1Long ago when she was a student, Tania was spending the weekend along with a Buryat-Mongolian friend, Ayuna, at the country cottage (dacha) near Ulan-Ude of a senior female Buryat academic, who was away at the time. One morning they went out into the garden and there was an extraordinary happening: they spotted a hoopoe (R. udod, B. oboljin, M. bövööljin), which is a very rare bird in these parts, alighting by the door. Ayuna gave a cry of horror. In Buryat culture, a hoopoe when it lands near human habitation is a terrible omen presaging disaster or even a death. To Tania, who is Russian, the bird meant nothing (“This was not my culture”) but Ayuna was frightened. Her fear was not on her own behalf but on that of the professor, because it was the latter’s garden that the bird had invaded. Ayuna made a libation of milk to the four cardinal directions, hoping this would avert the misfortune. According to Buryat recollections of this event – which went on circulating in the city for years, this professor died unexpectedly two months later.

3Eduardo Kohn has argued persuasively that anthropology should reorient itself to transspecies engagement and open itself to “the many ways in which people are, in fact, connected to a broader world of life and the ways in which this [reorientation] changes what it might mean to be human” (2007: 5). The relations he is concerned with are those between humans and that category of nonhumans held to be “selves”, that is, animals, the co-inhabitants of the forest ecology, which are capable of intention, representation, and communication on their own account. I am very much in sympathy with this “anthropology of life” for its emphasis on the dynamics of embodied processes and its consideration of how different kinds of sign (symbol, icon, and index) constitute transspecies communications. Omens, however, present a wider world of engagement. True, portents very often do take the form of actions by birds or animals. But in Mongol regions, many other phenomena can be omens, from the eclipse of the sun to an image in a nightmarish dream. In fact, it is difficult to think of anything that could not become an omen, given the necessary condition that it can be conceptually singularized and identified as a particular “thing” whose usual pattern of activity has switched, with seeming implication for the future. The fact that “un-self-like” nonhumans, such as lightning, are often experienced as omens extends our investigation beyond the more reciprocal, mutual “noticing” of human–animal relations discussed by Kohn. But, for Mongols, as I describe later, even a lightning-omen or an artefact-omen (such as a jacket left lying on the ground) is understood to have its history, and I shall argue that the form of the omen involves the antecedents as well as the consequences of the event itself. This point coincides with the more general argument Kohn makes that it is only the belief that we can know the patterns of the things in this world that allows us to act, and that it is dangerous to fail to recognize that they have inherent abilities and horizons which we humans do not have (2007: 7). In this context, it can be said that the omen-perceiver and the omen-maker are both temporal agents in a common cosmos.

4Mongolian regions provide a rich field for analysing this cosmological aspect of omens. In the precarious and finely tuned practice of pastoralism, it is vitally important for people to observe the behaviour of animals, birds, the weather, strangers, insects, in fact everything around them, for indications (shinj) of what will happen in the future. Vigilance does not cease when people move to towns. Decades of socialist education and the tumultuous post-1990 “age of the market” have culminated in a Mongolian practice of modernity in which “science” and “the occult” are combined; people routinely consult diviners who use calculators and attribute seers with hyper-scientific abilities, while shamanism is a boom industry (including shamans whose spirits make an appearance in blurry images on their laptops). Few people resolutely ignore completely sayings, affects, and emotions that surround omens, while some people are beset with one occult warning after another. In the countryside, a herder from Hövsgöl said that he was not at all religious, but it was “the nature of a Mongol to fear omens”, meaning that life for him was lived in constant awareness of omens (Oberfalzerova 2006: 22).

5Practical observations have been culturally standardized into countless “indications of the future” that are handed down through the generations: for example, the Buryat adage “If the mice build their nest high up in the hay stack it is a bad sign: there will be deep snow in the coming winter” (Khangalov 1960: 67). The mice already know something we do not know. Understandably, they are building their nest high above the snow to come. This kind of low-key portent acknowledges that the mice are more attuned, or more capable, than us humans in their awareness of the forces that determine atmospheric changes. However, a traditionally educated Mongol will also know countless “signs” (beleg) that can be related to one another more abstractly, in the fashion of astrology, also indicating a future outcome for his//her activity for good or ill. These signs can be auspicious/inauspicious numbers, sounds, colours, graphic shapes, spoken words, bodily gestures, artefacts, atmospheric events, animal behaviours, or bird calls. When encountered in life, they are not seen as absolutes with invariant meaning but interpreted relationally. This is done by calibrating the characteristics of the perceiver and his/her activity, such as a hunt, journey, or business deal, with the characteristics of the sign, by means of relating each to cosmic alignments such as the cardinal directions, the path of the sun, the time of day, or the date in the lunar calendar cycle. A Mongol ethnographer has devoted at least six books, including a five-volume set (Dulam 1999–2002), to categorizing and describing these beleg, but even so his work could never be complete, for in practice the actuality is the countless singularities consequent on aligning the diverse bundle of characteristics that compose the perceiver with those of the sign and with the nature of the encounter (seeing, touching, hearing, etc.).

6Yet the beleg, especially those given by animals, birds, or other living creatures, are not divorced from ordinary human life and they have an ethical dimension. In explaining this, I am reminded of what the chess-master Garry Kasparov said when explaining how he had to adjust to playing against a computer: “As a professional chess-player you are used to identifying moves with personality. This time the moves were coming out of a black hole – literally a black hole. It’s like there’s something there, but you don’t know exactly what it is” (Appleyard 2017, emphasis added). In the everyday understandable world of Mongol herders, the activities of wild animals and birds are like the chess moves of a human opponent: characteristic of the “personalities” of those species, which are judged by the same criteria as applied to people, such as bravery, fierceness, weakness, intelligence, and so forth. As Bernard Charlier (2015: 81–87) has shown for the fox and the wolf, rural communities in western Mongolia have crystallized highly complex mental portraits from observing and judging the behaviour of these animals, and these have become items of culture that can be deployed in advice to children, poetry, rituals, and so forth. In the case of foxes, identification of their habits as cunning, destructive, self-serving, and thieving has established them as muu sanaatai (having bad intentions), and this colours any incident of sighting a fox, which, as Charlier discovered, is considered a bad sign in any circumstances, but especially so for someone setting out on a hunt (2015: 81–82). These animal temperaments are reinforced by proverbs, legends, parables, and so forth, such as the story of a fox that tricked a wolf by making it feel so ashamed it withdrew and thus the fox was able to grab the piece of fat they were both after (2015: 83). I suggest that such stories are important to the analysis of transspecies omen encounters because they form the back-story or the genealogy of the animal and establish it in this sense as equivalent to the human actor with his or her character and life-story. The disturbing perception that a meeting of the two is a portent thus brings into play two parallel “histories”, or motivated complexes of on-going activities, whose clash has implications for the future. I will argue later that one way to think about such an encounter is in terms of dynamic interference patterns, though this idea does not capture all of the implications of omens.

8This article will be concerned with the doom-laden presages that Mongols call muu yor (“bad omens”). These stand out among the other signs in that they are so freakish, or so scary, or so absolutely identified with a forbidden action that people sense an instantaneous aversion, fear, and foreboding (sejigle-). Mongols say, “A thing that should not happen [is] a bad omen thing” (bolohgüi yum – muu yoriin yum, Oberfalzerova 2006: 18). Since this formula is reversible and the word bolohgüi can equally be translated “cannot happen” or “impossible”, we can also understand the idea in the following way: a “bad omen” is an impossible thing, that is, not possible in the accepted understanding of how things should happen. To witness a terrible traffic accident is a muu yor, so is physical contact with any item associated with a death; so is seeing a forbidden/impossible act like an animal mating with another species. It has been recorded among Buryats in the mid-twentieth century that even seemingly trivial acts, like a dog urinating while standing on all four paws, is a very bad omen (Natsov 1995: 114). Any human behaviour that is not intentional but “by itself” makes others feel horror, disgust, or disquiet is also a muu yor (Humphrey 2013). One Mongol told me that if her mother noticed her moping, she would say sharply, “Stop it! Don’t be a bad omen!” In such ways, a person’s “behaviour” or dream image can also seem to come from “beyond” rather than from him/herself. Mongols call this disquieting behaviour a “thing” (yum) that “reveals itself by images” (Oberfalzerova 2005: 16), which is an indication of something wrong with the world and a herald of impending disaster. There is no universally accepted unifying explanation for such “shouldn’t happen” events: there is no Satan, no Providence, nor an all-seeing God who “sends” omens; nor are omens thought to be created by malevolent human powers such as witchcraft or sorcery. People turn to shamans, Buddhist lamas, local sages, seers, and prophets for diagnosis of the agency, the “hidden thing” (daldyn yum) that lurks within the omen and resonates with the incorporeal elements of other beings (“souls”, “spirits”, “destiny”, hii yum “immaterial things”). But it is known that these specialists disagree with one another. There is always an essential doubt (Delaplace 2014). Someone who is suddenly “assaulted” (halda- , as Mongols say) by a muu yor is in a state of existential suspension, even if he or she hopes that the right way to “correct” it or “repel” it has been found.

Studies of omens and comparable phenomena

10Anthropologists have rarely focused on omens as a way of relating to the world. Thomas Beidelman’s essay on Kaguru omens (1963) was perhaps the first to consider this category of unusual event as an aspect of cosmology and to attempt to differentiate omens from other supernormal phenomena. The Kaguru distinguished omens from large-scale catastrophes (famine, flood, epidemics, etc.), which were thought to be sent from God, and also from misfortune caused by malevolent humanoid agents (witchcraft, magic, ghosts, and evil spirits). They were, rather, the coincidence of some unusual natural phenomenon with “a certain social situation in which a person is highly sensitive to how his affairs will go” (1963: 47). They included the calls of various birds, monkey shrieks, insect activities, a wild animal such as a zebra or giraffe entering a human house or cattle-pen, human accidents and physical abnormalities, and also human immoral acts such as bestiality that were thought to depart from the basic order of the world. All omens required immediate remedial action by the people involved or witnessing such events (1963: 51). In general terms, Kaguru omens seem remarkably similar to Mongolian ones. However, Beidelman’s essay, which is basically descriptive, was not followed up to my knowledge, and other studies of the period tended to dissolve omens in broader categories of analysis, such as religion, divination, or rationality.

12In this brief discussion of how ideas akin to the omen have made an appearance in the work of historians, we should also consider the “clue”. In the micro-historical approach of Carlo Ginzburg, the clue was a method. “It meant starting an investigation with something that did not quite fit, something odd that needs to be explained. This peculiar event or phenomenon is taken as a sign of a larger, but hidden or unknown, structure. A strange detail is made to represent a wider totality” (Peltonen 2001: 349). The very oddity of the clue leads thought in an unexpected direction, reveals connections, exposes some secret, a crime, or a fracture in society. The clue seen in this way has similarities with the omen as I try to frame it here. First, there is spatial condensation: Ginzburg (1979) argued that it is only on the micro scale of a single event or object that the historian can draw together the relations between systems of values, beliefs, representations, and social affiliations. And then there is the temporal tension, which is created by “the collision of an exceptional event with the long historical structure of popular culture” (Peltonen 2001: 350). However, if this comparison with the clue helps suggest a broader vision for historical analysis of omens, it also shows that for the actual practitioners the two are different. For detectives, one clue leads to the next, a chain that must be pieced together by the forensic rationality of the sleuth attempting to imagine the intention of the criminal, whereas when a Mongolian herder encounters one omen after another, these are conceived not as the traces left by a single intelligence but as separate moments whereby dislocations in the world are communicated to him.

13Most relevant for this paper, however, is the figure of the “sentinel”, which has become central to diagnosis of potentially catastrophic environmental futures that are invisible to humans, such as climate change, radiation, or emerging epidemics. Sentinels are animals or detection devices that provide the first warning signs – the canary in the mine that is extra-sensitive to gas, or the polar bear perched on shrinking ice. Frédéric Keck conceptualizes the sentinel as a “form/event”: it folds “event”, referring to an image that contains within itself both past and future, into “form”, defined as stabilized ways of raising problems, producing affects, and defining subjects of responsibility (Keck 2014: 47–49). The omen is in many ways similar to a sentinel thus defined: it involves both receivers and senders of a warning; like the sentinel it brings up the issue of perception (how do we judge if it is one or not?) and it produces a characteristic subjectivity; it involves the crossing of barriers; and it poses an uncertain temporality. In both cases, there is the predicament of the appropriate response when a threat has been made visible (Keck & Lakoff 2013). Still, if we consider the sentinel in the practice of the life-sciences, there is a difference: sentinels are like symptoms, in that they are located at the border of the suspected disease, and the connection between them and the impending effect is in principle scientifically traceable or conceivable virtually in statistical projections or phylogenetic trees; whereas omens are not an edge-index of some larger whole but are radically different from the impending threat. They are practices in an “ontology” (Descola 2013) in which relations are drawn on different principles. Before I describe this, it is appropriate to situate this article in relation to earlier work on Mongolian omens.

The context of the present study

14By comparison with other regions, there has been fertile discussion of omens in Mongolia, no doubt because the phenomenon is important to people there. In an early paper on Buryats (Humphrey 1976), I made a critique of Roberto da Matta’s structuralist analysis (1970) of omens among the Apinayé, arguing that it is overly reductionist to understand omens as he did in terms of an opposition between two blocks, Nature and Culture, that events are only recognized to be omens in respect of the interests/anxieties of a specific recipient, and that the connection between the omen-event and its eventual consequence is not causal in any simple way. I suggested that discursive explanations – in effect the threads across the “void” mentioned above – are created in multifarious ways, of which my article discussed six. Here I mention just one of these social temporalities that link unrelated things – synchronicity – in order to explain why I saw such relations as non-causal. Mongols have the idea that the arrival of migrating birds should coincide with the time when the herds begin to give their milk in spring. Therefore, the springtime sight of geese (or for that matter a hoopoe) flying in from the south is a good omen, because it heralds what should be the synchronic appearance of the longed-for milk. My paper received a riposte from Roberte Hamayon (1978), who made some just criticisms but basically defended the Nature/Culture opposition, arguing further that omens have an intrinsic sign value and that they are causal in the sense that the sign entails its consequence, to which I replied (Humphrey 1978). After this debate, it was not until the 2000s that there was a flush of excellent monographs, many of them mentioning omens though without focusing on the topic. Three publications with more direct relevance then appeared: da Col and Humphrey (2012) on conceptions and practices of fortune and luck; Empson (2006) on the political and didactic role of “prophets”, exceptional people who may deploy omens among other occult phenomena to influence a population; and my own study (Humphrey 2013) on the management of fear in Mongolia, one section of which was devoted to the emotional significance of omen-events and the ways in which people use them in retrospectively reorganizing their understanding of their misfortunes.

15These existing studies provide the context in which this paper attempts something different: to understand omens as a specific framework for dealing with the enigmas of life in the world and the implications of this for social relations and moral judgements. In Mongolian regions, the omen occurs in an utterly different cosmological configuration from those of Christianity and Islam; instead of a single divine cause, the cosmos is made up ab initio of separate and ontologically different elements (Humphrey 1996: 76–77; Pederson 2007) that can be associated in diverse ways. Even one element, let us say a mountain or a human being, has multiple components and these are combined in heterogeneous ways, including physical continuity or reproduction, but also metamorphosis, rebirth, and a-temporal reiterations. An example is the herdsman who said, “As for myself, I consider myself to be the soul of the mother of my father” (Oberfalzerova 2006: 20). In general, this mode of cosmological thinking is “analogical” in Philippe Descola’s terms: that is, composed of self-sufficient entities whose intrinsic differences are retained but are “attenuated by multiple relations of correspondence and cooperation” (2013: 279) and in which cosmos and society are equivalents, almost to the extent of being indiscernible (2013: 268). Mongolian regions provide ample evidence of the “obsessive particularism” of analogism, in which people attempt to recompose entities in always incomplete networks of partial similarities (2013: 376). This paper suggests that the omen not only has an affinity with this way of thinking but can also be considered a technique characteristic of it. And further, examination of the omen as a form implying certain practices and concerns opens dimensions beyond the purely conceptual recomposing of entities: the real-life interface in which cosmological spatialities/temporalities are recognized (or not), the delineation of permissible and impermissible kinds of boundary-crossing, and the revelation of ethical dilemmas – as I hope will become clear in the ethnography to follow.

Interference patterns

16As an example of spatio-temporal dynamics, let us return to the hoopoe and the professor. We can picture this omen as occurring at the point where the field comprised of the hoopoe’s activities crossed the quite different field of the professor’s activities. It would be quite possible to represent the event by the old image used by Mary Douglas (1966) and Edmund Leach (2000: 231), where the overlapping section becomes ambiguous and hence a source of purity/danger.

17But if we are to understand how omens disrupt social life, the idea I am trying to grasp is better illustrated by the notion of “interference pattern”. In physics, this is what occurs when two energy sources of light, sound, or liquid, and so on, produce waves that superimpose to form new patterns of greater or lesser amplitude and intensity. The advantage of this as a heuristic – and of course, it is only intended as an initial aid to my explanation – is (a) that waves, like birds’ and humans’ activities, are in motion, and (b) that waves of different substance deform when they meet one another, thus producing unexpected effects. Further, even in the case of two similar waves, both let us say sound waves of equal force, their superimposition results in numerous contact events, and not just one static space as in the Douglas–Leach image:

18This is a distant analogy of course because the human/bird activities I refer to do not have a wave structure, and furthermore they are composed not of homogeneous, equally weighted energy sources but of variegated components. Still, the idea is helpful in clarifying that when such different and jaggedly composed orders collide there may be several points of contact of varying intensity, as Illustration 2 suggests; as I will show later, omens affect not just one person but whole social arrangements.

19With this in mind, we can picture one “energy source”, that of the hoopoe, as consisting of an imagined pattern of nesting, laying eggs, feeding, hopping around, calling, flight, migration, and so forth, all having their proper space and time, which for the Buryat is that of the dry steppe and the summer months. The other “energy source”, the professor, is equally complex: she has a number of patterns, one of which consists of visiting her countryside dacha, its kitchen, veranda, garden, fence, and so forth. The omen happens when a forceful element of the hoopoe ensemble intrudes into the centre of the professor pattern, jolting it into a new shape. The professor was a powerful senior academic with a strong personality and several enemies. The immediate consequence of the hoopoe’s visit was that suddenly, in Buryat eyes, the ensemble of her elements was transformed into vulnerabilities: her problematic past was remembered, she might die or become ill, her dacha might burn down, her political status might be attacked, and so forth.

20Let me pursue the spatial aspect further. Birds and humans are co-present in the world, but in principle un-joined to one another – they may be in relation, but that relation is one of discontinuity. Mongols tend to exemplify conceptual separateness in real spatial terms. This means that each notional element, which may be conceived at different scales, is understood to exist in its appropriate space that should not be violated by others. Touching another person, for example, can only correctly take place as part of a pre-understood scenario of interaction; if you accidentally touch your foot against a Mongol person’s foot, you should immediately offer your hand in apology/friendship, as even a mere grazing is a violation of his/her spatial order. If you sit in a car seat still warm from a previous occupant, something of their bodily presence will seep into you – which is felt to be polluting. Understanding these everyday norms enables us to picture a rough division of space into “mine”, “no one’s”, and “theirs”. If we return to our case, a distancing zone should obtain between the hoopoe’s steppe domain and human settlement. Let us say the professor ventures into it: she exits her garden gate and goes for a walk, and she comes across a tiny element of the hoopoe ensemble, a black and white feather. She would not regard this as an omen (if she noticed it at all), as the pastures are a neutral space intermittently shared by people and birds. But if she came across a hoopoe nest, she would have already strayed into its space, where it is absolutely forbidden for humans to touch the nest, or take or destroy the eggs. In fact, this is a strict injunction with regard to all wild birds. Birds can be dangerous adversaries. With their “horn mouths” (beaks), which Mongols find unpleasant, they are apt to issue curses (haraal) that cause untold misfortune to humans.

22To see how this relates to cosmological thinking, we have to imagine this situation multiplied by the vast number of different beings thought to exist in the world.

Crossing the void: the omen of the marmot man

26There are several interesting points that can be derived from this case. First, the tiny man was assumed to be real and the driver was not credited with any supernormal abilities of perception. Second, although the incident was taken to be a bad omen – a human, but the wrong size, and behaving like an animal – the subject to be afflicted was not the driver in particular; he was merely the conduit. This point can be related to the two separate cases of perceiving “little humans” described by Grégory Delaplace and Rebecca Empson (2007), where in each region it was a daughter-in-law who both saw and dealt with the strange tiny creatures. The resulting effect, harm (in one case) and good (in the other), came not so much to the daughter-in-law herself as to the whole family. What Delaplace and Empson emphasize is the interstitial place occupied by the daughter-in-law in social organization – between her natal and her husband’s families – which gives her the possibility of alternating perspectives on familial relations (2007: 207) and makes her a suitable conduit for transmission of the omen. Perhaps it can be deduced that just as the daughter-in-law moves between two spatially separate families, the lorry driver has a similar intermediary role at a wider scale, since he pursues the mobile profession of moving between towns and cities. At any rate, it was these geographical dimensions that provided the subject of the omen: as Mongols said, it was feared that what the marmot man omen presaged was that the city itself, or at least the district of Zaisan, was becoming a güideltei gazar – a haunted place running with weird creatures.

27A very small person in Mongolia is known as a dwarf (odoi) or marmot person (tarvagan hün). At one level, people say that they use the epithet “marmot” simply because the person is unusually small, marmot-sized. But maybe there is more to it than that. For a start, the origin of the diminutive size can be derived from a wrongful mixture of natural elements.

Cosmological imagination

30In Mongolia, human knowledge about cosmological disturbances is not sequestered in a mythic time-warp, but is constantly added to in the present. One main source is the genre of “it really happened” (bolson yavdal) stories. These appear in newspapers, radio broadcasts, blogs on the Internet, social media, and so forth, and can spread rapidly in social assemblages. They are especially popular among young people, who sit around and discuss them as extraordinary but real happenings. Their temporality refers to something that has just happened, and the stories are always supplied with names, places, dates, and realistic starting scenes. Bolson yavdal are exciting because they tie the everyday to the unknown dimensions of the cosmos – to what really might be the case, given the strange and improbable things that happen to people. The story of the lorry driver at Zaisan is typical of the genre, and the rush of crowds to spot the marmot man shows that many people accept such eye-witnessing as valid. Bolson yavdal bring into play all kinds of animals, ghosts, zombies, demons, metamorphoses, and transmigration of souls, but to keep some continuity in this article I now cite one concerning marmots.

32If we think of this story through the analogy of the interference pattern, we see how Lhasbat’s entire family was shattered (above all by the wife’s transmogrification) when he ignored the omen and intruded murderously into the marmot domain. But the tale is revealing also of Mongolian suppositions about human domestic order: that its members are differentiated and do not share the same attitudes, that nevertheless a patriarch will prevail over a wife, that one person’s act will affect the others, and, more cosmologically, that a family has a provisional character and is vulnerable to merging with animals and a loss of humanity. Meanwhile, the event also transformed the marmots. This is disclosed by the re-configuration of the marmot population into a kind of society – they are no longer haphazardly encountered game but achieve the unison of a human-like tragic chorus, bewailing their persecution.

Omens that turn on the self

35As in the previous story about the marmot, the omen expresses a clash between two histories, but in this case the moral faults on the human side are most clearly exposed. Consider the following: the man who could not resist picking up and wearing the jacket, the lie he told his wife, the faithful dog that barks to warn the household that something is terribly wrong, the misrecognition of the two omens, the wanton killing of the dog, the disaster befalling the daughter and son, the recourse to a lama, and the use of fire and sanctified water to restore things to their proper order. The story assumes readers who will be able to decode the morality and relate it to a regrettably common family scenario with which they might identify: husbands who are thieves and liars, wives who turn a blind eye, profitless male violence. Yet the message in both omen-stories is that these people have it coming to them anyway – otherwise, why would the weird striped marmot/haunted jacket have presented themselves to male protagonists? The initial human failings are very ordinary ones: greed for delicious marmot meat and picking up something from the street. Perhaps Mongol thinking concerning this situation would be less about moral faults and more akin to what Ghérasim Luca wrote about the “objectively offered object” (an idea that I borrow here at some remove from Luca’s own argument that such an object, with its weird attraction, allows the introduction of an active collective unconscious into conscious relationships), namely to see the omen as an object not merely found but freely offered to a person by the world/cosmos. “For a found or made object to be transformed into an offered object, and for it to be able to transform its nature in line with the new relationships established in the interior life of the individual seeking a new balance between the internal and external, the setting-pretext to this transformation must have an interpretative value that is, if not always negligible, at least very limited” (2008 [1945]: 26). In line with this, we could see the omen form as providing such a setting-pretext. Like the “offered object”, it has the fortuitous character of an encounter brought about by objective chance, yet its actuality as an object/animal/bird delimits chance in a dramatic way through its lyrical and emotional appearance, through its interpretation, and above all through the total transformation it imposes on relationships between people (2008 [1945]: 31). If the fundamental attitude to an omen is not “I found it” but rather “it found me”, the real fault is not to recognize it as an offering, not to respond.

36This paper has tried to explain omens as a particular technique of self-transformation, in which both human subjects and “things” of the world are assumed to be active agents. As distinct from interpretations that would see omens as delusional, or imposed by a religious theology, the argument here is that omens make good sense for people dependent on, and constantly interacting with, unpredictable others, and surrounded by the classifications, signs, boundaries, and prohibitions that are their means of imposing order on phenomena. These very means, however, imply the existence of escapees from ordering. Understanding the device of the omen as cosmological sheds light on the underlying philosophy. For omens are not just escapees, unexpected anomalous events such as the flight of the hoopoe into a forbidden spot. Recognition of such a moment as an omen is at the same time to recognize the incorporeal aspects of all things – that they cannot be comprehended separately from their time (this moment as a point in their timing), their places, their singularity, and whatever forces impelled them to this point.

I am very grateful to Baasanjav Terbish, Sayana Namsaraeva, and Uranchimeg Borjigin for their help in preparing this paper, to Giovanni da Col, Laurent Legrain, Vera Skvirskaya, and other participants in the “Hiccups” workshop for comments on the first version of this paper, and to the reviewers for Terrain for their very helpful guidance in revising it.