Possibly Oriental Elements in Slavonic F PDF
Possibly Oriental Elements in Slavonic F PDF
Possibly Oriental Elements in Slavonic F PDF
OF LANGUAGES
AND LINGUISTICS
Dedicated to Marek Stachowski
on the occasion of his 60th birthday
Edited by
Michał Németh
Barbara Podolak
Mateusz Urban
Kraków 2017
Reviewers
Jens Peter Laut (University of Göttingen)
Georges Pinault (École pratique des hautes études ♦ Paris)
Edited and proofread by Michał Németh, Barbara Podolak, and Mateusz Urban,
with the assistance of Tomasz Majtczak, Dariusz Piwowarczyk, and Kamil Stachowski.
Laid out and typeset by Kamil Stachowski. Final proofreading by Tomasz Babnis.
ISBN 978-83-7638-861-8
The two creatures in the title have attracted the attention of ethnographers and
etymologists alike for more than a century now, resulting in several theories,
more than twenty etymologies, and no consensus. The present paper evaluates
these proposals and adds to them yet another one. It also presents linguistic
and extra-linguistic data that strengthens some of them and weakens others.
The proposal favoured by the authors is presented in more detail, and with new
supporting evidence.
etymology, Slavonic, Turkic, vampire, vampirism
0. Introduction
The vampire, as we see him today, has two defining properties: he is undead,
and he drinks blood. Both ideas are millennia old and can be found in entirely
disconnected cultures all over the world; their combination is less common but
*
This paper was originally intended as a translation of K. Stachowski (2005). As the work
progressed, however, it possessed a co-author, was updated, expanded, restructured, and
often rephrased, so that the whole grew to nearly three times the size of the original.
The final conclusion, while different in two details, remains essentially the same.
Several scholars contributed to the creation of this paper in smaller or greater ways.
We want to express our particular graditude to Michał Németh, PhD (Cracow, Mainz)
and Mateusz Urban, PhD (Cracow) for their dedication and help. All the remaining
errors and shortcomings, needless to say, are ours alone.
644 Kamil Stachowski & Olaf Stachowski
certainly not rare. As for Europe, Graeco-Roman Empusae, Lamia, striges, and
Gello are cited as the oldest known implementations; accounts more similar to the
modern image can be found in 12th century Anglo-Latin literature under the name
sanguisuga (Olivares Merino 2006). The term vampire did not appear in Western
Europe until the 18th century (reports of it featuring in the name of a mediaeval
English poem are not true; see Olivares Merino 2005) when cases of alleged
vampirism were reported in East Prussia and the Habsburg Empire, and kindled
the general interest. Bram Stoker’s Dracula of 1897 sealed the 19th century image
of the creature and secured for him a permanent place in the European folklore
and popular culture. More recently, Stephanie Meyer’s teenage series Twilight
(2005–2008) fueled another spike in his popularity.
The name vampire came from the Slavonic languages, in which it appears in
a host of phonetic variants most of which are similar either to wampir or to upiór.
The Slavonic beliefs can be traced back to around the 10th century but those early
vampires were not the aristocratic, elegant, lofty creatures that we know today.
In their early days in the Slavonic folk tales of the Middle Ages, vampires were
probably body-possessing evil spirits rather than actual persons, and their image
was certainly more down to earth, gruesome, and more terrifying (see e.g. Novič
kova 1995). It is only in the 19th century that the word vampire, together with the
romanticized image it represents, returned to the Slavonic languages and created an
etymologically interesting pair with its largely forgotten forefather upiór.
Certainly, etymologists were not immune to the other-worldly allure of this
doublet, and set off more than a century ago to investigate the words and their
cognates. Many suggestions have been put forward during this long time, which
obscured both the path by which the words reached Europe, and their ultimate
source. The present paper aims to evaluate these proposals, to reinforce the foun-
dations of what we believe to be the most probable one, and also to slightly
complicate the matter by simultenously proposing yet another etymology.
*
In all probability, the two Polish words in the title, upiór ‘phantom, spectre’ and
wampir ‘vampire’, are eventually one. Similar pairs exist in most if not all Slavonic
languages, having entered them through very similar routes. The latter (wampir)
are relatively late borrowings from the languages of Western Europe (see fn. 2),
and of lesser interest to us here. The former (upiór), which was itself the source
for the Western European shapes, is the one whose origin has for more than
a century resisted the efforts of etymologists. The creatures themselves were
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 645
often confused in Slavonic folklore, and accordingly, the word vampire will be
used below to denote both simultaneously. But see also section 1.
The literature on vampires and vampirism is positively vast; a bibliography
collected by Hövelmann (2007) contains more than four hundred positions, and it
is definitely not exhaustive. This paper is concerned with just two of their names,
and it makes use of non-linguistic works only inasmuch as it is required for the
etymology; but even in this limited field, we were not able to find or obtain all
the relevant publications. A list of works that we are aware of but could not
access, is given in 2.3.
In its entirety, the linguistic aspect of the problem can be reduced to three ques-
tions: 1. How did the word spread across Europe?; 2.1. Is the word native to the
Slavonic languages?; 2.2. Or is it a borrowing, and if so, whence?; and 3. What is
the ultimate source of the belief? All of these questions have been tackled more
than once, and in more than one way. The focus of this paper is on question two,
because it is central to the problem.
As for question one, a partial answer to it will be incorporated in the final
conclusion, but only in the form of what appears to us to be the most probable
solution, patchworked from already exisitng analyses and without a detailed com-
mentary. The reader is referred to works such as Anikin (2007–), Bielfeldt (1982),
Kiss (1970), or Wilson (1985) where further bibliography can be found. This answer
is: Serbo-Croat > Hungarian, German (> Swedish, Finnish) > French > English,
Italian, Spanish.1 We will not follow the further wanderings of our word since many,
if not all of them, can be relatively easily explained as recent, 19th or 20th century
borrowings from English, French, or German.
An answer to question number three, it will be shown below, is not necessary
to resolve question number two, and it will be perhaps better left to a council of
ethnographers, historians, archaeologists, and researchers of myths. The belief in
vampirism is surprisingly widespread all over the world (see e.g. MacCulloch 1921),
1 English: Onions (1966) proposes two sources for the English word, French and German.
Given that the time frame is the 18th century, and that the English spelling with -e is
consistent with French and inconsistent with German, we lean towards the former.
So also DHLF and OED (1916? The entry vampire is marked as “not yet […] fully up-
dated (first published 1916)”); Finnish: SSAES; French: DHLF; German: Kluge (2011);
Hungarian: TESz, EWU; cf. also Kiss (1970) who derives Hung. vámpir from German;
Italian: DEI, Migliorini (1960: 582); Spanish: DRAE; Swedish: Hellquist (1970).
646 Kamil Stachowski & Olaf Stachowski
and it is not unlikely that it appeared independently more than once in more than
one place. Since, however, the etymology we champion here assumes a Turkic
origin of the European word, we will briefly return in section 1 to an apparently
forgotten or overlooked proposal of K. Moszyński (11934: 666, 21967: 660) to seek
the origin of the myth in China, for Chinese culture had indeed exerted a strong
influence on the neighbouring peoples in the first millennium, including the
Turks, and the possibility of borrowing cannot be rebutted too lightly.
Let us first acquaint ourselves with the ethnographic background (section 1),
and then examine questions 2.1 and 2.2 in more detail (section 2).
1. Ethnography
The discussion of the spread of the vampire myth as it is understood today in
popular culture is better left to linguists with access to data which are more
conclusive and definite in dating than scholars engaged in purely anthropolog-
ical research. However, the reasons behind this spread as well as the possible
ultimate source can be to some extent explored on the basis of pre-existing re-
ligious beliefs of the peoples discussed in the linguistic analysis as well as the
psychological significance of those beliefs. This section of the paper will then be
devoted to the comparative religious and anthropological discussion of the main
points of the text as put forward in the Introduction: the question of the concept
being native to the Slavonic peoples; if not native, the problem of the source of
borrowing; and the ultimate source of the belief.
*
First, the general background of the vampire belief will be examined, discussing
the physiological, psychological and religious bases for such a concept to arise
and spread so broadly. Proposed theories of the source of the belief, attempts to
rationalize medical conditions, will be discussed in relation to the anthropological
approach proposed as the general mode of explanation in this section.
There is a certain trend, especially among scholars from fields unrelated to
the study of cultures or religion, to propose naturalistic explanations for religious
phenomena and things that go bump in the night as rationalizations of naturally-
occurring conditions. Such theories are quite often presented in the popular media
due to their convincing appearance of empiricism, but under closer scrutiny are
in most cases deeply flawed methodologically. Their main weak point is the as-
sumption that phenomena described in religious or folkloristic literature are to be
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 647
understood literally and uniformly in the way they are described in the available
accounts, as a sort of primitive pseudo-science; possibility for the inclusion of
psychological realities, alternative states of consciousness, metaphor or just en-
tertaining fiction used to illustrate metaphysical concepts in the way of a parable
are usually neglected. The fact that mythologies have other functions apart from
the explanatory one, such as initiation, social cohesion or the formation of the
so-called Sacred Cosmos as it was defined by Luckmann (1967), are usually not
addressed at all. This creates a major misunderstanding of the source material,
which is usually extracted from pre-scientific or non-Western cultures and world-
views based on perceptions of reality different than those seen as obvious and
axiomatic by contemporary Western scientists. While it is true that some beliefs
in the supernatural may indeed stem from attempts at understanding the natural
world, it has to be stressed that the connection of a belief to its proposed natural
source is not a satisfactory model, for it does not provide any explanation of the
particular form of the belief, its social function, philosophical and theological
significance or its interaction with other cultures, such as syncretization or spread.
To provide a succinct example, while it is to some extent true that deities such
as Demeter, Osiris and Xipe Totec are, among other functions, corn deities and
their mythological cycles parallel the vegetative cycles, such a statement does not
provide any clue to the reasons behind their widely diverging appearances, other
functions, associated rituals etc. For these reasons, medical explanations will be
cited to provide the reader with both an overview of the diversity of opinions
on the vampire complex and a listing of phenomena that may have contributed
to the vampire lore; we do not deny that some of the physiological conditions
described below may have been adapted to the mythological complex, but we are
very reluctant to see them as either definite sources or satisfactory explanations
of the belief in question.
Barber (1996) suggested that the vampire folklore was an attempt to explain
the process of decomposition otherwise inexplicable to pre-scientific societies.
Some of the bodily characteristics associated with vampirism may be explained
physiologically. The swelling of the corpse is due to the accumulation of gas-
es in the torso, and the ruddy appearance (along with blood oozing from the
mouth and nose) is an effect of increased vascular pressure. The gases escaping
through the bodily orifices may produce a sound similar to groaning; and the
desiccation of the gums and skin results in contraction of the tissue, revealing
more of the teeth and nails than was visible when the person was still alive. It is
notable however that if the body was left in the grave, these conditions could
648 Kamil Stachowski & Olaf Stachowski
not have been noticed; also, while the process of decomposition was certainly
ill-understood for most of the human history, we must note that the exposition
to death and corpses was much greater before modernity. It seems disputable to
us if the people who lived through famine, plague and war would be shocked
to see common signs of decomposition of a body and feel the need to interpret
them in a supernatural manner. However, these conditions seem to us to form
a believable explanation of the proliferation of the vampire hysteria in 18th and
19th century Europe, where exhumations of individuals considered potential
vampires were conducted and the aforementioned signs were indeed interpreted
as proofs of the condition. For this reason, the “argument from decomposition”
should, in our opinion, be mentioned in the discussions of the modern spread
of the vampire belief, but not of its early formation and possible borrowing
between mediaeval cultures.
It has also been hypothesized that individuals who were buried alive due to
an erroneous diagnosis of death and later regained consciousness could be the
source of the vampire belief; the sounds heard in the vicinity of the grave could
be explained as their screams, and upon the exhumation fingernail marks would
be found on the coffin. The victim could also hit their head on the wood and
therefore have blood stains on their face (Marigny 1993). In general, this theory
seems to be, similarly to the previous one, applicable rather to the later spread
of the belief when exhumations were more common. It is nevertheless plausible
that such cases of the dead “coming back to life” occasionally could be interpreted
as supernatural, if only for the much lesser frequency of such occurrences than
the conditions put forth by Barber. This theory is however still inconclusive due
to the impossibility of exclusion of other factors contributing to the formation
of the belief.
Vampirism has also been explained as a form of folklore linked to the outbreaks
of certain epidemic diseases, causing clusters of deaths in families or communi-
ties (Sledzik, Bellantoni 1994). The reported cases come from New England and
are associated with tuberculosis in particular, which may cause blood to appear
on the lips due to the damage done to lung tissue (Barber 1996: 115). The cases of
Petar Blagojević and Arnold Paole are also cited. The earliest of those accounts
come from the first half of the 18th century, when the belief in vampirism had
already widespread and was causing mass hysteria. While the examples cited may
indeed be connected to the mentioned epidemic diseases, the belief in the vampire
complex had already been well-defined at the time, and therefore the hypothesis
tells us nothing about its early sources.
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 649
being however connected explicitly to sexuality and repression thereof, and less
to violence per se.
While the aforementioned psychodynamic approaches constitute only a small
minority in contemporary psychology in general, they tend to be the ones most
commonly used in psychological interpretations of cultural and religious beliefs
and phenomena, and for this reason we find few works on the subject of vam-
pires which stem from other psychological perspectives. While similar to the
medical approach, these models do not provide an explanation of the source
culture of the vampire; they do however serve a useful purpose in explaining
the wide popularity of the phenomenon both in modern times and the Middle
Ages, allowing us to see the linguistic data presented in this paper in a wider
anthropological perspective.
*
Next, the specific forms the vampire belief has taken in the Slavonic, Turkic and
Chinese folklore will be examined against the background of the belief complexes
most closely linked to the vampire mythologem. The views on death, the afterlife
and the concept of life-force contained in blood are of greatest interest in this
respect; however, since those themes are crucial to most religious cultures of the
world this analysis must rely on a very general outline in an attempt to facili-
tate the understanding of the cultural context in which the discussed linguistic
processes took place.
As with all matters pertaining to the pre-Christian Slavonic religion, the
data is scarce and rather inconclusive, lacking any original written sources and
therefore being reconstructed from folklore, religious relicts, dispersed notes by
foreign chroniclers and comparative religious analysis. We will then proceed to
describe the vampire as it appears in Slavonic sources as well as the background
of Slavonic beliefs on funerary rituals and beliefs about revenants. Little is known
for certain about Slavonic pneumatology, the sources however seem to point to
a belief in at least two souls (Szyjewski 2003: 204) – and such a case would be
quite plausible in the context of other Indo-European religions, as well as most
traditional religions in general. This belief is an important part of the background
of the vampire belief – the power that animates the corpse to move and interfere
with the living is usually understood to be the animal soul (Szyjewski 2003: 205),
which explains why it is the people who died a violent death as well as deceased
sorcerers and transgressors that seem to be most affected by vampirism. Indeed,
among the Slavs a belief was widespread that the soul of a sorcerer or medicine
652 Kamil Stachowski & Olaf Stachowski
man is more prone to wander the earth after his death, sometimes becoming
a ghost or animating the corpse (Szyjewski 2003: 207). The funerary practices of
the Slavs seem to be composed of two earlier religious streams: cremation was
widespread, but even in pre-Christian times burial was also practiced (Szyjewski
2003: 204). This tradition of earth burial, however marginal, is important for our
case, because if the only known method of funeral were cremation, then the
belief in corpses rising to haunt the living due to an incorrect funerary ritual
or a confusing state of bodily decay would be rather difficult to spread, and if
so, the most probable conditions would be during wartime or plague; and such
conditions are not linked to the vampire in any consistent manner.
The vampire belief is quite well documented in the Slavonic folklore, however
it is sometimes conflated with other spirits of the dead. The Slavonic vampire
is always primarily said to be an undead being who sucks the blood of the liv-
ing to keep himself alive, leaving its victims apathetic and usually haunted by
nightmares. It also tends to be mentioned to possess two hearts and sometimes
two rows of teeth, both features being linked to the belief of the vampire having
two souls (Pełka 1987: 164, 166). A commonly reported trait is also a red or ruddy
face and neck, a hirsute appearance or an abnormally large head (Pełka 1987: 165).
In these folkloric accounts it is also mentioned that a vampire is not always ma-
levolent and may come to his house to help or visit his family (Pełka 1987: 169f)
or that a dead mother may try to come and nurse her children (Pełka 1987: 166).
Those concepts would remind us of the psychological interpretation of the phe-
nomenon as stemming from projection, not necessarily negative in character.
Among non-human shapes the vampire can take are a multitude of zoomorphic
forms (a ram, a dog, a cat, a horse or a bird among others; Pełka 1987: 167), or the
shape of a skeleton. The accounts mentioned by Pełka are rather varied and seem
to have little order to them other than that of local traditions; vampire stories
also overlap significantly with stories about ghosts or various other harmful
spirits (especially werewolves; the two mythical complexes have been confused
at various times in history (Petoia 2003: 18, 23–24), the only core trait strongly
and consistently linked to the vampire being his blood-drinking. This particular
choice of the most salient feature, it will be shown below, contrasts with the
Turkic belief complex.
The Turkic obur is described as a malevolent being, very large in posture and with
a very big head, tailed; able to fly and breathe fire, and shapeshift into different
creatures, especially wolves, dogs, cats and other animals, or a ball of fire. It is
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 653
marginal concept of the creature, the belief in qi being still widespread and
commonly recognized in the East Asian cultural sphere. While the draining of qi
could easily be culturally translated to the drinking of blood during a borrowing,
we should note that the blood-drinking mythologem is quite marginal to the
Turkic obur, and so its transmission to the Slavonic religion from the possible
Chinese source seems to be a far-fetched hypothesis. Its appearance also differs
dramatically; since the jiangshi is an animated corpse, it is strikingly dissimilar
to both the Turkic fire-breathing, flying giants, and the Slavonic two-hearted
revenants. Shapeshifting, the common trait of the Slavonic and Turkic vampire,
is also completely absent from Chinese accounts. For these reasons we consider
the hypothesis seeking the ultimate source of the vampire belief in China to
be very implausible and, unless new data explaining such a drastic change
of crucial characteristics on the Chinese-Turkic borrowing route, are found,
scientifically worthless.
*
In conclusion, on the basis of anthropological data presented in the above over-
view, we would propose that the word, being borrowed from Turkic languages as
described in section 2, has been appropriated to earlier Slavonic beliefs concerning
the revenant dead. The essential trait of blood-drinking seems to have gained much
more weight in Slavonic regions compared to the aforementioned Turkic sources
where it is present but not central. Such a syncretization is plausible also from
the historical point of view. Contacts between various Turkic and Slavonic tribes
have continued for centuries and resulted in a number of cultural and linguistic
borrowings. Tryjarski (1991: 45–48) uses three pages just to list the similarities
in burial customs, including such items as fear of the dead, together with the
will to mollify or satisfy them, the practice of binding the corpse, and of piercing
the corpses of people who died an unnatural death. While those customs are not
shared between all the Turkic and all the Slavonic peoples, and similar practices
are in general relatively widespread in worlds’ religions, they are certainly not
universal and should be taken into account as suggesting a common cultural
background of the discussed peoples which allows for a deeper religious exchange.
The hypothesis of possible Chinese source is highly implausible, however, due to
the extreme differences between the jiangshi and obur that appear to be impossible
to reconcile on the basis of information available to us. Further research into the
obur, especially its early forms and its possible roots in Turkic shamanism could
also show more light on the topic in the future.
656 Kamil Stachowski & Olaf Stachowski
2. Etymology
The multitude of ideas that have been offered as explanations of the origin of our
word, results from its wide spread across the Slavonic languages (and from them
across the entire Europe, and then, the globe), and of the many forms in which
they appear, some of which are impossible to justify through regular phonetic
changes (see e.g. Kiss 1970: 84). The most common ones are:2 Blrs. úpir, úpiŕ |
Bulg. vъpír, vapír, vepír, vampír | Cash. łuṕi | Cz. upír | Pol. upiór | Russ. upýŕ |
SCr. vàmpīr | Slvk. upír | Ukr. upýŕ | and many dialectal shapes, sometimes as far
removed as Bulg.dial. ljapír, Pol.dial. wąpierz, or SCr.dial. lampijer which, with
one exception, will be ignored here as they seem to be inconsequential for the
ultimate source of our word.3 All mean ‘spectre’, ‘vampire’, ‘ghoul’, or some
other fantastical creature, at times even ‘sorcerer’ or ‘werewolf’ (see e.g. Hobzej
2002: 144f; Leschber 2013); it seems that the ideas became somewhat confused
by various Slavonic peoples (Moszyński K. 1967: 658; also Leschber 2013: 191).
The common theme tends to be that of a dead person rising from the grave and
harming the living in one way or another. The Russian shape is often linked with
ORuss. Upiŕ, dated 1047, and cited as the oldest attestation of our word; this might
likely not be the case; see 2.4.1.
Below, presented and discussed are all the previous etymological proposals
that we were able to find, those that consider the word native in 2.1, those that
2 We ignore here those shapes that are apparently missing from native dictionaries
and can only be found in etymological papers or in dictionaries of other languages.
Some may be phonetic variants, dialectal or archaic forms, but many are, we are
afraid, simply erroneous citations or misprints. We also omit here forms belonging
to the second, romantic stage of the history of vampires, as they are but late, 18th and
19th century Rückwanderers from Western European languages (probably French or
German), such as Blrs. vampír | Cz. vampýr | Pol. wampir | Russ. vampír | Slk. vampír |
Ukr. vampír, i.e. Northern Slavonic forms beginning with vam-. In the same category,
only considerably more delayed, are modern Turkic shapes of the type vampir; see e.g.
Sariyannis (2013: 199).
3 See e.g. Cooper (2005: 252), Dukova (1997: 97f), Dźwigoł (2004: 67), Hobzej (2002: 142),
or Podgórska, Podgórski (2000). In particular, Cashubian has an unusually large number
of forms associated with our word, e.g. håṕi, ńełåp, połap, uṕȯn; see Popowska-Tabor
ska ([1999]), Rytter (1986: 123f), and SEK s.v. łuṕi. In some cases, the similarity is so
faint that one might even consider questioning that those words are indeed direct
cognates of our pair; see fn. 4.
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 657
assume a borrowing in 2.2, and several of which we only know second-hand in 2.3.
In 2.4, we attend briefly to several loose ends such as dating, and semantic and
phonetic peculiarities.
2.1. Native
This subsection discusses seventeen attempts at explaining upiór &c. as a native
Slavonic formation. They are presented in the chronological order and followed
by a short summary with a list of proposed reconstructions in 2.1.18.
One, rather important, general remark that pertains to all of the native pro-
posals and will not be repeated seventeen times below, is the question of Slavonic
burial customs. More about this in 2.4.1.
would be called ‘non-birds’. This is one of its two main weak points. The other is
that the meaning one would expect from a combination of the privative *ǫ- with
*per- ‘to fly’ is actually *‘flightless’ rather than *‘non-bird’.
as “Old Czech” which, given that the beginning of Old Czech is the 14th century
(see e.g. Horálek 1958: 7), seems rather too late. Overall, this proposal does not
seem to be particularly plausible.
2. *ǫpurъ, *ǫpirъ, *ǫporъ; 3. *upenъ; and 4. *op-, the latter two based solely on Cashu-
bian variants. Eventually, this number is reduced to five, of which three, Rytter says,
have not been preserved anywhere outside of Cashubian. In fact, there probably
exists a different explanation for those words; we are not certain that they even
belong to the same etymological family.4 At any rate, Rytter’s conclusion not only
raises doubts about what may seem like her disproportionate reliance on Cashu-
bian, but also does not in any way explain the occasional nasal element.
Rytter reaches equally original conclusions regarding the semantics of the
shapes she reconstructs. Their Indo-European phonetic equivalents would be **up-,
**ep-, **op-/**ōp-, and **ap-/**āp-, all with the meaning ‘water’ (after Pokorny
1949–1959: 51–52, 1149); added to them would be the archaic formant -r/n- with the
general meaning ‘to rise, to emanate’. Next, Rytter argues that the so reconstructed
nominal type is not too unusual for the Slavonic languages, if one accepts that it is
continued in various geographical names, appellatives connected with water, names
4 See e.g. Dźwigoł (2004: 69), Popowska-Taborska ([1999]), Rytter (1986: 123f), SEK
(s.v. łupi), Podgórska, Podgórski (2000). The lack of final -r in Cashubian forms is
puzzling. There are at least three ways to explain this absence.
Firstly, according to AJK I (27, s.v. opi; 95), Cashubian words should be derived
directly from some Germanic shapes, such as LG api ‘ape, monkey’, whence also OCz.
opice, ORuss. opica id., and related forms – among which AJK also counts Ukr. ópyŕ
‘vampire’. This last word, should it indeed prove to be of Germanic origin, could very
much complicate the study of the etymology of upiór &c. But for semantic, phonetic,
and historical reasons, we are more inclined to see in it a variant, perhaps dialectal,
of the ‘proper’ upýŕ id., and to derive it from Turkic.
The second, new possibility lies in MLG open ‘to kiss’ (> Plb. öpåk ‘a kiss’, see
SEJDP). The semantic connection between ‘to kiss’ and ‘vampire’ seems to be entirely
natural, if the latter were a blood-sucking creature, i.e. one that looked as if it kissed
its prey; see also fn. 13.
There is also a third possibility, suggested in Popowska-Taborska ([1999]: 348),
which derives the Cashubian word from the root -pir- and assumes that the -r was
quite simply dropped, giving the root an adjectival character. Phonetic changes in
inlaut are explained through a secondary association with an interjection, łup, łup,
łup! which was supposed to imitate the steps and the voice of the vampire (the shape
łupi), or with the verb łapac (the shapes ńełåp and połap). The changes proposed for the
inlaut appear to us to be considerably more plausible than that for the auslaut.
Overall, it seems that a detailed analysis of the Cashubian forms in unnecessary
for the discovery of the eventual origin of the pan-Slavonic word for ‘vampire’, and
as such, lies beyond the scope of this paper.
662 Kamil Stachowski & Olaf Stachowski
of mountains, hills and hillocks, and some hydronyms. She proceeds to list a large
number of examples; a detailed discussion is beyond the scope of this paper but we
should like to note that, while they all do indeed contain the element (-)Vp(-), it is
only in some of them that it is followed by an r or an n, and their meanings are
sometimes connected with ‘water’, sometimes with ‘rising’, sometimes with both,
and sometimes with neither – e.g. Wapno (lit. ‘lime’), a village in Greater Poland
known for its lime and gypsum quarries. We must also object to Rytter’s practice
of defining village names as ‘a village situated on a hill by a stream’. It has always
been much more a rule than a special case that settlements were located close
to a source of fresh water and, if only possible, on an elevation, and this fact can
hardly be expected to have become the motivation for the village’s name.
Next, Rytter tries to find a connection between her reconstruction and eth-
nographic data.
She argues that water played a very significant role in the Slavonic mythology,
that it was often considered to be the home of spirits and souls, and almost the
essence of the world of the dead. She mentions the opinion of Tomicka, Tomicki
(1975: 104f, 121f), that it is not only the souls of those who drowned that would
become topielce (water spirits that lure people into swamps, lakes, &c.), but that it
could be the soul of anyone who died a premature, unnatural death. K. Moszyński
(1964: 679) also remarks on the importance of water, though he does seem to be
more reserved in the matter. As for topielce, he simply writes about daemons that
are “quite meaningfully called simply topielce or topce”.5
Further, Rytter asserts that dialectal meanings show that upiory are not nec-
essarily bloodthirsty creatures, or even material beings. They can be daemons
inhabiting secluded places, most frequently ones connected with water. This last
piece of information is difficult to confirm because the meanings she adduces do
not in fact mention water at all. It also needs to be pointed out that the rituals she
describes, save one Cashubian one and one whose location she does not specify,
are not actually related to upiory.
Overall, Rytter’s proposal is more an exercise in teleology than etymology
and must be discarded as rather implausible.
Possibly a similar idea, at least so far as the meaning of ‘rising’ is concerned, was
put forward by Trubačev, though only in a less than rudimentary form; see 2.1.14.
6 The author of this proposal also mentions the shape *ǫpyŕь. The wording is not entirely
clear, but the forms with -jь appear to be the preferred ones.
7 Where *ǫ- is a regular phonetic variant of the verbal prefix *vъn-.
664 Kamil Stachowski & Olaf Stachowski
Maybe another theory or idea is closer to the truth. In 1998, the Musée cantonal
d’histoire naturelle in Sion (Switzerland) published a huge two-volume study
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 665
on bird and bat names in Europe. […] the second volume puts the names into
paradigms with a similar phonetic structure and compares them with other
words of the same structure. Under the current number 6.2.24 names with the
consonantic structure “b-mp”, “p-mp” are listed. […] Here we will find our
“vampire” and some of its regional variants as for example “vapir”. […] And it
is really astonishing to recognize how many words describing elements of
the folkloric vampire figure have a similar phonetic attitude. For example the
word flame: “vampa” in Sardinia, “vapë” in Albanian […] Or the word steam:
“pampore” in Romanian and “vapore” in Italian. […] The butterfly is called in
Galician “pamuprriña” and in Basque “pinpirin”. […] Do we have here may-
be something like a genetic relationship of phonetic structures? We must be
aware of jumping to conclusions. Nearly all of the languages mentioned above
are are [sic] of Indoeuropean origin so we have to expect a certain similarity
of many terms. But is that all? Isn’t it possible that “vampire” has something
to do with the words for steam and flame and butterfly? We have to hope that
the editorial board will publish a volume with explicative articles. And again
we have to wait…
2.1.18. Summary
Of the above, the most plausible appear to be proposals 2.1.6 (*piriti, ‘bulging,
bloated, swollen’), 2.1.9 (*vъ-pirati, ‘that which pierces in’), and 2.1.15 (*ǫ- + *per-,
‘unrotten’), the last two more so than the first one as they appear to rely on fewer
uncertain assumptions.
Should either of them be accepted, two major issues would remain to be
explained. One, easier to dismiss, is the remarkable similarity of *ǫpiŕь to the
Turkic word opyr &c.; see 2.2.3 The other, which we believe to be more resilient,
is the question of Slavonic burial customs; see 2.4.1.
Perhaps the most probable explanation is one that combines a foreign origin
with native reinterpretation, and thus avoids the chronological trap. See the final
conclusion in 3.
12. *ǫpyrь (2.1.2, 2.1.3, 2.1.10, 2.1.13, 2.1.14; Korsch 1886; Vasmer 1953–1958, 1986–1987;
Budziszewska 1983–1885; Lukinova 1984: 123; Trubačev [1994]: 424; Kurki-
na [1994–1996]: 199; Snoj 1997, 2003: -; Rejzek 2001; Tjapkina 2006: 105;
Šapošnikov 201011);
13. *ǫpyrъ (2.1.13; Gluhak 1993);
14. *pyrь (2.1.2; Ilьinskij 1911);
15. *upirь (2.1.2–2.1.4; Vaillant 1931; Rejzek 2001);
16. *vъ-pěrъ (2.1.8; Machek 1957, 1968).
2.2. Borrowing
We are aware of five attempts to find the source of upiór &c. outside Slavonic. Three
of them do not seem to be highly plausible, one is more so, and one considerably
more so. Below, they are presented in the chronological order and followed by
a short summary in 2.2.7.
11 Šapošnikov (2010) supports this reconstruction but already in the meaning ‘vampire’,
without a further deconstruction.
668 Kamil Stachowski & Olaf Stachowski
2.2.3.1. In its original form, the Turkic path leads to Tat. ubyr ‘witch; evil spirit’
(we ignore here the various forms that can be found in the literature, for they are
all clearly this word, only in a strange, or simply erroneous, version). The word
has a fairly straightforward etymology on the Turkic ground: < PTkc. *ōp- ‘to suck,
to swallow, to greedily catch with mouth’ + aorist participle -(V)r, lit. *‘that which
sucks, sucker’ (see e.g. Sevortjan 1974–; Şirin User 2010; and also Berta, Róna-Tas
2002: 58; and Róna-Tas, Berta 2011, s.v. àpol;12 see also 2.4.4 for an alternative
reconstruction). A combination, in one word, of the meanings ‘guzzler, glutton’
and ‘evil spirit’ is not unusual in the Turkic languages, see e.g. ič- ‘to drink’ →
Uigh. ič.käk ‘sot, drunkard’, then ‘daemon; vampire’ (Zajączkowski 1932: 100, Şirin
12 In both works authored by Berta and Róna-Tas, the root *op- [sic] is linked with *öp-
‘to kiss’, and suggested to be the eventual source of Hung. ápol ‘to look after someone
or something, to nurse’. Were this suggestion true, it would entail a slightly amusing
semantic correspondence, ‘vampire’ ≪ ‘to suck, to swallow’ ≫ ‘to look after’. But in
reality ápol, while attested for Hungarian with the meaning ‘to kiss’ (> ‘to caress’ >
1822 ‘to look after’; see EWU), cannot be derived from any Turkic language due to
the length of the initial vowel. Its older, 17th or 18th century shape is apol, with a short
vowel, and the Turkic reflexes point clearly to an original long one, see fn. 16. Indeed,
one might wonder to what degree Berta/Róna-Tas’s reconstruction of PTkc. *op- was
inspired by the short apol. TESz, EWU, and Gombocz (1912: 209) all speak against
a Turkic origin of the Hungarian word.
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 669
User 2010). Among the reflexes of *ōpyr are Tat. ubyr,13 Chuv. văbăr and similar
‘witch, hag (very fat)’ (Ceylan 1997: 176),14 Tksh. obur ‘glutton’,15 and many others,
see e.g. Radlov (1893–1911), Şirin User (2010), Ragagnin (2013: 64f), or Zaripova
Çetin (2007: 23f).
The idea is also sound from the point of view of semantics and ethnography.
Daemons are certainly not unknown to the pre-Islamic culture of the Turkic
peoples, and some of them display considerable similarities to upiory. More on
this can be found in section 1; see also Şirin User (2010) and Yaltırık (2013).
But in this form our proposal entails, as also do all the other ones, a neces-
sity to explain a large number of phonetic variants that our word has across the
Slavonic languages, not through regular developments because this is not possible,
but apparently through some conspiracy to contort it beyond recognition. We are
not aware of this reservation having been raised as yet – except by G. Rytter
(1986: 126; see also 2.1.11), who argued that any theory about borrowing is difficult
to accept in the light of the wide spread and old age of the notion of vampirism
throughout Slavdom. We fail to understand this reasoning.
Effectively, it is only Vaillant (1931: 676f) who presented an actual criticism
of the Turkic idea. His argumentation, however, is somewhat unclear and quite
impossible to accept. Using a key known only to himself, he divides the Turkic
forms into three groups: 1. Bshk., Tat. ubyr ‘(homme, femme à pouvoir de) sorcier,
sorcière’; 2. Chuv. wubur [sic] ‘démon qui dévore la lune ou le soleil’, Cauca-
sus-Karachai16 obur ‘être malfaisant, démon qui dévore les nouveau-nés’, Ott. obur
‘glouton’; and 3. Uigh. opur ~ obur ‘nourrice’. Next, Vaillant speaks of “l’unité des
faits slaves”, a grossly over-optimistic assessment, and rejects Deny’s proposal
13 The intervocalic voicing in Tatar requires further explanation because usually, Ta-
tar does tolerate voicelessness in this position. Nonetheless, the existence of this
derivative in Tatar, with and without the voicing, is beyond doubt. See also 2.4.4 on
Tatar phonetics.
14 With a regular phonetic development and a protetic v- + non-labial vowel in the place
of the original labial one.
15 With a regular Oghuz voicing of a stop following the shortening of an originally
long vowel, and with labial harmony. Regarding the length of the Proto-Turkic vowel,
see Dolg. uobū ~ ōbū ‘Bissen, Happen’ (defectively attested with -u instead of ū, see
Stachowski M. 1993: 243) = Yak. uobū ‘Biß’ < uop- ‘einen Bissen in den Mund neh-
men’ (Stachowski M. 1993: 243) < PTkc. *ōp-; see also Tekin (1995: 177).
16 A slightly surprising name, for we do not know of any Karachai spoken outside of
the Caucasus.
670 Kamil Stachowski & Olaf Stachowski
to derive the Turkic forms from a common root op-, up- ‘aspirer, sucer’ (making,
however, an exception for the Uighur forms) as he believes that this would be an
overestimation of the importance of the notion of vampirism which among the
Slavs, he continues, appears to be secondary or late, and derivative. We do not see
the logical connection here. Vaillant concludes (1931: 677): “Le mot obur, wubur
ressemble fort au persan awbār, obār ‘ravisseur’, verbe awburden ‘ravir’, qui sert
à désigner divers dragons ou monstres qui dévorent les hommes, les astres, etc.,
par ex. māhī-i-merdüm-obār ‘poisson qui dévore les hommes’.”, and does not
return to the remaining Turkic shapes.
We fail to see either the logic or the purpose behind Vaillant’s grouping of
the Turkic words. The phonetic similarity to the Persian word, while it indeed
exists, is coincidental and irrelevant, since Pers. oubār, oubāštan is as native to
Persian17 as Tat. ubyr &c. is to Turkic. There is neither basis nor need to take
borrowing into consideration.
Thus, the Turkic etymology rests for now in limbo, with multiple counter-
proposals but no serious counterargument.
2.2.3.2. There is also a newer variant of the Turkic proposal. In its original shape,
the idea was limited to the Tatar word ubyr. But this word is not a Tatar innovation;
it is a part of a sizeable family across several Turkic languages. Naturally, some of
the cognates have different phonetic shapes, and particularly interesting here is the
Bolghar branch, with forms such as Chuv. văbăr, vobăr, vubăr ‘evil spirit’ (Ašmarin
1994–2000), or vubăr, vybăr ‘witch, hag (very fat)’ (Ceylan 1997: 176).
In K. Stachowski (2005), which is the basis for this paper, it was proposed that
only Eastern Slavonic forms be derived from Tatar ubyr, while the Bulgarian ones
from ubyr’s Bolghar counterpart. This would shift part of the burden of explaining
the phonetic variation across Slavdom onto Turkic historical phonology, and so
at least in some cases free our etymology from the necessity to invoke arbitrary,
one-time phonetic changes ascribed, for lack of a better explanation, to taboo
and other irregular and unpredictable phenomena.
The idea to derive northern Slavonic shapes from a different source than the
southern ones is not new. It was already mentioned e.g. in Polívka (1901), but in an
untenable form which binds our word with OBactr. vyāmbura (see 2.2.5). During
the preparation of this paper, we learned that also a different, considerably more
defensible variant of this idea had been in fact proposed before K. Stachowski (2005).
Dukova (1997: 100) reports it was already Boev (1970: 905–906, unavailable to us)
who suggested Chuvash as the source of the Bulgarian word; she proceeds to support
his two-path solution, and let us too look at this proposal in more detail.
In practice, the history of contact between the Slavonic and Turkic peoples,
and also the phonetics of the Turkic shapes, leave us with but two possible sourc-
es: the Kipchak languages (among them, Tatar), and Bolghar. This variant of the
Turkic proposal assumes that our word was borrowed independently from both,
and spread across the Slavonic languages along two separate paths. Possibly,
a small amount of intertwining in the form of intra-Slavonic borrrowings would
need to be allowed for in order to fully account for all of the phonetic shapes
(regarding the most likely routes, cf. Boček 2010: 24f).
The northern path, in its core, is basically the original proposal (2.2.3.1).
The source is Kipch. *opyr (or perhaps *opyr ~ *upyr, see 2.4.4) ‘1. witch; evil spirit;
2. glutton’ (> Tat. ubyr ‘1. misfit, weirdo; changeling; werewolf; shape-shifter; 2. glut-
ton’) > Eastern Slavonic > Western Slavonic. Later, ephemerally, to French from
one of the latter two: 18th c. oupire, upire (NDEH).
The Kipchak word is not attested for this period, but it is maybe only because
the oldest written record of any Kipchak language is from the 13th century. It is the
Codex Comanicus, and it does in fact contain the word opmac ‘osculum’ (Kuun 1880,
though missing from K. Grønbech’s 1942 edition). It is also known that the root is
common throughout the Turkic languages, and one can assume with a relatively
high probability that a formation of the type *opyr should have existed.
The southern path assumes a borrowing from Bolghar, cf. Chuv. văbăr, vobăr,
vubăr ‘evil spirit’ (Ašmarin 1994–2000), or vubăr, vybăr ‘witch, hag (very fat)’
(Ceylan 1997: 176) > PBulg. *vъpъr (Boev 1970: 905–906) > Bulg. vъpír, vapír, vepír
‘vampire’ (shapes after BER) whence the word would spread across the Balkans.
As for the nasalization, Boev proposes OBulg. *вѫпъръ but Dukova (1997: 100)
sees the issue as unsolved; see 2.4.4 for some possibilities, and also Zoltán (2013b)
on the preservation of nasal vowels in Bulgarian dialects.
It should be noted that the assumption of a Bolghar etymon, with a reduced
vowel in the initial syllable, renders the Bulgarian alternation a ~ e ~ ъ immedi-
ately more understandable.
672 Kamil Stachowski & Olaf Stachowski
2.2.3.3. Of the two variants of the Turkic etymology, the newer one appears to
be more likely, in that it better accounts for Slavonic phonetics and thus partially
removes the necessity to assume a high number of later internal borrowings be-
tween various Slavonic languages. But the details are unclear. Most vital appears
to be the question of the time of borrowing, but this requires a slightly longer
commentary; see 2.4.1. Also potentially dangerous for this proposal is the only
dialectal form that this paper will look into more closely, which is Pol. wąpierz
with what appears to be a suspiciously ancient phonetic shape; see 2.4.3.
If the first syllable of *ẹmpuru [sic, -u pro -ŭ] was pronounced with a distinct
front vowel, as might be supposed, given that Romanian now has impur, then
a Slavonic borrowing […] would be expected to have initial ę-, with prothesis
leading to ję- (> ja- in Russian […]). (Cooper 2005: 263)
which is followed by the information that classical Lat. in- has actually two out-
comes in Romanian, in- and în-. Cooper had just said that it was the former in the
case of impūrus, but now he proceeds to fill two pages with an analysis of what
could happen if it were the latter: “it might then have resulted in the reflex ǫ-”,
leading eventually to *ǫpyrŭ/[*]ǫpirŭ | *vǫpyrŭ/*vǫpirŭ, and yielding u-/o- in Eastern
Slavonic, OBulg. *vǫ- (possibly later borrowed to Ukrainian dialects as vo- and
similar), &c. Essentially, all modern forms are to be explained by the alternation
*ǫ- : *vǫ-, either in Proto-Slavonic or later, and by internal borrowings.
On the whole, this proposal is not unlikely as such, but it does rely on a fairly
high number of assumptions, and entails a rather intricate web of borrowings
which, we believe, is quite unnecessary.
2.2.7. Summary
Of these five proposals, only two are in any way probable, the Turkic one (2.2.3) and
the Dacian one (2.2.6); the former substantially more so. It exists in two variants of
which the newer is more complex and also more plausible. This variant assumes
two independent borrowings along two separate paths. Some of the details remain
to be established; see in particular 2.4.1 and 2.4.4, and a summary in 3.
2.3. Others
Apart from the above, there exist several proposals about which we only have
second-hand knowledge. Some were mentioned without specifying the source,
674 Kamil Stachowski & Olaf Stachowski
sometimes without even crediting the originator, and some are in works which
are unavailable to us, and which we can merely report after other authors.
Publications that we found reported as relevant, or the ones we ourselves
suspect of being relevant, but which we could not access include: Bălteanu (2000),
Boev (1970; see 2.2.3.2), Burkhart (1989: 65f), Dmitriev (1962), Haefs (2001), Hock
(1900), Istrate (1987), Knobloch (1989), Kunstmann (1992), Memova-Sjulejmanova
(1981), Menges (1969/1970), Naylor (1983), Perkowski (1989), Schürmann (1990),
and Vakarelski (1969: 232f, 240).
Perhaps the oldest of those proposals about which we do have any knowledge,
is the one mentioned by Ralston (1872: 410), apparently as somebody else’s idea.
Şirin User (2010) and Cooper (2005: 260) attribute it to Afanasьev (Aleksandr
Nikolaevič?) but an exact location is never specified. It connects upiór with Lith.
wemptî ‘to drink’ and wempti, wampiti ‘to growl, to mutter’ < √pî ‘to drink’ with
the prefix u = av, va, assuming that “the characteristic of the vampire is a kind of
blood-drunkenness” (Ralston 1872: 410). See 2.1.17 for what appears to be a more
recent version of the idea.
Piger (1901) says in his review of Hock (1900) that our word is native to the
Slavonic languages, that it means ‘to suck, to suck out’, and has sound correspond-
ences with Greek and Hebrew, as was found by 18th century scholars. Unfortunately,
Piger neglects to specify the source of his knowledge.
Next, Kreuter mentions Haefs (2001), ridicules him and immediately dismisses
his proposal (Kreuter 2006: 57; by all means see 2.1.16 for his own idea). As for the
details, he only says that “the main part […] consists of a nearly word-by-word
rendering of […] Kunstmann […] 1992” (Kreuter 2006: 58). Kunstmann’s idea,
also after Kreuter, is that “the absolute origin of the vampire myth is the mantic
Greek god Amphiaraos (Ἀμφιάραος)” (Kreuter 2006: 58). The name would be
borrowed by the Slavs twice; for the first time, in the seventh century, yielding
*ǫpyŕ (≫ Cz. upír, Pol. upiór, Russ. upiŕ/upyŕ); and for the second time, after the
seventh century, yielding the shapes with vam-. Apparently, Kunstmann is also
not very clear about the difference between upiór and wampir: “Bei Serben und
Kroaten ist vampir zum Beispiel die schriftsprachliche, upir hingegen die munda-
rtliche Version.” (Kunstmann 1992: 183, after Kreuter 2006: 60). All in all, it seems
that, if Kreuter’s account is accurate, Kunstmann’s idea has to be rejected on
phonetic, semantic, and historical grounds.
Lastly, Žuravlev mentions (2005: 863) a proposal by A. Sobolevskij and A. Vajan
which connects our word with PSlav. *pariti; unfortunately he does not provide
any details, or a bibliographical address.
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 675
2.4. Loose ends
We came across several pieces of information during the preparation of this pa-
per that share a common theme but are not quite sufficient to form a complete
picture of their own. What appear to be the important ones among them revolve
around the dating of our word (2.4.1), its use as a given name (2.4.2), the meaning
of ‘bulging’ (2.4.3), and the nasal element in its first syllable (2.4.4). A summary
is given in 2.4.5.
2.4.1. Dating
2.4.1.1 The native proposals (see 2.1) generally derive our word from Proto-Slavon-
ic, and reconstructions are justified using Proto-Slavonic elements and Proto-
Slavonic word-formative methods. Thus, they imply that the word was coined
between about the 5th and about the 10th century.
The Turkic proposal, in its original form (2.2.3.1), points to a Tatar word which
might suggest the 13th century and the Mongol invasions, but it might also be
that Tatar was meant in it as more of an umbrella term for north-western Turkic
peoples in general, as was not rarely the case in 19th c. Turkological literature;
then, the time of borrowing could be anything between about the 6th (SSS Ⅵ: 210f)
and, in theory, the 18th century when the word appears in western European
press to describe cases of alleged vampirism in East Prussia and the Habsburg
Empire. In the two-path variant of the Turkic proposal (2.2.3.2) the northern
path is essentially the same as the original propostion, while on the southern
path the borrowing would have to occur some time between the 6th and the
13th century when Bolghars gradually lost their identity (SSS Ⅵ: 210f; Waldman,
Mason 2006: 106f).
2.4.1.2 It was mentioned above that the earliest known attestation of what ap-
pears to be our word is in Russian and dated 1047. It features in the colophon to
the Book of the Prophets as the name of the copyist, one Оупирь Лихыи, usu-
ally translated as ‘foul/wicked vampire’. This is an unusual name, and it has not
gone unnoticed. To explain it, A. Sjöberg proposed that it was in fact the same
person as the Upplandic rune-carver Upir Ofeigr. This allows to derive the name
Оупирь eventually from the Swedish verb öpa ‘to cry, to shout’ with the mean-
ing ‘a screamer, a noisy person’ or, as Sjöberg explains (1982: 112), ‘someone
with a strong voice’. In his 1982 paper, Sjöberg presents linguistic and historical
arguments to support his idea; in the 1985 one, he focuses on the historical aspect.
676 Kamil Stachowski & Olaf Stachowski
This, we will not attempt to judge; the linguistic part can be retold in brief as
below (cf. also Skrzypek 2011: 49f).
The runic signature ᚤᛒᛁᚱ (e.g. Fv1976 107; other variants also exist) can be read
ubir, upir, übir and üpir; the initial vowel would have to be rendered in Cyrillic as
‹оу›, and the palatality of -rь is the same as in Игорь or Гунарь. The rune-carver’s
full name was Ofeigr Upir, lit. ‘bold/daring screamer/shouter’; ORuss. lihyj meant
‘surplus, extra’, but also ‘bold, daring’. This points to a more conceivable name
than ‘foul/wicked vampire’. Texts carved by Ofeigr Upir’s contain several places
which can be neatly explained assuming the author’s background as a Novgorod
priest, but are unclear otherwise. A detailed discussion can be found in Sjöberg
(1982: 113f), here we will merely list them: the inscription iRma:k (Sö 11), possibly
ieromonach; the inscription kriki or kiriki ‘church’ (U 687), possibly under the
influence of OCS/ORuss. црькы ~ цьркы ~ цьрькы id.; Upir’s difficulties with
the rune ᚼ h, consistent with the Orthodox tradition; the spelling of Halfdan as
‹alfntan›, consistent with Gk. Αλφνταν.
Sjöberg makes a strong case for Ofeigr Upir to have had an Eastern Slavonic
clerical background. The supposition that he had been the same man as our Оупирь
Лихыи before he moved to Uppland, seems rather plausible. But the first part
of the argument, that Оупирь had had a Nordic background, is less well argued,
at least from the linguistic side, and this is what interests us here most. We are,
however, willing to take Sjöberg at his word – especially that this theory, also
supported by Lind (2004, 2012: 348, and p.c.), can explain the shift from и in Оупирь
to ы in modern Russ. упырь, which is difficult to account for otherwise. Sadly,
it does not quite explain late mediaeval attestations with an и (see Sreznevskij
1893), Popowska-Taborska ([1999]: 347), and also Odesskij (2011), who strongly
criticizes Sjöberg but without addressing his actual arguments at all).
Although there might exist another argument in favour of the use of ‘vam-
pire’ as a given name (see 2.4.2) it appears that 1047 is not the date of the ear-
liest attestation of our word and in consequence, the time window for its bor-
rowing would seem to remain as wide open as it was at the beginning of this
subsubsection.
might have been rare and generally limited to singular cases, but it was certainly
not unknown. We imagine it is quite unlikely that any community would choose
to administer inhumation, against the usual custom, if they had feared that the
corpse could then come back to haunt and possibly even kill them. In some
territories skeletal cemeteries actually appear even before Christianity (see e.g.
Gąssowski 1992: 138f), but the method only became prevalent together with the
spread of the new religion which demanded that bodies be buried whole. Certain-
ly, the introduction of a new religion did not always go unopposed as is shown
by several pagan rebellions that occurred in the 10th and 11th century, but to the
best of our knowledge, those reactions were inspired by religious and political
causes, and they did not place any particular stress on burial customs – which
would be expected had the belief in vampires really existed before the spread of
inhumation. It appears, then, that the earliest the Slavs could have begun to fear
vampires, is after their Christianization.
The official dates are: Croatia and Serbia – 7th century, Slovakia – around 830,
Moravia – 831, Bulgaria – 864, Kievan Rus – 867, Bohemia – 884, Poland – 966 and,
in view of its considerable Slavonic population, Hungary – early 11th century. Nat-
urally, the baptism of the ruler meant at best that the nobility would also convert,
and that missionaries would be let in. It would be sometimes more than a century
before the religion spread among the people and its customs were accepted.
It is surprising, then, that anti-vampire burials appear in Poland already in
the 10th century. Perhaps, Gardeła, Duma (2013) and Gardeła, Kajkowski (2013) are
right in supposing that not all of the measures that were traditionally interpreted
as protection against vampires (decapitations, prone burials, bodies covered with
stones, &c.) are indeed just that. Perhaps they really are the result of judicial
practices or some other beliefs.
Be that as it may, the above has significant consequences for the native et-
ymologies of upiór as, effectively, it sets the terminus post quem to at least the
9th century. Almost all of these proposals rely on an initial *ǫ- (see 2.1.18), and
the 9th century is just when nasal vowels are transforming and denasalizing
(Schuster-Šewc 2014: 1162, also Zoltán 2013b, who suggests a period of even up
to the 12th century, but only for Bulgarian-type dialects which is not sufficient
for our cause). With native etymologies uncertain, the only acceptable proposal
left is the two-path variant of the Turkic origin (2.2.3.2); notably, it is also fitting
from the ethnographic perspective (section 1).
The terminus ante quem is more difficult to establish. If the traditional inter-
pretation of atypical burials from Poland is correct, it would be the 10th century
678 Kamil Stachowski & Olaf Stachowski
for Poland and, perhaps, appropriately earlier for other Slavonic peoples. In such
case, the appearance of upiór on the Slavonic ground would have to be dated to the
9th–10th century. Later only if the archaeological interpretation were to change.
2.4.2. Given name
In the light of what was said in 2.4.1, it seems that it is of little importance whether
the Оупирь of 1047 is or is not our word. Most likely it is not, but we have to act as
the devil’s advocate and mention two pieces of information that may potentially
authenticate the use of upiór &c. as a given name.
One is the case of Képes Krónika, a Hungarian chronicle written in Latin in
1358, featuring (Geréb 1964: 36) a captain by the name of opour. Geréb (1964: 96)
reads the word as Apor, a surname found in Hungary to this day, but Tarnai (1992:
130) proposes the reading of Opur. To the best of our knowledge, the only attempt
to claim the name for Hungarian was made by Ladó, Bíró (1998) but, being limited
to the statement ‘an old Hungarian proper name’, it must be considered empty.
If one chooses to accept Tarnai’s reading, the name could be a borrowing from
Slav. upyr or upyŕ (see Bárczi [1967: 155f] for Slav. y > Hung. u, and Helimskij
[2000: 422] for OHung. u- > Hung. o-). On the other hand, the Hungarian shape
could also be derived from the Slavonic name Opor (attested for Polish since 1265,
see SSNO) < opora ‘benefit’ (see Rymut 1999–2001; Cieślikowa 2000) or opora
‘obstacle’ (?) (see Rymut 1999–2001).
The other piece of information is related to the meaning of ‘bulging’ and
discussed separately in 2.4.3.
2.4.3. ‘bulging’
The connection between our word and the meaning of ‘bulging’ has been somewhat
elusive. It was not completely convincing when stated explicitly in 2.1.6 (≪ *piriti,
structured like *ǫ-tъk-ъ, with the meaning ‘bulging, bloated, swollen’), but it does
reappear in several pieces of ethnographic data (see 1).
In particular, K. Moszyński (1967: 608) mentions that southern Russians use
the word upýŕ to talk about children with large heads (with hydrocephalus?). It is
not entirely clear to us how this might have come to be. Possibly, there is some
link here to the fact that Tatars say of children with hydrocephalus, who tend to
have trouble falling asleep, that they have been changed by ubyr (Zaripova Çetin
2007: 24), and that Turkic ubyr itself is often described as having a large head; see
section 1. However, one might also suspect a later evolution based on the shared
physical feature of ‘bulging’, in which case this information should probably be
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 679
18 NHG: woerterbuchnetz.de/cgi-bin/WBNetz/wbgui_py?sigle=Adelung&lemid=
DW00566; MHG: woerterbuchnetz.de/Lexer/?sigle=Lexer&mode=Vernetzung&lemid=
LW00497; OHG: www.koeblergerhard.de/ahd/ahd_w.html.
680 Kamil Stachowski & Olaf Stachowski
Figure 1a. The origin of Pol. wąpierz assuming a Proto-Slavonic rendering with *ǫ-.
2.4.4. Nasality
A weak point of the Turkic proposal, whether it assumes one or two paths of bor-
rowing, is that it does not at any step explain the nasality attested in wampir &c. –
or in Pol.dial. wąpierz, but for this see also 2.4.3. We can think of four ways to
address this issue.
One way in which the Slavs could have changed u- into ǫ- is an insertion of
a nasal infix, such as was fairly often added to Indo-European zero grade roots,
namely in 34 out of 56 cases of the Slavonic u ~ ǫ alternation (e.g. zǫbr ~ zubr
‘urus’, čęb ~ čub ‘satureia’; Sławski 1939–1947: 286).
This possibility requires that the Kipchak shape begins with *u-. Actually,
PTkc. *o-/*ō- > MKipch. *o- > Tat. u-, which is to say that the change occurred
about the 13th/14th century (see Berta 1993, 1998 for more details), but it seems that
in some dialects the raising might have happened earlier. If that were the case,
this scenario would simultaneously explain the Slavonic (mostly, Ukrainian) and
Slavo-Hungarian (?; see 2.4.3) alternation u- ~ o-. Unfortunately, lack of pre-13th c.
written sources for Kipchak makes it very difficult to definitively accept or reject
this possibility.
Another possibility is that the -m- was inserted secondarily to ease the pro-
nunciation. Prenasalization is certainly not an unknown phenomenon (see e.g.
Flemming 2005: 165f; Ohala 1983: 200f), though it is true that it tends to occur
before voiced stops rather than unvoiced ones; cf. nonetheless the phonetically
quite similar Arom. pampore ‘vapour’, Bulg. вампóръ id. < It. vapore id. (Mlad-
enov 1941). Similarly Serb.dial. tȃmbor ‘camp’ (Karadžić 1935 s.v. tȃbor; see also
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 681
Németh 2014 for the history of the word), and perhaps also Mac.dial. fambrika,
junguslavija | Rom. sâmbătă ‘Saturday’, G Samstag id., Pers. šanbe id., &c. ≪ Hebr.
šabbat id. (see Şirin User 2010) | E harbinger < ME harbergere, -geour (Onions
1966)19 | Pol. cmentarz ‘cemetery’ (ca. 1500 cmyntarz) with the -n- inserted in
Polish or borrowed together with the word itself from OF cimentire < Lat. cime-
tarium ~ cimeterium | Gk. Τσιγγάνος (-ng-) ‘Gypsy’ < Slav. Cygan id.20 (< Tkc.;
Stachowski M. 2002: 160f).21
The phenomenon is of course irregular, but this might explain why Bulgarian
has both shapes, with and without the nasal consonant. The apparent appeal of
the -mb- sequence would be the reason why it is the shape with this addition
that spread across the Southern Slavonic languages, not the other one.
Next, the excess -m- could be explained by resorting to folk etymology. We saw
in 2.1 that our word raises more than one association with the Slavonic material.
Perhaps Tkc. *ōpyr was reinterpreted by the Slavs in such a way as was suggested
in 2.1.6, 2.1.9, or 2.1.15, and then its phonetic evolution simply proceeded as if it
were a native word?
19 This is Onions’s flagship example where he collects references to all the other similar
words. But actually, this particular word might be not so much a case of insertion
of -m-, as of dissimilation of -ergere > -ingere.
20 This last example may prove to be irrelevant here. Gk. g > γ in most positions by
the 2nd century bc, remaining unchanged only after a nasal consonant (Horrocks
2010: 170). When it came much later to rendering the Slavonic word Cygan, Greeks
were effectively reduced to the choice between γ and ng (cf. also Arvaniti, Joseph
2004: 77), and may have settled for the letter simply because they found it sounded
more like the etymon than the former.
21 Though attributed to a different phonetic mechanism, the insertion of -p- or -b- after a
nasal consonant is also a not uncommon occurrence and, since it results in the same
sequence, contributes to its frequency and hence, perceived attractiveness. Cf Serb.dial.
amberika (Sawicka 2005) | Mac.dial. mbleko ‘milk’ | MG zimber ‘1. Wohnung; 2. Bauholz’ =
OEng. timber ‘1. building, edifice; 2. building material, wood for building’ < PGerm.
*temra- ‘Bauholz’ (Kluge 2011; Onions 1966) | F nombre < Lat. numerus (Dauzat 1938)
| G colloquial [kɔmpt] for kommt and [kɔmpst] for kommst (Ramers, Vater 1995: 51)
or old sampt for samt (e.g. in a 1633 chalcography by Merian d.Ä.: “Wahre Bildnuß
der Statt Maintz, sampt den newen Schantzen, Schiffbrucken und Leger […]”; www.
regionalgeschichte.net/rheinhessen/mainz/bilder.html). While there are also examples
of dropping of a b in phonetically similar words, see e.g. Knüppel (2009) for the rather
complex case of Ott. tambur(a) ~ damur(a) &c., they appear to be on the whole rarer,
and do not anyway negate the existence of examples which support our case.
682 Kamil Stachowski & Olaf Stachowski
Lastly, Ragagnin (2013: 66) suggests that the Turkic etymon was not *ōp-
‘to suck, to swallow, to greedily catch with mouth’ + aorist participle -(V)r, but the
same verb with the deverbal nominal suffix -gur, i.e. *ōpkur. This would ultimately
yield on the southern path *vupkor > *vapkir > *vappir > vampir. Ragagnin does
not say so explicitly, but we understand from the context that she offers this re-
construction as a way to easier account for the -m-. It is theoretically possible but
we are afraid that in fact explaining the loss of -k-, in all of the Slavonic reflexes,
might prove even more difficult.
It is not possible to tell which of these possibilities is the most plausible. The first
one appears to be better suited to the northern, Kipchak path; the second one
to the southern, Chuvash path; the third one could have happened on both;
the fourth would have to have partially happened on both. In theory, one might
even imagine the first and the third, the second and the third, or perhaps even
some other combination, happening simultaneously in approximately the same
area. But for now this is fantasy. More Slavicist work is necessary.
2.4.5. Summary
The traditional, pagan burial of the Slavonic peoples typically involved cremation,
but inhumation was also sometimes practiced. It is unlikely that the belief in
vampires could take hold among the Slavs before the wide spread of the latter
method (enforced together with Christianity). The terminus post quem should
probably be set at the 9th century, which significantly weakens all of the native
etymologies. A plausible terminus ante quem would be the 10th century. The 1047
attestation of Оупирь Лихыи, while chronologically possible, is more likely
a misattribution. There are more serious reasons behind this view than merely
the fact that it would be odd for a person to bear a name that means ‘vampire’.
(2.4.1) It does, nonetheless, seem that there are no traces of the word ever being
used as a given name. (2.4.2) It was, however, used to describe people, possibly
as a consequence of the belief that bodies of vampires were bulging, swollen
with the blood they had drunk. This property of theirs might help explain the
Polish dialectal shapes wąpierz &c. Otherwise, we would need to assume that
a nasal element was inserted into our word twice, once by southern Slavs, and
once probably by Poles. (2.4.3) We can think of four ways how and why such
an insertion could occur but we lack a way to precisely evaluate these ideas.
We can but suspect folk etymology to be a more plausible explanation than
others (2.4.4)
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 683
3. Conclusions
The wide spread of the word vampire is a fairly late, 18th and 19th century develop-
ment, due to alleged cases of vampirism which received considerable attention in
German-speaking countries. The word itself is of Southern Slavonic origin, and
eventually cognate to northern Slavonic shapes upiór and similar (although Pol.
dial. wąpierz and others also exist). Originally, in the Slavonic folklore, the word
referred to a much more gruesome creature than we picture today as a vampire,
but which too was most probably a dead person who rose from the grave to harm
the living in one way or another. (See 1.)
The word exists in the Slavonic languages in a great multitude of phonetic
shapes which cannot be easily explained. More than seventeen etymologies have
been proposed to interpret it as a native word (see 2.1), and more than five that
assume a borrowing (2.2). Of the first group, we deem three to be more plausible
than others (2.1.18), and of the second just one, in its newer and more complex
version (2.2.7). As for the dating, the often cited attestation of 1047 appears to be
a misattribution; a more plausible terminus post quem is the 9th century when
Christianity is introduced to the Slavonic peoples together with a new burial
custom (2.4.1). This dating is an important argument against all of the native
solutions simultaneously, effectively leaving us with just the Turkic etymology.
To account for the Slavonic phonetics, it needs to be supplemented with a nasal
element added once or perhaps twice (2.4.3 and 2.4.4).
Overall, what appears to be the most probable history of the words upiór and
wampir can be summarized in the following way: a reconstructed Proto-Turkic
form *ōpyr ‘that which sucks, that which swallows’ has reflexes in several Tur-
kic languages; in particular, northern (Kipchak) shapes are similar to opyr, while
southern (Bolghar) shapes are similar to văbăr, all with meanings ‘evil spirit’
and alike. The northern shapes were borrowed by Eastern Slavs and transmitted
to Western Slavs; the southern shapes by Bulgarians and transmitted to South-
ern Slavs.22 Both borrowings occurred most likely in the 9th or the 10th century.
This was followed by the rise of forms with an unetymological -m- (vampir &c.)
in the south, and possibly some internal borrowings between various Slavonic
dialects. It seems rather probable, if unprovable, that folk etymology has more
22 Incidentally, Tkc. *ōpyr may likely also be the ultimate source of Russ. vurdalák
‘vampire, werewolf’ (Kajtoch [forthcoming]), but this word merits a separate study
(in preparation).
684 Kamil Stachowski & Olaf Stachowski
than once altered the phonetic shapes. In this sense, both upiór and wampir –
or perhaps only some of their cognates – can be viewed as simultenously native
and borrowings. This may sound like a cheap attempt at reconciling the opposing
camps, but having two origins is not in fact a self-exclusive idea; see Laakso (2001)
for examples and commentary. The entire scheme is illustrated in fig. 2.
The above does not mean that all Slavonic shapes have now been explained.
Still unclear are such forms as Bulg.dial. ljapír, Cash. niełåp, uṕȯn, opi, &c.
(see also fn. 4), Pol.dial. lupirz, łupior, upierz (Lublin area; Karłowicz 1900–1911:
vol. 4: 32), wąpierz &c. (see also 2.4.3), SCr.dial. lampijer, and others. Many are
probably no more than dialectal innovations, but for now we know of no way
to verify this.
Slav.W Slav.E
)Pol. upiór( )Russ. upíŕ( Kipch. *opyr
Fr. vampire G Vampir )Pol. wampir( )Russ. vampír(
Gk. βαμπύρος
Barber P. 1996. Staking Claims: The Vampires of Folklore and Fiction. – Skeptical Inquirer
20.2. [www.csicop.org/si/show/staking_claims_the_vampires_of_folklore_and_fiction;
accessed June 29, 2015].
Bárczi G. 1967. Hangtörténet. – Bárczi G., Benkő L., Berrár J. (eds.). A magyar nyelv tör
ténete. Budapest: 95–180.
Bartholomae Ch. 1961. Altiranisches Wörterbuch. Berlin.
BER = Георгиев В. et al. (ed.). 1971–. Български етимологичен речник. София.
Berta Á. 1993. Graphische Probleme des Tatarischen. – Werner O. (ed.). Probleme der
Graphie. Tübingen: 107–120.
Berta Á. 1998. Middle Kipchak. – Johanson L., Csató É. (eds.). The Turkic Languages. Lon-
don, New York: 157–165.
Berta Á., Róna-Tas A. 2002. Old Turkic Loan Words in Hungarian. Overview and Sam-
ples. – Acta Orientalia Academiae Scientiarum Hungaricae 55.1–3: 46–67.
Beydili C. 2005. Türk Mitolojisi Ansiklopedik Sözlük. Ankara.
Bielfeldt H.H. 1982. Die slawischen Wörter im Deutschen. Ausgewählte Schriften 1950–1978.
Leipzig.
Boček V. 2010. Studie k nejstarším romanismům ve slovanských jazycích. Praha.
Boev 1970 = Боев Е. 1970. Казахски и български речникови успоредици. – Известия
на Института за български език 19: 899–906.
Böhm C. 1870. Vampire. – Notes & Queries 120: 522.
Boryś W. 1975. Prefiksacja imienna w językach słowiańskich. Ossolineum.
Boryś W. 2005. Słownik etymologiczny języka polskiego. Kraków.
BoS = Archive for Research in Archetypal Symbolism (ARAS). 2010. The Book of Symbols.
Taschen.
Brückner A. 1927. Słownik etymologiczny języka polskiego. Warszawa.
Brückner A. 1934. Etymologien. – Slavia 13: 272–280.
Budagov 1869–1871 = Будаговъ Л. 1869–1871. Сравнительный словарь турецко-
татарскихъ нарѣчій. Санктпетербургъ.
Budziszewska W. 1983–1985. Z nazw motyla w językach słowiańskich na tle porównaw-
czym. – Rocznik Slawistyczny 43: 13–19.
Burkhart D. 1989. Kulturraum Balkan. Studien zur Volkskunde und Literatur Südosteuropas.
Berlin, Hamburg.
Ceylan E. 1997. Çuvaşça Çok Zamanlı Ses Bilgisi. [= Türk Dil Kurumu Yayınları 675]. Ankara.
Charnock R.S. 1870. Vampire. – Notes & Queries 120: 378.
Cheung T. 2013. The Element Encyclopedia of Vampires. London.
Cieślikowa A. [with the assistance of J. Szymowa and K. Rymut]. 2000. Słownik etymo-
logiczno-motywacyjny staropolskich nazw osobowych. [vol. 1: Odapelatywne nazwy
osobowe]. Kraków.
Cooper B. 2005. The Word vampire: Its Slavonic Form and Origin. – Journal of Slavic
Linguistics 13.2: 251–270.
Dauzat A. 111938. Dictionnaire étymologique. Paris.
Possibly Oriental elements in Slavonic folklore. Upiór ~ wampir 687
DEI = Cortelazzo M., Zolli P. 1990. Dizionario etimologico della lingua italiana. Bologna.
DHLF = Rey A. (ed.) 1992. Dictionnaire historique de la langue française. Paris.
Dmitriev 1962 = Дмитриев Н.К. 1962. О тюркских элементах русского словаря. –
Дмитриев Н.К. (ed.). Строй тюркских языков. Москва: 503–569.
DRAE = Real Academia Española. 222012. Diccionario de la lengua española. [www.rae.es/
recursos/diccionarios/drae].
Dukova U. 1997. Die Bezeichnungen der Dämonen im Bulgarischen. [= Specimina Philologiae
Slavicae 115]. München.
Dźwigoł R. 2004. Polskie ludowe słownictwo mitologiczne. Kraków.
Edelspacher A. 1876. Denevér. Lidércz. Vampir. – Magyar Nyelvőr 5: 151–159.
ESUM = Melьnyčyk et al. 1982– = Мельничук О.С. et al. (eds.). 1982–. Етимологічний слов
ник української мови. Київ.
EWU = Benkő L. (ed.). 1993–1994. Etymologisches Wörterbuch des Ungarischen. Budapest.
Flemming E. 2005. Speech perception and phonological contrast. – Pisoni D.B., Remez R.E.
(eds.). The handbook of speech perception. Malden: 156–181.
Gardeła L., Duma P. 2013. Untimely death: atypical burials of children in early and late
medieval Poland. – World Archaeology 45.2: 314–332.
Gardeła L., Kajkowski K. 2013. Vampires, criminals or slaves? Reinterpreting ‘deviant
burials’ in early medieval Poland. – World Archaeology 45.5: 780–796.
Gąssowski J. 1992. Archeologia o schyłku pogaństwa. – Archeologia Polski 37.1–2: 137–157.
Geréb L. 1964. A Krónika magyar szövege. Budapest.
Gluhak A. 1993. Hrvatski etimološki rječnik. Zagreb.
Gombocz Z. 1912. Die bulgarisch-türkischen Lehnwörter in der ungarischen Sprache. Helsinki.
Gómez-Alonso J. 1998. Rabies: a possible explanation for the vampire legend. – Neurology
51.3: 856–9.
Grønbech K. 1942. Komanisches Wörterbuch. Türkischer Wortindex zu Codex Cumanicus.
[= Monumenta Linguarum Asiæ Minoris, Subsidia 1]. København.
Haefs H. 2001. Neues nutzloses Wissen für die Westentasche: zunebst der Geschichte vom
Vampir Graf Dracula. München.
Hanusz J. 1885. O wpływie języków wschodnich na słownik języka polskiego. – Prace
Filologiczne 1: 458–466.
Harenberg J.C. 1733. Vernünftige und Chriſtliche Gedancken Uber die VAMPIRS Oder Bluht
ſaugende Todten (…). Wolffenbüttel.
Helimskij 2000 = Хелимский Е.А. 2000. Венгерский язык как источник для прасла-
вянской реконструкции славянского языка Панонии. – Хелимский Е.А. (ed.).
Компаративистика, уралистика. Лекции и статьи. Москва: 416–432. [reprint of
the original from 1988].
Hellquist E. 1970. Svensk etymologisk ordbok. Lund.
Hobzej 2002 = Хобзей Н.В. 2002. Гуцульська міфологія. Етнолiнгвiстичний словник.
Львів.
Hock F.S. 1900. Die Vampyrsagen und ihre Verwerthung in der deutschen Literatur. Berlin.
688 Kamil Stachowski & Olaf Stachowski