassifies eight major traditions through this framework and closes by arguing that recursive discernment does what no previous comparative category has managed: it treats every tradition as a coherent solution to a real problem, rather than as a more or less adequate approximation of one tradition’s answer.
The Crisis of Comparison in the Study of Religion
The comparative study of religion has always struggled with a deceptively simple question: what exactly are we comparing? The question sounds as though it should have a straightforward answer, and every generation of theorists has proposed one, with genuine confidence that this time the answer will hold.
Belief was the first and remains the most persistent candidate. Edward Tylor’s foundational definition of religion as belief in spiritual beings made propositional content , what a person holds to be true about the invisible order of things , the basic unit against which traditions could be catalogued and ranked. The difficulty, pointed out with increasing force across the twentieth century and most sharply by Talal Asad, is that belief as an inner, propositional, individually held mental state is not a neutral analytic category waiting to be applied evenhandedly across traditions. It is itself a specifically Protestant Christian achievement, the product of a Reformation that had, for its own theological reasons, relocated the essence of religious life from practice and participation into private assent to doctrine. Measuring Catholic sacramental practice, Confucian ritual propriety, or Shinto shrine attendance by the propositional beliefs their practitioners privately hold is not comparing traditions on a shared axis. It is comparing every tradition to Protestantism’s own selfdescription and grading the others on their distance from it.
A second bias has done nearly as much damage while attracting far less scrutiny, because it presents itself not as a theological import but as a neutral empirical question: is the entity a given tradition addresses a person, or is it an impersonal force? The question has organised more than a century of theorising about the origins and varieties of religion. Edward Tylor’s animism proposed that religion begins with belief in personal spiritual beings, souls and ghosts modelled on the human self. Robert Marett’s animatism proposed, against Tylor, that an earlier and more basic religious orientation addresses not persons at all but a diffuse impersonal power, mana, that precedes and does not require personification. The disagreement between animism and animatism has never been resolved, and this article argues that it could not have been resolved, because both positions accept a framework in which the decisive question about any attributed power is whether or not it has been modelled as a person. That framework is not neutral. The insistence that the fundamental religious question is personal-or-impersonal is itself most at home in a theological tradition that has spent centuries elaborating exactly what it means for God to be a person , three persons, one substance , and that reached, in its Protestant elaboration, an unprecedented emphasis on the believer’s direct, personal relationship with a personal God. Importing that question into the comparative study of traditions that have never organised their own selfunderstanding around personhood at all does not neutrally describe those traditions. It measures them against a category that belongs to one tradition’s specific theological history.
Ritual fares no better as the universal unit than belief does, and for the mirror-image reason. Traditions organised centrally around correct doctrinal formulation , much of postReformation Protestant Christianity, much of modern Buddhist apologetics addressed to Western audiences , turn out to be poor fits for a comparative category built from traditions in which orthopraxy rather than orthodoxy is the primary concern. Myth privileges traditions with rich narrative cosmogonies over traditions whose central material is legal, doctrinal, or contemplative. Symbol, in the Geertzian formulation, is capacious enough to describe almost anything and therefore distinguishes almost nothing; it tells us that traditions produce systems of meaning without telling us what, structurally, differentiates one system of meaning from another. Institution privileges traditions with clergy, canon, and organisational continuity over traditions whose religious life is diffusely distributed across households, lineages, and land. Ontology, the most recent and in some ways most promising candidate, associated with the so-called ontological turn in anthropology, still tends to reproduce a background distinction , nature versus culture, object versus person , that is itself a specifically Western philosophical inheritance, however much the ontological turn’s own proponents intend to escape it.
What every one of these candidates shares, belief and the personal/impersonal binary most of all, is a specific and recurring failure mode. Each proposes an object , a kind of mental content, a kind of practice, a kind of narrative, a kind of institutional form, a kind of interior life attributed to the addressed power , and then asks how much of that object each tradition contains. But different traditions do not contain different amounts of the same object. They have configured an entirely different problem differently, and the search for a common object mistakes structural variation for a scale of more or less.
The proposal advanced here begins from a different starting point. The universal does not lie in an object that traditions share to different degrees. It lies in a task that every tradition, whatever its historical content, cannot avoid facing: living beings coordinating a life must continually discern the recursive organisation of the reality they inhabit. They must determine which entities respond to being acted upon and which do not; which entities are self-organising in ways that must be respected rather than merely predicted; which entities respond specifically to them, modelling their modelling in return; and what, given this recursive landscape, must be done to coordinate successfully within it. Science, magic, and religion, and every specific tradition within the third of these categories, are different answers to this one shared task, not different amounts of a shared substance. This becomes the article’s foundational claim, and everything that follows works out its consequences.
Recursive Discernment as the Missing TertiumComparationis
Recursive discernment, in the sense developed across this series, is the continual activity through which living beings determine, of any aspect of an encountered situation, what kind of recursive organisation it exhibits and how much further discernment is warranted before action can proceed. It is not knowledge in the propositional sense, because it does not terminate in a settled true-or-false claim about the world. It is not ritual in the narrow behavioural sense, because it operates prior to and beneath any specific practice, as the ongoing coordinative activity that specific practices exist to serve. Discernment is what a mesocosm , a lived, coordinated world , must continually accomplish simply in order to remain a coordinated world at all, and every tradition of thought and practice that has ever organised human collective life has had to accomplish it somehow.
Living Value Theory distinguishes three domains across which this discernment operates, and the distinction is decisive for everything the rest of this article argues. nonrecursive entities are those that do not respond to being discerned, classified, or engaged: they behave according to regularities that hold regardless of whether, or how, they are understood. selfrecursive entities are those that organise and reorganise themselves according to their own internal processes: they change, but not primarily in response to the discerner. interrecursive entities are those that model the discerner in turn, so that the discerner’s own conduct becomes one input among the many the other entity is itself processing, and neither party fully controls or predicts the resulting exchange. Every coherent tradition of practice , scientific, magical, or religious , has had to decide, across the whole of reality it addresses, which entities belong to which domain, and its entire subsequent architecture of practice follows from that decision.
The second concept required is recursive relevance. Not every recursive process that could, in principle, be discerned is treated as mattering equally by a given tradition. A tradition allocates its attention, its ritual labour, and its institutional resources unevenly across the recursive field it has discerned, elevating some processes to central, load-bearing importance and leaving others as background conditions that receive only occasional acknowledgment. Two traditions can agree, in principle, that ancestors are interrecursive beings and radically disagree about how much of ordinary life should be organised around maintaining that relationship; the disagreement is not about domain but about relevance. This is the comparative unit this article proposes: not what a tradition believes, and not what rituals it performs, and not whether the beings it addresses are personified, but how it has allocated recursive relevance across a discerned field of nonrecursive, selfrecursive, and interrecursive entities, and which of the load-bearing processes within that field it has organised its collective life around.
Most of this discernment, across most of a life, is never noticed. It is carried at the most basic level of coordination, in posture, timing, and unreflective attention, precisely because successful discernment is discernment that has stopped requiring conscious effort. A person does not consciously verify, each time they sit in a chair, that the chair will not respond to being sat on; they do not consciously verify, each time they greet a colleague, that the colleague is a being with intentions rather than a fixture. The overwhelming majority of a tradition’s recursive discernment is settled discernment, inherited rather than performed anew, and this is precisely why the moments at which it is not settled , the moments this article turns to next , are so disproportionately revealing of what a tradition actually is.
Disturbance and the Occasion of Discernment
If recursive discernment ordinarily operates below notice, then something must explain why it is ever noticed at all, and why entire traditions of practice have developed to manage it explicitly. The answer is disturbance: the disruption of settled coordination that occurs when an entity’s recursive status can no longer be taken for granted, or when a process whose ordinary operation was invisible suddenly threatens the conditions of coordination itself. Illness disturbs the settled discernment of one’s own body as a reliably self-organising process. Bereavement disturbs the settled discernment of another as a continuing presence in the coordinated field of a life. The failure of an expected harvest disturbs the settled discernment of land and season as regular and provident. The experience many traditions describe as a crisis of meaning disturbs the settled discernment not of any single entity but of the entire symbolic apparatus through which discernment itself had previously been organised.
Disturbance is not an occasional visitor to an otherwise settled discernment. It is the generative condition under which explicit discernment , discernment that rises to conscious attention, that becomes the subject of practice, prayer, ritual, and doctrine , becomes both necessary and possible. A tradition’s practices are not primarily deployed to celebrate a discernment that was already stable; they are developed, tested, and refined precisely at the points where the ordinary machinery of settled attribution has failed and must be actively restored, renegotiated, or endured. This is why the highest-stakes practices of every tradition considered in this article cluster so consistently around birth, illness, death, and calamity: these are the points at which the recursive status of what matters most , one’s own continuity, another’s presence, the reliability of the world one depends on , is placed under maximum strain, and at which a tradition’s accumulated technology for discernment is called upon to do its most demanding work.
This has an immediate consequence for how load-bearing recursive processes, introduced in Section IV below, come to occupy the central place they do within a given tradition. A load-bearing process becomes load-bearing not because a community decided, in the abstract, that it mattered most, but because it names the specific recursive uncertainty that recurring disturbance has forced into view across generations of a community’s collective life. Sin becomes load-bearing where the disturbance most persistently forced into view is the rupture of a specific, personal, bilateral coordination. Karma becomes load-bearing where the disturbance most persistently forced into view is the diffusion of consequence across an unbounded field of others whose suffering one’s own action continually implicates. A tradition’s central concepts are not arbitrary cultural inventions layered onto an otherwise settled world. They are the accumulated record of exactly where that tradition’s settled discernment has, again and again, needed to be actively restored.
Load-Bearing Recursive Processes
Every religious tradition organises the whole of its collective life around one or more processes that function, within that tradition’s own architecture, as load-bearing in the specific sense already established in this series: remove the process, and the tradition’s entire coordinative apparatus of practice, prohibition, and repair collapses into incoherence. These processes are not doctrines in the propositional sense that belief-centred comparison has always sought. They are not claims to be assented to. They are recursively organising processes: ongoing dynamics that a tradition’s practices exist to track, participate in, or correct.
Sin, in the Abrahamic traditions, is not primarily a proposition about moral failure but the recursively organising process of a bilateral relationship’s rupture, and the entire technology of confession, atonement, and repentance exists to track and repair that specific rupture. Grace is the recursively organising process of a benefit flowing from an autonomous source that cannot be procedurally secured. Karma, in the traditions that elaborate it, is the recursively organising process by which action accumulates consequence across an unbounded field of affected others, largely independent of any single addressee’s personal will. Covenant is the recursively organising process of a bilateral commitment maintained across historical time through memory, law, and renewed obligation on both sides.
Submission, in the Islamic tradition’s central term, names the recursively organising process of aligning human will with a decisive, non-compellable divine will whose full reasoning exceeds human access. Liberation and enlightenment, in their various South and East Asian elaborations, name the recursively organising process of exiting a self-reproducing cycle rather than of satisfying, or being reconciled to, any single addressed being. Ancestor continuity names the recursively organising process by which the dead remain active participants in the coordination of the living. Harmony, in traditions that centre it, names the recursively organising process of maintaining correct proportion among the many recursive relations , familial, cosmic, political , that a well-ordered life must simultaneously sustain.
None of these terms describes a proposition to be believed. Each describes a process to be tracked, and the whole of a tradition’s practical apparatus , its rites, its calendars, its moral instruction, its institutions of authority , can be read as the accumulated technology for tracking, correcting, or completing that specific process. This is the article’s principal analytical tool, and Section IX applies it systematically to eight major traditions.
Orientation, Not Resolution
A comparative framework built from disturbance and load-bearing process must confront an observation that has puzzled the field for as long as it has studied ritual and prayer: these practices very often continue, sometimes with undiminished intensity, in circumstances where they demonstrably do not resolve the disturbance they address. The mourner continues the practice of remembrance long after any hope of the mourned one’s return has passed. The community continues the rain ritual through a drought the ritual does not end. The practitioner continues a daily discipline through a period of suffering the discipline does not relieve. If a tradition’s practices were primarily technologies of resolution, judged by whether they reliably produce the outcome they are aimed at, this persistence would be difficult to explain except as failure of nerve, social pressure, or self-deception.
Recursive discernment supplies a better account. What most religious practice supplies is not resolution of a disturbance but orientation within it: the maintenance of a coherent, practised relation to an entity or process whose recursive status cannot, by the tradition’s own understanding, be fully settled from the human side. Where the entity addressed is interrecursive and irreducibly autonomous , a God who is free to answer or not, an ancestor whose continued attention cannot be guaranteed , no procedure could resolve the relationship once and for all, because resolution would require exactly the kind of unilateral control that autonomy, by definition, excludes. What can be maintained is the discerner’s own side of an open, unresolved coordination: continued address, continued attentiveness, continued readiness to receive whatever response may or may not come. This is why prayer that receives no discernible answer is not, on the tradition’s own terms, a failed technique in the way a spell that fails to produce its promised outcome is a failed technique. It is the correct and expected texture of address to a genuinely autonomous other.
Where the entity addressed belongs instead to a self-reproducing process rather than to any responsive other at all , the Buddhist analysis of suffering’s causal chain is the clearest case , orientation takes a different but structurally parallel form. The practice does not resolve the underlying conditioned process, because that process does not depend on any single addressed will whose favour could in principle be secured. What the practice cultivates is the capacity to remain correctly oriented toward a process that continues regardless, discerning its operation with enough precision that one’s own participation in it can be altered even though the process itself is not brought to a stop by any one act of discernment. In both cases , the autonomous other who may or may not answer, and the self-reproducing process that continues regardless , the practice’s success is not measured by whether the underlying uncertainty disappears. It is measured by whether the practitioner remains correctly oriented within it, across however many disturbances arise, for however long the disturbance persists.
Skill: The Practical Training of Recursive Discernment
If orientation rather than resolution is what most religious practice actually cultivates, then the traditions’ extraordinary investment in disciplined practice , fasting, prescribed prayer at fixed times, prolonged meditation, the rigours of pilgrimage, the ordeals built into rites of passage , becomes intelligible as something more specific than symbolic performance. It is the deliberate training of a skill: the practical capacity to discern recursive status correctly and to sustain the appropriate orientation under conditions, embodied, relational, spatial, material, and symbolic, that make correct discernment difficult.
Fasting across traditions as different as the Hindu vrata, the Jewish Yom Kippur fast, the Christian Lenten discipline, and the Islamic observance of Ramadan is not primarily a symbolic renunciation. It is a practical training regime for maintaining orientation toward what a tradition holds to be interrecursive or selfrecursively decisive, precisely when the body’s own nonrecursive and selfrecursive demands are pressing hardest against it. The five daily prayers of Islam, requiring fixed postures and orientations at fixed times regardless of what else is occurring, train the specific capacity to interrupt ordinary absorbed coordination and re-establish, at regular and non-negotiable intervals, an explicit orientation toward a decisive, non-compellable will. Sustained meditative discipline in Buddhist and Hindu contexts trains the capacity to discern, with increasing precision, which of the states arising in one’s own experience are selfrecursive process and which are misattributed as something more fixed or more other than they are , precisely the discernment error the tradition identifies as the deepest source of suffering.
Rites of passage deserve particular attention within this account, because they train discernment at the points where a person’s own recursive status, and the recursive status of their relations to others, is itself undergoing structural change. Birth, initiation, marriage, and death each reorganise the field of coordination a person inhabits: a new selfrecursive being enters a community’s field of care; a child’s recursive standing is renegotiated into that of an adult, with new authority to discern and new obligations attached to correct discernment; a marriage creates a new interrecursive bond of covenantal weight; death removes a being from the field of the living while, in traditions that attribute continuity to the dead, requiring the living to discern a new form of relation to a presence that is no longer available for ordinary address. The liminal phase that anthropological theory has long identified as central to these rites , the interval in which the initiate belongs fully to neither the old configuration nor the new , is best understood as the deliberately created condition under which new discernment can be trained at all, since the ordinary supports that made the old discernment automatic have been withdrawn and the new discernment has not yet had the chance to settle into the same invisibility.
Dharma, Revelation, and the Traditions’ Own Self-Understanding
The argument advanced in this article is not, on inspection, a claim imposed on religious traditions from outside their own self-understanding. Several of the most sophisticated traditions of religious reflection ever produced have already arrived at essentially this claim, in their own vocabulary, and the comparative study of religion has spent a great deal of scholarly energy failing to see it because it went looking for something else.
The clearest case is dharma. Generations of scholars working on Hindu and Buddhist textual traditions have struggled with a translation problem that has never been fully resolved:
dharma appears to name, in different contexts, both what is the case , the order of things, the way the cosmos is actually structured, the nature proper to a being or a station in life , and what ought to be the case , duty, righteousness, the correct action demanded of a person in a given situation. Paul Hacker, Wilhelm Halbfass, and successive generations of Indological scholarship have documented this apparent double reference at length, and much of the debate has proceeded as though the task were to determine which of the two senses is primary, with the other derivative, or to explain how a single term came, perhaps through historical accident, to carry two meanings that a philosophically careful vocabulary would keep separate.
The recursive discernment framework dissolves this as a puzzle rather than resolving it as a dispute, because it shows that the two apparent senses were never separate meanings needing reconciliation. Dharma, from the root dhr, to hold, sustain, or support, names what sustains the recursive organisation of a situation: what kind of process this is, whether it is the fixed and orderly regularity of ṛta inherited from the earliest Vedic cosmology, the selforganising nature, or svabhava, proper to a particular being or station, or the responsive obligation owed to another within a relationship whose terms are interrecursive. Once this discernment is correctly made, the question of what ought to be done is not a further, separate question requiring its own independent answer. It follows from the discernment itself, in exactly the way that discerning a support beam as load-bearing settles, without any further deliberation, that it must not be removed. This is why the varnashrama-dharma literature can move, without apparent strain, between describing what a person’s station in life is and prescribing what that person must therefore do: the description is doing the entire prescriptive work, because dharma is teaching the reader how to discern the recursive status of their situation correctly, and correct discernment of recursive status is what generates correct action, not a separate premise added to it. Krishna’s instruction to Arjuna on the battlefield of the Bhagavad Gita is not, on this reading, primarily a lesson in duty ethics requiring a subsequent argument about why duty should be performed. It is an extended lesson in recursive discernment: Arjuna has misdiscerned the recursive status of the kinsmen arrayed against him, of his own role as a warrior, and of the entire cosmic process within which the battle occurs, and Krishna’s teaching corrects the discernment, after which the question of what Arjuna ought to do no longer requires separate settlement.
Buddhist dharma, the Buddha’s teaching, admits precisely the same reading, and it clarifies rather than complicates the Hindu case. What the teaching discloses is the correct discernment of dependent origination: which states arising in experience are conditioned, selfrecursive processes rather than a fixed, continuing self; how craving and aversion generate the specific recursive loop by which suffering reproduces itself without requiring any external agent to sustain it; and what follows, by strict entailment rather than by additional moral argument, once this discernment has actually been made rather than merely accepted as a proposition. The Four Noble Truths are not four separate claims, one descriptive and three prescriptive, awkwardly bundled together. They are a single progressive act of recursive discernment , this is the situation, this is what generates it, this is what its cessation would be, this is the path that discernment itself constitutes , in which each further truth becomes visible only once the prior discernment has actually been achieved, not merely asserted.
Once this pattern is seen in dharma, it becomes difficult not to see it everywhere revelation is spoken of. Every tradition organised around a disclosed teaching, a revealed law, or the declared will of the divine is making a structurally identical claim, however different its vocabulary. Torah, from a root meaning to teach or instruct, discloses to its recipients which beings are interrecursive , God, and through God’s ordinance, the neighbour whose claims on one are not optional , which processes are selfrecursive and therefore a matter of the heart’s own ongoing discipline, and which regularities of creation are dependable and nonrecursive, obedient to an ordinance that does not require ongoing negotiation. The commandments that follow are not an arbitrary list of rules layered onto this discernment; they are what correct discernment of these three domains requires once it has actually been made. Sharia, meaning the path or the way to the water source, functions identically: it is not merely a code of prohibitions but a disclosed discernment of how the world is recursively organised under a single, non-compellable divine will, from which the specific requirements of conduct follow as a consequence rather than standing beside the discernment as an independent demand. Halakha, the Jewish legal tradition whose very name means the walking or the going, makes the same point through its own etymology with unusual directness: law, in this tradition, is not a static rule imposed from without but the way one walks once the recursive terrain has been correctly discerned.
The Christian tradition’s language of the Word carries the identical structure at a higher level of theological abstraction. The Word that was in the beginning, that reveals the Father, that becomes the criterion against which every claim to religious knowledge must be measured, is not, on this reading, primarily a body of propositions to be believed. It is the disclosure of exactly the recursive organisation this article has been describing throughout: which being is decisively interrecursive and irreducibly autonomous, what the human’s own selfrecursive condition actually is, and what nonrecursive regularities of creation can be relied upon as fixed rather than negotiated. Even traditions with no personal deity at their centre exhibit the same structure from the other direction. The Daoist teaching of the Way discloses, with unusual explicitness, that the deepest error available to a living being is to treat self-organising, non-compellable natural processes as though they were amenable to forceful intervention, and its central practical counsel, wu wei, non-forcing action, is nothing other than the correct mode of acting that follows once this specific discernment , that these processes are selfrecursive or nonrecursive rather than interrecursive and compellable , has actually been made.
What all of these traditions share, across theological content as different as monotheistic covenant, non-theistic dependent origination, and a cosmology without any personal centre at all, is not a shared belief and not a shared personification of the powers they address. It is a shared structural claim, stated in each tradition’s own idiom: here is how to correctly discern the recursive organisation of the world, and here is what follows, not as an additional demand but as the entailment of having discerned it correctly. The comparative study of religion has read this claim, wherever it appears, as either a metaphysical assertion to be assessed for truth or a moral code to be assessed for its social function, and has missed, in both cases, that the traditions were telling their interpreters exactly what they were doing all along.
Science, Magic and Religion Reconsidered
The classical evolutionary sequence, given its most influential statement by James Frazer, proposed that human thought progresses through three stages: magic, in which the world is treated as governed by sympathetic correspondences that ritual can manipulate; religion, in which those impersonal correspondences are replaced by propitiation of personal wills; and science, in which propitiation is in turn replaced by the discovery of impersonal law. Each stage, on this account, is a more accurate approximation of a single reality that all three are attempting, with increasing success, to describe. Bronisław Malinowski’s later and more sympathetic treatment retained the three-way distinction while rejecting the evolutionary ranking, arguing instead that magic, religion, and science coexist within the same societies because they answer different practical needs , magic where technical control is uncertain, science where it is reliable, religion where existential rather than technical problems are at stake.
It is worth noting explicitly, given the argument of Section I, that Frazer’s own formulation smuggles the personal/impersonal distinction directly into the definition of religion itself, treating the shift from impersonal correspondence to personal propitiation as the very hinge on which magic becomes religion. The framework proposed here rejects that hinge along with the evolutionary sequence it supports. What actually distinguishes magic from religion, on the present account, is not whether the addressed power has been personified. It is a specific and decisive discernment about whether the power, personified or not, can be operationally compelled through the correct performance of a procedure. A tradition can address an entirely impersonal cosmic order , ṛta, the Dao, the impersonal transmission of karmic consequence , with the structural posture of religion, treating that order as something to be correctly discerned and aligned with rather than forced, exactly as a tradition can address an intensely personified household spirit with the structural posture of magic, compelling its cooperation through correctly performed formulae regardless of its wishes in the matter. Marett’s animatism was reaching, without quite the vocabulary to state it, toward exactly this insight: that impersonal mana can be every bit as decisive a religious object as a personified god, and his instinct was correct even though the debate he entered, framed as animism against animatism, kept the question tethered to personhood rather than freeing it into the question of compellability that actually does the comparative work.
Science is the systematic cultivation of discernment directed primarily at nonrecursive processes. Its central operations , measurement, controlled observation, replication, mathematical modelling , are precisely the operations that achieve maximal purchase where the domain under study holds still while being studied, where prediction does not have to contend with a domain that responds to being predicted. Science does not deny that selfrecursive and interrecursive processes exist; it has increasingly extended its methods into psychology, ecology, and the social sciences, where such processes are the explicit object of study. But its founding and still paradigmatic successes , celestial mechanics, chemistry, much of physics and physiology , belong overwhelmingly to the nonrecursive domain, and its characteristic epistemic virtues, repeatability and falsifiability chief among them, are virtues that make sense as ideals precisely because nonrecursive domains permit them.
Magic is best distinguished from religion, then, not by personhood but by a specific discernment about compellability: magic assumes that interrecursive or selfrecursive powers, whatever their degree of personification, can be operationally compelled through the correct performance of a procedure. The efficacy of a spell, on the magical tradition’s own understanding, does not depend on the good will, mood, or free response of the power addressed. It depends on getting the procedure right. This is why magical failure is characteristically explained by procedural error , the wrong words, the wrong timing, an impurity in the performer or the materials , rather than by the addressed power’s refusal, and it is why magical systems, however cosmologically populated with spirits, ancestors, or forces, structurally resemble nonrecursive science in one crucial respect: correct procedure is expected to yield reliable results. Magic, in the vocabulary of this article, operationalises interrecursivity, treating beings or processes that are, by the tradition’s own account, capable of responsive agency as nonetheless compellable through the right technique, in something close to the way a nonrecursive mechanism is compellable through the right input.
Religion, on this account, begins precisely where compulsion ends, whether or not what is addressed has ever been personified in anything like the Western theological sense. The decisive recursive beings or processes addressed by a religious tradition are attributed genuine, irreducible autonomy: they can respond, refuse, delay, or surprise, and no procedure, however correctly performed, guarantees their response; or, where what is addressed is an impersonal self-reproducing process rather than a responsive will at all, no procedure secures its cessation, only correct discernment of it and correct orientation within it. This is the structural difference between a spell and a prayer that has been noted by generations of scholars without quite being given a mechanism, and it is the structural difference, equally, between a spell and the disciplined discernment of a self-reproducing process that Buddhist practice cultivates: a spell is a technique whose efficacy is procedural; a prayer is an address whose efficacy, if any, is the free response of another; a contemplative discipline is a training in discernment whose efficacy is neither procedural compulsion nor another’s free response but the practitioner’s own correctly altered participation in a process that was never going to be compelled by anyone. Grace, in the theological traditions that use the term, names exactly this structural feature: a benefit that could not have been secured by any correct performance and is not owed in return for one. Divine freedom, whatever its specific theological elaboration in a given tradition, is recursive autonomy elevated to its highest and least compellable degree, and it requires no personification whatsoever to be theologically decisive.
This distinction has an immediate consequence that the evolutionary sequence obscures: a
single historical tradition can, and typically does, contain all three orientations simultaneously, addressed to different entities or even to the same entity under different aspects. A tradition may treat weather as nonrecursive for the purposes of agricultural planning, as compellable through the correct sacrifice for the purposes of averting drought, and as the free gift of a deity to be petitioned rather than commanded for the purposes of ultimate providence , three recursive discernments operating side by side without contradiction, because they concern different aspects of what, described at the level of the entity alone, looks like a single undifferentiated phenomenon.
Reclassifying the World’s Religious Traditions
Each of the traditions considered here is examined through the same four questions: what is recursively decisive; what carries recursive authority; how coordination is achieved; and what constitutes successful recursive discernment within that tradition’s own terms. Two further, independent variables sharpen the comparison at several points below: the breadth of the field to which a tradition attributes recursive standing, and the density of recursive relations attributed among the non-human beings or processes themselves, independent of their relations to any human practitioner.
Catholic Christianity organises a densely distributed recursive ecology. What is recursively decisive is not access to a single addressed being alone but participation in a graded network of mediating recursive relations: the Church as an institution that carries recursive authority across historical time, the sacraments as the technology through which grace, an autonomous and non-compellable benefit, is nonetheless reliably made available through the Church’s stewardship, the saints as intermediate interrecursive beings whose intercession supplements direct address to God, and works as the practical demonstration of a life properly coordinated within this network. The breadth of the attributed field is wide , God, Christ, the Spirit, Mary, the saints, angels, and the souls of the dead in purgatory all carry some form of recursive standing , and the inter-relations among these beings, Christ’s mediation with the Father, the saints’ intercession through Christ, are themselves a significant object of theological elaboration. Coordination is achieved distributively: no single point of the network bears the entire weight of the relationship, and successful discernment consists in correctly locating oneself within a graded hierarchy of mediating relations rather than in achieving unmediated access to any one of them.
Protestant Christianity performs a systematic contraction of this same recursive field. What is recursively decisive contracts to a single, direct relation between the individual believer and God, with faith itself , not works, not sacramental participation, not saintly intercession , becoming the principal load-bearing recursive process. The breadth of the attributed field narrows sharply, and the dense inter-relations among intermediate beings that occupied so much of Catholic theological attention correspondingly lose their significance. Institutional mediation contracts as well: the Church retains authority to teach and to gather, but it no longer functions as an intermediate node in the relation itself. Coordination is achieved through direct address, primarily textual, and successful discernment consists in the individual’s own unmediated relation to a decisive divine will, verified internally rather than institutionally.
Eastern Orthodoxy configures the relation differently again, around the concept of theosis: recursive participation rather than either distributed mediation or direct address. What is recursively decisive is a transformative process in which the human participant comes to share, through liturgical and ascetic practice, in the very life of the divine energies without collapsing the distinction between creature and creator. Coordination is achieved through communion, understood as a real recursive participation rather than a symbolic or purely propositional act, and successful discernment consists in the deepening of this participatory relation across a lifetime rather than in either institutional location or a single verified moment of assent.
Judaism organises what is best described as covenantal recursion. What is recursively decisive is a specific historical covenant, maintained bilaterally across generations through law, memory, and community rather than through a single moment of individual assent. Recursive authority is carried jointly by the covenant’s textual record, its interpretive tradition, and the community’s continuous practice of the law that constitutes the covenant’s living side. The breadth of the attributed field is comparatively narrow , one God, engaged in a specific relationship with a specific people, with the inter-relations among other attributed beings receiving relatively little independent theological elaboration in the tradition’s dominant strand. Coordination is achieved through the ongoing, generationally transmitted maintenance of covenantal obligation, and successful discernment consists in the community’s continued faithfulness to a relationship whose terms are historical and specific rather than universal and abstract.
Islam organises what is most accurately described as radical divine interrecursivity combined with the strictest possible refusal of compellability. What is recursively decisive is submission to a divine will disclosed through revelation, a will that remains, by the tradition’s own insistence, absolutely non-compellable by any human procedure whatsoever. Recursive authority is carried by the revealed text and by the transmitted practice, the sunna, that specifies how submission is to be enacted across every domain of life. The breadth of the attributed field is deliberately kept minimal, with a theological insistence on divine unity that forecloses the kind of dense inter-divine recursive elaboration found elsewhere.
Coordination is achieved through obedience rather than through negotiation, intercession, or participatory transformation, and successful discernment consists in aligning human conduct with a will whose reasoning may exceed human access entirely, without thereby treating that will as anything other than free.
Hindu traditions, considered together as a family rather than as a single tradition, present the clearest case of multiple, overlapping recursive architectures operating without any single dominant hierarchy among them. Karma, dharma, the devas, ritual sacrifice, personal devotion to a chosen deity, and the pursuit of moksha each organise a different recursively decisive process, and a given practitioner’s life may be coordinated primarily around one or several of these simultaneously without contradiction. The breadth of the attributed field is exceptionally wide, and the inter-relations among the attributed beings , the devas’ sustained cosmic conflict with the asuras, Krishna’s position as simultaneously a participant in that conflict and its deepest metaphysical ground , receive some of the most extensive elaboration found in any tradition considered here. What carries recursive authority varies correspondingly: textual and ritual tradition for some processes, direct devotional relation to a chosen deity for others, disciplined practice aimed at liberation for others still.
Coordination is therefore plural rather than singular, and successful discernment consists in correctly navigating this plurality according to one’s own stage of life and disposition, rather than in achieving a single kind of relation that the tradition holds up as uniquely decisive.
Buddhism admits the most original reclassification this framework makes possible. What is recursively decisive is not, in the tradition’s own central diagnosis, relation to an ultimate interrecursive deity at all, but the recursive reproduction of suffering itself: saṃsāra as a self-reproducing process, and karma functioning within it largely as an impersonal recursive mechanism rather than as the currency of a bilateral relationship with a personal other. What carries recursive authority is the Buddha’s disclosed discernment of this selfreproducing process and the path specified for interrupting one’s own participation in it. Coordination is achieved not through address, propitiation, submission, or participation in another being’s life, but through disciplined discernment of the causal chain by which suffering continually regenerates itself, and successful discernment consists precisely in seeing through, and thereby interrupting, a recursively self-reproducing pattern rather than in securing any particular outcome from any addressed being. The breadth of the field the tradition attributes recursive standing to is, in its Mahayana elaborations, extremely wide , an enormous plurality of buddhas and bodhisattvas distributed across multiple realms , but even this wide field is subordinate to the tradition’s central diagnosis, which requires no addressed other at all in order to be correctly discerned and acted upon.
Indigenous and animist traditions, treated together while acknowledging the real differences among them, present the most thoroughly distributed recursive world of any category considered here, and the widest attributed breadth of any tradition in this survey. What is recursively decisive is not concentrated in one or a small number of addressed beings but dispersed across land, animals, ancestors, mountains, rivers, and weather, each attributed its own recursive standing within a shared and densely populated field, in which the inter-relations among non-human beings, a peccary community’s own internal leadership and its capacity for organised response to human conduct, for instance, are as practically consequential as any human-to-nonhuman relation. Recursive authority is correspondingly dispersed, carried by specific custodial relationships, ritual specialists, and inherited protocols specific to particular places and particular non-human partners rather than by a single textual or institutional centre. Coordination is achieved through the maintenance of a very large number of specific, often highly localised relational protocols rather than through a single overarching relation, and successful discernment consists in correctly reading the recursive standing of the specific entities that dwelling itself continuously presents, which is why these traditions are so tightly bound to the mediation of dwelling in a way none of the traditions considered above are bound to it in the same degree.
Recursive Relevance
The concept of recursive relevance, introduced in Section II, does the comparative work that belief and ritual, and the personal/impersonal binary examined in Section I, cannot. Two traditions may agree, at the level of bare domain classification, that ancestors are interrecursive beings, that a mountain has some form of recursive standing, or that a text carries authoritative weight, and yet organise dramatically different proportions of collective life around these agreements. What varies across traditions, and what a comparative framework built from recursive relevance can track with real precision, is exactly this differential allocation: how much recursive weight is placed on the moon’s phases, on the continued presence of ancestors, on the sanctity of a particular temple or landscape, on the authority of a specific text, on the individual conscience as a site of discernment, on the maintenance of a religious community as such, and on the body as itself a site of recursive participation.
A tradition that allocates heavy recursive relevance to sacred landscape will organise pilgrimage, land tenure, and ritual calendar around that allocation in ways a tradition that allocates minimal relevance to landscape, and correspondingly heavy relevance to text, will not. A tradition that allocates heavy recursive relevance to ancestor continuity will organise funerary practice, inheritance, and ongoing household ritual quite differently from a tradition that has, whether through theological decision or historical contingency, allocated minimal relevance to the continued participation of the dead. This is not a difference in what is believed about the reality of ancestors, landscapes, or texts, considered as isolated propositions. It is a difference in how much of collective coordinative life has been built around maintaining a relation with each, and this difference explains observable religious diversity with a precision that belief-centred comparison, forced to ask only whether a given content is affirmed or denied, has never achieved.
Beyond Belief: Symbolic Capture and the Limits of Doctrine
Belief has functioned, across more than a century of comparative scholarship, as the default unit of religious content: the thing a tradition is ultimately trying to get its adherents to hold as true. The critique offered in Section I is not that belief is unimportant to the traditions that emphasise it, but that belief, considered as a comparative category, is a symbolic description operating at a specific and relatively late stage of a much larger coordinative process , the stabilised, articulable formulation of a recursive relation that has already been organised, practised, and lived at more basic levels before it is ever stated as a proposition.
There is a specific and recurring historical process by which belief comes to occupy the outsized place it does within a tradition’s own self-presentation, and naming it clarifies why belief-centred comparison, however mistaken as a universal method, is not simply an outside imposition without any basis in traditions’ own history. Every tradition begins in practice: specific bodies, specific relationships, specific places, specific materials, and specific symbolic forms, jointly engaged in maintaining a discerned relation to a decisive recursive process. As traditions develop, the symbolic dimension of this practice characteristically elaborates into a substantial reflective apparatus , commentary, doctrine, systematic theology , that is itself a genuine achievement, refining and transmitting what the practice discerns with an increasing precision the practice alone could not have generated. The difficulty arises when this reflective apparatus comes to treat itself as the primary site of the tradition’s truth, and the underlying practice as a merely illustrative or derivative expression of doctrines that could, in principle, be fully stated without it. At that point, a specific and diagnosable inversion has occurred: the symbolic articulation that was a tool for discerning and transmitting the practice’s central recursive relation now claims to define, on its own authority, what that relation is and requires, and can come to treat the embodied, relational, spatial, and material dimensions of the practice as secondary illustrations of a truth the doctrine alone properly holds. Belief-centred theorising about religion is best understood as a comparative method that has, in effect, generalised this specific historical inversion, taking one moment in one tradition’s own internal development and mistaking it for the basic structure of religious life as such.
Recursive discernment operates across the full range of the five mediations that structure human coordination: embodiment, being-with, dwelling, multimateriality, and
multisymbolism. A tradition’s discernment of the recursive organisation of reality is enacted in bodily posture and gesture before it is ever stated as doctrine; it is enacted in the maintenance of relationships with other people and with non-human others; it is enacted in the specific configuration of sacred and ordinary space that a tradition’s dwelling arrangements assume; it is enacted in the treatment of objects, relics, and materials as more or less recursively charged; and only at the furthest and most abstracted reach of this whole architecture does it also, in some traditions more than others, crystallise into explicit propositional formulation. A comparative method that begins and ends with belief has selected, from this whole architecture, the single mediation in which one specific historical tradition happens to have invested the most elaborated symbolic effort, and has then measured every other tradition’s entire coordinative life by how much of that one mediation it, too, has produced. Living traditions organise recursive coordination across bodies, relationships, places, and materials; they do not merely assert truths, and a framework capable of comparing them must be built to see the whole architecture rather than only its most textually legible layer.
Comparative Religion after Recursive Discernment
Recursive discernment supplies what a century of proposed comparative units has failed to supply, and it is worth stating the gains together rather than distributing them across separate sections. It supplies a genuinely universal comparative principle, because every tradition, whatever its specific historical content, must discern the recursive organisation of the reality it inhabits, and no tradition is exempted from this task by virtue of lacking belief, ritual, myth, or institution in the specific form some other tradition happens to have elaborated, and no tradition is exempted from it by virtue of addressing powers that a Western theological vocabulary would call impersonal. It supplies a new and structural distinction between science, magic, and religion, grounded not in relative rationality, evolutionary stage, or personhood, but in which domain is primarily discerned and whether the recursive agency or process addressed is treated as compellable or free. It supplies a new basis for classifying religious traditions internally, through load-bearing recursive processes, attributed breadth and inter-relational density, and the allocation of recursive relevance, that does not require measuring every tradition against the specific profile of any one of them. It replaces a belief ontology with a process ontology: traditions are no longer compared as containers of more or less correct propositional content but as organisations of ongoing recursive coordination. And it grounds an anthropology of religion centred on coordination rather than doctrine, capable of taking a practitioner’s own account of a disturbance, a discerned autonomy, or a self-reproducing cycle of suffering seriously as a structural claim about the world’s recursive organisation, rather than reducing that account in advance to something else , psychology, sociology, or symbol , before the comparison has even begun.
Return, then, to the opening question. The comparative study of religion has searched, across more than a century, for a common object that religious traditions share: a content, a practice, a form, or a type of addressed being, that could be measured across all of them on a single scale. Recursive discernment shows that the search was misdirected from the outset. The commonality does not lie in a shared object. It lies in a shared task, discharged differently by every tradition that has ever undertaken it, and the differences between traditions are not differences of quantity along one axis but differences of structure across several: which domain is discerned as decisive, whether the recursive agency addressed is compellable or free, which processes are treated as load-bearing, how recursive relevance is allocated across the whole architecture of embodied, relational, spatial, material, and symbolic life, and , as Section VII has shown many of the traditions themselves already understood , what follows, not as a separate further demand but as the entailment of having discerned correctly.
Conclusion
This article has not proposed another definition of religion to add to the long list that has failed to hold. It has proposed a different foundation for comparison altogether, one that does not require choosing which historical tradition’s self-description becomes the yardstick for every other, and one that does not require settling, before comparison can even begin, whether the powers a tradition addresses are persons or forces. Science, magic, and religion are three different recursive discernments of reality, distinguished by which domain each takes as primary and by whether the recursive agency or process each addresses is treated as compellable through correct procedure or as irreducibly free. The world’s religious traditions, in turn, are best understood not as different amounts of belief, ritual, or institutional elaboration but as different organisations of recursive relevance, different configurations of recursive mediation, and different load-bearing recursive processes: sin and grace organising a bilateral relationship’s rupture and repair, karma and dharma organising an unbounded field of consequence and a correctly discerned order of action, submission organising alignment with a non-compellable will, saṃsāra and its interruption organising a self-reproducing cycle that requires no addressed other at all, ancestor continuity and land organising a distributed field with no single centre.
The disturbances that force this discernment into view , illness, bereavement, calamity, the collapse of an inherited symbolic world , are not occasions that religion addresses from outside, as medicine or law address the disturbances proper to their own domains. They are the generative ground on which every tradition’s load-bearing processes were first forced into explicit visibility, and the practices that traditions have refined across generations are best understood as the accumulated skill for maintaining correct orientation within a recursive relation that disturbance has placed, however briefly or however permanently, back into question. And several of the most sophisticated traditions of religious reflection ever produced have said something very close to this about themselves, in their own vocabulary, long before any comparative theorist arrived to ask what they were doing: dharma discloses how to discern correctly and, in that disclosure, already supplies what follows; revelation, teaching, and the disclosed will of the divine, across traditions that otherwise share almost nothing else, name the same structural act under different names.
The comparative study of religion has long searched for a universal religious substance, certain that if the search were only conducted carefully enough, the substance would eventually be found. What this article suggests is that no such substance exists, and that its absence is not a failure of the field but a discovery about its subject. What is universal is not belief, ritual, doctrine, or personhood, held or possessed in different amounts by different traditions, but the living necessity, faced by every coordinating community that has ever existed, of discerning the recursive organisation of the world it inhabits and coordinating its collective life accordingly. Recursive discernment, understood this way, is not one more candidate to add to the list of failed comparative units. It is the missing account of why every previous candidate, however carefully elaborated, was always going to fail: each mistook one tradition’s particular solution, or one tradition’s particular way of asking the question, for the shared problem that every tradition, including that one, was actually trying to solve.
References
Asad, Talal. Genealogies of Religion: Discipline and Reasons of Power in Christianity and Islam. Baltimore: Johns Hopkins University Press, 1993.
Durkheim, Émile. The Elementary Forms of Religious Life. Translated by Karen E. Fields. New York: Free Press, 1995 [1912].
Eliade, Mircea. The Sacred and the Profane: The Nature of Religion. Translated by Willard R. Trask. New York: Harcourt, Brace, 1959 [1957].
Frazer, James George. The Golden Bough: A Study in Magic and Religion. 3rd ed. London: Macmillan, 1911–1915.
Geertz, Clifford. “Religion as a Cultural System.” In The Interpretation of Cultures. New York: Basic Books, 1973.
Hacker, Paul. “Dharma in Hinduism.” Translated and reprinted in Journal of Indian Philosophy 34, no. 5 (2006 [1965]): 479–496.
Halbfass, Wilhelm. India and Europe: An Essay in Understanding. Albany: State University of New York Press, 1988.
Malinowski, Bronisław. Magic, Science and Religion, and Other Essays. Garden City, NY: Doubleday, 1954 [1925].
Marett, Robert Ranulph. The Threshold of Religion. London: Methuen, 1909.
Otto, Rudolf. The Idea of the Holy. Translated by John W. Harvey. Oxford: Oxford University Press, 1923 [1917].
Descola, Philippe. Beyond Nature and Culture. Translated by Janet Lloyd. Chicago:
University of Chicago Press, 2013 [2005].
Tylor, Edward Burnett. Primitive Culture: Researches into the Development of Mythology, Philosophy, Religion, Language, Art, and Custom. London: John Murray, 1871.