Surnames of Georgian Jews: Historical and Linguistic Aspects

Abstract

The article provides an analysis of several aspects of the corpus of surnames used by Jews who lived after the end of the Middle Ages in the territory that today corresponds to the Republic of Georgia. One section covers historical aspects: the earliest attestations and their exact status and the period when the use of surnames became stabilized. The next two sections discuss morphological aspects: the endings found in the surnames and historical, linguistic, and social explanations of the distribution observed, compound names, names with demonymic suffixes, and those based on hypocoristic forms of given names (a detailed coverage of methods of constructing such forms is also provided). In the remaining sections, the reader will find an analysis of phonetic peculiarities found in Georgian Jewish surnames, the types of surnames with their statistical distribution, as well as the description of surnames that were not created in Georgia but were brought as ready-made forms by Jews who migrated during the 19th–20th centuries to Georgia from other territories.

1. Introduction

Very few scholarly studies of surnames used by Jews who lived in the territory that today corresponds to the Republic of Georgia have been published until now. The earliest works are due to Gagulashvili, who in 1987 compiled the first representative list of surnames (Gagulashvili 1987) and in 1996 suggested etymologies for a few dozens of them (Gagulashvili 1996). Enoch (2014) analyszes the endings of Georgian Jewish surnames, provides a comprehensive list of surnames used by Georgian Jews, and discusses etymologies for a few of them. This article discusses several major questions of Georgian Jewish onomastics that were either outside of the scope of the above studies or received no answer in them. One of these questions deals with historical aspects: the earliest attestations of surnames and the period when the use of surnames became stabilized. Several topics addressed here are morphological. They concern the structure of surnames, endings used along with the explanation of the distribution observed, and the methods of constructing hypocoristic forms of given names that became the bases for numerous patronymic and matronymic surnames. Finally, this article discusses surnames brought during the 19th–20th centuries to Georgia by Jewish migrants from other regions.1

2. History of Names

The scarcity of historical documents dealing with Georgian Jews before the 17th century does not allow us to determine the period when Georgian Jews started to use hereditary family names. Several authors point to the Early Middle Ages. For example, Mamistvalishvili (2011, p. 77) implies this proceeding in two steps. Firstly, he quotes the opinion by Ǧlonṭi (1986, p. 48) about the 7th–8th centuries being the period of the mass inception of family names within various social groups of Georgian Christians. Secondly, he claims—without providing any argument to support his idea—that Georgian Jews most likely received their surnames simultaneously with Georgian Christians.2 The quote from Ǧlonṭi is accurate, but it is taken out of its context. In his book, Ǧlonṭi focuses on the inception of Georgian given names and personal nicknames. When discussing family names, he mainly deals with those known during the last centuries. He discusses at length the non-hereditary character of numerous family names and the gradual split of the descendants of the same male ancestor into branches bearing different family names (pp. 44–47). Ǧlonṭi also emphasizes (p. 47) that the roots of some modern surnames are very old. In this context, his remark about the inception of family names in the 7th–8th centuries is no more than a hypothesis. The period selected is the one that immediately precedes the time from which the oldest available Georgian written sources date. Most importantly, his hypothesis concerns the roots of the surnames, not the surnames themselves, which appear much later. For example, famous medieval Georgian authors such as the chronicler Leonti Mroveli (11th century) and the poet Shota Rustaveli (circa 1172–circa 1216) had no surnames.3 Today, a large majority of Georgian Christians have surnames ending in -dze or -shvili. The earliest references to names with these endings date from the 13th and the 14th centuries, respectively. Moreover, nothing implies that these early forms were hereditary. In both cases, they could also be patronymics, or names based on the given names of the grandfathers, not retained by the following generations (Nikonov 1988, pp. 152, 155). For much more recent times, we still have rich documented evidence about the non-fixed and non-hereditary last names used by Georgian Christians. For example, during the first half of the 16th century, the son of Gabriela Betiasshvili is called Ganona Gabrielasshvili (that is, ‘child of Gabriela’). A person with the given name Avtandil, the son of Iese and the grandson of Tamaz, appears in various documents from the end of the 17th century under three different last names: Iesesshvili ‘child of Iese’, Tamazisshvili ‘child of Tamaz’, and Baratashvili (a name used in several generations of the same family, apparently inherited from a more distant ancestor) (Ḳldiashvili et al. 1991, pp. 21, 394–95, 585). In documents from the 17th century, we find numerous Christians called by their given names only.4 This situation was possible because before the 19th century, there was no legal obligation to have a hereditary surname for any inhabitant of Georgia independently of the religion.

The oldest documents from Georgia in which Jews are called by names other than their given names provide references to Ioseb Buǧapaisdze (1260) and Eliozisdze (between 1519 and 1530, this person from Tskhinvali appears in the document without his given name).5 No element in our possession implies that the last names of these two persons were their hereditary surnames. For example, Eliozisdze just means ‘son of Elioz’ in Georgian. This way, his last name could be his patronymic, or a non-hereditary name based on the given name of his grandfather.6

The same dilemma is still valid for certain last names ending in -shvili ‘child of’ appearing until the mid-19th century. In scarce sources from the second half of the 17th century dealing with Georgian Jews, we find references to such last names as Ǩezerashvili, Khakhanashvili, and Matvalasshvili (Ḳldiashvili et al. 2004, p. 9; 2015, pp. 32, 372). In theory, all of these could be non-hereditary, designating sons or grandsons of men called Ǩezera, Khakhana, and Matvala, respectively. In 1737, documents from western Kartli mention three local Jews: Shabatas-shvili Balua, Isrelashvili Ǩobia, and Mosias-shvili Daniela (Berdzenishvili 1940, p. 174). Here, we can also be dealing with Jews without surnames, that is, just sons or grandsons of Shabata, Isr(a)ela, and Mosia. Surely, the modern Georgian language uses -dze ‘son of’ to form patronymics. Yet the pattern of adding -shvili to form patronymics was still operational, at least regionally, even in the middle of the 19th century. For example, a document compiled in 1845, most likely in western Kartli, describing the distribution of the legacy of the deceased Georgian prince Palavandov between his sons, refers, among others, to a group of male Jewish serfs.7 For some of them, their given name (Ioseba, Bato, Tsise, Ḳaḳa, Israela, and Elishaǩa) is followed immediately by their surname Ḳrikheli. In other cases, the whole name is composed of three elements that can appear in a different order: the given name, the patronymic, and the surname Ḳrikheli, with the word khakham(i) that can be added before the given name:8 Ḳaḳo Ḳezerashvili Ḳrikheli, Eliḳashvili khakhami Elishaǩa Ḳrikheli, Iskhaḳashvili khakham Iaḳobi Ḳrikheli, Eliḳashvili Ṭaṭo Ḳrikheli, Babalashvili Datua Ḳrikheli, Ǩezerashvili Ǩezera Ḳrikheli, Davitashvili Ḳrikheli khakhami Moshe, Svimonashvili Moshe Ḳrikheli, and Mardakhas-shvili Ḳrikheli Shalo. There is no doubt that in the above list, all forms ending in -shvili are not surnames but patronymics. For example, the names of the last three persons can be translated as Moses, the son of David Ḳrikheli; Moses, the son of Simon Ḳrikheli; and Shalo(m), the son of Mordecai Ḳrikheli.9 Other Jewish serfs mentioned in the same document are Babalashvili Iskhaḳi, Iskhaḳashvili Shamoela, Eliḳashvili Ḳaḳo, and Abrama gorishi ‘from Gori.’ For the first three of them, their name ending in -shvili is ambiguous. It could be a surname. Yet we cannot exclude the possibility of it being a patronymic, so we can be dealing with sons (or grandsons) of Babala, Iskhaḳa ‘Isaac’, and Eliḳa ‘Elijah.’ One can observe that the last person in the above set of names is listed without any surname. This fact can be interpreted in two ways. On the one hand, we cannot exclude a possibility that some serfs owned by Georgian noble landlords simply had no surname. On the other hand, some Jews could have surnames, but because of the absence of any official status of these surnames, they could be non-fixed and considered of little, if any, importance and omitted in legal documents. An example can illustrate this scenario. A document from 1789 refers to the donation of the Jewish serf Elia Gagulashvili to a monastery in Racha. In certain documents from the first third of the 19th century, the same Elia appears without his surname (David 1989, vol. 1, pp. 139, 254).

The census of Tskhinvali made in 1781 (Tabuashvili 2013, pp. 56–60) notes the presence in that town of fourteen Jewish households. The family heads bear the following last names: Aṭenelashvili, Davitashvili, Eliashvili, Elishaǩashvili, Israilashvili, Khukhashvili, Khundiashvili, Mamistvalashvili (two), and Ṗaṗisimedishvili. We also find households of Ḳatsoba Manashera, Binia Khakhami, Shaloma, and Gagula. Yet we cannot be sure that these four families had no surname. For the first two of them, the actual surnames could be Manasherashvili (known from other sources from 18th-century Tskhinvali) and Khakhmishvili, and the last two names appear in two lines of the document that are corrupted. Note, however, that nine Christian families are recorded in the same census with no surnames. In neighboring villages, we find references to Jews called by given names only: Babala, Daniela, Elia, and Ḳaḳia (Tabuashvili 2013, pp. 46, 47, 66).

Table 1. References to Jews with no surnames in Georgian Christian sources.

Name	Period	Place or Province	Source
Ioseba	1st third of the 17th century	Imereti	(Mamistvalishvili 2011, p. 133)
Ṗaṗia, Khatuna, Mardakha	1642	Largvisi (Tskhinvali area)	(Ḳldiashvili et al. 2015, p. 358)
Mardakha	1671	Tamarasheni	(Ḳldiashvili et al. 2004, p. 56)
Mardakha	Between 1676 and 1709	Mukhauri (Tskhinvali area)	(Ḳldiashvili et al. 2004, p. 56)
Ḳoba, Abrama, Ṗaṗua, Datuna	1723	Tskhinvali	(Berdzenishvili 1940, p. 168)
Manukha	1766	Tamarasheni	(Mamistvalishvili 2011, p. 224)
Ḳaḳia	1782	Kutaisi	(Berdzenishvili 1940, p. 191)
Shabata, Itskhak10	1791	Sujuna	(David 1989, vol. 1, p. 140)
Mardakha	End of the 18th century	Kartli	(David 1989, vol. 1, p. 140)
Shalom	Circa 1800	Unclear	(David 1989, vol. 1, p. 140)
Abrama, Ǩobo, Shaloma, and Elia, sons of Shamoela	1802	Imereti	(Berdzenishvili 1940, p. 218)
Shabata	1800s	Kartli	(David 1989, vol. 1, p. 140)
Elia	1809	Surami	(David 1989, vol. 1, p. 187)
Kobia	1838	Unclear	(David 1989, vol. 1, p. 188)

provides other examples of references to Georgian Jewish men with no surnames.

The marginal character of numerous surnames for the conscience of their own bearers and/or other members of their communities follows from certain Jewish documents compiled in Hebrew during the second half of the 19th century, when all Georgian Jews necessarily already had official hereditary surnames. In his book (Chorny 1884), the Ashkenazic traveler Joseph Judah Chorny often refers to Georgian Jews he meets by their traditional Jewish names only: the Rabbi Shalom in Breti (p. 265) and Rabbi Eliahu from Sachkhere in Surami (p. 149). In 1869, a letter was signed by thirteen leaders of the Jewish community of Atskhuri (Akhaltsikhe area). For three of them, their last names could be either surnames or Georgian patronymics (*Iaǩobashvili יעקב שוילי and two *Berishvili ברי שוילי). Two others have the last name *Cohen כהן, which could be their surname or an indication of the priestly origin. The eight remaining men are all called according to the traditional Jewish pattern X ben Y, that is, after their given names and patronymics (Chorny 1884, p. 256). A document published in the Hebrew press in 1870 was signed by twenty-six representatives of the community of Kulashi (Imereti), of whom twelve appear without surnames. Among the signatories of a similar document from the community of Sujuna (Mingrelia), we find three bearers of the surnames Mikhelashvili, one Israelashvili, and numerous persons called after the traditional Hebrew pattern X ben Y (David 1989, vol. 1, pp. 428–29). Note that both surnames in question are of patronymic origin meaning ‘child of Michael’ and ‘child of Israel’, respectively, and, therefore, they could appear rather recently. Surnames are exceptional in Jewish tombstones in Georgia before the end of the 19th century.11 It is also important to observe that in the Montefiore Jewish census of Jerusalem made in 1875, the majority of persons born in Georgia are listed without surnames, and among last names that look like surnames, the commonest are Mizrahi ‘Oriental’ and Gurji ‘Georgian’, both of which are unknown in Georgia and represent typical nicknames assigned already in the Land of Israel (Beider 2023).

The oldest known surnames whose hereditary character is doubtless appear in the 17th century: Danela Pichkhadze was a Jewish serf donated in 1644 to the Gelati monastery, and brothers Matvala and Shalia Jinjikhashvili were serfs living in Mdzovreti (near Kareli) circa 1670.12 Both surnames, Pichkhadze and Jinjikhashvili, survived until our days, and both are based on Georgian nicknames: compare pichkhi ‘branch cuttings’ and jinjikhi ‘lukewarm’, respectively.13

Jinjikhashvili appears in the legend recorded by Joseph Judah Chorny in the town of Kareli during the second half of the 19th century.14 Local Jewish leaders related to him a story about three Jewish boys who in the past were the only survivors of the massacre by Persians of the inhabitants of the village of Mdzovreti. These boys were found by Jews from Tskhinvali. As adults, they returned first to Mdzovreti and later settled in Kareli. All members of the community of Kareli are said to be their descendants. The three men received nicknames that later became surnames. One became Ṭsiṭsuashvili because he liked to sit under the conifer trees.15 Another was a tall red-haired fellow, and for this reason, he became Jinjikhashvili. The third one received his name, Dzorelashvili, after the name of their native village. We do not know the exact factual basis for this legend. Of the etymologies proposed for these three surnames, the last one, toponymic, is fully reliable. The text about Jinjikhashvili just represents an attempt, maybe of a relatively recent origin, to explain why a person could receive a nickname meaning ‘lukewarm.’ Note that the link to Mdzovreti appearing in this story is well correlated with the place of the earliest known reference to Jinjikhashvili, two centuries before the legend was recorded. The census of Kareli made in 1781 (Tabuashvili 2012, p. 176) sheds additional light on the above legend. It shows that five Jewish families dwelled there at that moment, those of Marḳoza Gzirisshvili, Babata16 Jinjikhashvili, Moshia Iosebasshvili, Batua Ṭsiṭsuashvili, and Mardakha Ṭsiṭsuashvili. For the first four families, the source indicates their provenance from Mdzovreti. This way, not only the census data corroborate the idea that Jews of Kareli originated in Mdzovreti, but also the timeframe of this resettlement becomes clear: the second half of the 18th century. Indirectly, this information implies that the surname Dzorelashvili was created after 1781. Perhaps it became the family name for either Iosebasshvili (which could be a patronymic rather than a hereditary surname at the moment of the census) or Gzirisshvili (this name does not appear in more recent documents).

For numerous surnames borne by Georgian Jews during the last two centuries, their earliest reference in available sources dates from the 18th century. For this reason, if we want to avoid speculative assertions, it would be logical to consider that it was precisely during that period that the Georgian Jewish surnames were mainly formed.17 This idea is compatible with the above analysis of the history of the surname Dzorelashvili. Globally, it is also well correlated with the fact—discussed above—that during the 19th century, surnames were not an integral part of the Georgian Jewish naming system yet.18 Several other indirect factors also point to a relatively recent adoption of surnames. One can observe a high proportion of surnames derived from male given names.19 These patronymic surnames are not sufficiently specific—especially if they are based on common biblical names—to serve as markers of belonging to specific families. Normally, for surnames formed well before the 18th century, one would expect a higher percentage of nickname-based names that are really distinguishing a family from others. Also, numerous patronymic surnames borne by Jews are based on hypocoristic forms of given names with diminutive suffixes including the consonant /ḳ/. Before the 18th century, such forms are rare for Georgian Christians, and there is no reason to consider that this pattern of forming hypocoristic forms had any Jewish specificity.20 Moreover, surnames used during the 19th century in areas where the Jewish presence is well attested in the 18th century are usually limited to certain geographic areas. Examples appear in

Table 2. Examples of surnames limited geographically in the 19th century.

Area/Place	Surnames
Western Kartli	Astanjelashvili, Aṭanelashvili, Beniashvili, Biniḳashvili, Datuashvili, Davarashvili, Dediashvili, Dzorelashvili, Iosebashvili, Jinjikhashvili, Khukhashvili, Ḳozhiashvili, Leviashvili, Mamistvalashvili, Manasherashvili, Naniḳashvili, Ṗaṗismedashvili, Ṭoriḳashvili, Tsitsuashvili, Zizovi
Oni	Beriḳashvili, Buzuḳashvili, Chachashvili, Chanchalashvili, Gagulashvili, Gorelishvili, Khakhiashvili, Khiṭibashvili, Konashvili, Kosashvili, Shimshilashvili, Ṭoṭiashvili
Kutaisi and its area	Babaliḳashvili, Bachilishvili, Batashvili, Biniaurishvili, Boterashvili, Buziashvili, Chakhosvhili, Chilashvili, Chuṭiashvili, Datiashvili, Eligulashvili, Eluashvili, Iaḳobishvili/Iaǩobishvili, Jǧuniashvili, Ḳaḳiṭelashvili, Ḳaṭaṗariashvili, Khikhinashvili, Khoṭeveli, Leḳviashvili, Mirilashvili, Rizhinashvili, Shamelashvili, Shatashvili, Tarunishvili, Tavdidishvili, Tetrokalasvhili, Tetruashvili, Topchiashvili, Tsitsiashvili, Ṭsveniashvili, Zhuṭiashvili
Akhaltsikhe and its area	Abajanovi, Aivazashvili, Aḳoshvili, Anṭoshvili, Bajoti, Baḳaloti, Baṭonjanashvili, Bibilashvili, Charukhchevi, Darchiashvili, Injabeli, Injashvili, Ḳatsobashvili, Ḳazhiloti, Khakhmishvili, Korpashvili, Kurkchishvili, Nanaziashvili, Piṭimashvili, Ṭetsoti
Lailashi	Batiashvili, Beberashvili, Beruchashvili, Januashvili, Ḳiḳalishvili, Mardakhiashvili, Meǧrelishvili, Sepiashvili, Zonenashvili
Sujuna	Khubelashvili, Mikhelashvili

.21

The above geographic distribution indicates that these surnames were most likely adopted locally. However, no element implies that various Georgian communities were isolated from each other. If surnames appearing in Table 2 would be old, since Jews represented a very small minority, one would expect many of these surnames to spread by migrants to various areas. Surely, this argument is not absolute. For example, Jinjikhashvili seems to remain restricted to western Kartli, and this surname was already used in the 17th century. Still, the argument remains cogent from the statistical point of view: the number of examples in Table 2 is large, and, moreover, this table is not exhaustive.

Some of the last names appearing in sources from the 18th century are derived from male given names. For them, these early references can correspond either to patronymics or to actual hereditary surnames. Examples include: Aronashvili, Biniashvili, Davitashvili, Iosebashvili, Janashvili, Ḳobiashvili, Ḳoḳiashvili, Mamistvalashvili, Manasherashvili, Shamuelashvili, and Ṭsiṭsuashvili. The exact status of these forms can be firmly determined only using genealogical data. However, some general information can be also useful. Firstly, one can compare the geographical distribution of these names in the 18th century and during the following period when their status as hereditary family names is already doubtless. A document compiled during the 1730s refers to Shamuelashvili Ǩobia (Berdzenishvili 1940, p. 169). No other mention of Shamuelashvili is found: it does not seem to be used in the 20th century as a surname. Here, the odds are high that the form appearing in the 1730s is a patronymic. The name Manasherashvili appears in Tskhinvali in 1709 borne by a family from Imereti. Later, we find numerous other references to Manasherashvili in Tskhinvali and its area, whereas in other regions, this name was rather unusual. Consequently, the odds are high that these later references correspond to descendants of the migrants from Imereti for whom, in turn, this name was either their surname or the patronymic that became a fixed family name for the next generations. Secondly, the less common the given name for Jews, the larger the chances for the name in -shvili based on it to be hereditary. According to this criterion, Mamistvalashvili (derived from a Georgian-based given name) is much more likely to be a surname already during the 18th century than, say, Aronashvili, Davitashvili, and Iosebashvili (based on local forms of biblical Aron, David, and Joseph, respectively). For last names known in the 18th century whose roots are not male given names, we are dealing with surnames. This is the case with several names of toponymic origin (Khoṭoveli and Ḳrikheli), names based on female given names (Tetruashvili), or those derived from nicknames (Khundiashvili).

3. Endings of Surnames

Table 3. Percentages of surnames with various endings.

-shvili	-dze	-oti	-eli	-ovi/-evi	Other	Total Number of Surnames
92	2	1	1	2	2	379

presents the statistical distribution of surnames used by Georgian Jews during the 19th–20th centuries by their endings.22

One can observe that a large majority of surnames end in -shvili. This word means ‘child’ in Georgian. It is applicable independently of the gender of the person and his/her age. The number of surnames with other endings is small. Among those ending in -dze ‘son’ are: Babaladze, Beridze, Jinjikhadze, Ḳatsobadze, Khisḳiadze, Manasheridze, Pichkhadze, and Shamashidze. In this list, Pichkhadze—as discussed in the previous section—is the oldest known Georgian Jewish surname, attested for the first time during the first half of the 17th century. For this reason, it is commonly used. Others are significantly younger, and all of them are unusual: the odds are high that many of them represent secondary forms derived from the original surnames ending in -shvili.23 This scenario is particularly plausible for Jinjikhadze. On the one hand, Jinjikhashvili is an old and common surname. On the other hand, the root of these two forms is not a male given name: this factor significantly diminishes the probability of their independent inception. Yet, in pairs such as Babaladze/Babalashvili and Ḳatsobadze/Ḳatsobashvili, the two elements can be, at least in theory, independent and taken by sons or grandsons of different Jews called Babala and Ḳatsoba.

Several surnames end in the Georgian demonymic suffix -eli. They are based on toponyms located in various regions: Khoṭeveli and Ḳrikheli in the province of Racha, Ksovreli in western Kartli, Diǧvireli in the area that today corresponds to northeastern Turkey.

Among Georgian Jewish surnames ending in -ovi/-evi, we find: Charukhchevi, Ieḳutielovi, Ḳobaivanovi, Ḳoḳielovi, Natanelovi, and Zizovi.24 This ending comes from the main Russian patronymic suffix -ov, the variant -ev being used in Russian after palatalized consonants (of which the Russian ̀Cyrillic ч transliterated in this article as ‘ch’ is an example). Their final -i represents the Georgian nominative singular suffix, automatically added in the Georgian spelling but absent from the Russian one. For this series of names, we can be sure that they appeared during the 19th century, after the incorporation of Georgia to the Russian Empire. Several alternative scenarios can be suggested for their inception, and only genealogical data could help to determine which one of them was valid. Firstly, these names could be simply created. For example, sons of bearers of biblical names Iekutiel ‘Jekuthiel’ and Natanel ‘Nathaniel’ could acquire the surnames Iekutielov(i) and Natanelov(i), respectively. Secondly, we cannot exclude a possibility that the first bearer was a Mountain Jewish or Bukharan Jewish migrant to Georgia. For example, Zizov(i) could be a corrupted form of Azizov, a surname used by Mountain Jews and based on a given name Aziz, borrowed from Muslims. Note that given names of this kind were much more common among Mountain Jews who lived in the territory where the majority was Muslim than among Georgian Jews whose non-Jewish neighbors in Kartli and Imereti were Christians. Finally, the surnames ending in -ov(i) or -ev(i) could be secondary, Russified forms of the previous, unattested forms with non-Russian endings. For example, *Charukhchi, *Ḳobaivani, and *Ḳoḳielashvili became Charukhchev(i), Ḳobaivanov(i), and Ḳoḳielov(i), respectively.

Contrary to the first two purely theoretical scenarios, the last one is corroborated by the existence of numerous secondary Russified surnames in which their final Georgian ending was replaced with -ov or -ev in Russian-language administrative documents. This phenomenon was commonplace for Georgian Christians (including nobles) for whom the Russian ending replaced one of the two most common Georgian endings, -dze or -shvili. Since Georgian Jewish surnames mainly end in -shvili, it was this suffix that was replaced. Thus, we have such pairs and Aivazashvili and Aivazov(i), Aṭanelashvili and Aṭanelov(i), Baazashvili and Baazov(i), Khakhanashvili and Khakhanov(i), Mamistvalashvili and Mamistvalov(i), Manasherashvili and Manasherov(i), Ṗaṗismedashvili and Ṗaṗismedov(i), Ṗaṭarḳatsishvili and Ṗaṭarḳatsov(i), Rapaelashvili and Rapaelov(i). The Russified forms were typical for northwestern Kartli, that is, Tskhinvali and its area.25 In theory, we cannot exclude a possibility that in some cases, bearers of the surnames ending in -shvili and -ov(i) belong to independent families. This phenomenon is particularly plausible for (potentially) polygenetic surnames based on common biblical male given names: Abramashvili and Abramov(i), Aronashvili and Aronov(i), Danielashvili and Danielov(i), Solomonishvili and Solomonov(i). If the same forms of given names were used by various Jewish groups, then the origin of specific branches becomes particularly ambiguous. For example, Abramov(i) and Aronov(i) in Tbilisi can be, in principle, of Mountain Jewish, Georgian, Ashkenazic, or Bukharan origins.26 In other cases, distinctions can be made due to differences in the pronunciation of Hebrew in various Jewish traditions. For example, biblical names ending in -el(i) in Georgian usually end in -il for Mountain and Bukharan Jews. For this reason, we have such doublets as Danielov(i) and Danilov(i), Rapaelov(i)/Rafaelov and Rafailov(i), in which the first forms most likely reveal Georgian Jews, whereas the second ones are typical for Mountain Jews. Iekutielov(i) cannot be of Ashkenazic origin, because biblical Jekuthiel is pronounced with internal /s/ in Yiddish (compare Standard and Lithuanian Yiddish Yekusiel). Not all Russified forms ending in -ov or -ev appearing in historical documents were valid for the same persons in other documents. Some spellings seem to be due to the Russification made by particular scribes: these forms do not appear in more recent sources. A Russian document from 1828 is an example. In it, all Georgian surnames, Jewish and non-Jewish, are Russified. We find there not only Krikhelov (instead of Krikheli), but also several forms in which the ending -shvili was changed to -shvilev: Khundiashvilev, Nanikashvilev, and Khakhanashvilev.27

In certain documents from Tskhinvali and its area, we find forms ending in -ant (or -ent): Biniaant, Davaraant, Mamistvalant, and Ṗaṭarḳatsient.28 In Georgian, forms with this ending have the meaning ‘belonging to the house/family of.’ For example, Biniaant can designate any member of the Biniashvili family. This way, such forms are no more than morphological variants of other surnames, those ending in -shvili. For this reason, the suffix -ant is ignored in Table 3. In Akhaltsikhe, a series of surnames end in -oti or -ati: Bajoti, Baḳaloti, Injabelati, Izhoti, Ḳazhiloti, Ǩorlati, Piṭimati, Shalkhoti, and Ṭetsoti. To the same group belong Ḳatsoṭi, Ḳezeroṭi, Shalolati, and, maybe, also Ṭeṭrodi.29 This suffix should represent a dialectal form of -ant.30 This idea would explain their morphological and phonetic similarities and the existence of such pairs as Injabelati and Injabeli, Izhoti and Injashvili, Ḳatsoṭi and Ḳatsoshvili, Ǩezeroṭi and Ǩezerashvili, Shalolati and Shalolashvili, Ṭetsoti and Ṭetsoshvili. The only surnames ending in -oti considered in Table 3 are those for which no other form with the same root was found.

Remaining Georgian Jewish surnames have various endings: Shishiani (of uncertain derivation), a few family names coinciding with full forms of male given names such as Gabrieli ‘Gabriel’ and Rakhavia, several forms ending in Georgian diminutive suffixes (Babaluḳi, Bibiluri, and Solomonia, with the suffixes -uḳi, -uri, and -ia), surnames coinciding with Georgian common nouns (Mashia ‘slipper’ and Poladi ‘steel’31), Khakhmigeri ‘stepson of khakham (a person having religious responsibilities in the Jewish congregation).’ For the last surname, we also find its morphological variant, Khakhmigerishvili, which most likely represents a secondary form created as a result of standardization, that is, influenced by the fact that a large majority of Georgian Jewish surnames end in -shvili. From the semantic point of view, the opposite process—the creation of Khakhmigeri after dropping -shvili in Kakhmigerishvili—is significantly less plausible: the expression ‘child of stepson (of khakham)’ could hardly be present already in the original form. This way, the creation of the surname Khakhmigeri, without adding -shvili ‘child’ and -dze ‘son’, is not a surprise: all three words, geri, shvili, and dze, belong to the same semantic group. Contrary to Khakhmigeri, all the other names in the above list are unusual. Some of them could be secondary forms that appeared after the shortening of the original forms. This idea is particularly attractive for Gabrieli: the form Gabrielashvili exists as well. It is a good candidate for being the source for Gabrieli. Similarly, Mashia could be drawn from Mashashvili. For Solomonia and Mashia, we cannot exclude dealing with forms that originated in Mingrelia, where -ia is the most standard ending of surnames used by Christians.

For Georgian Christians, surnames ending in -eli appear in various regions, but their total number is small. Other endings depend primarily on the region. In western provinces such as Imereti and Guria, surnames ending in -dze are by far the most commonly used (about two thirds of the total), though those ending in -shvili cover 20–25 percent. In eastern areas such as Kakheti and the eastern part of Kartli, almost all surnames end in -shvili. The transitory area between the -dze and the -shvili territories corresponds, in the north, to Racha (the region of the town of Oni, with comparable proportions of both groups of surnames) and, in the south, to western Kartli (the Gori area). In the latter, in towns situated on the Kura River, the proportion between names with these two endings changes dramatically from west to east: in Khashuri, 76 percent end in -dze and 24 percent in -shvili; in Kareli, 40 percent and 59 percent; in Gori, 28 percent and 66 percent; and in Kaspi, 10 percent and 85 percent. In Mingrelia, surnames mainly end in -ia, -ua, or -ava. In Svaneti (northwestern mountainous part of Georgia), surnames ending in –(i)ani ‘belonging to’ largely dominate. In northeastern Georgia, surnames usually end in -uri or -uli. In Guria, a small group of surnames has the ending –(n)ṭi, unknown in other regions.32

Comparing the above information about Jewish and Christian surnames, we can observe that both groups show similar behavior for surnames ending in -eli only. Exceptional Jewish surnames ending in -ia and -uri are unattested in the areas where Christian surnames with these endings were frequently used. A few Jewish surnames ending in -oṭi/-oti are not from Guria: as explained above in this section, they have no link to Gurian Christian surnames ending in –(n)ṭi.

The fact that a large majority of surnames end in -shvili represents the main peculiarity of the Jewish corpus. These surnames were found not only in areas that were transitory for Christians between the -dze and -shvili territories (near such localities as Oni and Kareli), but also in Imereti, where such surnames were significantly less common for Christians than those ending in -dze. Available data from the 1886 census allow for a detailed comparison between these two religious groups for the regions of Racha and Lechkhumi. In the town of Oni, the administrative center of Racha, all seventeen surnames borne by Jews end in -shvili. Among surnames in the same town based on the Georgian language that were borne by non-Jews (mainly Georgian Christians and also possibly some Armenians), 16 end in -shvili (27 percent), 32 in -dze (54 percent), and 11 have other endings (19 percent). For Georgian Christians who were the only inhabitants of the village of Ḳvatskhuti situated 25 km west to Oni, the distribution is similar enough as for non-Jews of Oni: 11 surnames end in -shvili (30 percent), 23 in -dze (62 percent), and 3 in other endings (8 percent).33 One can observe that Georgian Christians of Racha used both endings, -dze and -shvili, with a clear preference for -dze. In Lechkhumi, a region situated west of Racha and north of Imereti, the difference between Jews and non-Jews is even more striking. In the town of Lailashi, the only place in Lechkhumi where the population was not exclusively Georgian, all twelve Jewish surnames end in -shvili. Among surnames borne by Christians, only one ends in -shvili, six in -dze, one in -ani, one in another Georgian ending, and three (most likely borne by Armenians) in the Russian suffix -ov. In the neighboring village with the same name of Lailashi, all inhabitants were Georgian Christians. Local families had names with the following endings: eleven -dze, ten -ani, one -shvili, and three others.34 One can observe that for Christians of Lechkhumi, surnames ending in -shvili were marginal. In this way, it is unlikely that Jews of Lailashi acquired their surnames locally during the same period as Christians. More likely, either they brought them as ready-made forms when moving there from another area, or they adopted them locally during a different period.

Certain authors who are knowledgeable about the main features of the geographic distribution of Georgian Christian surnames and the fact that almost all Georgian Jewish surnames end in -shvili suggest the following scenario: initially, Jewish surnames originated in eastern Georgia, and later, Jews bearing these surnames migrated to western Georgian territories.35 Formulated this way, this scenario is too simplistic and implausible. It should surely be nuanced. As discussed in Section 3, historical sources show that for the last few centuries, the cradle of Georgian Jewry was situated in a compact area covering western Kartli (Tskhinvali, Kareli, and their areas), Meskheti (Akhaltsikhe and its area), and Imereti (Kutaisi and its area). In Tskhinvali, the ending -shvili was typical for Georgian Christian surnames. For example, the census of 1781 lists surnames for 108 Christian families. Among them, 88 end in -shvili (80 percent) and 22 in -dze (20 percent).36 As indicated above in this section, for Christians of Kareli, -shvili was also more commonly used than -dze. Moreover, even in 1845, the ending -shvili was used in Georgian Christian documents to indicate patronymics in western Kartli.37 These factors show that in this area, the Jewish pattern of forming surnames was clearly influenced by the local Christian culture. They do not explain, however, the absence of Jewish surnames ending in -dze in that area. Two additional factors could be important here.

The first of them is related to chronology rather than geography. The adoption of surnames by Georgian Jews seems to be recent enough, and in Georgia, surnames ending in -shvili appear to be precisely more recent than those ending in -dze. For Jews, the last assertion is corroborated by the fact that the three oldest known Jewish last names ending in a patronymic suffix end in -dze: Buǧapaisdze, Eliozisdze, and Pichkhadze. For Christians, several indirect corroborations of the same idea can be found in areas where both categories of surnames were present. Firstly, as it is true for numerous other cultures, it is logical to consider that the classes situated at the top of the social hierarchy acquired surnames before other social groups. The Georgian society of the 19th century distinguished three classes: tavadebi ‘princes’, aznaurebi ‘gentry’, and glekhebi ‘peasants.’ In the census of the province of Racha (1886), only eleven different surnames were borne by tavadebi: five ending in -dze, six ending in various other suffixes, and none ending in -shvili. The distribution for aznaurebi was quite different: 35 ending in -dze (49 percent), 19 in -shvili (26 percent), and 18 in other endings (25 percent).38 For peasants, as discussed above, it was similar enough to that of aznaurebi. Secondly, in Tbilisi, the number of different surnames ending in -shvili is larger than that ending in -dze: this is not a surprise, because Tbilisi is situated deeply in the -shvili territory. Yet 45 percent of inhabitants have surnames ending in -dze, and only 30 percent have surnames ending in -shvili (Nikonov 1988, p. 154). Consequently, the frequency of use of surnames ending in -dze is significantly larger than the same parameter for surnames ending in -shvili. A conjecture about surnames ending in -dze being older than those ending in -shvili can explain this phenomenon.39 In other words, Jews could acquire their family names when the addition of -shvili was already the main pattern. Moreover, the same conjecture can help to explain the fact that 20–25 percent of Georgian Christians have names ending in -shvili in Imereti. In this province, these names ending in -shvili could, at least in theory, cover the younger layer of surnames, those assigned following the new pattern relatively recently to people—including local Jews—having no surnames at that moment. For Meskheti, the chronological factor could be decisive. All surnames from Akhaltsikhe and its neighboring villages known to us appear in sources compiled after 1829, when the area was already cut from the Ottoman empire and attached to Georgia, which in turn was already a part of the Russian Empire. These surnames mainly end in -shvili, and none of them end in -dze.40 It is plausible that all names ending in -shvili appeared there already after the end of the Ottoman period following the Georgian standard pattern at that time.41 It is also possible that in the Georgian dialect spoken in Akhatsikhe, -shvili rather than -dze was the standard ending for patronymics.

The second factor explaining the large prevalence of names ending in -shvili for Jews could be social. For Christians in central Georgia, it does not seem to have a significant role: names ending in -shvili and -dze were commonly used by peasants and the gentry, and in western Kartli, some princes bore names ending in -shvili too.42 Yet it is possible that it became customary for Christian clerks to record Jews—a special religious sub-group within that of serfs—with last names ending -shvili.43

4. Other Morphological Peculiarities

A large majority of surnames of Georgian Jews are based on the Georgian language. For this reason, an understanding of their structure would be impossible without indicating basic morphological features of this idiom.44 Globally speaking, morphological peculiarities of Jewish surnames in Georgia are not specifically Jewish. We find them in the surnames of Christians too.45

Surnames and their roots mainly have the form of a noun or an adjective. The exact forms of Georgian nouns and adjectives depend on the grammatical case.46 If the stem ends in a consonant, the nominative singular form acquires the ending -i. If the stem ends in -a, -e, -o, or -u, this form coincides with the stem. Genitive singular forms acquire the ending -is for stems ending in a consonant. If the stem ends in -a or -e, this vowel is dropped when adding -is. We find an illustration for this rule in mam-is-tval-i ‘father’s eye’ (compare mama ‘father’ and tval-i ‘eye’) and ṗaṗ-is-imed-i ‘grandfather’s hope’ (compare papa ‘grandfather’ and imed-i ‘hope’), the bases for the surnames Mamistvalishvili and Ṗaṗisimedishvili, respectively. However, in genitive forms of personal names ending in -a or -e, this final vowel is usually kept, and the /i/ of the suffix -is is dropped.

Etymologically, the most common patterns of forming last names in Georgia—those with the endings -dze ‘son of’ and -shvil-i ‘child of’—involve the genitive case of the preceding word. For this reason, these forms can serve as a good illustration of the development of the above rules. The oldest known examples include the genitive suffix -is: compare Buǧapa-is-dze (1260, with the final stem vowel -a not dropped) and Elioz-is-dze (circa 1520). In certain forms appearing in the 17th century, we can already observe the elision of the initial vowel of the suffix -is: Matavala-s-shvil-i and Tsetsela-s-shvil-i. Numerous examples of forms including the suffix –is or, much more commonly, its shortened form -s appear in the 18th century. Among them are: Batina-s-shvil-i, Besiḳa-s-shvil-i, Chlaḳ-is-shvil-i, Davita-s-shvil-i, Guguna-s-shvil-i, Jana-s-shvil-i, Ḳobia-s-shvil-i, Ḳoḳia-s-shvil-i, Mosia-s-shvil-i, Shabata-s-shvili, Shalela-s-shvil-i, Shedana-s-shvil-i, and Ṭsiṭsoa-s-shvil-i. A few examples are known in the first half of the 19th century too: Ṭsamala-s-shvil-i and Tetrokala-s-shvil-i. Gradually, this internal genitive-case-related /s/ disappeared with only the forms in which -shvil-i or -dze follow a kept vowel.47 For Jews, the earliest examples are known already in the 17th century: compare Pichkha-dze and Khakhana-shvil-i. During the second half of the 19th century, the internal /s/ was dropped altogether in Georgian surnames (independently of the religion), remaining in certain written sources only inside of patronymics.48

In genitive forms, during the addition of -is, the elision of the last vowel present in the stem can take place. For stems ending in a sonorant preceded by a vowel such as -al, -am, -an, -ar, -el, -ol, or -or, the disappearing of this vowel is regular.49 Multiple examples can be found in Jewish surnames: bagdadel-i > Bagdadl-i-shvil-i, kachal-i > Kachl-i-shvil-i, khakham-i > Khakhm-i-shvil-i, mṭsḳeṗeli > Maṭsḳeṗlishvili, modzǧvar-i > Modzǧvr-i-shvil-i, natsval-i > Natsvl-i-shvil-i, potol-i > Potl-i-shvil-i, roḳeṭel-i > Roḳeṭl-i-shvil-i, and ṭsitel-i > Ṭsitl-i-shvil-i.

The stem of several surnames ending in -shvili represents a compound word composed of two stems. In Georgian, several methods are used to form such words. One of them was already discussed above, namely, a combination of two nouns in which the first one appears in a genitive case: Mamistvalishvili and Ṗaṗisimedishvili. Another method uses a simple concatenation of the two stems, without any connecting element. In this case, the stem determining the semantic category of the compound can be in either the first or the second position (Chikobava 1967, pp. 35–36). Examples of stems combining two nouns or an adjective and a noun include: akhal-i50 ‘new’ + ḳats-i ‘man’ > Akhal-ḳats-i-shvil-i, ḳaṭa ‘cat’ + ṗaria ‘thief’ (the combination of these roots also means ‘valerian’) > Ḳaṭa-ṗaria-shvil-i, ṗaṭara ‘small’ + ḳats-i ‘man’ > Ṗaṭara-ḳats-i-shvil-i, tav-i ‘head’ + did-i ‘big’ > Tav-did-i-shvil-i. In surnames, the last vowel of the first stem or the first vowel of the second one can be dropped: the variants Ṗaṭarḳatsishvili and Ṗaṗismedashvili are more common than Ṗaṭaraḳatsishvili and Ṗaṗisimedashvili. A small set of names based on nicknames combine a noun (ḳaṭa ‘cat’, khakhvi ‘onion’, tapli ‘honey’) and ċhamia ‘eater’: Ḳaṭaċhamiashvili, Khakhviċhamiashvili, and Tapliċhamiashvili.

Georgian participles are formed by adding the prefix m- and the suffix -el-i to the verbal stem (Chikobava 1967, p. 52). The surname Maṭsḳeṗlishvili is an example. Its initial part, m-ṭsḳeṗ-el-i, is a participle of the verb ṭsḳeṗa ‘to cane’. Names of certain occupations are constructed by adding the prefix me- and the suffix -e to the object of the work (Chikobava 1967, p. 34). Meḳinulashvili illustrates this pattern: its root *me-ǩinul-e is related to ǩinul-i ‘ice’.51 The suffix -el-i is also used to create demonyms: the names of inhabitants of various places. It is usually added to the toponym stem: Al-i > Al-el-i-shvil-i, Aṭen-i > Aṭen-el-a-shvil-i, Diǧvir-i > Diǧvir-el-i, Gor-i > Gor-el-i-shvil-i, Khoṭev-i > Khoṭev-el-i, and Ḳrikh-i > Ḳrikh-el-i. For toponyms ending in -a or -e, this final vowel of their stem is dropped: Guria > Guri-el-i-shvil-i. If a place name ends in the suffix -et-i ‘land of’ or -is-i, these endings are also dropped: Ajam-et-i > Ajam-el-a-shvil-i, Ksovr-is-i (or Ksovr-et-i) > Ksovr-e-l-i, (M)dzor-et-i > Dzor-el-a-shvil-i.

The surname Ḳobaivanov(i) is a Russified form, with the Russian suffix -ov added to an original name. Its initial letters coincide with the male given name Ḳoba used by Georgian Jews. The remaining part, -ivan-, looks like a suffix that could have here either a patronymic or a diminutive role.52

A large number of surnames are based on diminutive forms. Modern Georgian uses the following suffixes to construct such forms: -a, -ilo, several suffixes with /ḳ/ (such as -aḳ-i, -iḳ-i, -iḳo, and -uḳa), -una and -unia, -uchuna, -utsuna, and -utsana.53 The suffix -a is particularly common. When it is added to base forms having a stem ending in a consonant, the elision of the final -i (present in the nominative case) is optional. It is in surnames drawn from given names that -a and other diminutive suffixes are particularly commonly present.

Table 4. Diminutive forms of Christian and Jewish given names.

Suffix	Christian Examples	Jewish Examples
a	Davit-a, Dat-a; Gabriel-a, *Gab-a, *Gabr-a; Giorg-a, Gi-a, Gig-a, Gog-a	Abner-a, Abram-a, Dat-a, Davit-a, Eliǩazar-a, Eliǩezer-a, Khaim-a, Ioseb-a, Iskhaḳ-a, Israel-a, Levi-a, *Manasher-a, Mardakh-a, Sepi-a, Shat-a, *Shamuel-a
i-a	Dat-i-a; Gabel-i-a; Gog-i-a	*Aḳ-i-a, *Aron-i-a, *Bat-i-a, *Ber-i-a, Dat-ia, Davit-i-a, Jan-i-a, Ṗaṗ-i-a, *Sar-i-a, *Shash-i-a, *Shat-i-a, *Shimsh-i-a, *Solomon-i-a, *Tsits-i-a, *Ṭurp-i-a
o	Dat-o; *Gabro; *Gogo	*Aḳ-o, Bin-o, El-o, Ḳats-o, Shal-o
o-a	Dat-o-a; Gig-o-a	*Tetr-o-a
u-a, vi-a, v-a	Dat-u-a, Dat-vi-a; Gabu-a; Gigu-a, Gogua, *Gogva	Bal-u-a, Bat-u-a, *Chiḳ-v-a, Dat-u-a, El-u-a, *Lel-u-a, Shab-u-a, *Shash-u-a, *Tetr-u-a, *Tsits-u-a
an-a, en-a, in-a, on-a, un(-i)-a	Dat-in-a, Dat-on-a, Dat-un-a; Gab-un-i-a; *Gog-an-a, Gog-in-a	Bat-in-a, Bat-un-a, Ber-un-a, Bin-in-a, *Mosh-en-a, *Ṗaṗ-in-a, Ṗaṗ-un-a, *Shash-un-a
aṭ-a, eṭ-a, iṭ-a, oṭ-a, uṭ-a	Gab-iṭ-a; Gi-guṭ-a, Gog-aṭ-a, Gog-iṭ-a, *Gog-oṭ-a	No example found
aḳ-a, eḳ-a, iḳ-a, oḳ-a, uḳ-a, uḳ-i	*Dat-iḳ-a, Dat-uḳ-a, Dat-uḳ-i	Babal-iḳ-a, Babal-uḳ-i, Bach-iḳ-a, Bat-oḳ-a, Ber-iḳ-a, Bin-iḳ-a, Dat-iḳ-a, Dat-uḳ-i, *El-iḳ-i, *Mikh-aḳ-a, *Nan-iḳ-a, *Shal-iḳ-a, *Shaml-iḳ-a, *Susun-iḳ-a, *Ṭaṭ-iḳ-a
al-a, el-a, il-a, ol-a, ul-a	Gab-il-a;55 Gig-ol-a, Gog-el-a, Gog-il-a	Bab-al-a, *Ǩob-el-a, *Ḳoḳi-el-a, *Mir-il-a, *Shal-el-a, *Shal-ol-a, *Shapat-el-a
ur-i	Gig-a-ur-i, Goga-l-a-ur-i	*Bab-ur-i, *Bibil-ur-i, *Binia-uri56
ab-a, eb-a, ib-a, ob-a	*Gig-ab-a, *Gog-eb-a, Gog-ib-a	Ḳats-ob-a, *Tot-ob-a
ach-i, ich-a, och-a, uch-a	Gig-ich-a, Gog-ich-a, *Gog-och-a, *Gog-uch-a	*Ber-ich-a, *Ber-uch-a, *Ḳaḳ-ach-i, *Lel-uch-a

presents a list of suffixes used in hypocoristic forms of Georgian given names. Data for the second column, the one dealing with Christian examples, correspond to forms of only three full names: biblical Davit-i ‘David’ and Gabriel-i ‘Gabriel’ (they start with Da- and Gab-, respectively) and Giorg-i ‘George’, one of the most commonly used names for Georgian Christians (all other forms).54 In the second and the third columns, given names reconstructed from surnames are preceded by the asterisk sign (*). For almost all personal names appearing in the last column, known Jewish surnames end in -shvili. The only exceptions—surnames coinciding with hypocoristic forms of given names appearing in this column—are underlined. The given names of the last column put in italics are female. Numerous Jewish forms from Table 4 were not limited to Jews but were used by Christians too.

The addition of a diminutive suffix represents the most common method of constructing hypocoristic forms of personal names. For some names in Table 4, the addition of the suffix was concomitant to the root truncation. The parts dropped can be internal (Data from Davita, *Shata from Shabata), initial (*Bata from Shabata, Ḳoba from Iaḳobi, *Rami from Abrami), final (*Shimshia from Shimshoni), or initial and final simultaneously (*Aḳia and *Aḳo from Iaḳobi). The reduplication of one of the syllables or just its consonants is another method. For Christians, we find it, for example, in *Gaga from Gabriel.57 Among initial parts of the Jewish surnames ending in -shvili, we find such examples as *Jajana, Ḳaḳa, Ḳoḳo, *Khakhua, *Khukha, and *Shashua.

As it can be seen from Table 4, the majority of hypocoristic forms of given names end in -a. For this reason, surnames based on such forms end in -ashvili (for Georgian Christians and Jews) or -adze (almost exclusively for Georgian Christians). For Jews, such surnames are particularly common. A statistical analysis of 184 Jewish patronymic surnames ending in -shvili shows the following distribution by the vowel that precedes this ending: 78 percent of ‘a’, 7 percent of ‘i’, 6 percent of ‘o’, and 11 percent of surnames for which two variants, one with ‘a’ and another with ‘o’, exist.58

We also often find the suffix -a in surnames based on sobriquets having a form of a diminutive noun based on physical, moral, or other characteristics (Khundiashvili from khund-i ‘small dove’, Maǧalashvili from maǧali ‘tall’, Pichkhadze from pichkh-i ‘branch cuttings’), occupational terms (Topchiashvili from topch-i ‘gunsmith’), and even toponyms (Ajamelashvili, Aṭenelashvili, Digurashvili, Dzorelashvili). In total, among 82 Jewish surnames of these types ending in -shvili, the preceding vowel is: ‘a’ (58 percent), ‘i’ (24 percent), ‘o’ (2 percent), and one variant with ‘a’ and another with ‘o’ (16 percent).59

In theory, the existence in Georgia of numerous patronymic surnames ending in -ashvili or -adze could contribute to the creation of a standardization pattern (a) either provoking a change of the vowel in certain surnames that were originally ending in -ishvili or -idze, or (b) used directly to construct surnames, as if the actual patronymic suffix were not -shvili and -dze, but -ashvili and -adze. The last method could have been reinforced by the fact that the final -i of the word preceding the patronymic suffix is not a part of the stem: this -i disappears in grammatical cases other than the nominative. For example, Maǧalashvili could appear as (1) a patronymic surname based on the personal nickname Maǧala (that, in turn, represents a diminutive form of maǧal-i ‘tall’), (2) a secondary form of the original surname *Maǧalishvili, or (3) a patronymic surname based on the personal nickname Maǧali. The first scenario sounds the most plausible, because it is the simplest one. To corroborate the second scenario, one would need to find references to the presumed primary form *Maǧalishvili. The absence of such reference makes this scenario implausible for this specific surname. Yet we have an example of the change in the opposite direction. The form Mamistvalashvili is present in numerous sources from the 17th–19th centuries. It is only in the 20th century that we find the first reference to Mamistvalishvili, and the odds are high that the last form is secondary.

It is in the 15th century that hypocoristic forms of male given names start to appear as the roots of surnames of Georgian Christians. At that moment, such names are restricted to low social strata. During the next two centuries, they became commonly used in all social groups.60 Yet, even during that period, hypocoristic forms ending in suffixes including /ḳ/ were rare. All such forms present in the second column of Table 4 are drawn from Davita ‘David’, and no example derived from Gabriel or Giorgi is known. Yet surnames based on such forms are quite common for Georgian Jews. There is no reason to consider that we face here a Jewish peculiarity. Indeed, as indicated above, -aḳ-i, -iḳ-i, -iḳo, and -uḳa appear in a relatively short list of diminutive suffixes commonly used in modern Georgian. Since the second, Christian, column of Table 4 is based on sources from the 15th–17th centuries, the explanation of the observed difference is chronological. Jewish names based on hypocoristic forms with /ḳ/ are more recent: they are likely to be formed after the 17th century.

As explained in the previous section, the suffixes -at(i) and -ot(i) are dialectal variants of the standard Georgian suffix -ant ‘belonging to a family/house of’, and a series of forms ending in these suffixes (including Ḳazhiloti) is attested in Akhaltsikhe. In Tbilisi, we find a few examples of surnames ending in -oṭi: Ḳatsoṭi, Ḳazhiloṭi, and Ḳezeroṭi. The change from /t/ to /ṭ/ could be operated locally under the influence of the existence of various Georgian diminutive suffixes including this consonant (compare the line with aṭ-a, eṭ-a, iṭ-a, oṭ-a, uṭ-a in Table 4) and because in Tbilisi’s Georgian, the suffix -ot(i) is unknown.

5. Phonetic Peculiarities

The Georgian alphabet is phonetic: one letter corresponds to one sound, and vice versa. Distinctive phonetic features mainly concern consonants. One of them is the absence of /f/ in Georgian, the closest sound being the aspirated p. For example, biblical Ephraim is Epremi in Georgian, this form being the basis for the surname Epremishvili.61 The letter f—present in the Russian variant spelling of the same surname, Efremashvili (Eфpeмaшвили)—is either related to the Russian colloquial form Efrem ‘Ephraim’ or to the direct influence of the Hebrew form. The same absence of /f/ in Georgian also explains the form Pizitsḳi for the Ashkenazic surname spelled Fizycki in Polish and Fizitskiy (Φизицкий) in Russian.

Georgian has two velar fricatives: voiced ღ (transliterated in this article as ǧ) and its unvoiced equivalent ხ (kh). The first of them does not exist in standard Russian, in which the closest sound is the stop г (g), the direct equivalent to the Georgian letter გ (g).

The existence of two sets of stops and affricates, ejective (pronounced with a glottalic airstream) and aspirated, represents one of the main peculiarities of the Georgian phonology.

Table 5. Two sets of Georgian unvoiced stops and affricates.

	Stop				Affricates
	Labial	Dental	Velar	Uvular	Alveolar	Post-Alveolar
Aspirated	ფ (p)	თ (t)	ქ (k)	-	ც (ts)	ჩ (ch)
Ejective	პ (ṗ)	ტ (ṭ)	კ (ḳ)	ყ (ǩ)	წ (ṭs)	ჭ (ċh)
Russian transcription	п (p)	т (t)	к (k)	к (k)	ц (ts)	ч (ch)

presents both sets, indicating in every cell the Georgian letter followed by the Latin-based character used in this article for its transliteration. The last line indicates the Russian transliteration of these consonants, followed, in the parentheses, by the transliteration of these Cyrillic letters to the Latin characters used in this article. Note that the same Russian letters are used in various columns for both the aspirated and the ejective Georgian consonants. Moreover, the same Cyrillic к /k/ is used for three different Georgian consonants, the velar and the uvular stops.

In the 20th century, for some Jewish surnames, several alternate Georgian spellings are attested. Their existence can be related to several independent phenomena. An intermediary role of Russian could be one of the reasons for the variation. As explained in the two previous paragraphs, several contrasts existing in Georgian do not exist in Russian. Since Russian was the official language of the Russian Empire and—later, between 1922 and 1991—the USSR, the Russian spelling was the basic one for certain surnames in Georgia. Some alternative Georgian spellings could appear after the incorrect back transcription from Russian to Georgian—for example, from Georgian Aṭanelashvili (ატანელაშვილი) to Russian Atanelashvili (Aтaнeлaшвили) to Georgian Atanelashvili (ათანელაშვილი). Several other pairs could appear in a similar way: Natanelovi and Naṭanelovi, Taronishvili and Ṭaronishvili, Apriamashvili and Aṗriamashvili, Tsitsuashvili and Ṭsiṭsuashvili, Tsotsolashvili and Ṭsoṭsolashvili, (Eli)ǩezerashvili and (Eli)ḳezerashvili.

Certain other reasons for variation are phonetic. In Georgian, the assimilation of consonants is a common phenomenon (Chikobava 1967, p. 28). A voiced consonant placed before an unvoiced one also becomes unvoiced. We observe this regressive assimilation in Ekhisḳelashvili derived from Ekhizḳelashvili62 and Shaptoshvili from Shabtoshvili. The assimilation of consonants can also occur at a distance: Chanchanashvili from Chanchalashvili. A similar phenomenon, with a loss of voice, can be responsible for the given name Shapata, a phonetic variant of Shabata. These two forms gave rise to the surnames Shapatashvili and Shabatashvili, respectively. In the vicinity of an ejective consonant, an aspirated one can be replaced with its ejective counterpart: compare Roḳeṭlishvili from Roḳetlishvili (the change from t to ṭ being related to the presence of ḳ)63 and Ṗiṭimashvili from Piṭimashvili (the initial consonant changes because of the presence of ṭ). Yet, in Pitimashvili from the same Piṭimashvili, we observe the replacement of the internal ejective by its aspirated equivalent because the initial consonant is aspirated. Zhuṭiashvili gave rise to the variant Ċhuṭiashvili after the replacement of the initial consonant by an aspirated unvoiced affricate because of the presence of ṭ (another aspirated unvoiced consonant) in the following syllable. Several other phonetic changes can be observed for consonants in the position before another consonant: interchanges between /m/ and /v/ (Namtalashvili, a variant of Navtolishvili, Khemsurishvili from khevsuri ‘one from Khevsureti’, Shavliḳashvili from Shamliḳashvili) and the elision of /r/ (Ṗalagashvili from Ṗarlagashvili). In several forms, we observe the interchange between /sh/ and /ch/ (Chelelashvili from Shalelashvili in Sachkhere (Imereti) and Lelushashvili from Leluchashvili),64 /ch/ and /ts/ (Tsiḳvashvili from Chiḳvashvili, Chkhvirashvili from Tskhvirashvili). Kurchishvili results from Kurkchishvili after the simplification of the consonantal cluster.

The inception of Khanuḳashvili as a variant of Khanukashvili is likely to have a morphological basis. Its root Khanuka, a male given name of Hebrew origin, was apparently re-interpreted as ending in the Georgian diminutive suffix -uḳa. Elishaḳashvili, a variant of Elishaǩashvili, could appear in a similar way, after the ending of the root -aǩa replaced with the Georgian diminutive suffix -aḳa.

The vowel /u/, when placed before /a/, sounds close to the consonant /v/. For this reason, we find such pairs as Baluashvili and Balvashvili, Chakhvashvili and Chakhuashvili, Chiḳvashvili and Chiḳuashvili, Tsitsuashvili and Tsitsvashvili.

In several cases, we observe an assimilation of vowels. For example, Aṭenelashvili is the oldest attested form, and, moreover, the /e/ in the second syllable conforms to the idea that this name is derived from the toponym Aṭeni. In more recent sources, we find Aṭanelashvili or its Russified variant Aṭanelov(i). The change was influenced by the stressed /a/ of the first syllable.65 The same phenomenon explains the variant Khoṭoveli instead of the etymological Khoṭeveli.66

Several surnames exist in two variants: one with /o/and another with /a/, or one with /i/ and another with /e/: Kosashvili and Kasashvili, Moṭsonashvili and Moṭsanashvili, Tsatsiashvili and Tsotsiashvili, Taplishvili and Toplishvili, Mirilashvili and Mirelashvili. Here, we are dealing with the intermediary role of Russian in which, in an unstressed syllable, a contrast exists neither between /a/ and /o/ nor between /i/ and /e/, and in the Russian pronunciation, in all these surnames, the accented vowel appears in the penultimate syllable -shvi-.

An additional group of variants is related to forms of biblical names coming from different biblical traditions: Judeo-Georgian and Georgian Orthodox Christian.

Table 6. Judeo-Georgian pronunciation of Hebrew.

Hebrew	Judeo-Georgian	Examples
ב (v)	ბ (b)	Abramashvili (אברהם), Gabrielashvili (גבריאל), Ribashvili (רבקה)
ח (ḥ), כ (ḵ)	ხ (kh)	Khaimashvili (חיים), Khakhmishvili (חכם), Khasidashvili (חסיד), Iskhaḳashvili (יצחק), Bekhorashvili (בכור), Mikhaelashvili (מיכאל), Iekhisḳelashvili (יחזקאל)
כּ (k)	ქ (k)	Koenishvili (כּהן), Khanukashvili (חנוכּה)
ו (w)	ვ (v)	Levishvili (לוי), Davidashvili (דוד)
ק (q)	კ (ḳ)	Ieḳutielovi (יקותיאל), Iskhaḳashvili (יצחק)
תּ ,ת (t)	თ (t)	Navtolishvili (נפתלי), Ieḳutielovi (יקותיאל)
צ (ṣ), שׂ (ś)	ს (s)	Iskhaḳashvili (יצחק), Sariashvili (שׂרה)
עּ (ʕ)	ყ (ǩ)	Eliǩezerashvili (אליעזר), Iaǩobishvili (יַעֲקֹב), Shimǩonashvili (שמעון), Elishaǩashvili (אלישׁע)
שּׁ (š)	შ (sh)	Moshashvili (משה), Shimǩonashvili (שמעון)

presents the main peculiarities of the Judeo-Georgian pronunciation of Hebrew.67 The surnames in its last column are followed in the parentheses by the Hebrew word or given name from which the surname is drawn.

For example, ṭ in Namṭalashvili can be related to the Georgian Napṭali ‘Naphtali’, and t in Namtalashvili can result from the Judeo-Georgian form of the same male name. Davitashvili and Davidashvili are based on the Georgian Christian and the Jewish forms of David, respectively. Iaḳobishvili has as its root Iaḳobi, the Georgian form of Jacob. Yet, in Iaǩobishvili, the internal ǩ comes from the Judeo-Georgian reading of the letter ayin present in the Hebrew spelling of Jacob (יעקב).68 The same alternation between the Jewish ǩ and the Christian ḳ can be observed in surnames based on hypocoristic forms of Jacob: Ǩoboshvili and Ḳoboshvili, Ǩoǩuashvili and Ḳoḳuashvili. Moshiashvili and Mosiashvili are based on hypocoristic forms of the Jewish (Moshe) and Georgian Christian (Mose) variants of Moses.

Similar to the Sephardic and different from the Ashkenazic pronunciation, the shewa under the first consonant is pronounced in Judeo-Georgian as /e/: compare Bekhorashvili from Hebrew bḵôr בְכוֹר ‘firstling’, Iekhisḳelashvili from יְחֶזְקֵאל ‘Ezekiel’, and Zebulashvili from a hypocoristic form of biblical zvûlun זְבוּלֻן ‘Zebulon.’

6. Types of Surnames

Table 7. Types of Georgian Jewish surnames.

Patronymic	Matronymic	Nickname-Based	Occupational	Toponymic	Total Number of Surnames
56%	8%	23%	7%	6%	367

provides the percentages covered by Georgian Jewish surnames of various types. It ignores names brought by Jewish migrants to Georgia and secondary surnames obtained after the change of the ending in the original, primary, surname.

One can observe that the number of surnames based on male given names is larger than that of all other types taken together. This feature is usually valid for ethno-cultural groups whose surnames were adopted relatively recently, when numerous personal patronymics acquired the status of hereditary surnames. During the period when the possession of a surname is not required by the local administration, surnames develop naturally from nicknames that distinguish their bearers from other members of the group. Yet patronymics, especially those based on common given names, can hardly serve as individual nicknames. For example, for the son of Abrami ‘Abraham’, his patronymic, Abramishvili, is not a distinctive feature if there are several men called Abrami in the same place. Surely, for tiny congregations, even this biblical name can be individual. Moreover, if a given name is specifically Jewish in a place where Jews represent a small minority, then this name can be a distinctive feature for all neighbors, Jews and non-Jews, and so a potential basis for an individual nickname. Nevertheless, even in such cases, chances for it to become hereditary are not high, since for the next generations, the memory about Abrami or his children who used the patronymic Abramishvili could be lost already.

Male given names that became sources for patronymic surnames belong to several categories. One of them encompasses biblical names that represent either Jewish or Christian forms. For this reason, we have such pairs as Davidashvili and Davitashvili, Iaǩobishvili and Iaḳobishvili, Iskhaḳashvili and Isaaḳashvili, Shimǩonashvili and Simonishvili, Moshiashvili and Mosiashvili.69 Some other surnames are drawn from forms of biblical names that were both Jewish and Christian (Abramishvili, Aronishvili, Danielashvili, Eliashvili, Gabrielishvili, Iosebashvili from Joseph, Israelashvili, Mordekhashvili from Mordecai), specifically Jewish (Iekhisḳielishvili from Ezekiel, Shimshonashvili from Samson),70 or Georgian Christian (Eliozishvili from Elijah). Numerous surnames are based on hypocoristic forms of biblical names.71 This is true for all those ending in -ashvili present in the list above, as well as multiple other names including, for example, Biniashvili and Beniashvili from Benjamin; Davitiashvili, Dat(i)ashvili, Datuashvili, and Datiḳashvili, all related to David; Ǩazarashvili from Eleazar; Ǩezerashvili from Eliezer; and Ḳob(i)ashvili, Ǩoboshvili, Ḳoḳoshvili, and Ǩoǩuashvili, all related to Jacob. A small category covers surnames based on non-biblical Hebrew given names: Bekhorishvili, Khaimashvili, and Khanukashvili. Abramkhaimashvili is derived from a double given name. A large category includes surnames based on Georgian Christian non-biblical given names. Examples include: Beridze, Darchiashvili, Gagulashvili, Ḳatsoshvili, Khakhanashvili, Mamistvalashvili, Matvalashvili, Ṗaaṭashvili, Ṗaṗiashvili, Ṗaṗunashvili, Sepiashvili, Shotashvili, and Tsitsiashvili. Often, we find in historical sources references to Jews bearing corresponding given names. For example, Jews called Matvala and Ṗaṗia appear in the second half of the 17th century (Ḳldiashvili et al. 2015, pp. 32, 373). These references show an important degree of cultural assimilation of Georgian Jews at that period. Indirectly, they also imply that these families were not recent migrants to Georgia. For some other Georgian given names, no reference of their use by Georgian Jews is found in available sources. For this reason, we cannot exclude the possibility of surnames drawn from them to be borrowed by Jews from Christians as ready-made forms. This scenario, however, remains purely theoretical. We do not find any factual corroboration for its validity. It is also unclear under what conditions such a borrowing could take place. We know no examples of Georgian Orthodox Christian masters who assigned their own names to their serfs. It would be even less plausible that before the 20th century, traditional Georgian Jews could themselves adopt the surnames of their owners72 or those of their other Christian neighbors. The idea that the surnames in question indirectly reveal the use by Georgian Jews of the corresponding male given names appears the most logically attractive. Certain Georgian Christian given names that became sources for Jewish surnames are of Turkish origin—for example, Aivaz(i), the base for Aivazashvili. A small group of surnames—Abajanashvili and Abatkhanashvili—have Muslim given names as their sources. They can reveal descendants of migrants from Persia or the Ottoman Empire. Alternatively, they can represent traces of the Persian and the Ottoman rules in the Caucasus, including the territory of Georgia. Some names could be due to Mountain Jewish ancestors. For example, Zizov(i), known in western Kartli during the second half of the 19th century, can be either a corrupted form of Azizov, a surname borne by Mountain Jews, or it can be directly based on a variant of the given name Aziz, of ultimate Arabic origin (from which the surname Azizov is derived). The surname Manasherashvili is based on Manashera, a form related to biblical Manasseh. This given name, with /r/ that is absent from the biblical form, does not appear in available sources from Georgia. Yet a close form, Manashir, was commonly used by Mountain Jews (Danilova 2000, p. 195). For these reasons, the progenitor of the Manasherashvili family could be a Mountain Jew. Alternatively, this surname can reveal common ancestry of one part of Georgian and Mountain Jews, with the form of the given name with /r/ inherited from these common ancestors.

Matronymic surnames are significantly less frequently found than the patronymic ones. This fact is perfectly in line with the traditional naming of people in Georgia by their patronymics. For Christians, surnames derived from female given names are marginal. For Jews, as it can be seen from Table 7, their part is significant enough. The analysis of female given names and family names based on them is difficult because of the extreme scarcity of references to women—other than queens—in Georgian sources written before the 19th century. The patriarchal character of the Georgian society, both Christian and Jewish, is directly responsible for this situation. The census of Kutaisi made in 1850 (Berdzenishvili 1945, pp. 315–28) is the earliest available source that provides a representative list of female given names used by Georgian Jews. It includes, among others, several names that became sources for surnames, either directly or via their hypocoristic forms: Khana ‘Hannah’ for Khan(i)ashvili, Ḳona (‘bunch of flowers’ in Georgian) for Ḳon(i)ashvili, Miro ‘Miriam’ for Mirilashvili, Ribḳa ‘Revecca’ for Ribashvili, Sara for Sariashvili, and Tetra (‘white’ in Georgian) for Tetr(u)ashvili. Some other Jewish surnames seem to be based on hypocoristic forms of given names that are commonly used today in Georgia: Eterashvili from Eteri, Laliashvili from Lali or Lala, and Lelu(ch)ashvili from Lela. These given names could be used by Jews because they sound close to biblical Esther and Lea. One surname, Rizhinashvili, could be related to Regina, a given name typical for Jewish migrants to the Ottoman Empire from medieval Iberia and Italy. No trace of any other Sephardic or Italian name is found in other matronymic or patronymic surnames.

The number of surnames of toponymic origin is small. It is only in this category that surnames having no patronymic endings are relatively numerous: Khoṭeveli, Ḳrikheli, Ksovreli, Diǧvireli, and possibly Ḳanzaveli. All of them end in the Georgian demonymic suffix -eli. Other surnames including this suffix in their structure end in -shvili: Ajamelashvili, Alelishvili, Aṭenelashvili, Baǧdadlishvili, Dzorelashvili, Gaznelishvili, Gorelishvili, Gurielishvili, and Roḳeṭlishvili. In a few cases, -shvili was added directly to the toponym: Baǧdadishvili, Digurashvili, and Ḳasṗishvili. Only Baǧdad(l)ishvili is likely to point to the foreign origin of the first bearer. All other toponyms are local. Several names are based on ethnonyms: Megrelishvili from megreli ‘Mingrelian’, Khemsurishvili from khevsuri ‘one from Khevsureti (northeastern Georgia)’, and Lezgishvili from lezgi ‘Lezgin’. For the Jewish families, such names either indicate their provenance from the corresponding regions, or, less likely, they are based on nicknames that relate them—according to some feature—to the corresponding ethnic groups.

Occupational names are not numerous either. Moreover, even those known often originated in the same area: Akhaltsikhe and its region. It is in that city that we find the earliest references to Charukhchev (from the Ottoman Turkish noun meaning ‘maker and/or seller of rawhide sandals’, compare modern Turkish çarıkçı), Kurkchishvili (from kurkchi ‘furrier’), and Meḳinulashvili (from meǩinule ‘one who works/deals with ice’). Two brothers, Osman and Abdulla Topchiashvili (topchi ‘gunsmith’), appear at the end of the 18th century in western Georgia. Their typical Muslim given names and especially the first one, Osman, of doubtless Turkish origin, reveal migrants from the Ottoman territories, that is, most likely, the Akhaltsikhe area. All these names are related to crafts that apparently were not typical for Georgian Christians; note that two of the three Georgian nouns mentioned above (kurkchi and topchi) are both of Turkic origin (compare Turkish kürkçü and topçu, respectively). Baḳaloti—also from Akhaltsikhe—seems to be based on Georgian baǩali (or directly Turkish bakal) ‘grocer’. Several names are derived from Judeo-Georgian nouns of Hebrew origin: Khakhmishvili from khakhami ‘rabbi or any person regularly involved in the religious life of the Jewish community’, Shaliakhishvili from shaliakhi ‘messenger (from the Land of Israel)’, and Shamashidze from shamashi ‘prayer leader.’

As in almost all other Jewish communities, we find surnames designating the belonging of their bearers to Jewish castes: Koenishvili (Cohen) and Levishvili. Both names are rarely used. Moreover, both can be patronymic, since Georgian Jews used the given names Koen and Levi. The surname Leviashvili, based on Levia, a hypocoristic form of Levi, is almost surely patronymic.

Multiple surnames are based on nicknames related to some non-professional characteristics of their bearers. In this group, we find roots having a meaning that can be positive (‘pure in spirit, bright’—Natliashvili), neutral (‘newcomer’—Akhalḳatsishvili, ‘slim waisted’—Injabeli, ‘thin’—Injashvili,73 ‘lukewarm’—Jinjikhashvili, ‘bald’—Kachlishvili, ‘beardless’—Kosashvili, ‘tall’—Maǧalashvili, ‘big-headed’—Tavdidishvili, ‘tiny little’—Tsutsunashvili, ‘with long nose’—Tskhvirashvili, ‘plump’—Bot(v)erashvili, a series related to colors including ‘black’—Shavishvili, ‘white’—Tetrashvili,74 ‘red’—Tsiṭlishvili, and ‘yellow’—Ǩvitelashvili), or negative, or, at least, that can be interpreted as derogatory or ridiculing (‘headache’—Davarashvili, ‘little fly’—Buziashvili, ‘ugly’—Jǧuniashvili, ‘cat-eater’—Ḳaṭaċhamiashvili, ‘valerian, cat-stealer’—Ḳaṭaṗariashvili). Such names are based on sobriquets that could be assigned by Jewish or non-Jewish neighbors. Their large number is related to the natural way of the inception of surnames.

There are no doubtless examples of surnames borne by Georgian Jews before the 20th century that would be borrowed from Georgian Christian. Yet more than 150 surnames (that is, about forty percent of all Jewish surnames) are shared by both religious groups. Patronymic surnames of this kind were discussed above in this section: they cover about two thirds of the shared names. Numerous occupational, nickname-based, and toponymic surnames are also not specifically Jewish, because the same characteristics were valid for the first bearers independently of the religion. Several categories of surnames are unknown among non-Jews. One of them encompasses those derived from given names used by Jews only. Numerous examples are patronymic: Abnerashvili, Abramkhaimashvili, Aroniashvili, Bekhorishvili, Eligulashvli, Eliǩazarashvili, Eliǩezerashvili, Elishaǩashvili, Iekhisḳielishvili, Iskhaḳashvili, Israelashvili, Khaimashvili, Khanukashvili, Khiskiadze, Leviashvili, Manasherashvili, Mardakhiashvili, and Shimshonashvili. A few examples (including Mirilashvili) are matronymic. All names with Hebrew roots (Khakhmishvili, Khasidashvili, Koenishvili, Shaliakhishvili) and some based on Turkish words (Charukhchevi, Injabeli, Injashvili) are also specifically Jewish. By a combination of circumstances, only Jewish bearers are known for certain surnames whose etymology does not preclude their use by non-Jews. Among them are some toponymic names (Ajamelashvili, Aṭenelashvili, Diǧvireli, Dzorelashvili, Ḳasṗishvili, Khoṭeveli), matronymics (Ṭurpiashvili), nickname-based names (Jinjikhashvili, Khakhviċhamiashvili, Naskhleṭashvili, Pichkhadze, and Potlishvili), and occupational names (Kurkchishvili and Meḳinulashvili).

In the corpus of surnames borne by Georgian Jews, we do not find doubtless examples of migrated surnames, that is, names brought as ready-made by migrants from other countries.75 Several factors can be responsible for this phenomenon. Firstly, it is clear that Georgia has never been the destination of mass migrations of Jews belonging to communities using surnames. According to various elements discussed in this paper, it appears that Akhaltsikhe (Turkish Ahıska), the administrative center of an eyalet of the Ottoman Empire, was an important source for the Jewish congregations present in the territory of modern Georgia during the 17th–20th centuries. The total absence of examples of surnames of Sephardic or Italian origins indicates that the Jewish population of Akhaltsikhe (before Jewish migrations from that city) was not formed by migrants from major Ottoman Jewish centers.76 Apparently, it was constituted by other groups of Jews, mainly local: those who moved to this city from various parts of modern Georgia and/or those with Romaniote and/or Mizrahi roots from the territories of modern eastern Turkey, Syria, Iraq, and Iran. If individual Sephardic Jews (usually with hereditary surnames) were joining the communities of Georgian Jews during this period, they would lose their surnames, because for local Jews, surnames were either of little importance or even absent.

7. Surnames of Jewish Migrants to Georgia (19th–20th Centuries)

During the 19th–20th centuries, numerous Jewish migrants came to Georgia from other regions of the Russian Empire and, several years after the Bolshevik Revolution, from the USSR. Ashkenazic Jews represent by far the largest group. Their main destination was Tbilisi, the capital city of Georgia. For example, archival documents from that city indicate the presence in 1836 of the following Jews: Brodskiy, Eydel’son, Krimskiy, Ladizhinskiy (Ladyzhinskiy), Magilevskiy (Mogilevskiy), Meshingiser, Sakharov, Sal’man, Shteyn, Tsmelyanskiy (Smelyanskiy), and Zaydenberg.77 Gradually, the number of these migrants—primarily originating from the Pale of Settlement—exceeded the number of local Jews. In certain places, the Ashkenazic newcomers were the only Jews present. For example, in the list of 47 Jews of Sukhumi (Abkhazia) who donated some money in 1901 to Zionist activities (David 1989, vol. 2, p. 373), we find 45 bearers of Ashkenazic names and two men with the surname Kag’ya (Kaгья in Russian), a specifically Krymchak surname. Documents dealing with Jews from Batumi (Adjara) from the same period also refer to Ashkenazim only.78 The presence of 427 Jews is attested in Tbilisi in 1864–1865. Yet only one of these families was local: others were recent migrants, mainly of Ashkenazic origin (Ter-Oganov 2019, p. 170).

Table 8. Percentages of surnames belonging to various Jewish groups in Tbilisi (20th century).

Georgian	Ashkenazic	Mountain Jewish or Bukharan	Total Number of Surnames
12%	83%	5%	2411

provides statistical data based on tombstone inscriptions in the Jewish cemeteries of 20th-century Tbilisi. This calculation—based on JCG 2016—considers the numbers of surnames, not that of persons.79

Ashkenazic migrants came to Tbilisi with ready-made surnames, mainly based on Germanic (German, Yiddish) and Slavic (Ukrainian, Belarusian, Polish, and Russian) languages: we find the same surnames used in Ukraine, Belorussia, and/or Lithuania. In a few cases, new variants seem to appear already in Georgia. For example, Musashchikov could exist in eastern Belorussia because it is in that region that we find its variant forms Musyashchikov and Masashchikov (Beider 2008). Yet the form Musastikov, derived from Musashchikov, seems to appear in Georgia. The name Razamat is unattested in available sources for the Pale of Settlement: it is known only in Tbilisi (JCG 2016) and Baku. Either all its bearers moved to the Caucasus, or it represents a variant of some other surname, for example, Rozmait, with the corruption that took place already after the migration or immediately before it. Pozhamchi (Пожaмчи) is unknown outside of Tbilisi. The odds are high that this name represents a corrupted form of Podzamcze (Polish spelling), known as a Jewish surname in Galicia. No reference was found outside of Georgia for Melukha, another family of Ashkenazic stock. During the second half of the 20th century, certain bearers of the last surname were already considered Georgian Jews.80 The same merging process was valid at the same period for members of several other originally Ashkenazic families including Mesengiser,81 Minovich, and Fizitskiy. Apparently the last of these families came from the Kingdom of Poland, where the surname was spelled Fizycki in Polish. In any case, no reference to it is known in the Russian Pale of Settlement. No Jewish bearer of Vantsovskiy, also used by Georgian Jews during the second half of the 20th century, is known in Eastern Europe. This situation is either due to the non-exhaustive character of the collected corpus of Jewish names of Eastern Europe, or the family descends, on its paternal line, from a Slavic Christian.82

A significant number of Mountain Jews came to Georgia, primarily Tbilisi. These families originated in Azerbaijan and North Caucasus. In the areas in question, local Jews mainly acquired surnames only after the annexation of these territories by the Russian Empire. Typically, these names have forms of Russian patronymics ending in the suffixes -ov or -ev and having a male given name as their root. Among examples known in Tbilisi are: Abramov, Agaronov, Amirov, Aronov, Avdeev, Azizov, Babaev, Babizhaev, Badalov, Bakhshiev, Bashirov, Budagov, Danilov, Davidov, Gavrilov, Gililov, Ibragimov, Ikhaev, Irmiyaev, Isaev, Isakov, Israilov, Izmaylov, Khanukaev, Khudadatov, Leviev, Magaseev, Mardakhaev, Mekhtiev, Mierov, Mikhaylov, Mishiev, Naftaliev, Nasimov, Nisanov, Pisakhov, Rafailov, Ragimov, Rakhamimov, Rakhmanov, Ruvinov, Safarov, Salimov, Sasunov, Shabanov, Shabataev, Shalumov, Shamailov, Shamilov, Shayaev, Shuminov, Simanduev, Simkhaev, Suleymanov, Uzilov, and Zakharov. Some of the above names—as, for example, Abramov, Aronov, Babaev, Badalov, Danielov, Davidov, Gavrilov, Simkhaev, and Suleymanov—are also used by Bukharan Jews. Consequently, the corresponding families in Tbilisi could, in theory, originate in Central Asia too. However, because of the geographic proximity, the Mountain Jewish origin of these families is much more plausible. The Bukharan origin is plausible only for a few surnames (for example, Alaev) for which references are well known in the Jewish communities of Central Asia but absent from available sources dealing with Mountain Jews.

Not all Mountain Jews coming to Tbilisi had already fixed surnames. Moreover, the patronymic-based names ending in -ov/-ev assigned by the Russian administration were marginal for the conscience of their bearers who continued to use in daily life traditional naming patterns that do not include surnames. Outside of the purely administrative context, surnames were often non-existing. For example, in the list of Jewish migrants who came to Tbilisi from Vartashen (now Oğuz, western Azerbaijan) during the 20th century compiled by a local Jewish historian, surnames are provided only for physicians and authors who lived at the end of the century, but not for tradesmen who moved there during the first two decades (Shirin 2010, pp. 131–34). For this reason, we cannot exclude a possibility that some Mountain Jews acquired their surnames already after the migration to Georgia. This is particularly plausible in two cases. Firstly, this is applicable for forms that do not appear in available lists of surnames used by Mountain Jews outside of Georgia as, for example, patronymic forms Saraydarov, Nagdiev, and Yunisov,83 as well as Ilou and Ninish coinciding with given names. Secondly, it is the case when the surname root is identical to the father’s given name—for example, Iosif Shamailovich (that is, Joseph, the son of Samuel) Shamailov, born in 1880. Several last names known in 20th-century Tbilisi have patronymic endings typical for inhabitants of former Persian and/or Ottoman provinces. One group ends in -zade: David-Zade, Isaḳ-Zade, and Zavlun-Zade. This element, commonly known in Azerbaijan, ultimately comes from Persian zâde ‘offspring’, and it was also borrowed by Ottoman Turkish. Another larger group ends in -oğli: Biniamin-Oğli, Chiraoğli, Daniel-Oğli, Iairoğli, Ifraim-Oğli, Isaoğli, Levioğli, Pasha-Oğli, and Shamiloğli. This Turkic element is the Azeri patronymic suffix -oğlı ‘son of’ (spelled -oğlu in standard Azeri). For both known bearers of the last name Levioğli (born in 1883 and 1902), Levi was the given name of their father(s). Similarly, for the earliest known bearers of the surname Pasha-Oğli in Tbilisi (born in 1895, 1904, and 1910, apparently three brothers), Pasha was the given name of their father. Most likely, in both cases, the Azeri patronymic became a hereditary family name during the first half of the 20th century only. The non-fixed character of surnames can be illustrated by the use of several different forms in the same family: the grandfather Khudad-Oğli (born 1898), apparently his sons Khudodat-ogli (born 1935) and Khudadov (born 1942), and his grandson Khudadatov (born 1961). The ending gızı, the Azeri word for ‘daughter of’ (spelled qızı in modern literary Azeri), represents the female equivalent of oğlı. In Russian-language tombstone inscriptions of Tbilisi, we find Bylkha Shamaykizi (born 1901), the daughter of Shamay; Mariam Bilyalgyzi (born 1905); Lia Mekhtikizi (born 1891), the daughter of Moshe; Jeirani Mekhtikizi (born 1909), the daughter of Mikhael; and Mariya Mekhtikidze (born 1918), the daughter of Abram. For the first four women, their name ending in -kizi can be just their patronymic rather than a hereditary surname.84 In the last case, Mekhtikidze seems to be a Georgianized form of Mekhtikizi, with the ending replaced by -dze, a common ending of Georgian surnames. Zavlunishivili is another example of Georgianizing of an original non-Georgian surname. Indeed, its root, Zavlun, represents a form of the biblical Zebulon that was not used by Georgian Jews. Yet it was used by Mountain Jews: compare the surname Zavlun-Zade mentioned above that could be the source for Zavlunishvili; note that both names are known in Tbilisi only.

Achkinazi, Bakshi, and Konfino are surnames of Krymchak origin known in Tbilisi (JCG 2016). One surname, Davitian, ends in the patronymic suffix -ian typical for Armenian and Iranian surnames. Several tombstones belong to ’subbotniki’, the families of ethnic East Slavs who abandoned Christianity and converted to Judaism. All these families already had their surnames before their conversion and their expulsion by the Russian government to the Caucasus. Luk’yanchenko (Lukyanchenko), Bashkarev, and Chaplygin are examples of these originally East Slavic Christian surnames.

8. Conclusions

As discussed in this article, more than 350 different surnames were used by Georgian Jews. More than half of them are patronymic, often based on hypocoristic rather than full forms of given names. It was during the 18th century that the use of hereditary family names became standard for Georgian Jews, and the earliest attestations date from the 17th century. The fact that a large majority of Georgian Jewish surnames end in -shvili can be explained by a combination of geographical, chronological, and social factors. The number of surnames brought to Georgia as ready-made forms by Jewish migrants from other regions of the Russian Empire and later the USSR is significantly larger than that of surnames borne by local Jews.