Varieties of Tulpa Experiences: Sentient Imaginary Friends, Embodied Joint Attention, and Hypnotic Sociality in a Wired World

“The intention to know”, from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.
“The intention to know”, from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.


This article presents a summary and discussion of key findings from ten months of experimental cyberethnography among tulpamancers.[i] Tulpas, a term reportedly borrowed from Tibetan Buddhism, are imaginary companions who are said to have achieved full sentience after being conjured through ‘thought-form’ meditative practice. Human ‘hosts’, or tulpamancers, mediate their practice through open-ended how-to guides and discussion forums on the Internet and experience their Tulpas as semi-permanent auditory and somatic hallucinations.

Studying Tulpas and their hosts is fascinating on many counts, not least because it provides an opportunity to observe an emerging culture and the mediation of new kinds of persons – in this case, that of multiple humanoid and non-human persons ‘hosted’ in single bodies and a large-scale sociocultural matrix of ‘healing’ generated without physical interaction between members. As an anthropologist who underwent retraining in cognitive science, however, I am less concerned with the seemingly ‘strange’ and ‘exotic’ aspects of Tulpamancy and am most interested in what the practice can reveal about fundamentally human mechanisms and processes. Thus, I seek to investigate (but in no pre-determined order) how neurocognitive, attentional, and narrative processes invariably shape all forms of sociality and experiences of personhood on the one hand, but also how social, political, and technological processes invariably shape mechanisms of attention, cognition, and perception. I gravitate toward sociocognitive, enactive models of hypnosis as ways of mediating sociality and personhood.

My investigation is grounded in the study of interactions between environment, cognition, and culture. In this model, mind is understood as embedded, embodied (Kirmayer, 1992a; Csordas & Masquelier, 1997), enactive (Varela, Thompson & Rosch, 1991), and extended (Clark & Chalmers, 1998) to an organism’s whole interactive environment. Just like, as Evan Thompson elegantly puts it, the flight of a bird isn’t an intrinsic property of its wings but exists as a relation between the organism and its whole environment, thinking isn’t ‘inside’ the brain, but distributed in a broader ecology of interacting sense modalities and environmental matrices (Thompson, 2015; Bateson, 1972; 1980). Here, I opt for a working definition of ‘culture’ borrowed from the natural sciences: when clusters of individuals within a similar species engage in cumulative social learning and develop relatively stable ways of doing things that differ from the ways of other groups, we can speak of culture (see Tomasello, 2009). Here, I add sensory experiences, modes of affect, joint attention, and hypnosis to the set of cumulative, iterative, differentiated phenomena that arise through social learning and give us forms of life we call ‘culture’.

Before presenting aspects of Tulpamancy practice in greater detail, I begin with an old question: how can highly similar sets of embodied mental representations, experiences, and behaviours come to be shared by large groups of individuals who never interact physically with one another? Are socialities mediated online paradigmatically different from ‘physical’ ones, or is a fundamentally similar process at stake?

“Sudden fright”, from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.
“Sudden fright”, from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.

1. The language of invisibility and the invisibility of language

“Sometimes people get logically conscience-stricken […] and like to have some criteria of ‘real’ things, e.g. entities occupying space, and will then say things like ‘boundaries are imaginary lines’. They seem to think that countries occupying territory are real but the lines separating them are somehow imaginary.”

Ernest Gellner, Language and Solitude, 1998, p54

“No one, wise Kubla, knows better than you that the city must never be confused with the words that describe it.”

Marco Polo, addressing the Great Kubla Khan

“Memory’s images, once they are fixed in words, are erased,” Polo said, “Perhaps I am afraid of losing Venice all at once, if I speak of it. Or perhaps, speaking of other cities, I have already lost it, little by little.”

Italo Calvino, Invisible Cities, Harvest books, 1974

“I wonder if the Internet is like a city”, Ian Gold told me one morning over our third round of double-espressos. Ian is a philosopher of psychiatry who is investigating why certain migrant and minority groups living in cities experience higher rates of psychosis than they do in their home communities (Gold & Gold, 2014). Discrimination, adversity, stigmatization, and living in fragmented polities are increasingly understood as important causal variables in the mediation of mental illness (Heinz, Deserno, & Reinighaus, 2013), but the question of how such differentiated trends become distributed and experienced with such violent stability and precision remains open. Large cities and their polities, after all, like ‘societies’, are difficult entities to handle physically and cognitively. “What kind of imagined community is a city”, Ian went on, “when most people’s daily routines are limited to bounded spheres like home and work, or impersonal interaction with strangers and a few shop owners?”

This is an old question: how can societies be understood, ‘internalized’, or embodied – how can societies hold – when the vast majority of the people, ideas, and infrastructure that make up these totalities are invisible to individual members? One might as well propose that, given the non-physicality of interaction between members, it is cities and societies that are like the Internet. Invisibility and physical in-interaction, thus, are important pieces in this puzzle.

For Erving Goffman, who championed studies of face-to-face interaction in modern societies, the “anonymized”, “surface character” of life in cities is routinized through what he called “civic inattention” (Goffman, 1971, p385) – that is, through the many ways in which strangers avert their gazes, avoid conversations or physical contact, and reinforce private boundaries in the public sphere. Loneliness and invisibility, as Goffman saw it, are logical outcomes of civic inattention as a “mode of personal territoriality” (ibid, p359). As a theoretician of sociality, I am particularly interested in how different regimes of joint-attention mediate lived experiences of personhood with distinct sensory, somatic, ‘embodied’ qualities. Civil inattention for example, is a specific regime of attention, but it is certainly not an absence of attention. In Goffman’s Invisible City, attentional resources are being mobilized to not pay attention to certain features of the world – particularly people caught in a symbolically-marked game of allegiances. Those that come to feel most generally inattended-to, thus, will come to embody their invisibility in physically unbearable ways. This is a terrible problem, but the general question remains: given the infinitesimally narrow possibilities of horizontal interaction between members of any given polity, how can it come to hold at all with such violently predictable experiential quality? What is the minimal physical requirement for any scheme of sociality – for any imagined community – to be embodied? What is the maximal spatial and cognitive capacity for joint-attention – usually understood as being limited to dyadic, or spatially-bounded, interaction between two or a few more actors? Could it be that Calvino got the City wrong in his anti-representationalist fable? Isn’t it, rather, that language does not so much fail to capture the city, but instead brings it into being?

Steven Levinson at the Max Planck Institute for Psycholinguistics has taken this hypothesis seriously, and has led a series of elegant experiments to revise our current understanding of linguistic relativity, first proposed by Benjamin Whorf in the early 20th century and subsequently dismissed by most social and cognitive scientists. In an experimental study of the Senses in Language and Culture, Levinson and colleagues attempted to correlate the richness and diversity of sensory experiences across cultures with the grammatical categories and specific terms attributed to the sensorium in different languages. They found that speakers of languages (like American English) that lack gradient olfactory terms, for example, performed very poorly at identifying common scents from their environment (like cinnamon) when presented with scratch-and-sniff cards. The Jahai of the Malay Peninsula, conversely, possess a very rich olfactory vocabulary and could identify an equal amount of smells and shapes.

Could it be, then, that immersion in new narrative practices with terms like ‘tulpa-forcing’, ‘possession’, or ‘wonderland’ spreading through internet technology is a sufficient condition for the mediation of new ways of experiencing touch, voice, pleasure, and synaesthesia, to name but a few of the ‘senses’ mobilized by Tulpamancy?

 “Sympathy and love for all” , from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.
“Sympathy and love for all”, from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.

2. Varieties of Tulpa Experiences Origins

The term Tulpa began circulating in the West in 1929 following the publication of Magic and Mystery in Tibet [Mystiques et magiciens du Tibet] by the Belgian-French explorer Alexandra David-Néel. The author, who reported observing the practice in Tibet, claimed to have created a Tulpa of her own in the image of Friar Tuck. Often fully transcribed as sprul pa’i sku from the Tibetan སྤྲུལ་པ, the term can be translated as ‘emanation’ or ‘incarnation’, and is associated with the physical body (Nirmanakaya). A Tulpa, as presently understood in the tulpamancer community, is a sentient being who becomes incarnate, or embodied through thought-form.

Tulpas and the senses

Drawn from primarily urban, middle class, Euro-American adolescent and young adult demographics, most Tulpamancers cite loneliness and social anxiety as an incentive to pick up the practice and report overwhelmingly positive changes in their individual and offline social lives, in addition to an array of new, ‘unusual’, but largely positive sensory experiences. These include (in order of frequency) auditory, tactile, visual, and olfactory sensations. “Raw thought”, “intuitive thinking”, “speaking with no words” and “communicating with images, feelings and music” are also reported along with other non-verbal, non-narrative forms of interaction. One informant, for example, a Caucasian-American young woman majoring in Cognitive Science at Midwestern University, reports being underdressed and cold as she was walking to class one morning. She explains that upon sensing that her host was cold, the Tulpa took off his [Tulpa] coat to place it on her [the host’s] shoulders, producing a feeling of warmth and the distinct sensation that she was wearing another layer of clothing. Such reports of spontaneous help from Tulpas in social, environmental, and professional situations abound and, indeed, seem to characterize the practice.

Sexual and romantic interactions are controversial topics in the community, with a growing consensus tending to converge toward a taboo on the latter. Because Tulpas are imagined, experienced, interacted with, and collectively validated as sentient persons with mental states, propositional attitudes, feelings, bodily sensations, biases, and preferences of their own, the issue of mutual consent is deemed crucial. Creating a Tulpa for one’s selfish enjoyment, as such, is understood to be just as unethical as seeking a one-sided, power-imbalanced relationship of any kind. General possibilities of tactile and multi-sensory experiences inherent in the practice, however, indicate that the ‘taboo’ was put in place to establish norms around a common, or, at the very least, possible practice.

In addition to imagined agents, tulpamancers’ mental constructs include spaces for Tulpa-host interaction usually termed “mindscape” or “wonderland”. Tulpas often assume human form, but many are imagined within a continuum of humanoid variations with gender-fluid, gender-neutral, or pan-ethnic traits. Fandom culture drawn from fantasy-oriented genres also frequently prompts the forcing of non-human Tulpas such as elves, dragons, or ‘imaginary creatures’. A sizeable, but non-majoritarian section of the community seems to have emerged from Internet forums dedicated to Bronies – the so-called ‘highly unexpected adult male fans of [children’s cartoon series] My Little Pony’. Many tulpamancers, thus, report creating one or more pony Tulpa.

Tulpa folk theory

The community is primarily divided between so-called psychological and metaphysical explanatory principles. In the psychological community, neuroscience (or folk neuroscience) is the explanation of choice. Tulpas are understood as mental constructs that have achieved sentience. The metaphysical explanation holds that Tulpas are agents of supernatural origins that exist outside the hosts’ minds, and who come to communicate with them. Of 118 respondents queried on the question, 76.5% identified with the psychological explanation, 8.5% with the metaphysical, and 14% with a variety of “other” explanations, such as a mixture of psychological and metaphysical.

Several Tulpamancers (from both psychological and metaphysical communities) report having had sentient imaginary friends for up to several years before finding out about Tulpamancy. For one informant, the practice had been established in her family for several generations. Many Tulpas from the psychological tradition, when interviewed separately from their hosts, also claim to have ‘been around’ in their hosts’ consciousness before their hosts became aware of them through Tulpamancy.

Of 73 Tulpamancers tested on this question, 37% reported that their Tulpas felt “as real as a physical person”, while 50.6 % described their mental companions as “somewhat real – distinct from physical persons, but distinct from [their] own thoughts”. 4.6% claimed “extremely real” phenomena, where Tulpas were “indistinguisable from any other agent or person”. Only those 4.6% claimed to hear and see their Tulpas “outside” their heads. The median length of Tulpamancy experience for these respondents was one year. Tulpamancers with 2+ years of experience reported higher degrees of somatic experience.

Demographic, social, and psychological profiles

The age range of interviewed Tulpamancers in another survey (n=141) was 14 to 34 years, with most falling in the 19 to 23 range. The male to female ratio is approximately 75/25 (male/female), though up to 10% identify as gender-fluid, and explore further ‘creative’ gender and ethnic variations through their humanoid Tulpas.

Tulpamancers are predominantly white, middle to upper-middle class urban youth. Of 141 respondents surveyed in September 2014, only two described themselves as “African American”, with two more reporting being “half black”. Four respondents described themselves as Asian, 4 more as “half Asian”, and one as “one quarter Asian”. All others described themselves as “white”, or by a variety of euro-American ethnic labels (Irish, German, Russian, etc.). One identified as “Siberian”. Most are undergraduate university students, but up to a third are fully employed. The IT field is the most commonly reported sector of employment.

The majority of Tulpamancers are located in urban areas in the US, Canada, the UK, Australia, Western Europe, and Russia. The breakdown from a survey taken by 141 tulpamancers in September 2014 was as follows:

163 Tulpamancers by Geolocation

The only known groups of tulpamancers to meet in person at the time (Sept, 2014) were located in Moscow and Omsk, where weekly gatherings were held with Skype-conferencing capacities for other Russian-speaking tulpamancers located outside these locations. Other group meetings have since emerged in St Petersburg, Volgograd, and other parts of Siberia. English and Russian seem to be the two dominant languages for the diffusion of Tulpa culture. As of February 2015, the Reddit forum through which most tulpamancy conversations in English converge had 7740 members, but less than 200 active posters. The Russian social networking site Vkontakte also listed 6000+ members, with a smaller ratio of active posters.

From coding of qualitative interviews collected in large surveys, the most common tulpamancer profile to emerge is one of a highly cerebral, imaginative, highly articulate, upper-middle class, formally educated person with many consistently pursued interests, talents, and hobbies, but limited channels of physical social interaction.

Typical tulpamancers are confident about their talents, but are quite modest and socially shy. They possess – or have cultivated – a high propensity for concentration, absorption, hypnotisability, and non-psychotic sensory hallucinations. Their limited social life and social anxieties, however, are not correlated with impaired levels of empathy and interest in other people. They score average or above average on empathy and Theory of Mind tests, indicating that their ability to relate to other humans is either optimal or enhanced (NB: I used my own revised version of Baron-Cohen’s empathy and ASD quotients tests. As these primarily rely on explicit mindreading that can be passed logically, more testing of implicit cognition is required.)

Loneliness is overwhelmingly reported as a common factor for creating Tulpas, who are described as “most loyal” and “perfect” kinds of companions. Of 74 tulpamancers tested, the majority scored higher than average on shyness scales and lower than average on sociability scales for comparable population sets (note: I used my own scales, revised from Asendorpf et al.). Most respondents reported some degrees of social anxiety. Their ‘happiness’ levels were assessed through a variety of qualitative interview tools and correlated with the Positive and Negative Affect Schedule Scale (Watson et al., 1988), on which all scored very highly (n=74, m=35.5, sd=7.5, r14-49)

High scores (n=74, m=21.35, sd=6.7, r1-33) on the Tellegen Absorption Scale (to measure capacity for hypnotisability, synaesthesia, and ‘trance’ states) seem to reflect practice as much as proclivity. In other words, respondents reported improvements on their ability to concentrate, visualize, and experience sensory ‘hallucinations’ since taking up Tulpamancy. Among the most interesting results is the negative correlation between low sociability and high empathy. Further ethnographic findings from forum discussions and interview data also indicate a moderate-to-high prevalence of tulpamancers who identify with, or have been diagnosed with, Asperger’s syndrome. No significant findings of impairment were found for either of the two respondents who took the Theory of Mind test in the first survey.

Relationship with mental illness

A subsequent survey was designed to target tulpamancers who had been diagnosed with or identified with mental illnesses or DSM-type psychopathologies. The most common ‘conditions’ reported by respondents (n=24), excluding social anxiety, were, in order of frequency, Asperger’s Syndrome (25%), Attention Deficit Disorders (21.4%), general anxiety (17.8%), depression (14.4%), and Obsessive Compulsive Disorder (10.7%).

The survey revealed a similar trend of overall reported improvement. 93.7% of respondents (n=33) expressed that taking up Tulpamancy had “made their condition better”.

54.5% of the respondents who identified with Asperger’s or Autistic Spectrum Disorder (n=11) claimed that their ability to read physical humans had improved with Tulpamancy, while 45.5% reported being unsure about changes in mindreading, despite overall positive changes in their social lives.

“I would say that it [my ability to read other humans] has improved quite a lot since I have been with my tulpa”, claimed one informant. “Although, at this point, its [sic] difficult to say if it’s my ability that is improving, or if I am relying on my tulpa to recognize things that I miss.”

This prompted further research on how Tulpas perceive and transcend their hosts’ limitations. When queried individually via email or specific questionnaire entries, Tulpas reported overall cognitive and affective difference from their hosts’ ‘baseline’ and often claimed relative or total independence from the hosts’ conditions. Mixed Tulpa responses on ASD-type conditions, however, indicated that most, but not all, Tulpas shared some aspects of their hosts’ autism, but were generally able to benefit from their position of ‘observer’ free of ‘participant’ obligation (see the Tulpas’ full responses here).

Inner voices: language, narrativity and episodicity

The role of narrative in the mediation of tulpa experiences – and by extension, to any experience of what it is like to be conscious – demands careful examination. Tulpamancy, as we have seen, entails explicit efforts (but only in the forcing stage, which typically lasts up to four months) at narrating the Self, in addition to initially conscious cognitive costs in the harnessing of absorption and the training of hypnotizable proclivity. The ‘self’ in this case is initially narrated as ‘different’ kinds of multiple selves within single bodies, and subsequently operates automatically once the practice is successfully enskilled.

This raises specific and general question about the role of language and inner narration in the mediation of conscious experiences. In Against Narrativity, an important essay in the philosophy of mind and language, Galen Strawson (2004) challenged what he took to be the naïve celebration of narrative as a linchpin of conscious experience. How literally, he asked, should we take the trope that we become the autobiographical stories we tell ourselves (Bruner, 1987) or perceive our lives as an explicitly unfolding narrative through which our sense of Self is constructed (Taylor, 1989)? Some people, he argued, are not particularly drawn to inner narration, and do not perceive their ‘Self’ as a continuous unit that persists over time and change. These types of Selves, which he termed ‘episodics’, tend to think of themselves as different persons in different moments and stages of their lives. He contrasted epidodics with ‘diachronic’ Selves, who tend to actively narrate the authorship of their life as a unified, continuous project. Strawson identified diachronicity and episodicity as personality types, and hypothesized that while both modes can co-exist and fluctuate within a single person, diachronicity seemed to be dominant in most contemporary experiences of selfhood. Anthropologist Maurice Bloch (2014a), in turn, recently proposed that while core neuro-phenomenal elements of sentience are universally shared by humans and other animals, cultural and historical differences were likely to be found at the level of narrative aspects of consciousness. He concluded, building on Strawson, that diachronicity might have become dominant in the West, and may be the locus of superficial difference that is too often extrapolated to the clichéd anthropological notion that the Self is an exclusively Western, post-reformation construct.

Tulpamancy offers an interesting case study to verify Strawson and Bloch’s claims, particularly in light of the central role of narrative in the practice. If a strong emphasis on inner-monologue is thought to lead to continuity and diachronicity, what to make of multiple selves enacted through narrative? Could different modes of narrativity be conducive to episodicity? Could episodic proclivities remain dominant in spite of the narratively intensive modes of alphabetic literacy that shape our subjectivities (see Collins, 1995, for a review of debates on literacy and cognition)? How much do we know about these differences within and across populations?

The distribution of diachronicity and episodicity, as it turns out, has yet to be empirically examined on any large scale. Expanding on the rare experimental tools devised to assess this question (Chandler et al., 2003; Hertler et al., 2015), I designed a questionnaire that weights people’s experience and intensity of inner-narration with their perceived continuity of conscious experience (you can take the test and see your results here) Respondents were matched with one of four points on a diachronic-to-episodic scale, and were later grouped as belonging to either one of two spectrums.


The same questionnaire was given to tulpamancers (n=113) and a group of non-tulpamancers (n=93). While 59% of non-tulpamancers fell in the diachronic spectrum, 70.8% of tulpamancers tended toward episodocity. In debriefing sessions with both control groups (a tulpamancer forum, and two groups of undergraduate and graduate students), many informants reported feeling a strong sense of multiplicity and discontinuity in their lives against the otherwise strong presence of an inner-narrative voice. We concluded that episodic proclivities may be more prevalent than previously assumed, and that more comparative data from non-academic, less hyper-verbal population sets were required to make better arguments. Overall questions remained on the place of narration in ‘thinking’ (see Bloch, 2014b, for arguments on how thinking is not ‘language-like’). As a trilingual, triliterate person with strong episodic tendencies, for example, I am rarely aware of the language (if any) I am thinking in, unless I am working on an explicitly narrative task like rehearsing arguments for a lecture, talk, imaginary conversation, or paper. Neurolinguists and clinicians, however, have found that psychotic manifestations in multilingual patients can occur in any of the patients’ languages (Paradis, 2010). When queried on the question, several multilingual tulpamancers explained that different tulpas within a single host could display distinct linguistic identities (e.g. one Spanish-speaking tulpa, and one English-speaking tulpa), while others reported code-switching with their tulpas (e.g. English, or Spanish, or Spanglish between tulpa and host). Others described having tulpas with foreign accents from languages in which the hosts were not proficient (e.g. Anglophone host with a tulpa who speaks English with a Japanese accent).

While inner-voice and phenomenal aspects of consciousness are likely to remain hard problems to study with any populations, my current claims about tulpamancy’s therapeutic effects will need to be supplemented with further face-to-face ethnographic, behavioural, and neuroscientific findings.

I now turn to a discussion of the interactive mechanisms that make tulpamancy – and, I argue, any experience of human personhood – possible.

 “An aspiration to unfold all”, from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.
“An aspiration to unfold all”, from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.

3. Theorizing Tulpas: Personhood in Shared, Embodied, and Hypnotic Perspectives

The kinds of neurological, sociocognitive, political, linguistic, and technological mechanisms that enable tulpamancers (and, as we will see, members of any formal-enough ‘culture’) to experience such a stable embodied sense of personhood (in this case that of multiple and ‘healing’ forms of personhoods) warrants careful discussion. This requires detours through such disciplines as cognitive psychology, ethnology, ethnobiology, linguistic anthropology, the neuroscience of attention, and social approaches to hypnosis.

Tulpamancy is a new cultural phenomenon that has yet to be studied ethnographically and scientifically. Psychological anthropologist Tanya Luhrmann mentioned the community in a 2013 New York Times op-ed and offered preliminary comments about links with the Cognitive Science of Religion (CSR), in which the perceived presence of supernatural agents in most human cultures is understood as an evolutionary ‘by-product’, or maladaptive properties of mind. Luhrmann, as I explain below, draws on her own studies of ‘hallucinations’ and ‘unusual sensory experiences’ among Pentecostal Christians to depart from these evolutionary models and emphasizes the learning-dependent, absorption-and-practice-intensive, ‘healing’ quality of interaction with imaginarily conjured agents.

Here, a brief review of the cognitive literature on ‘religion’ (or human belief in and interaction with ‘supernatural’ agents) and ‘animism’ will give us further clues to theorize Tulpa and Tulpa-like experiences.

In the first wave of CSR theorizing, the inference of supernatural agents from the world around us is explained as more or less inevitable features of cognition; namely a tendency to attribute anthropomorphic animacy and agency to living things and inanimate objects alike. This is why, in Steward Guthrie’s famous formula, we see “faces in the clouds” (Guthrie, 1993). A second current of CSR theory, championed by scholars like Pascal Boyer, Justin Barrett, Harvey Whitehouse, and Scott Atran, draws on evolutionary, cognitive, and experimental psychology, ethnography, and ethnobiology to expand on the insight that humans across cultures tend to project fundamentally human mental characteristics on supernatural agents. In this model, humans are said to reason about supernatural agents by expecting them to reason like humans, particularly in terms of goal-directedness, shared intentionality, intuitive physics, naïve psychology, and semantic and episodic memory. We expect a spirit who would return each night at midnight to torment us in our bedroom, for example, to know and remember that we will be in our bedroom at the same time each day, to understand and expect that and how we will be afraid of it, and to know just how to torment us in universally human and culturally specific ways. At the same time that we intuitively accept that the spirit can go through walls but not fall through the floor, we assume that we can read its mind as much as it can read ours. This propensity to attribute human-like intentionality (that is to say, ‘aboutness’, or the property of minds to be about, or represent things, events, and states of affairs) to non-human entities is posited to have evolved in predator-prey environments, when the need to detect the presence and predict the behaviour of dangerous agents would have been a crucial survival mechanism. Evolutionary psychologists working from a domain-specific, or ‘modularist’, hypothesis explain the emergence of specific cognitive modules to handle such specific problems in our environment. This ‘agent-detection’ cognitive module – or device – thus, is understood as going on overdrive, or agent-hyperdetection when we incorrectly infer the presence of agents. A major finding of second-wave CSR, however, is that agent-hyperdetection resulting in formal systems of ‘religious’ belief may also be universally counter-intuitive. The presence of roughly similar folk taxonomies of animals, plants, and kinds of objects across cultures, and most particularly of grammatical categories to account for animate vs. inanimate objects and agents seems to indicate a universal sense of intuitive physics in humans. The kinds of objects and entities to which human infants seem inclined to attribute animacy, however, are still the subject of debates among developmental psychologists. Baillargeon and Luoy (2005), for example, have argued from experimental evidence in a looking time study that 5-month-old infants are likely to attribute goals to any entity, living or not, that they identify as an agent. According to the authors, any moving thing (such as a toy car or self-propelled box) that may appear to be self-directed can be interpreted as an intentional agent. A similar study by Mahajan and Woodward (2007), however, offered that 7-month-old infants respond visually to the movement of both animate and inanimate objects, but only reproduce the goals of the former.

‘Animism’ applied to other animals and living species, in any case, appears to be much more intuitive, and is found in the cosmologies and practices of many cultures, from Amazonia and Melanesia to Siberia and the Canadian Arctic (see Descola, 2005). As biological anthropologist Agustín Fuentes explains, the similar sense modalities, central nervous systems, and cognitive architecture shared by all mammals are most noticeable in similar physiological responses to fear, pain, and suffering found across species. If humans can read highly stable indexical cues signifying fear, pain, or suffering (like squeaking, wailing, twitching, fleeing, or others signifying anger or threat) in members of other species, then it follows that we can recognize members of these species as sentient beings, or as persons.

A capacity for shared empathy and intersubjective recognition that extends beyond the boundaries of our own species, thus, may hint at a good recipe for the bounds and possibilities of agent ‘hyper-projection’. We may not know precisely what it is like to be Thomas Nagel’s bat, but we need no conscious cognitive effort or internalized cultural script to recognize that a bat is in pain. This is a good start. Revising Nagel’s famous thought experiment will shed more light on the ‘naturalness’ of the kinds of ideas which, when elaborated upon and frequently shared and practiced in a formal set of cosmological narratives, may lead to animist ontologies in which animals are recognized as full persons – or indeed, where Tulpas think on their own as full persons. Ask yourself whether, and to what extent you may be able to recognize that each of the following ‘animal’ is in pain: a bear, a dog, a dolphin, a raven, a salmon, a spider, and an earthworm. We may infer from a bird’s broken wing that it is in pain, or we may form semi-reflective beliefs about a twitching fish ‘gasping for water’ as we would gasp for air. We can most definitely recognize suffering in any mammal. But what about an ant or a clam?

The Cree, a historically hunting and gathering ‘animist’ people living in the Northern Boreal forest region of sub-arctic Canada, speak an Algonquian language that marks nouns as being animate or inanimate. Unlike gendered nouns in Romance languages, there are no ‘obvious’ rules for distinguishing the animacy of a noun. To complicate things further, word order is also very flexible, and subjects and objects are usually expressed by means of agglutinative inflection with a verb: this typically produces long words in which objects or agents are described in the context of an action. To speak of a particular kind of bird, for example, one may say yuuskahiiu, which literally translates as “it (marks the animated noun ‘partridge’) perches on a tree and does not fly away as the hunter goes near to shoot it”. Such complex, ‘covert’ grammatical categories were first described and labelled ‘cryptotypes’ by Benjamin Whorf, who pioneered the study of linguistic anthropology in the early 20th century. Since the rules of cryptotypes are unknown to native speakers, Whorf showed that they can only be identified when they are broken.

In my work with the James Bay Cree, I have asked Cree speakers if the word awesiis, which is usually translated as ‘wild animal’, corresponds exactly to the English word ‘animal’. My informants usually answer that it does, until I proceed by elimination to ask whether, say, a bear, a wolf, a moose, a human, a raven, or a spider can be an awesiis. While younger Cree hunters almost always contend that a human cannot be an awesiis, all agree that spiders, ants, bugs, insects, earthworms and mollusks do not belong to the class of ‘wild animals’. I take the finding that the Cree, a people with a well-documented sense of deep empathy, friendship, and intersubjectivity with many animal forms (see Scott, 2006) do not attribute personhood or readability to insects and mollusks to be added evidence that full-fledged empathetic animism becomes more counter-intuitive with phylogenetic distance between species.

For Boyer and others, the minimally counter-intuitive attribution of full-fledged intentionality and anthropomorphized personhood to non-human and inanimate entities is precisely what makes ‘religious’ narratives catchy, easy to recall, and efficient to transmit culturally. Add to this what Harvey Whitehouse calls a ‘doctrinal’ mode of religiosity with a hierarchy of ‘experts’, formal narratives, and frequently repeated rituals, and you have the recipe for the efficient, rapid spread of religious ‘beliefs’ and practices.

When my 7-year-old son tells me that his penguin friend at the Montreal Biodome “misses him”, or that the lump in his throat “doesn’t want to let [him] eat”, he is making a minimally counter-intuitive anthropomorphic inference about the agency of animals and living things. I, as his father and ‘expert’ purveyor or relevant doctrinal knowledge in a secular polity, would normally proceed to ‘correct’ him, thereby continuing to ensure that he is becoming more proficient at playing our particular language game. Were I to reward his inferences with rich narratives about Penguin-and-Lump-Personhoods within a broader social context in which everyone believes in and interacts with penguin-friends and lump-agents, my son would soon start having full conversations with his ‘imaginary’ friends.

Could it be, then, that ‘entirely imaginary’ agents are, in a sense, more intuitively imaginable, and so precisely because we can conjure them in the absence of the marks of illegibility found in what we readily recognize as inanimate or inpersonal entities – or in other words that our agent detection and projection abilities enable us to recreate personhood attributes with more intuitive precision in the absence of physical designata?

What, then, of the somatic quality of ‘belief’?

In contrast with the evolutionary literature, Tanya Luhrmann’s work with evangelical Christians has shown that somatically experienced religious practices (like hearing the voice of God) take ‘hard work’ and require a proclivity for and training in absorption, in addition to a broader socio-cultural context that is permissive of and conducive to such experiences. She also showed that, in such a context, these experiences could be highly rewarding and conducive to healing.

My work with Tulpamancers, which owes a lot to Luhrmann’s theorizing of absorption and learning, invited me to revise central questions in the problem of physicality and invisibility in the study of sociality, and pointed to more cumulative feedback loops between proclivity and practice. The social and cumulative nature of learning, the doctrinality of enculturation, and the sensory grounding of narrative practice have added further clues to this puzzle and pointed me in the direction of regimes of attention as a possible linchpin of socially mediated experiences and ways of being a person.

A good account of attention-mediated sociality will entail a revision of current sociocognitive models of joint-attention – usually understood as occurring between agents in direct interactional spheres of gaze-following, finger-pointing, or other verbal or non-verbal cues. In addition to demonstrating how non-indexical, narrative forms of doctrinality can allow shared intentionality and ‘joint’ attention to rise far beyond dyadic and spatially-bounded spheres in the process of forming joint goals and achieving jointly-mediated focus, more connections will need to be established with theories of active imagination. Just like attention in the Invisible City can be jointly focused away from individuals, so too can attention be jointly focused inward within individuals, thereby giving life and sensory grounding to individually imagined but collectively scripted agents. The bounded, invisible selves of modern cities, but also the healing, God-hearing selves of Pentecostal polities, or the multiple humanoid selves of Tulpamancy, thus, are best explained as produced hypnotically.

For neuroscientist Amir Raz, whose work on neural correlates of attention departs from reductionist models that present dissociation and trance as distinct (or ‘altered’) states of consciousness, hypnosis is simply any intense, or ‘atypical’ form of attention (Raz, 2004). Attention, in more anthropological terms, is socially shaped as much as it shapes sociality; or as the cultural psychiatrist Laurence Kirmayer puts it, “social discourse and narratives shape hypnotic experience, but they are themselves influenced by mechanisms of attention” (Kirmayer, 1992, p276; see also Spanos, 1996 for a more socioconstructivist view; Kirmayer, 1998 for comments on Spanos).

As an anthropologist, I am inclined to think of the ‘typical’ as any dominant normative scheme governing the expected order of states of affairs in any given context. But ‘typical’ regimes of attention, seen from other perspectives, will appear just as strange as any variation easily recognized as ‘hypnosis’ from the perspective of the dominant. If we strip all social schemes and ways of being a person of perspectival exoticism, they become equally strange, or equally banal.

Whitehouse’s mode of religiosity theory will offer further clue to explain the social grounding of these mechanisms. Whitehouse has hypothesized that the emergence of doctrinal modes of religiosity characterized by frequently repeated rituals and expert-led, formal exegetic, behavioural, and cosmological prescriptions played an important role in the rise of large-scale polities after the Neolithic, particularly because they tend to elicit widely-spread and conformist forms of semantic memory. He contrasts this mode with the historically older ‘imagistic’ mode found in many small-scale societies, in which rarely performed, intense, often dangerous rites and rituals tend to elicit high emotional arousal, which in turn facilitate episodic recall and strengthen social bonds between participants. While the doctrinal mode affords efficient and large-scale spread of similar mental representations and practices, imagistic modes can only be sustained in small groups and lead to highly personal exegetic reflection that rarely amount to a collective consensus on the ‘meaning’ and content of visions and experiences that arise in ritual. Whitehouse’s theory is most useful to my own theorizing of sociality outside and beyond religious contexts. The doctrinal and imagistic modes, here, are best described as modes of social learning and joint attention.

Tulpamancy provides a fascinating case of sequentially unusual co-existence between both modes. The hard-work of initial visualization, thought-form and forcing invariably affords a high-frequency, low-arousal, relatively formal set of prescriptions that structurally resembles the modes of doctrinality of our contemporary social, educational, economic, religious and emotional lives – but with more conscious degrees of discipline. The counter-normative, ‘atypical’ nature of the focus, however, and the gradual success in conjuring ‘unusual’ sensory experiences eventually leads to a highly arousing set of deeply personal interior imageries and sensations that triggers imagistic modalities. That these highly arousing, hard-to-reach experiences are mediated socially by a growing number of individuals working toward common goals consequently leads to a deep sense of reward validated in a common Tulpamancer ‘identity’, but one which affords a broader degree of improvisation from what is culturally and ecologically available to the hosts. Thus, a relatively formal script and a doctrinal modality (“visualize, concentrate, build shape and personality traits and wait until you experience voices and touch from sentient Tulpas”), when successfully endoctrinated, leads to human hosts who interact with such automatic processes as elvish, pony, dragon, or other bodiless minds and voices. The very hard work reported by Tulpamancers who attempt to undo their Tulpas points to the high degrees of automaticity achieved by mature practitioners. Getting rid of a Tulpa for a seasoned -mancer, thus, could be analogically situated somewhere between unlearning the piano or correcting one’s posture. Should the practice survive, gain public acceptance, and formalize itself for another decade, it will be as hard as willing onesself to forget how to read or completely unlearn a language in which one is fully fluent. But such examples, once more, pertain to scales of degrees, but not kind.


Classical anthropological insights from Mauss and Whorf to Bourdieu have shown us that ‘culture’ and ‘automaticity’ are in many ways synonymous. Turning to the absorptive, somatic quality of ‘belief’, Tanya Luhrmann demonstrated that religious experiences were Tulpa-like. I hope to show, in turn, that ways of being social and of being a person are also hypnotic and Tulpa-like. Tulpa and Human, indeed, as terms used to describe persons embodied and enacted through narration and joint attention, may well turn out to be synonymous.

 “Thought form of the music of Gounod” , from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.
“Thought form of the music of Gounod” , from Annie Besant & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.



[i]REB approval for this project was granted through McGill University. Please contact Lynda McNeil, Research Ethics Officer, with any questions or concerns. Note that the REB was concerned with the anonymity and protection of Tulpa persons as well as that of their hosts. This is a rather hopeful development in legal definitions of personhood.


Samuel Veissière is Assistant Professor of Anthropology at the University College of the North, and currently holds a Visiting Professorship in the Culture, Mind, and Brain program at McGill University where he is affiliated with the Division of Social and Transcultural Psychiatry and the Department of Anthropology. He has conducted fieldwork on emergent modes of sociality and cooperation in the context of street children livelihoods and transnational sex work in Brazil, ‘clandestine’ migration in the Pan-Amazon, indigenous revivalism, and various new forms of Internet cultures. His current work investigates the phylogeny and ontogeny of social cognition and cooperation with an emphasis on how children across cultures learn to reason about ‘kinds’ of humans and persons.



I am deeply grateful for all the support provided by Laurence Kirmayer and Ian Gold at McGill University. Tanya Lurhrmann’s work on absorption, voices, and unusual sensory experiences provided the initial inspiration for this study, and I am very thankful for all her pointers and comments on earlier drafts of this paper. I’d also like to thank Eugene Raikhel and Elle Nurmi for the generous editorial work, and Ahmed Soliman for his help with statistics. I am indebted to my undergraduate students in the Theories of Culture and Society class at McGill for their insightful comments and questions on the links between narration and consciousness. I’d also like to acknowledge the great work of Deanna Day, managing editor at Somatosphere, for pulling it all together.


Works cited

Aken, M. A. G., & Asendorpf, J. B. (1996). “Continuity of the prototypes of social competence and shyness over the life span and across life transitions”. Journal of Adult Development, 3, 4, 205-216.

Atran, S. (2002). In gods we trust: The evolutionary landscape of religion. Oxford: Oxford University Press.

Baron-Cohen, S., & Wheelwright, S. (2004). “The empathy quotient: an investigation of adults with Asperger syndrome or high functioning autism, and normal sex differences”. Journal of Autism and Developmental Disorders, 34, 2, 163-75.

Barrett, J. L. (2011). Cognitive science, religion, and theology: From human minds to divine minds. West Conshohocken, PA: Templeton Press.

Bateson, G. (1972). Steps to an ecology of mind. New York: Ballantine Books.

Bateson, G. (1980). Mind and nature: A necessary unity. Toronto: Bantam Books.

Besant, A & C. W. Leadbeater (1901) Thought-Forms. London: The Theosophical Publishing House.

Bloch, M. (2014a). “Reconciling social science and cognitive science notions of the “‘self'”: In, Bloch, M. Anthropology and the Cognitive Challenge. Cambridge, UK: Cambridge University Press.

Bloch, M. (2014b). “What goes without saying”. In, Bloch, M. Anthropology and the Cognitive Challenge. Cambridge, UK: Cambridge University Press.

Bruner, J. (1989). “Life as Narrative”. Social Research New York, 71, 3, 691-710

Csordas & Masquelier. (1997). “Embodiment and experience: The existential ground of culture and self”. American Ethnologist, 24, 4, 940.

Chandler, M. J., Lalonde, C. E., Sokol, B. W., & Hallett, D. (2003). “Personal persistence, identity development, and suicide: a study of Native and Non-native North American adolescents”. Monographs of the Society for Research in Child Development, 68, 2.)

Calvino, I. (1974). Invisible cities. New York: Harcourt Brace Jovanovich.

Clark, A., & Chalmers, D. (1998). “The extended mind”. Analysis Oxford, 58, 1, 7-19.

Collins, J. (1995). “Literacy and literacies”. Annual Review of Anthropology, 24, 75-93.

David-Néel, A. (1932). Magic and mystery in Tibet. New York: C. Kendall.

Descola, P. (2005). Par-delà nature et culture. Paris: NRF.

Fuentes, A. (2006). “The Humanity of Animals and the Animality of Humans: A View from Biological Anthropology Inspired by J. M. Coetzee’s ‘Elizabeth Costello'”. American Anthropologist, 108, 1, 124-132.

Gellner, E. (1998). Language and solitude: Wittgenstein, Malinowski, and the Habsburg dilemma. Cambridge: Cambridge University Press.

Goffman, E. (1971). Relations in public: Microstudies of the public order. New York: Basic Books.

Gold, J., & Gold, I. (2014). Suspicious minds: How culture shapes madness. New York: Free Press.

Guthrie, S. (1995). Faces in the clouds: A new theory of religion. New York: Oxford University Press.

Heinz, A., Deserno, L., & Reininghaus, U. (2013). “Urbanicity, social adversity and psychosis”. World Psychiatry, 12, 3, 187-197.

Hertler, S. C., Krauss, H., & Ward, A. (2015). “Assessing diachronic reasoning: exploratory measures of perceived self-change in young adults”. Psychological Reports, 116, 1, 176-93.

Kirmayer, L. J. (1992a). “The Body’s Insistence on Meaning: Metaphor as Presentation and Representation in Illness Experience”. Medical Anthropology Quarterly, 6, 4, 323-34

Kirmayer, L. J. (1992b). “Social constructions of hypnosis”. The International Journal of Clinical and Experimental Hypnosis, 40, 4, 276-300

Kirmayer, L. J. (1987). “Hypnosis and the limits of socialpsychological reductionism”. Behavioral and Brain Sciences, 10, 3, 521.

Luhrmann, T. M., Nusbaum, H., & Thisted, R. (2010). “The Absorption Hypothesis: Learning to Hear God in Evangelical Christianity”. American Anthropologist, 112, 1, 66-78.

Luhrmann, T. M. (2011). “Hallucinations and Sensory Overrides*”. Annual Review of Anthropology, 40, 1, 71-85.

Luhrmann, T. M. (2012). When God talks back: Understanding the American evangelical relationship with God. New York: Vintage Books.

Luo, Y., & Baillargeon, R. (2005). “Can a self-propelled box have a goal? Psychological reasoning in 5-month-old infants”. Psychological Science, 16, 8, 601-8.

Majid, A., & Levinson, S. C. (2011). “The senses in language and culture”. Senses & Society, 6, 5-18.

Mahajan, N., & Woodward, A. L. (2009). “Seven-Month-Old Infants Selectively Reproduce the Goals of Animate But Not Inanimate Agents”. Infancy, 14, 6, 667-679.

Nagel, T. (1974). “What Is It Like to Be a Bat?”. The Philosophical Review, 83, 4, 435-450.

Paradis, M. (2008). “Bilingualism and neuropsychiatric disorders”. Journal of Neurolinguistics, 21, 3, 199-230.

Raz, A. (2004). “Anatomy of attentional networks”.The Anatomical Record Part B: the New Anatomist, 1, 21-36.

Raz, A. (2005). “Attention and Hypnosis: Neural Substrates and Genetic Associations of Two Converging Processes”. International Journal of Clinical and Experimental Hypnosis, 53, 3, 237-258.

Scott, C. (2006). “Spirit and practical knowledge in the person of the bear among Wemindji Cree hunters”. Ethnos, 71, 1, 51-66.

Spanos, N. P. (1996). Multiple identities & false memories: A sociocognitive perspective. Washington, DC: American Psychological Association.

Strawson, G. (2004). “Against Narrativity”. Ratio, 17, 4, 428-452

Taylor, C. (1989). Sources of the self: The making of the modern identity. Cambridge, Mass: Harvard University Press.

Thompson, E. (2015). Waking, dreaming, being: Self and consciousness in neuroscience, meditation, and philosophy. New York: Columbia Univ. Press.

Tomasello, M. (2009). Why we cooperate. Cambridge, Mass: MIT Press.

Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind: Cognitive science and human experience. Cambridge, Mass: MIT Press.

Whitehouse, H. (2004). Modes of religiosity: A cognitive theory of religious transmission. Walnut Creek, CA: AltaMira Press.

Whorf, B. L., Carroll, J. B., Levinson, S. C., & Lee, P. (2012). Language, thought, and reality: Selected writings of Benjamin Lee Whorf. Cambridge, Mass: The MIT Press.

14 replies on “Varieties of Tulpa Experiences: Sentient Imaginary Friends, Embodied Joint Attention, and Hypnotic Sociality in a Wired World”

I haven’t finished reading this paper… But after the first few pages I already felt compelled to comment; thank you for writing such an accessible paper as well conducting thorough research.

[…] Pour autant, les tulpamanciens ne visent pas à une réalisation spirituelle, ou un rapport dévotionnel à une divinité. Ils ne souhaitent pas non plus utiliser leur tulpa pour réaliser leurs désirs, comme le font aujourd’hui les occultistes occidentaux. Leur objectif est plus simple et plus touchant : ils veulent seulement se faire un ami. Samuel Veissière, anthropologue à l’université McGill, a enquêté sur les tulpamanciens et dressé un profil type des adeptes basé sur une enquête impli…. […]

[…] Pour autant, les tulpamanciens ne visent pas à une réalisation spirituelle, ou un rapport dévotionnel à une divinité. Ils ne souhaitent pas non plus utiliser leur tulpa pour réaliser leurs désirs, comme le font aujourd’hui les occultistes occidentaux. Leur objectif est plus simple et plus touchant : ils veulent seulement se faire un ami. Samuel Veissière, anthropologue à l’université McGill, a enquêté sur les tulpamanciens et dressé un profil type des adeptes basé sur une enquête impli…. […]

[…] Pour autant, les tulpamanciens ne visent pas à une réalisation spirituelle, ou un rapport dévotionnel à une divinité. Ils ne souhaitent pas non plus utiliser leur tulpa pour réaliser leurs désirs, comme le font aujourd’hui les occultistes occidentaux. Leur objectif est plus simple et plus touchant : ils veulent seulement se faire un ami. Samuel Veissière, anthropologue à l’université McGill, a enquêté sur les tulpamanciens et dressé un profil type des adeptes basé…. […]

[…] ◆神経科学では解明しきれないのが心の声  脳が私たちに話しかけるよう仕向ける方法は他にもある。目隠しや耳当てなどで特定の感覚をブロックする“感覚遮断”には、内なる声を呼び出す効果がある程度はある。祈りや瞑想のようなでも内なる声を聞くことができる。実際、タルパマンシーを実践する人たちは、瞑想によって明らかに感じられる存在を呼び出せると主張している。 […]

Comments are closed.