When someone asks you this rather straightforward question, many of us can come up with a place name based on our family’s oral traditions – Keghi, Evereg, Kharpert, Dikranagerd, Aleppo, Bardizag, Bitlis, Kars, whatever. But can we actually find that place on a modern map or prove the family’s origins with primary documentary evidence?
The first challenge is defining the geographic term “Armenia” in the context of the 1800s through, say, 1930. Are we talking about the boundaries of the current Republic of Armenia (as we find it on today’s maps), or historic Armenia, which also included a much larger area encompassing parts of today’s Turkey, Iran and various countries in the Middle East? At the time of the onset of World War I in 1914 (and the Armenian Genocide, which soon followed), there was not a country (recognized by other nations) called Armenia. The traditional Armenian homeland was distributed across three Empires – the Ottoman Empire (throughout eastern Turkey and the region of Cilicia near the Mediterranean coast), the Russian Empire (in the Caucasus region) and the Persian Empire (in the northwestern part of today’s Iran). The First Republic of Armenia, established by a decree of independence on May 28, 1918, was the first modern Armenian state since the loss of Armenian statehood in the Middle Ages. In the tumultuous years following this decree, state boundaries would shift, resulting in the loss of the Kars region to Turkey. (Battles over the Artsakh region continue to this day.)
An Armenian immigrant born prior to 1918 living in the diaspora in later years might refer to their place of origin as Armenia, but they were referring to a region, not an actual state in existence at the time of their birth. Within this region, there was a mix of ethnic groups living interspersed with each other – Turks, Armenians, Assyrians, Kurds, Greeks, Jews, Russians, Persians, etc. You might have a predominantly Armenian village on a hillside overlooking a river with a predominantly Kurdish village on the opposite hillside on the other side of the river. A town would often be partitioned into neighborhoods, or quarters, with an Armenian church in the Armenian quarter and a mosque in the Turkish quarter. The military and administrative structures in this region called Armenia were controlled by the central government – Russian, Turkish or Persian. But, to our immigrant ancestors, this was their Armenia.
How can you use primary documentary sources and oral traditions to pinpoint the actual village, town or city where they were born?
Geographic Place Names & Administrative Structures
A place in historic Armenia might be known by a number of different names depending on the time frame, whether you were using the traditional Armenian name or the name given by the then-ruling empire (Ottoman, Russian or Persia), and to what level in the administrative hierarchical structure you were referring.
If someone said, in later years, that they were from Kharpert (Խարբերդ in Armenian), they might have been referring to a city of that name in Turkey that would appear on most maps of that time under the Turkish name Harput. In the 1800s, much of Kharpert (located high on a hill) was moved downhill to another nearby location called Mezeré. (In the 1920s, the Republic of Turkey under Mustafa Kemal Atatürk would rename this city to Elazig, along with wholesale name changes to most other traditional Armenian places.) On the plain to the south of Kharpert (Mezeré) were dozens of small villages. An Armenian from one of these small villages might have given his birth place as Kharpert/Harput or might have used the actual village name.
The Ottoman Empire divided its realm into a hierarchical administrative structure. At the top of the hierarchy, the empire was divided into vilayets (what we in the United States or Canada would call states or provinces). Each vilayet was then divided into sanjaks (subprovinces), typically three or four sanjaks per vilayet. Each sanjak was divided into a number of smaller areas called kazas (districts). Kazas were further subdivided into nahiyes (subdistricts). Kazas are roughly equivalent to counties, with nahiyes as townships.
A map is helpful in illustrating this structure. In the early 1890s, a French geographer and ethnographer named Vital Cuinet published a four-volume work entitled La Turquie d’Asie – Géographie Administrative, Statistique Descriptive et Raisonnée de Chaque Province de l’Asie-Minor (Asiatic Turkey – Administrative Geography, Statistical Description & Catalog of Each Province of Asia Minor). While the veracity of the statistics themselves proved problematic to Cuinet and his critics, the maps and descriptions of the administrative structures of each Ottoman vilayet are very helpful to researchers of Armenian genealogy and immigration history. Many of Cuinet’s maps can be found here.
Here is Cuinet’s 1892 map of the vilayet of Mamuretulaziz (where Harput is located). (The place names in these maps are the French version, which again illustrates the variability of the place names in historic Armenia.)
In the key to this map, we can see the administrative hierarchy to this vilayet. It was divided into three sanjaks and 18 kazas. The sanjak colored in yellow in this map is Kharpout-Mezeré (Harput). Within this sanjak were four kazas, one of which had the same name (Harput). The city of Harput and its nearby villages were all located within the kaza of that name.
That’s the problem. When someone said they were from Harput, were they referring to the city (Harput/Mezeré, now Elazig), one of its dozens of nearby villages, the kaza or the sanjak? And the vilayet of Mamuretulaziz was also referred to by another name – Harput! In order to figure out which Harput they were referring to, it is often necessary to analyze many different types of primary records for an individual (more on this later).
As another example, let’s turn to a different vilayet – Ankara (Angora) (from Cuinet’s 1890 map), and look at the sanjak of Kayseri (Césarée) in its southeastern corner.
There were several kazas in the sanjak of Kayseri. One of these (in its northeastern section) is the kaza of the same name, which contains the principal town (you guessed it – Kayseri). If someone said they were from Kayseri, were they referring to the town, the kaza, or the sanjak?
Another kaza in the sanjak of Kayseri (in its southeastern section) was Evereg (Everek), enclosed in red in the map above. Its principal town was Evereg. To the northeast of Evereg (the town) was a village called Tomarza (Tomarzou). Many immigrants from Tomarza came to America.
Someone born in the village of Tomarza could have stated their place of birth in several different ways: Tomarza (the village), Evereg (the kaza), Kayseri (the sanjak), or Ankara (the vilayet), all of which were correct. If you were to look for a Tomarza native in primary source records in the United States (ship manifest, World War I and World War II draft registrations, naturalization applications, passports, etc.), along with oral family traditions, you would probably find each of those places mentioned. You might conclude that they lived in different places and moved around a lot, when in fact they remained in the village of Tomarza the entire time before immigrating.
A common mistake by researchers of Armenian genealogy is to assume that a place name refers to the town or city, when it could refer to the kaza, sanjak, or vilayet. The Ottoman practice of giving the same name to the vilayet, sanjak, kaza, and principal town of the kaza was pervasive in that timeframe. Examples include the vilayets of Erzurum (Garin), Van, Bitlis (Paghesh), Diyarbekir (Dikranagerd), Sivas (Sebastia), etc.
There were a few areas in Ottoman-era Turkey that are a bit different. Marash and Urfa were autonomous sanjaks (not tied to a vilayet). Izmit and Jerusalem were mutesarrifates (also independent from vilayets). Each contained kazas.
In the part of historic Armenia that was located in the Russian Empire (prior to 1917), there was a different administrative hierarchy, but the concepts are similar. Here is a 1903 map (in Russian) showing the Caucasus region. (This work is in the public domain in Russia according to article 1256 of Book IV of the Civil Code of the Russian Federation No. 230-FZ of December 18, 2006.)
We can see in this map two different structures. To the west is the oblast of Kars, divided into four okrugs. One of those okrugs (called Kars) contained the city of Kars. Someone in the diaspora who gave Kars as their birthplace could have been referring to the city itself, the okrug, or the oblast. (Much of this oblast later became a part of the Republic of Turkey.)
The governorate of Yerevan (to the east) was comprised of several divisions called uyezds. The uyezd of Alexandropol (to the north of the governorate) contained the city of Alexandropol. (This city was renamed to Leninakan during the Soviet period, and is now known as Gyumri.) Someone stating their birthplace as Alexandropol may have been referring to the city or to a village somewhere in the uyezd of the same name.
The names of most of the Armenian towns and villages in Turkey were renamed by the Turkish government starting in the 1920s, as a part of their efforts to “Turkify” the country. My grandfather’s ancestral village of Sergevil (also known as Sivgelik) in the kaza of Keghi, sanjak of Erzurum (Garin), vilayet of Erzurum (Garin) is now known as Açıkgüney. (It took me many years to figure out where Sergevil is now located.) George Aghjayan has created a very helpful website using Google maps for several parts of “western Armenia” (eastern Turkey), showing the locations of many Armenian towns and villages with their historic and modern (Turkish) place names.
Another excellent source for finding Armenian place names in Turkey is The Armenian Genocide – A Complete History, by Raymond Kévorkian (London: I. B. Taurus & Co., Ltd., 2011). In Part IV (pages 265-621), he describes the events during the initial phase of the genocide (1914-1915) in each of the vilayets. For each sanjak and kaza, Kévorkian provides the names and populations of many of the Armenian towns and villages. This book is available as an e-book through Google. (Search “Kevorkian” and “genocide.”)
Using Primary Sources in American Records to Find Place of Origin
Armenian immigrants to America (United States and Canada) in the late 1800s and early 1900s often left behind much valuable genealogical information in primary source records. I have spent much of the past 20 years abstracting information from these sources and putting them online in a free, searchable database.
The database can be searched through the Queries & Reports page.
The following types of primary sources are included:
- ship manifests
- military records
- naturalization applications
- passport applications
- birth records
- marriage records
- death records
- missing person ads (in Armenian newspapers)
As of this writing, the database contains abstracts of over 125,000 records relating to Armenians in America, most of whom arrived prior to 1930, as well as their children and grandchildren born in the United States and Canada. More records are being abstracted each month and added to the database. Images for these records are not a part of the database, but can be found online at web sites like Ancestry.com and FamilySearch.org, using the citations contained in the abstracts.
Places of origin can be found in all of these records, but some sources are better than others for identifying the exact city, town, or village. Best results are obtained by finding your immigrant relative in as many of these records as possible, and then analyzing and interpreting the evidence as a whole to come to a conclusion regarding the most likely place of origin. Also, look at records for that person’s close relatives to see where they were from. Start with the Armenian Immigration Project database, but recognize that it is based on a small subset of the total records publicly available through the Internet and in churches, courthouses, and archives.
The best primary source records for identifying the exact place of a person’s birth are ship manifests (1907 and later), military draft registrations (World War I and World War II), naturalization and passport applications, and missing persons ads (especially those in the years immediately following the end of WW1). As described earlier in this paper, the places of birth listed in these records are not consistent with respect to their level in the origin country’s administrative hierarchy. In the case of the place of birth in the Ottoman Empire (Turkey), it could be the village, town/city, kaza, sanjak, country, or just “Armenia,” many of which have the exact same name.
The other types of records in the database typically list just the country of a person’s birth (or parents’ birth in the case of vital records – births, marriages, and deaths).
Let’s look at an example. Here is an abstract of a 1912 ship manifest for Soghomon Depigian.
Three places of origin are mentioned on ship manifests of 1907 and later – birth place, place of last permanent residence (at least one year), and the place “of nearest relative or friend in country whence alien came”. Any of these may be helpful. In this particular manifest, we see “Keghi” in all three fields, so we known that he was born in the kaza of Keghi (in the sanjak of Erzurum, vilayet of Erzurum, in Ottoman Turkey). But, was he born in the principal town of the kaza by the same name as the kaza – Keghi (or Keghi Kasaba) or in an outlying village?
At the bottom of the ship manifest abstract page are entries for other records found for this person.
Soghomon had another arrival in 1919, which states his birth place as Erzurum, Turkey (referring to either the city, sanjak, or vilayet). But, we knew that already, as Keghi is within the sanjak and vilayet of Erzurum. We can see that he returned to Turkey sometime after his 1912 arrival, and escaped the genocide by traveling east through Russia, across Siberia, to Vladivostok. However, this does not narrow down his place of birth to a particular village within the kaza of Keghi. Likewise, the 1930 and 1940 censuses, as well as his World War II draft registration all state Turkey. No help there. Fortunately, we have found three naturalization applications for him. He filed a Declaration of Intention (also called First Papers) immediately upon arriving in Granite City, Illinois in 1919. In this application, he gave his birth place as Erzurum. He didn’t follow up on this Declaration of Intention (which expired after seven years) and then moved to Detroit, Michigan (as shown by the 1930 census). In 1935, he again filed his Declaration of Intention, this time at the Circuit Court in Detroit. In this record, he finally named a village – Sergevil! His 1938 Petition for Naturalization also named Sergevil.
In this example, we had to locate eight different records for this individual until we found the village of his birth in those last two records. Had we not found the naturalization applications filed in Detroit in 1935 and 1938, the most we could have concluded was that he was born in the kaza of Keghi. We would not have had enough evidence (yet) to determine if he was born in the principal town (Keghi Kasaba) or in a village somewhere in the kaza. We would have had to resist the temptation to conclude that he was born in the town and be content with some level of uncertainty (not a bad thing, by the way).
Often, you are not able to find enough records for an individual to get a definitive answer. It is often a good strategy to look at records for close relatives (often a brother, uncle or cousin) to see if there is enough evidence in that person’s records to establish a birth place with a high level of certainty. In the case of my paternal grandfather Dikran Arslanian (who arrived in America in 1906), anecdotal evidence from my discussions and correspondence with his nephews and cousins in the 1970s and 1980s said that he and his entire family (aunts, uncles, and cousins) were all from the village of Sergevil. Of the 40 individuals remaining in Turkey at the time of the genocide, only two survived. My grandfather (who died in 1965) was mentioned in a number of primary sources, but none gave a birth place more precise than Erzurum (which could have meant the city, sanjak, or vilayet). I found records for several of his brothers and cousins that narrowed it down to Keghi (the kaza). Eventually, I found a naturalization application for a son of his first cousin that confirmed Sergevil as the village where this family lived. I also found that he lived with a couple of other men from Sergevil in the 1910 census, adding further corroboration.
This same technique can be used to determine a specific kaza, town/city, or village of origin for your own family. Some of the records may already be in the Armenian Immigration Project database; many will not. You will probably need to dig hard to find the “golden record” with the information you are seeking. Don’t be content with just family anecdotes, published genealogies, or a single primary source record. Look for strong corroboration by thoroughly searching for the person in all records you can find. Avoid the temptation to assign the place of birth to a town/city when they could be referring to a kaza, sanjak, or vilayet by the same name. Without the anecdotal evidence I obtained from elder relatives when I was a teenager, I might have concluded that my grandfather was from the city of Erzurum (70 miles away from his actual home village). (I asked these relatives the same questions independently to see if I’d get the same answers.)
One of the biggest challenges with primary sources is that the spelling of place names (as with personal names) is extremely variable. Here is just a small sampling of the different spellings I have encountered for the city and kaza of Harput/Kharpert (the place contributing the most Armenian immigrants to America in the late 1800s and early 1900s):
When you encounter names of small villages or obscure kazas, sometimes the only way to figure where they are located is to find other records for that person that provide the name of the sanjak or vilayet. The same applies for villages of the same name that appear in multiple vilayets of Turkey (like Ichme, Bardizag, or Orta Koy). Which one are they from? Find other primary source records to sort it out.
Doing an exhaustive search for your relatives in primary documentary sources takes a lot of time and patience, but it can be very rewarding and gratifying. You may find a personal signature or even a photograph that you’ve never seen before (in naturalization applications starting in 1930 and passport applications starting in 1915). Ship manifests may reveal new names and connections, as well as show how your relatives got to America. The decennial censuses of the United States and Canada often show extended family groupings soon after they arrived. Be prepared for surprises, as family anecdotes are almost never completely accurate. Keep an open mind. Have fun.