RR Marett’s “Pre-Animistic Religion”

Over Spring Break, I went diving in the Caribbean dove into the often neglected work of Robert Ranulph Marett (1866-1943), yet another classicist (like James George Frazer) who ditched literature for a new discipline. Marett, though little known today, eventually succeeded Edward Burnett Tylor as the Chair of Anthropology at Oxford. Marett’s primary interest was religion and he wrote a great deal about the subject. I had assumed that my dive would be brief, more like snorkeling. Much to my surprise, this wasn’t the case. It was deep and deeply satisfying.

Although I had previously read Marett’s famous 1900 paper titled “Pre-Animistic Religion” (which you can access here – it is the first chapter), I apparently was not prepared or equipped – on the first reading – to understand what he was saying and doing. I’ve now read the paper four times. Suffice it to say that I now consider “Pre-Animistic Religion” to be of fundamental importance to evolutionary religious studies. It is seminal and should be read by everyone who claims or suspects that cognition is the key to religion. It should also be read by everyone who thinks, wrongly, that Tylor’s minimum definition of “religion” (as belief in spirits) is either sufficient or workable. What follows is my re-statement and analysis of Marett’s claims, along with some musings about his place in anthropological history.


In 1900, when James George Frazer published his expanded second edition of the Bough, he was joining a long line of cultural evolutionary theorizing that had begun with Lubbock, continued with Tylor, and been pursued with a perversely detailed and speculative vengeance by Herbert Spencer and Lewis Henry Morgan. While neither Spencer nor Morgan fixated on religion or considered it central (as did Lubbock, Tylor, and Frazer), the cumulative effect of these schemes was negative. They were too positivist, rationalist, and normative. In 1896, Franz Boas published the first of his many well-taken broadsides against evolutionary theorizing and speculation. As the standard history of anthropology is usually told, this Boasian backlash led, in relatively short order, to a focus on fieldwork and particularism. While this standard history may not amount to myth, it comes perilously close. Evolutionary theorizing, initially in revised and later in transformed guises, did not simply come to an end. It not only continued – it intensified. As had been the case from the beginning (of anthropology as a discipline), religion was central to the ensuing debates. The essential bridge between the old and new forms of evolutionary theorizing was Robert Ranulph Marett (1866-1943), Tylor’s successor at Oxford.

In 1900, the relatively little-known Marett published an article titled “Pre-Animistic Religion” in the journal Folklore. It was based on lecture he had given, under unusual circumstances, the previous year. Marett had been asked to speak frankly and originally about the then-prevailing evolutionary schemes of religion. Working under a tight deadline and with limited sources (circumstances which ironically may explain his brilliant flash of insight), Marett managed to produce something that simultaneously ended an entire era of evolutionary theorizing about religion, inaugurated a new era that paraded under a non-evolutionary banner, and instigated (many decades later) a new approach to animism that would bear enormous fruit. In doing all this, Marett also delivered a foundational critique to all purely cognitive approaches to religion. Given Marett’s vastly underestimated importance, I will consider “Pre-Animistic Religion” in detail. It is, in my estimation, a complex and subtle piece that holds several keys to understanding the history and future of scholarship in evolutionary religious studies.

Marett begins by noting that Tylor’s minimum definition of religion – belief in “spirits” (a category that includes souls) – does not begin the capture the complexity of religious experience. His empathetic reading of the ethnographic record leads him to believe there is something else that is sensed or felt “religiously” before people form concrete ideas about spirits and express those ideas through language. By the time people actually get to this point (in their thinking), they have already done a great deal of emotional and cognitive work. Spirit beliefs, in other words, are relative late-comers both to individuals who have “religious” experiences and to societies that tell stories about those experiences. There is a deeper substratum, something earlier in individual experience or cultural time, that causes people instinctively (or subconsciously) to sense or feel what he calls “religion.” Accordingly, Marett rejects Tylor’s minimum definition of religion as being too restricted. In its place, he offers a definition that on its face appears rather inchoate and perhaps unpromising: “[Religion] stands for a certain composite or concrete state of mind wherein various emotions and ideas are together directly provocative of action” (164). It is important to note that Marett still defines religion as a state of mind (i.e., cognition), but this state of mind includes emotions first and ideas second.

Marett identifies these “emotions” as fear, admiration, wonder, and most importantly, awe. Other than fear, these are not emotions in a strict scientific or modern sense. They are instead what neuroscientific authorities on evolutionary emotions would call “feelings.” These are not simple or basic emotions – such as anger, disgust, fear, and desire – that govern arousal, approach, and avoidance. Feelings are more complex – they arise only after basic emotions have been felt and processed at higher levels into language. They are, in other words, a combination of emotions and ideas – just as Marett has suggested. But regardless of what we call them or how we categorize them, nearly everyone has these experiences and they are often difficult to describe with language. We sometimes struggle to find words which can re-create the feeling or convey the experience.

It is important to recognize that Marett is not suggesting these emotions or feelings are in fact “religious” or that they are sui generis, ineffable, or numinous. Rather, they are common experiences that nearly everyone has, even if the experience itself is uncommon. Marett uses a variety of words, most drawn from ethnography, to convey this amorphous category of sensed – rather than thought – experience: wonderful, surprising, uncanny, strange, great, unexpected, extraordinary, awful, marvelous, incomprehensible, unknown, powerful, startling, arresting, shocking, and unusual. He places greatest emphasis on awe and awful, the latter indicating that which causes “awe” and is not anything negative. As should be evident, this is not the realm of ideas, cognition, inference, logic, or expression. It is rather a feeling or sense that there is a force, energy, or power in and about the world. In other ethnographic contexts, something similar is rendered as potency.

It is critically important (for evolutionary religious studies) to understand that this feeling or sense need not be connected in any way to souls, spirits, or invisible agents (though this can certainly happen, which is something that Marett acknowledges). Nor is it always sensed or conceived in anthropomorphic terms. While it can manifest in spirit beliefs, invisible agents, or in anthropomorphic terms, it is not limited to these. It would be more accurate to say that it subserves and surrounds these ideas.

Stated another way, this is a feeling or sense (which Marett renders, a bit confusingly, as “emotion”), that there is a force, energy, power, or potency that causes things to happen, makes the world cohere, binds things together, and breaks them apart. Whatever “this” or “it” is, it often operates as expected, in ways that are routine or normal. Ordinary day-to-day events or things are not associated with it. But it can also operate not as expected, in ways that are non-routine or abnormal. It is the latter which gives rise to the “emotional” feeling or sense that there is a non-agentive or non-anthropomorphic something, vaguely conceived and hard to express, that undergirds and surrounds everything. The parallels to modern cosmology and physics should be apparent. There is something that brings order and disorder. Something causes things to happen, often unexpectedly, for better or worse. There is a vast tissue of causation that cannot be fully understood; this tissue or matrix is too complex for ordinary understanding or straightforward logic. This is not a failure to understand causation (though it can be this) so much as it is a recognition that causation can be uncertain, strange, eccentric, incredible, bizarre, mysterious, paradoxical, and serendipitous. None of this causation is seen, heard, touched, or smelled – but it is sensed and felt. This sensing and feeling, it should be noted, occurs where the physical meets the seemingly non-physical: in our minds.

Marett groups these all ideas, concepts, and descriptions together under an unfortunate label: the “Supernatural.” It is an unfortunate choice because we, as materialist or empiricist or positivist heirs of the Enlightenment, associate this word with a completely different set of ideas. We operate with the idea that the “natural” world is the real world of substance and material that can be empirically sensed and instrumentally recorded. It is governed by laws. Everything else is an imaginary or metaphysical world, lacking reality or groundedness. We associate the “supernatural” with the non-verifiable and non-falsifiable – it is the imaginary or visionary realm of religion and the paranormal. In our understanding, the “supernatural” violates known physical laws and carries negative connotations. This is not, however, what Marett means by the term and it is not how the difficult idea is expressed by other cultures.

What Marett means by “supernatural” goes by different names: among the Lakota it is called wakan and among the Bushmen it is n/um. The hunter-gatherer ethnohistoric record is replete with similar references to this pervasive force, power, energy, or potency: manitou for the Algonquians, paha for the Shoshoneans, maxpe for the Crow, orenda for the Iroquois, and kalou for the Fijians. It has various names among the Aborigines and is woven deeply into the warp and woof of Dreaming. Whatever this force, energy, or potency is called, it has a dual aspect: in its positive guise – bringing benefits, luck, or wellbeing, Marett borrows the Melanesian term and calls it “mana.”[i] In its negative guise – causing harm, misfortune, or ill-health, it is taboo. Most rituals accentuate the positive and thus might be said to seek or ensure mana; other rituals are designed to ward off the negative and often entail taboos. The corollary of this dual aspect is that the “supernatural” is neither inherently good nor bad. In its base-state or form, it is neutral.

Marett proposes to unite all these ideas under the banner of a new word, “animatism,”  that has some relationship to Tylor’s animism. Animatism, for Marett, is the sense of supernaturalism that people have before they form or express animist spirit ideas. It is not entirely clear whether this sense is something that arises in people before they even conceive of spirits, or whether “pre-animistic religion” is something more like an evolutionary stage – a reference to something more primordial in human prehistory. These are not mutually exclusive ideas, and it may be that he is speaking of both. It is clear that Marett discards the standard evolutionary paradigm along with much of its normative baggage. There is no progression through stages, and no suggestion that he is tracing mental, intellectual, or cultural evolution. The triumphal march toward rationalism and science is gone. There is nothing infantile or childlike about the minds that Marett discusses. They are fully formed and capable.

Marett does not, however, reject evolution or evolutionary paradigms. He uses the word twice, and both times indicates that his entire discussion bears on the issue of “religious evolution.” Though he never mentions stages, epochs, or types that are typically used in these contexts, Marett uses an alternative device – the “savage” or “primitive” – which both implies and suggests either a stage or a type. Though Marett insists that “savage” means nothing more than technologically simple or small-scale, and implies that he has abandoned the comparative method which considers “savages” to be living fossils or static representatives of prehistory, he focuses on them exclusively to illustrate the “supernatural” at work and the principles of animatism. In doing so, he makes any such distinction difficult to maintain. At some point, the “best proxy” or “closest analogue” caution is forgotten, and the category merges (or even collapses) into the idea that “savages” are in fact exemplars of the past. An incessant focus on the “savage” and “primitive” gives rise not only to the idea that these societies are the key to unlocking the evolutionary past, but also to the idea – which will dominate sociology and anthropology for decades to come – that studying such societies will reveal the inner workings (or “structures” and “functions”) applicable to all societies. Both ideas will have baleful effects and seep so deeply into anthropological and popular discourse that all efforts to disrupt them, including my own, will prove extremely difficult.

While Marett’s 1900 article brought him fame and stands as his most important contribution, he was just getting started. In 1909 Marett published The Threshold of Religion (open e-book), a collection of articles published during the previous decade. It is not, however, simply a re-print of previous work – Marett begins each chapter (the first of which is “Pre-Animistic Religion”) with a summary of the article and additional commentary that reveals the common thread running through each. Marett was, in other words, constructing a complete theory of religion and building on ideas that had only been synoptically treated in “Pre-Animistic Religion.” In the preface, Marett claims – in response to apparent confusion – that “pre-animism” or animatism was neither a chronological concept nor an evolutionary era, though some had interpreted it that way. He did not think that the “genetic” or origins problem could ever be solved: “the first chapter of the history of religion remains in large part indecipherable” (1914:viii). His intent all along had been to paint a broader psychological and emotional portrait of religious experience. He repeatedly criticizes previous evolutionary schemes (and singles out Frazer for special abuse) for being too rigid, systematic, rationalist, and “intellectualist.” In his estimation, the emphasis had been misplaced – religious concepts come after (and are usually subordinate to) religious experiences – thus, Tylor’s minimum definition of “religion” as spirit beliefs is far too restricted. He then situates his ideas within the history of the field:

[W]hen I began to write, certain representative theories dominated the entire field of Comparative Religion, and had to be forcibly induced to relax their claims before a “place in the sun” could be found for a new interpretation. I need not here refer to those theories specifically, but may describe them generally as in my judgment too intellectualistic, too prone to identify religion with this or that doctrine or system of ideas. My own view is that savage religion is something not so much thought out as danced out; that, in other words, it develops under conditions, psychological and sociological, which favour emotional and motor processes, where ideation remains relatively in abeyance (xxxi).

Here, for perhaps the first time, we find an evolutionary theorist acknowledging that religion is not merely an individual phenomena – it is also a collective or social one. Marett emphatically asserts: “I hold that religion in its psychological aspect is, fundamentally, a mode of social behavior” (ix, emphasis added). So not only has Marett provided a necessary corrective to previous evolutionary theorizing and critiqued standard cognitive accounts, he has also recognized that individuals do not generate religion on their own. This recognition would sweep the field a few years after Marett wrote these words, with the publication of Emile Durkheim’s Elementary Forms of the Religious Life (1912).

[i] It should be apparent that Marett’s concept of the “supernatural” goes far beyond western or monotheist ideas about “holy” or “sacred” things. While holy and sacred may sometimes capture elements of Marett’s “supernatural,” these terms are loaded with millenia of Christian baggage. They are best avoided.

Did you like this? Share it:

Leave a Reply