Turkish Patronymic Surnames Ending with -oğlu ‘Son of’: A Corpus Linguistic Investigation

This study examines Turkish patronymic surnames ending with the suffix -oğlu ‘son of.’ For this examination, this investigation compiled three corpora of Turkish surnames (N = 10,369) representative of the Turkish population. The aim of this corpus analysis was the identification of the conceptual and grammatical functional elements of these surnames as markers of linguistic and cultural diversity within Turkey. The findings reveal that the Turkish patronymic surnames did not merely denote the name of someone’s father. They also reflect manifold cognitive diversity and express meta-language awareness of Turkey’s intercultural richness. Consequently, this article contends that integrating insights gained from surname research into the foreign language classroom may assist learners of Turkish in overcoming language and cultural barriers. As this study concludes, the findings presented here and in analogous research can help language learners gain a better understanding of the meanings and cultural symbolism embedded in names.


Introduction
People's identities and heritages are often denoted through specific forms of language, such as their names. In this article, the construct of "identity" is used to refer to who people are or the qualities that make them different from others, whereas "heritage" refers to the socio-cultural features (i.e., traditions, attitudes, beliefs, and behaviors, etc.) particular to a group. Both identity and heritage can help us understand the complex processes of identity formation embedded within various intersecting cultural contexts (Aksholakova 2014). Their critical analysis can provide key contextual information about history, politics, and literature as well as identity as it relates to culture, society, religion, ethnicity, and language (465).
For the purposes of this study, "linguistic identity" is defined as "a set of linguistic characteristics of people belonging to a group, consisting of proficiency in a language" (Aksholakova 2014, 467). By comparison, the construct of "cultural identity" is defined as "a conscious acceptance of cultural norms and patterns of behavior, valuable orientations, and language" (Aksholakova 2014, 467). Accordingly, as Uca (2004) states, each society has its own naming traditions: names reflect a way of thinking, a social structure. As Sarıtaş (2009, 423) states, the names of individuals not only regulate people's cultural and individual identity, they also regulate their relationships by reducing confusion. Additionally, Calp (2014) emphasizes the relationship between names and governmental administration in both private and public life. This article explores these points as they relate to surnames and their power to reveal a country's unique historico-cultural heritage and impact on the cognitions underlying language use.
In Turkey, surnames have officially been given to newborns since 1923, when the country became a republic. The rules for choosing and registering surnames were established by the Surname Law Act No. 2525 in 1934. This law was one of the last reforms undertaken by Mustafa Kemal (Atatürk) in his program to promote a Turkish national identity as a foundation of state-building during the 1930s. This law stipulated that 1) every Turk must bear a surname in addition to a proper first name; 2) this surname must follow the proper first name in signing, speaking, and writing; and 3) names related to military rank and civil officialdom, tribes, foreign races, and ethnicity are forbidden as are surnames that are deemed either unsuitable for Turkish customs or are considered "disgusting or ridiculous" (Türköz 2007, 895). Through this law, "a multi-ethnic population came to be constructed in the image of a national idea" (Türköz 2007, 893). This law also required all Turkish citizens to adopt the use of hereditary and fixed surnames by which they could be known locally. The registration period lasted for two years (Türköz 2004) because there was great initial resistance to the regulation. Many Turkish citizens were unprepared to find or create surnames for their families. In part, the reticence to conform to this legislation came because some citizens felt the most desirable names had already been taken by the elite (Fındıkoğlu 1941). Despite this initial difficulty, eventually Turkish residents officially adopted last names. Özcan and Üzen (2019, 137) describe four ways in which many of these surnames were chosen after the decree: 1) some Turks took on completely new surnames having never had surnames before; 2) some changed their previous surnames; 3) some discarded their previous names; and 4) some adapted their ancestors' surnames by affixing the suffix -oğlu meaning 'son of.' This patronymic surname type is the focus of this study.

Previous Onomastic Studies of Turkish Surnames
Since 1934, when last names were first officially required on Turkish birth certificates, there have been several studies on onomastics in Turkey. Some of the trailblazing scholars who made great contributions to the study of Turkish surnames include the following: Behnan Şapolyo 1935;Orbay 1935;Turfan and Bayar 1935;Banarlı 1950;Önder 1968;and Sakaoğlu 1984. These studies cover several sub-topics such as language revolution, legalization of names, name-giving customs and conventions, naming practices in Islamic eras, name-giving to women, and botanical surnames. With regard to Turkish onomastic legalization, four scholars have particular prominence in the literature. Fındıkoğlu (1941) examines the application of the Surname Law in Çankırı Province in Turkey. Tezmen (1949) investigates the court cases involving name changes, while Turinay (2012) and Akı (2015) scrutinize legal bans on of the use of foreign surnames.
On the subject of cultural traditions and Turkish surnames, there has also been much important research. Sakallı (2016), for example, explores traditional customs of Turkish name-giving and analyzes recent trends in commemorative names. Erol (1992) explores names through songs, poems, and folk songs. Similarly, Rásonyi and Baski (2007) describe how Turkish name-giving conventions highlight national myths and religion through alliteration, metaphor, and polysemy. Despite this diversity of research objectives, there are still many questions about Turkish surnames that have yet to be explored. One such area involves the social aspects of patronymic surnaming conventions-such as the relationship between language and power as reflected in the grammaticality and the coherence of word-formation. As Motschenbacher (2020) states, a powerful tool for the investigation of these socio-onomastic conventions is corpus linguistics. Using this tool, this study endeavors to shed light on the formation of Turkish surnames using a cognitive-pragmatic theoretical approach.

Cognitive-Pragmatic Theory and Onomastics
Cognitive-pragmatic theory focuses on different aspects of competence/performance, the nature of linguistic knowledge, and its relationship to language use (Nuyts 1992). Explained briefly, "competence" refers to the system of linguistic knowledge possessed by native speakers, whereas "performance" is the way a language system is used in communication. As Chomsky (1965) famously explained, there is a difference between the competence of "knowing" a language and the performance of "doing" a language when interlocutors converse. From this theoretical perspective, onomastics is enormously important because it identifies morphological and semantic aspects of linguistic knowledge.
The cognitive-pragmatic approach was developed by Coates (2009), who distinguished between "onymic" and "semantic" referential modes. The former might be interpreted as the proper meaning, and merely a convenient label for the unique identification of an individual. The latter, however, might represent an internal elaboration of meaning from nouns to other functional grammatical aspects regarding the referent's meaning. This meaning therefore denotes cultural, social, and linguistic features of a particular society. Although both onymic and semantic features exist in the personal names of every language, the form and meaning of these features are unique. Along with the theoretical distinctions made by Chomsky and Coates, Aksan (1987) also distinguishes between "conceptual" and "functional" aspects of onomastics (Sarıtaş 2009). The former reveals how meaning is composed through language, while the latter reveals the rich and varied linguistic and cultural context behind words. Below, the conceptual and functional aspects of Turkish surnames are examined in some detail.

Conceptual and Functional Aspects of Turkish Surnames
Turkish surnames reflect a wide range of concepts-from patriotism, religion, art, and precious metals to professions, flora and fauna, and family and kinship relations (Aykut 2017, 665). Surname choices may also mirror the education of families as well as their experiences of bureaucracy during the population registry (Türköz 2007, 899). In the table below, the conceptual diversity of Turkish names is presented by category. Kahraman 'hero', Cenksever 'combat-lover', Cihangir 'conqueror'

Metals/Precious Stones
The category Culture refers to surnames not belonging to a subordinate category (e.g., Art/Music). The category Nature refers to surnames not belonging to the subordinate categories Flora, Fauna, and Metals/Precious Stones.
As displayed in Table 1, Turkish surnames often identify conceptual aspects of importance to the name-bearer (e.g., the professions of ancestors, physical appearance, faith in the Islamic prophet and his disciples, the names of their grandparents and ancestors, totems, desiderata, and personal values). Turkish surnames may also provide information about the various choices made by bureaucrats involved in onomastic legislation at the time the last name was adopted. For instance, when the Surname Law Act No. 2525 was instituted, surnames could not be duplicated within the same province. Hence bureaucrats often attached prefixes (e . g . , e r -, ö z -) to existing names to clearly differentiate between people for the purposes of registration. For example, Er 'soldier' was sometimes added to 'ancestor' to make the surname Ersoy 'soldier-ancestor;' and Öz-'kernel/core/self' could be attached to demir 'iron, the metal' to form Özdemir kernel/core-iron). In addition, -er 'soldier' and -öz 'kernel' might also be used as suffixes, as in Soyer 'ancestorsoldier' and Demiröz 'iron-the-kernel.' Nouns could also be attached to avoid duplicating surnames for unrelated residents in the same province. For instance, the noun ay 'moon' could be used to form a compound noun as in the Turkish surname Ayyıldız 'moon-star,' referencing the Turkish flag. The noun büyük 'big' could be used to create the surname Büyükyılmaz 'big+dauntless,' and küçük 'little' was compounded to produce the last name Küçüker 'little-soldier.' These and other examples reveal how numerous conceptual features were not only successively used to create the Turkish surname store. They also hint at how these names can be deconstructed like a puzzle to reveal information about people's sociocultural identity.
Functional surnames in Turkish have a wide variety of morphological and syntactic structures. These structures are presented in observable patterns of "linguistic complexity in names" (Matushansky 2009). As in other agglutinative languages, suffixation plays an important role in Turkish surnames. In the table below, several examples are given of grammatical markers that may be found in functional Turkish surnames.

Turkish Surnames Ending with -oğlu
There are many equivalents to Turkish patronymic surnames ending with -oğlu 'son of' in several languages, one of which is English (-son as in Johnson). These elements build a bridge between lexis and syntax (Dunifa 2019). However, Turkish patronymic surnames can do more than merely denote the first name of an individual's father. For instance, Demir 'iron' is a common surname in Turkey.
If the occupational suffix -cI is affixed, it becomes Demirci, a metonymic surname 'ironmonger.' If then the suffix -oğlu is attached to Demirci, the surname Demircioğlu 'son of ironmonger' is formed. In this way, Turkish surnames were historically modified to produce different meaning(s) to distinguish between different families. If, for example, Demirci had already been registered in a district, an individual subsequently requesting this surname might have received the alternative surnames Özdemir 'iron-the-kernel' or Özdemirci 'kernel-ironmonger' by adding the prefix öz-'kernel/self.' The progeny of persons with those last names might then receive surname like Özdemiroğlu 'son of iron-kernel.' Alternatively, they might carry a case-inflected plural surname variant such as Özdemiroğulları 'sons of iron-kernel,' Özdemircioğulları 'sons of kernel-ironmonger,' or Özdemircileroğlu 'son of kernelironmongers.' More examples that illustrate the diversity of Turkish surnames featuring -oğlu can be found in the tables below.   Given the great sociocultural informational store of Turkish surnames like those illustrated above, this study investigates how Turkish patronymic surnames ending with -oğlu accentuate the conceptual and functional aspects of these identity markers.

Methodology
A mixed research method was utilized for this study. After obtaining permission to use official onomastic data, I computerized three corpora containing the surnames of Turkish citizens. There were three sources for these corpora: 1.) The Dictionary of Turkish Names and Surnames (ATL; n = 9,699) published by the Association of Turkish Language; 2.) the current list of Turkish members of parliament (MP; n = 589); and 3.) a registry of surnames listed for the academics (AC) employed at one randomly chosen government university (n = 2,284). The total number of surnames used in this study was 9,964 after all of the first and middle names were deleted and the maiden names of the married women were excluded. The final overall corpus contained one surname per person. The representativeness of these three corpora combined helps to generalize the findings to Turkey's approximately 83 million inhabitants (Association of Turkish Statistics 2020). After the onomastic data was assembled, NVivo 12, a qualitative data analysis tool, was utilized to create categories and analyze the data. This study was restricted to surnames and patronymic surnames taken out of the input data. The selected names were then placed into conceptual and functional categories. This onomastic coding process presented some challenges. Some names and surnames, for example, are identical in Turkish language. For example, Demir is given as both a first name and a surname in Turkey. In order to address this problem, three Turkish language instructors were asked to recode the corpora of 9,964 names. The final classifications were based upon agreement between three coders. The same procedure was followed to determine the conceptual and functional classifications of the surnames. The following categories were developed: Abstract Nouns, Anatomy, Astronomy, Art/Music, Festivity, Culture, Colors, Emotions, Ethnicity, Metals/Precious Stones, Nature, Numbers, Patriotism, Flora, Fauna, Place of Origin/Region, Profession, Religion, Verbs, Adjectives, and Father's Name. In ambiguous cases, the coders used the category "Other".
There was a super diversity in the nature-denoted surnames. After discussing this hetergeneity, the three coders agreed to use four categories: Nature, Fauna, Flora, and Minerals/Precious Stones. In the study, nature was used for surnames containing descriptors of environmental features and processes that exist independently of people (e.g., storms, the sea, rain). The category Fauna was used to classify names that contained zoonyms. Turkish surnames of this type often originally expressed the desire of the namebearer to exhibit characteristics associated with certain animals such as bravery. The category Flora was used for surnames that were based on plant names. Like the category Nature, there are a great diversity of surnames that reflect some aspect of Turkish culture. To reduce the size of this group, separate categories were developed for different aspects of Culture (e.g., Art/Music, Ethnicity, and Religion). Taken together, these Culture surnames frequently mirror a family's desire identification with some valued feature or symbol of Turkish society (e.g., Cirit 'javelin,' Efe 'swashbuckler'). Once the classifications were complete, the percentages and frequencies values of surnames were calculated. The Chi-Square test was used to determine whether the surnames in the three corpora differed. Table 5 displays data gathered from the three corpora gathered from the lists of ATL, MP, and AC.

Results for the Three Sub-Corpora The Results for the Association of Turkish Language (ATL) Sub-Corpus
The distribution of the Turkish surnames reveals significant differences between the categories Patriotism, Flora, Metals/Precious Stones, Religion, Astronomy, Verbs, and Other (χ 2 = 6914.887; df= 20; p= 0.000) regarding both conceptual and functional aspects. Table 5 displays the analyses of the conceptual and functional aspects of Turkish surnames from the ATL sub-corpus. According to the results shown in Table 5, the highest proportion of surnames in the ATL sub-corpus is from the Patriotism category (14.73%). This finding might indicate citizens' awareness of their country's unique historical and cultural heritage. The second highest proportion belongs to surnames of Flora (12.39%). This finding might reflect the emotional importance of plants in Turkish culture. Then, the surnames with Metals/Precious Stones might draw attention to families who deserve high quality as an exhibit material (9.60%). Surnames falling within the category Religion might reflect piety and relationship to the external spiritual world (6.89%). The other significant difference observed relates to the category Astronomy. These surnames might mirror a family's interest in the scientific happenings of their environment (4.92%). The Other category is also prominently represented (10.35%) although these names had no unifying theme. However, there was not a significant difference detected for these patronymic surnames.

The Results for the Turkish Members of Parliament (MP) Sub-Corpus
The findings from the MP sub-corpus indicate that the most significant differences are found between the conceptual categories Metals/Precious Stones, Patriotism, Astronomy, Colors, Profession; as well as for the functional categories Verbs, Others, and Adjectives (χ2 =412.384; df=20; p=0.000). Table 6 shows the results of the analysis of the conceptual and functional categories of Turkish surnames found in the Turkish MP sub-corpus. According to the results, the highest proportion of surnames belongs to Metals/Precious Stones (11.54%), making this grouping the most represented of the 22 categories. These surnames might reflect familial connections to valuable mines or their association with the characteristics associated with these precious resources. With regard to the second most common surname category, Patriotism, this result might reflect respect for or adoration of rulers in Turkish history (9.67%). Among the other common surname types in this sub-corpus, Astronomy (7.39%) might signpost familial belief that celestial powers can give name-bearers character-building experiences or assistance in successfully managing their affairs (e.g., in harvesting their crops). Where the functional categories are concerned, surnames utilizing Verbs make up a very large proportion (12.73%). As also shown in the Table 7, surnames featuring Adjectives (7.47%) are also quite common. The Other category, however, forms an even higher proportion (8.99%). The Chi-Square test verified no significant difference in the patronymic surnames of this sub-corpus.

The Results for the University Academics (AC) Sub-Corpus
Of the 2,284 surnames in the AC sub-sample, a significant difference exists between the categories Patriotism, Metals/Precious Stones, Fauna, Verbs, Adjectives, and Other (χ 2 =2280.712; df=20; p = 0.000). Table 7 displays the classifications from the list of surnames carried by academics who work at a government university in Turkey. As shown above, the most common surname category is again Patriotism (9.54%). This finding might reflect Turkish citizens' desire to show their respect towards their national history across the generations. By contrast, the category Fauna is not as common. This result could mean that the drive for familial identification with the nation of Turkey is greater than it is for associating one's family with prized animals. When regard to the functional aspects, surnames utilizing Verbs (14.18%) are quite common. These surnames display verbs across different tenses, aspects, modalities, and voice, etc. Surnames constructed from Adjectives were also prominent (11.51%). The Other category is again strongly represented (17.03%) which is understandable considering that the super diversity of this group.

The Results for the Overall Corpus of Turkish Surnames
In the table below, the distribution of the Turkish surnames by functional and conceptual category across all three sub-corpora is displayed. The analysis of the entire sample of 9,964 names in all three sub-corpora reveals that surnames related to Patriotism have the highest frequencies and percentages: ATL (1,045) 14.73%; MP (57) 9.67%; AC (218) 9.57%. Also of interest here are the surnames in the category of Metals/Precious Stones (MP 68) 11.54%; ATL (681), 9.60%; AC (155); 6.78%. The Chi-squared test confirms that these two categories have a significant difference in terms of their frequency in relation to the other categories investigated. This result may reflect the cultural importance of having strong connections to the Turkish history as well as the national value of natural resources in Turkey.
An examination of the surnames also reveals differences between the three corpora. For instance, for the category Profession, only two sub-samples have a relatively high proportion: ATL (6.89%) and MP (5.09%). In the same vein, Astronomy is quite common in the two sub-corpora MP (7.39%) and ATL (4.92%). The category Flora, however, is comparatively common in the ATL sub-corpus (12.95%), but relatively uncommon for the sub-corpora MP (1.35%) and AC (2.05%). This finding was the same for Religion (ATL 15.39%, MP 1.18%, and AC 1.00%), as well as the category Colors, which was only common in the MP sub-corpus (6.11%).
Where the functional aspects are concerned, there is an important difference in the categories Verbs and Adjectives. The analysis reveals that only two sub-corpora make up a comparatively large proportion of these surnames: the MP (12.73%) and the AC (14.18%). Similarly, analysis of the Adjectives category shows that the two sub-categories make up a substantial share of this type: AC (11.51%) and MP (7.69%). The Other category has the highest percentage in the AC sub-corpus (17.03%), a relatively high percentage in ATL sub-corpus (10.35%), and the lowest proportion for MP (8.99%).

Conclusion
The findings of the study reveal how varied Turkish surnames featuring the patronymic suffix -oğlu, are, marking familial identification from categories as diverse as Turkish patriotism to regionalism. The Turkish onomastic system represents a broad and nuanced diversity of meanings that reflect the nation's culture, ideology, geography, literature, history, sociology, heritage, and namegiving traditions. As a result, the onomastic analysis presented here may provide valuable insights for cross-cultural pragmatics and socio-onomastics that may be relevant for many different languages. Similar conclusions have been reached by other studies (Acıpayamlı 1992;Kibar 2005;Şahin 2016;Uca 2004).
The surnames analyzed in this study also display numerous functional elements of Turkish grammar (e.g., simple, compound, and complex nouns; verbs in various tenses, voices, and aspects; and adjectives that modify and enumerate). These embedded features constitute logic puzzles that offer clues about not only Turkish culture, but also Turkish syntax, vocabulary, and morphology. Examining Turkish surnames can offer individuals a means for accessing the numerous cognitive and cultural elements of the Turkish language (Aykut 2017;Ergene 2016;Gülensoy 2012;Örnek 2015). This benefit may be especially great for foreign learners of Turkish, who often find Turkish morphology to be difficult (Benzer 2020;Çalışkan 2020;Erdem & Bahşi 2019;Kardaş & Koç 2019;Örnek 2015).
Furthermore, in this age of globalization, by learning about Turkish surnames, foreign language learners such as immigrants and refugees may obtain a powerful tool for overcoming cognitive barriers by attaining knowledge important for developing their mental lexicon about the Turkish society. In this respect, Turkish surnames have potential pedagogical benefit for those learning Turkish as a foreign language. This study will hopefully help to raise awareness of how onomastics can help language learners gain a deeper understanding about names and societies: a point also made by other international scholars (Luchtenberg 2010;Motschenbacher 2020;Pau Jordà et al. 2018;Romanova & Spiridonov 2018). For future research, it would be interesting to study how Turkish names and surnames could be utilized in cross-linguistic learning environments to remove barriers for cross-cultural learners.
As with all studies, this investigation was not without challenges. Aside from the difficulty of obtaining permission to access otherwise confidential name lists, I also faced three other major difficulties. The classification of the three corpora required considerable time and very careful analysis because of the previously mentioned difficulty presented by names that could function as either first names or surnames. It was also time-consuming to classify the surnames that belonged to married women who used both their maiden names and marital surnames. The last limitation concerned determining the national origin of the Turkish surnames. However, since all the surnames were gathered from official lists, the decision was made to treat them all as Turkish surnames without delving into their individual etymology. Researchers in the future, however, may wish to examine the historical roots of Turkish surnames. Despite these limitations and challenges, this study offers a fascinating look at the real-life diversity and potential utility of conducting research into Turkish names.