BY 4.0 license Open Access Published by De Gruyter Mouton June 30, 2021

The sound of gender – correlations of name phonology and gender across languages

Tanja Ackermann and Christian Zimmer
From the journal Linguistics


Our article is dedicated to the relation of a given name’s phonological structure and the gender of the referent. Phonology has been shown to play an important role with regard to gender marking on a name in some (Germanic) languages. For example, studies on English and on German have shown in detail that female and male names have significantly different phonological structures. However, little is known whether these phonological patterns are valid beyond (closely related) individual languages. This study, therefore, sets out to assess the relation of gender and the phonological structures of names across different languages/cultures. In order to do so, we analyzed a sample of popular given names from 13 countries. Our results indicate that there are both language/culture-overarching similarities between names used for people of the same gender and language/culture-specific correlations. Finally, our results are interpreted against the backdrop of conventional and synesthetic sound symbolism.

1 Gender marking on names

Names are usually said to have no lexical meaning. However, there is one piece of information, which is quite commonly coded in personal names: the name bearer’s gender (cf. Alford 1988: 66–68). Depending on the language/culture, different types of personal names are involved in the marking of gender, e.g., given names, the family name, and/or unofficial names.[1] Among these types of personal names gender marking is assigned most often to the given name (cf. Oelkers 2003: 134), which we focus on in the following. As we will outline in the next section, there are different ways to mark gender on a given name including specific phonological structures (Section 1.1). This will be exemplified by studies on German and English (Section 1.2). These studies show that – in these particular languages – there are strong correlations between semantic (e.g., femininity) and phonological properties (see below). For both German and English, it has been discussed how these correlations relate to the concept of sound symbolism (cf. Section 1.2). However, there has been little agreement yet: While some scholars interpret these correlations as instances of synesthetic sound symbolism (e.g., Cutler et al. 1990; Oelkers 2003, 2004; Pitcher et al. 2013; van de Weijer et al. 2020; Whissell 2001) others argue in favor of conventional sound symbolism (e.g., Cassidy et al. 1999; Hough 2000; Nübling 2018).[2] With our study, we want to add an important perspective to this discussion by considering a range of disparate languages/cultures. This promises to shed new light on this topic as one important controversy centers around the question whether we are dealing with arbitrary and language/culture-specific cues to gender (i.e., conventional sound symbolism; see, e.g., Cassidy et al. 1999: 378) or with non-arbitrary correlations, that are valid beyond individual languages/cultures (i.e., synesthetic sound symbolism; see, e.g., Oelkers 2003: 228). The basic questions we want to answer are: Are correlations of name phonology and gender language/culture-specific or are there language/culture-overarching similarities? Which phonological properties indicate (at least by tendency) gender? Hence, we add a contrastive perspective to the discussion on sound symbolism in the domain of onymic gender marking (and we investigate name giving practices in some countries which have not yet been studied in this respect, see below).

The remainder of this paper is structured as follows: After the short overviews of gender marking on names in general (Section 1.1) and by phonological means in particular (Section 1.2), we summarize two central positions on how phonological gender marking on names relates to the concept of sound symbolism (Section 2). Section 3 is dedicated to our own study: We describe the sample selection, the transcription, the annotation, and the multifactorial modelling of the data. The findings are discussed in Section 4 and Section 5 draws a brief conclusion.

1.1 Types of gender marking

According to Alford (1988: 66–68), gender is the most frequently marked information on proper names. Gender marking languages are traditionally assigned to one of three types – the semantic, the formal, or the conventional type:[3]

  1. (i)

    Gender can be marked by the semantics of the name. In this case, desirable characteristics are assigned via a name according to gender stereotypes. E.g., the Turkish male name Yılmaz is etymologically rooted in an adjective meaning ‘fearless’ and thus fits a male gender stereotype. The same holds for female names such as Gül (‘rose’) which is, for example, associated with beauty and thereby connected to central components of the female gender stereotype. Hence, we are dealing with a case of ‘doing gender’. Languages/Cultures belonging to this type of gender marking on names comprise, for example, Japanese and Chinese.

  2. (ii)

    Gender can also be marked on the basis of formal means (e.g., suffixes). In languages/cultures belonging to this type (e.g., Italian, Ojibwa, or Garo), gender is overtly marked and can be deduced from the form of (a specific part of) a name. For example, Italian names ending in -a are (with some exceptions, such as Luca) female, while names ending in -o are male.

  3. (iii)

    Finally, there can also be separate inventories in a language/culture (Alford 1988: 65–68). In this case, gender is associated with a name by convention. For example, one simply has to learn that certain names, which are commonly used in Germany (e.g., Doris), are used for females while others (including phonologically very similar ones) are used for males (e.g., Boris).

However, phonology also seems to play an important role – even in languages/cultures that are assigned to the conventional or the semantic type of gender marking (i/iii).[4] Studies on English (cf., e.g., Cassidy et al. 1999; Slater and Feinman 1985) and German (cf., e.g., Nübling 2009; Oelkers 2003) – languages/cultures which are usually assigned to type (iii) – have shown in detail that female and male names show significantly different phonological structures, which are discussed in the next section.[5]

1.2 Prosodic-phonological cues to gender in German (and English)

In her study on German given names Oelkers (2003) showed that female and male names differ significantly with regard to the prosodic-phonological features ‘number of syllable’, ‘main stress’, ‘portion of vowels/ consonants’, and ‘quality of stressed vowel’. The main results are that female names contain on average more syllables than male names, female names show initial stress less often than male names, for which initial stress is the dominant pattern, and – most importantly – female names prove to be more sonorous in general. This can partly be explained by two facts: female names have more vowels than male names and female names show significantly more final vowels on average. The final sound of a personal name has in general proved to be a structural position which particularly contributes to the gender differentiation of given names. A further result is that female names have more stressed front vowels than back vowels on average. According to these results, the German given name Katharina [ka.ta.ˈʁiː.na] can be classified as exhibiting predominately ‘female’ phonological structures (four syllables, penultimate stress, balanced portion of vowels and consonants, final vowel, stressed front vowel) while Rolf [ˈʁɔlf] exhibits predominantly ‘male’ structures (one syllable, higher portion of consonants, final obstruent, stressed back vowel). Table 1 summarizes these findings on prosodic-phonological differences between female and male names in German (for similar results obtained with other samples and a slightly different method, see Nübling 2009 and Nübling et al. 2015: 131–137) and lists possible phono-semantic explanations for these differences (we will comment on these explanations in Section 2).

Table 1:

Prosodic-phonological patterns of German given names and their relation to sound symbolism (cf. Oelkers 2003: 220).


Female NamesMale NamesPhonological ‘meaning(reported for female names)
Number of syllablesMore syllablesFewer syllablesLonger words = higher sonority;

Euphony = female stereotype
Main stressMore non-initial stressMore initial stressDeviation from unmarked pattern;

Exotic = female stereotype
Portion of vowels/consonantsMore vowelsMore consonantsVowels = sonority/ soft sound structure;
Final soundVowel finalConsonant finalSoftness = female stereotype
Quality of stressed vowelMore front vowelsFewer front vowelsFront vowels = smaller size;

Small size = female stereotype

In German, these phonological patterns seem to be so firmly established and associated with the respective gender that they can be transferred to other types of names, such as brand names, and are thus used for gender marketing.[6] In this area, the association of phonological properties and a certain gender is used for the marketing of products that have either a male or a female target group (e.g., deodorants; cf. Ackermann 2011). Here, phonological properties are part of other marketing aspects such as the coloring of the products.[7]

Those correlations of (targeted) gender and phonological structure of a name cannot only be found in German. Cassidy et al. (1999) found very similar patterns for (product and) given names in English (see also Cutler et al. 1990; Fredrickson 2007; Pitcher et al. 2013; Slater and Feinman 1985; Whissell 2001; Wright et al. 2005; for an overview cf. Elsen 2016: 120–126).[8] Just as in German, English female names have more syllables than male names, show more often non-initial stress as well as a greater ratio of open to closed syllables on average and predominantly end in a final vowel. The results are depicted in Table 2.[9]

Table 2:

Prosodic-phonological patterns of English given names (cf. Cassidy et al. 1999).

Prosodic-Phon. featureFemale namesMale names
Number of syllablesLarger numberSmaller number
Main stressNon-initial stressInitial stress
Final soundVowel finalConsonant final
Ratio of open to closed syllablesGreater

To date there has been little agreement in the literature on how these findings relate to the concepts of iconicity and sound symbolism: are we dealing with synesthetic or with conventional sound symbolism? This controversy will be addressed in the next section.

2 Synesthetic versus conventional sound symbolism

According to Oelkers (2003), specific prosodic-phonological structures are not distributed randomly over female and male names. Instead, she interprets these differences phono-semantically. The main idea is that certain phonological properties (e.g., the vowel /i/) tend to be associated with semantic concepts (e.g., smallness), some of which are associated with gender stereotypes (e.g., femininity; sexual dimorphism plays a role here). Oelkers (2003) argues that phonological differences between male and female names reflect these associations of phonological property, semantic concept, and gender stereotype. Parents seem to “give names to children in a manner that is aurally (i.e., through sound) metaphorically congruent with gender stereotypes (i.e., shared beliefs about the traits of women and men)” (Slepian and Galinsky 2016: 512; see also Section 4).

Oelkers’ (2003) main claims concerning specific phonological properties are briefly summarized in the following.[10] The finding that female names have more syllables is explained in such a way that longer words are perceived as more melodious than shorter words. Euphony again is connected to central components of the female gender stereotype (according to Oelkers 2003: 144). A similar explanation applies to the two factors ‘higher portion of vowels’ and ‘vowel final’: vowels make a name more sonorous, which is usually perceived as sounding softer, and softness fits better to the female gender stereotype than to the male one.[11] As far as the position of the main stress is concerned, exoticism is the decisive characteristic: Since initial stress is characteristic of Germanic languages, male names represent the unmarked case more often, which, according to Oelkers (2003: 160), can be associated with the male gender stereotype. Female names deviate from the unmarked stress pattern more often and thus correspond to the female gender stereotype due to their markedness and exoticism. Finally, the distribution of the quality of the stressed vowel can be associated with the famous size-sound symbolism (also known as frequency code, cf., e.g., Hinton et al. 1994: 10; Ohala 1983, 1984). Oelkers (2003: 227) argues that sounds that have been shown to be associated with bigger size (e.g., dark vowels such as /a/, /o/, /u/) are connected with masculinity (for an early study on the mapping of size and sound cf. Sapir 1929). This size-sound mapping is one of the best studied relations in the field of sound symbolism. For instance, Knoeferle et al. (2017) have shown in a recent study which acoustic cues best characterize ‘large’ and ‘small’ sounding phonemes. Consistent with the predictions, their experiment has shown that size judgements were indeed higher for sounds with a higher F1 (which reflects the progressive opening of the jaw) and simultaneously a lower F2 (which increases with vowel frontness) and for sounds with a longer duration.

In sum, Oelkers (2003: 227) claims that German female names are characterized by prosodic-phonological features that make them sound softer and more melodious compared to male names. Additionally, female names sound rather ‘exotic’ compared to typical German prosodic-phonological structures. Male given names, on the other hand, sound ‘harder’, and have rather unremarkable structures (Oelkers 2003: 199). Thus, the prosodic-phonological structures of given names do not only determine the gender of a name but also transport stereotypical ideas of gender (as they exist in many other areas of life) according to Oelkers (2003).

Oelkers (2003: 227–228) assumes a non-arbitrary mapping between phonetic properties of speech sounds and their meaning – namely gender in the case of personal names. Hence, she argues for a case of sound symbolism where semantic concepts – such as femininity – that have no audible characteristics are coded phonologically. This case of sound symbolism has been called synesthetic sound symbolism by Hinton et al. (1994: 4–5). In the following, we adopt this term (although the more common use of the term synesthetic relates to sensory experience).

The hypothesis that ‘fe/male sounds’ exist implies that the prosodic-phonological patterns that we find in German given names must be valid beyond (closely related) individual languages if not universally (Oelkers 2003: 228). Accordingly, Oelkers (2003: 227–228) explicitly argues against Cassidy et al.’s (1999) claim that “phonological cues to gender appear to be language specific and psychologically arbitrary”. On the basis of data from countries where English is the majority language also Cutler et al. (1990), Whissell (2001), and Pitcher et al. (2013) reason in favor of synesthetic sound symbolism.[12]

By contrast, Nübling (2018) (and similarly also Cassidy et al. 1999 and Hough 2000), argues against synesthetic sound symbolism and states that no sound as such is male or female – it is only due to convention that some phonological patterns are associated with the information [± female] (i.e., conventional sound symbolism in the sense of Hinton et al. 1994). Furthermore, she claims that correlations between a name’s prosodic-phonological structure and the assigned gender can be explained from a diachronic perspective: In German -a (sometimes also -e) has attained the morphological status of a gender suffix by deriving female names from many male names for centuries (cf. Example (1)) – the other direction is blocked in German.

a./ˈmaɐ.tin/male name → / maɐ.ˈtiː.na/female name
b./ˈkʁıs.tjan/male name → /kʁıs.ˈtjaː.nǝ/female name

Thus, we are dealing with reanalyzes at the morphology/phonology interface: Morphologically (by suffixes such as -a) caused phonological effects (open final syllables, more syllables, shift to non-initial stress, etc.) result in specific sound patterns and are associated with femininity. By contrast, the sound structures of morphologically unmodified names (closed final syllables, fewer syllables, initial stress, etc.) are associated with masculinity. Hence, Nübling (2018) interprets correlations of gender and phonological structure as language/culture-specific conventions.

Against this backdrop, a study which takes different languages/cultures into account suggests itself. This is where our study takes its point of departure. In the next sections, we compare the relation of gender and phonological structures of names across different languages/cultures. By doing so, we want to assess whether the observed correlations of name phonology and gender are valid beyond single languages/cultures.

3 Empirical investigation: phonological cues to gender in 13 languages/cultures

3.1 Sample selection

In order to fill the outlined research gap, we analyzed two samples of popular given names. Sample 1 consists of names from linguistically rather disparate countries, namely China, France, Germany, Hungary, Israel, Japan, Poland, and Turkey.[13] These countries differ not only concerning their majority language but also with regard to gender marking on given names. All three types introduced in Section 2 are covered, i.e., the conventional type (which predominates, e.g., in Germany), the semantic type (which predominates, e.g., in Turkey), and countries where gender is often marked by formal means such as final -a for female names (e.g., in Poland).[14] With this sample, we want to find out whether there are language/culture-overarching phonological cues to gender.[15]

Sample 2, by contrast, consists of names from European countries where an Indo-European language is primarily spoken, namely Bulgaria, Denmark, England, France, Germany, Poland, Romania, and Spain.[16] This sample allows us to study possible language-family/culture-specific correlations of name structure and gender.

For each country, the 30 most popular names per gender were selected on the basis of recent (maximum five years old) and trustworthy statistics on the naming of newborns, such as those made available, for example, by public authorities.[17] A country was only selected for the study if such statistics were available – which excluded some interesting areas/countries.[18][19]

Table 3 gives an overview of the composition of the two samples. (See also the link to the data samples on Zenodo in the Supplement below).

Table 3:

Composition of heterogeneous and Indo-European name sample.

Sample 1 (Heterogeneous)Sample 2 (Indo-European)
CountryMajority languageLanguage FamilyNCountryMajority languageLanguage FamilyN
ChinaMandarinSino-Tibetan Sinitic60BulgariaBulgarianSlavic60
FranceFrenchIndo-European Romance60PolandPolishSlavic60





Common Turkic

For all countries, the official (and thus registered) full forms of the names were used. This is not without problems because it can be assumed that, for example, a Mevlüt is sometimes called Mevo in Turkish. The problem of the unofficial shortening or change of the name applies to many languages/cultures. However, for our study we decided to choose the official names since reliable and representative information on unofficial names in the selected countries can hardly be gathered. In addition, one can assume that official names have at least a certain relevance in the countries at hand. Only Russia, for which it is generally known that the full forms recorded in official statistics are exceptionally rarely in use and the unofficial names have little in common with the registered ones (e.g., AlexanderSascha; MariaMascha), was not included in the sample for exactly this reason.[20]

After sample composition, all 780 names were transcribed with the help of native speakers and were annotated for a range of prosodic-phonological features, such as the number of syllables, the average sonority, et cetera.[21] The transcription of the names and the annotation of factors such as sonority are the subject of the next section.

3.2 Transcription and annotation of the data

First, a few essential procedures regarding transcription have to be addressed. In general, for all names narrow phonetic transcriptions were made based on speech recordings or native speaker competence. This was always based on what our informants (of whom all are linguistically well informed) perceived as default pronunciation. Affricates and diphthongs were always transcribed and counted as a sequence of two segments each (e.g., James [dʒɛims] = six sounds).[22] Hiatuses were only counted as two syllables if they appear non-contracted as in the German female name Mia [ˈmiː.a]. Names that are usually contracted were transcribed accordingly (e.g., Julia [ˈjuːl.ja]). This method is arguable (see, e.g., Nübling 2009, who consistently counts hiatuses as two syllables). However, since the transcription used here treats female and male names equally, possible biases between the genders should be minimal.

After phonological transcription all names were annotated for prosodic-phonological features that have been mentioned in the literature on the correlation of name phonology and gender (Cassidy et al. 1999; Cutler et al. 1990; Lieberson and Mikelson 1995; Nübling 2009; Oelkers 2003; Pitcher et al. 2013; Slater and Feinman 1985; Whissell 2001). One factor refers to the entire name structure, namely the number of syllables. The other factors – namely sonority and the vowel quality – refer to specific structural positions within the name, such as the final sound.

Let us first come to the factor vowel quality. As mentioned in Section 2, the quality of vowels can give rise to different associations with regard to the size of a denotatum. Applied to gender, this means that smaller-sounding vowels correspond to a female gender stereotype and should therefore occur more frequently in female names. As Knoeferle et al. (2017) have shown the opening of the jaw as well as frontness play a significant role for the size-judgement. Accordingly, we made a distinction between palatal (e.g., [i], [e], [y]) and non-palatal vowels (e.g., [a], [o], [u]). Unlike, e.g., Oelkers (2003), who only considers the quality of the stressed vowel, we counted all palatal and non-palatal vowels in one name, but not the final sound. We excluded the final sound here because the name final position has been shown to be especially prominent with regard to gender coding in many (Indo-European) languages. It is often the most sonorous, but also ‘large’ sounding vowel /a/, which indicates femininity (cf., e.g., Kürschner 2018: 306; Lieberson and Mikelson 1995: 935; Nübling 2018: 244; Oelkers 2003: 195). This structural position should therefore be considered separately in order to avoid the possibility that the final sound obscures other more subtle phonological cues to gender.

The second factor that needs to be explained in more detail is sonority. As Nübling (2009: 81) notes in her analysis of German given names, it is too simplistic to assume a dichotomy between vowels and consonants and to put them in opposition to each other. It is well known that a continuum extends between vowels, which are highly sonorous, and consonants. This can be shown with the sonority scale in Figure 1 (Nübling 2009: 81; cf. also, e.g., Neef 2002; Vennemann 1982).

Figure 1: Sonority hierarchy and sonority values (cf. Nübling 2009: 81).

Figure 1:

Sonority hierarchy and sonority values (cf. Nübling 2009: 81).

According to the sonority scale, vowels are most sonorous. Within the group of vowels open vowels (/a/) are more sonorous than mid vowels (e.g., /e/, /o/) and mid vowels are more sonorous than close vowels (e.g., /i/, /u/). In addition to considering such sonority differences within a group of sounds the scale also incorporates the fact that not only vowels at the top of the scale, but also consonantal sonorants (and voiced fricatives) are highly euphonic and give names a sound that is perceived as soft (cf. Nübling 2009: 81). The sonority decreases continuously towards the right end of the scale. Accordingly, voiceless plosives have the lowest sonority and thus the highest degree of consonantality. In order to operationalize the factor sonority, values – based on Nübling (2009) – have been assigned to all sounds. These values are shown in Figure 1 in the top row.[23] In accordance with this scale, numerical sonority values were assigned to every name as a whole and to special structural positions of the name, i.e., the initial and the final sound.[24]

Table 4 lists all coded factors and their possible characteristics as well as the predictions regarding a male or female gender cue (as discussed in Section 2).

Table 4:

Relevant factors with possible characteristics and predictions regarding the gender cue.

FeatureValuesGender cue
Number of syllables1–5 syllablesHigher number of syllables = female gender cue
Number of non-palatal vowels (excluding final sound)0–3Higher number of non-palatal vowels = male gender cue
Sonority of first sound1–10 points (on sonority scale)Higher sonority = female gender cue
Average sonority (till final vowel)1–10 points (on sonority scale)Higher sonority = female gender cue
Sonority of final sound1–10 points (on sonority scale)Higher sonority = female gender cue

Finally, it should be briefly mentioned why certain factors that have been proven to be relevant in individual languages are not (or: cannot be) taken into account in our crosslinguistic study. These are stress position, the ratio of open versus closed syllables, and the number of consonant clusters. The main stress – which has been shown to be a significant factor in some Germanic languages – was not considered here due to different types of lexical stress realization and different stress positions in the languages of our sample. Thus, the stress types are not comparable and a derivation of hypotheses regarding gender cues seems questionable.[25]

The ratio of open and closed syllables in a name depends on the respective language: are we rather dealing with a syllable or with a word language (cf., e.g., Auer 1993; Caro Reina and Szczepaniak 2014)? German and English – languages in which the factor has been shown to be relevant – have traits of word languages, i.e., the prosodic domain of the phonological word is central. However, there are also syllable languages in our sample, i.e., languages in which the prosodic domain of the syllable is central, which, among other things, is reflected in an optimized CV-structure (e.g., Japanese). Therefore, it is a priori clear that the ratio of open and closed syllables cannot be relevant across all languages included.[26] The same holds for the number of consonant clusters per name.

In the next section, the crosslinguistic relevance of the factors discussed here will be examined in two multifactorial analyses.

3.3 Binary logistic regressions

3.3.1 Heterogeneous sample

In the previous section, various features were discussed that have been said to be typical for female or male names in German and English. As shown, there are some striking parallels between these two languages. However, this is not particularly surprising as English and German are geographically close and genetically related. In order to examine whether these factors are also influential when it comes to a sample of names from linguistically rather disparate countries, the data were analyzed with a binary logistic regression model.[27] This will be described in this section.

In the first multifactorial model the variables number of syllables, number of non-palatal vowels (excluding final sound), sonority of first sound, average sonority (till final vowel), and sonority of final sound were integrated. Hence, we started with a maximal model. Here (and in all following models), gender is the dependent variable. Independent variables that have no significant influence were removed from the model in a step-down procedure based on the Akaike Information Criterion (AIC). By this procedure, two out of five variables – i.e., number of syllables and sonority of first sound – were removed as they do not improve the quality of the model. The specification of the final model is given in (2).

GENDER ∼ non-palatal vowels without final sound + average sonority till final V + sonority final sound

Table 5 gives an overview of the significance of the predictors. In addition, the effect sizes (log odds), standard errors, Wald Z, and p-values are listed. The concordance index shows that the model discriminates acceptably: C = 0.784 (cf., e.g., Hosmer and Lemeshow 2000: 162).[28]

Table 5:

Predictors in the minimal adequate logistic regression model for name gender (reference level = female).

Coeff.S.E.Wald ZPr(>|Z|)

In contrast to p-values, effect coefficients (here we used log odds) do not indicate whether an influence is significant, but its direction and its strength. Log odds are centered around 0 and reach from +Infinity to −Infinity. Thus, a coefficient of 0 would mean that there is no difference between the levels of the predictor with regard to the choice between female and male names. Coefficients greater than 0 mean that the probability that the respective name is a male name is greater than the probability of a female name.[29] For example, number of non-palatal vowels (excluding final sound) has a positive coefficient and thus an increase in the number of non-palatal vowels (excluding the final sound) increases the odds for a male name.[30]

Based on the included variables, the model predicts the gender of 341 names (i.e., 71 % of the tokens) correctly (cf. Table 6).

Table 6:

Confusion matrix (heterogeneous sample).

ActualPredictedCorrectly predicted gender

Altogether, the coefficients confirm our hypotheses: Female names have a more sonorous final sound, a higher average sonority (till final vowel), and fewer non-palatal vowels than male names. To test whether these significant predictors are relevant for all languages/cultures or whether the significance is based on high values in only some languages/cultures we checked possible interactions with the factor country in a second step. In order to do so, we compared a first regression model where we included the factor country but no interaction term with regression models respectively including interaction terms for country*number of non-palatal vowels (excluding final sound), country*average sonority (till final vowel), and country*sonority of final sound. Then, we compared the model with no interaction term with the models with an interaction term via ANOVA and tested whether the interaction term improved the model significantly. This was the case for the sonority-factors. Especially, the factor sonority of final sound is highly language/culture-dependent (3):

Model 1: Gender ∼ sonority_final_sound + non.palatal_vowels_without_final_sound + average_sonority_till_final_V + country
Model 2: Gender ∼ sonority_final_sound * country + non.palatal_vowels_without_final_sound + average_sonority_till_final_V
ResidDfResid. DevDfDeviancePr(>Chi)
2462453.09777.3194.845e-14 ***

As Table 7 shows, the gender differences regarding the sonority of the final sound are particularly high in European countries. Contrarily, in Non-European countries, such as China and Japan, the final sound seems to be irrelevant with regard to gender.[31]

Table 7:

Gender differences in the sonority of final sound per country.


In contrast to the sonority of the final sound, the factor number of non-palatal vowels (excluding final sound) seems to be language-/culture-overarching as the comparison of the model with and without interaction term shows (4).

Model 1: Gender ∼ sonority_final_sound + non.palatal_vowels_without_final_sound + average_sonority_till_final_V + country
Model 2: Gender ∼ sonority_final_sound + non.palatal_vowels_without_final_sound * country + average_sonority_till_final_V
Resid.DfResid. DevDfDeviancePr(>Chi)

This result can be illustrated by the monofactorial analysis given in Table 8: In all countries in our sample male names have on average more non-palatal vowels (if one ignores the final sound) and the differences between the countries are much weaker compared to Table 7.

Table 8:

Gender differences in the number of non-palatal vowels per country.


The results described in this section showed that there are three groups of variables:

  1. (i)

    variables that did not have a significant effect at all (number of syllables and sonority of first sound),

  2. (ii)

    variables that have an effect but whose significance is based on only some countries in the sample (average sonority till final vowel and sonority of final sound), and

  3. (iii)

    one variable that proved to be important across all countries: number of non-palatal vowels (excluding final sound).

As there is only one variable in group (iii) it is not surprising that the model can only explain a moderate proportion of variance (Nagelkerke Pseudo-R2 = 0.300) – the vast majority of variables is irrelevant for some countries in the sample. In the next section, we test whether the model quality improves if a linguistically more homogeneous set of countries is analyzed. In order to do so we compiled a sample of names from countries where an Indo-European language is the majority language. Comparing the resulting model and the model described in this section will clarify whether we are dealing with language family internal tendencies.

3.3.2 Indo-European sample

The analysis of the second sample mostly matches the procedure described in Section 3.3.1. To avoid repetitions, not all steps are explained in detail again. We started with a maximum model and removed all variables that have no significant influence in a step-down procedure based on the Akaike Information Criterion (AIC). Only sonority of first sound was excluded. The specification of the final model is given in (5):

GENDER ∼ number of syllables + non-palatal vowels without final sound + average sonority till final V + sonority final sound

Table 9 contains relevant values resulting from the regression analysis.

Table 9:

Predictors in the minimal adequate logistic regression model for name gender (reference level = female).

CoefS.E.Wald ZPr(>|Z|)

The model closely resembles the one described in Section 3.3.1. The only additional predictor that reached significance is number of syllables. The coefficients of the other variables match their counterparts from the analysis of the other sample with regard to their direction. However, the sonority variables deviate more strongly from 0 which means that they have a stronger impact on the dependent variable. The additional variable and the higher impact of two variables are reflected in a better model quality, which is shown in Table 10.[32] According to Hosmer and Lemeshow (2000: 162) a C index above 0.8 indicates “excellent discrimination”.[33]

Table 10:

Comparison of model qualities (Heterogeneous vs. Indo-European sample).

Heterogeneous sampleIndo-European sample
Nagelkerke Pseudo-R20.3000.524
Index of discrimination (C)0.7840.885

Given that both samples comprise the same number of names these differences demonstrate that our variables can explain the variance in the comparatively homogenous Indo-European sample better than the variance in the much more heterogeneous sample analyzed in the previous section. This can also be shown by the confusion matrix in Table 11: The gender of more names is predicted correctly compared to the model described in the previous section (cf. Table 6 above). The following section discusses what our empirical results imply with regard to sound symbolism in the domain of phonological gender marking on names.

Table 11:

Confusion matrix (Indo-European sample).

ActualPredictedCorrectly predicted gender

4 Discussion – synesthetic versus conventional sound symbolism

As we have outlined in Section 1, there is no consensus in the literature whether correlations of phonological features and the name bearer’s gender are language/culture-specific or crosslinguistically valid. Our results indicate that the majority of such correlations are not valid beyond related languages in geographically close countries. Using the example of the final sound’s sonority, we have shown that there are important differences between the countries in our first sample. While the sonority of the final sound is a phonological cue to gender in some countries (e.g., Poland, Hungary, Germany), it is completely irrelevant in others (e.g., Japan, China). The way countries cluster in our analysis suggests that phonological gender marking strategies coincide with other cultural, geographical, and linguistic classifications. For example, China and Japan deviate from European countries with regard to the relevance of the final sound’s sonority. The analysis of our second sample supports this idea: Our variables explain the variance in the Indo-European sample much better than the variance in the heterogeneous sample. There are two explanations for this observation. Firstly, all countries in the second sample are not only closely related due to their majority language but they do also share cultural traditions and borrowing played and plays an important role – which results in overlapping/related onomasticons (cf., e.g., Gerhards 2010: 158). Hence, there are several etymologically related names in this sample. For instance, variants of Alexander (which is an extreme example) can be found in the sub-samples from Bulgaria, Denmark, England, Germany, Romania, and Spain (see Appendices A and B). Although these names are pronounced slightly differently in the respective countries, they make the sample more homogenous and this partly explains the difference between the two models. Secondly – and more importantly with regard to our research question – the names in the second sample are also more homogeneous with regard to the structure of the names in general and with regard to gender differences in particular. For example, final a is much more common for female names than for male. This holds irrespective of etymologically related names and applies also to female names without related variants in the sample (such as Bulgarian Iwaila).

Overall, it is not very surprising that our variables explain the variance in the Indo-European sample better than the variance in the heterogeneous sample since our variables are taken from studies on phonological gender marking on names in selected Indo-European languages. Still, this difference contradicts the idea that all or many correlations of phonological properties and the name bearer’s gender discovered by, e.g., Oelkers (2003) for German, are non-arbitrary and potentially universal.

However, we discovered a solid correlation of gender and the name’s sonority (regarding the final sound and the rest of name) in certain countries. In the Indo-European sample the number of syllables also had a significant impact. These correlations seem to be firmly established but linguistically and geographically limited. Therefore, they seem to qualify as cases of conventional sound symbolism. Originally, the diverging phonological properties of female and male names were motivated by, e.g., specific word/name formation patterns. For example, the formation of female names by -a suffixation led to numerous names with comparatively more syllables and a maximally sonorous final sound (cf. Section 2). Subsequently, the frequent co-occurrence of comparatively many syllables and/or a sonorous final sound on the one hand and the semantic property [+female] on the other hand resulted in associations of phonology and semantics in this domain. Since the word formation pattern that triggered this development traces to Latin and is only found in European languages these associations are not universal (cf. Cassidy et al. 1999: 362; Hough 2000: 6; Nübling 2009: 100, 2018: 242; Oelkers 2003: 144–145).

Though, our analyses also revealed that one correlation is valid across all countries in our samples: the number of non-palatal vowels (excluding the final sound). Thus, this variable qualifies as a candidate for synesthetic sound symbolism. Indeed, a corresponding explanation is obvious. One of the most robust findings in the literature on sound symbolism is the fact that non-palatal vowels are associated with bigger size (cf., e.g., Elsen 2017: 492) and bigger size is often associated with the masculine gender stereotype. Against this backdrop, it seems to be explicable that male names have more non-palatal vowels in word-initial and word-medial position than female names across all countries in our samples. The claim that synesthetic sound symbolism is at play here (as put forward, e.g., by Cutler et al. 1990; Oelkers 2003: 227–228; Pitcher et al. 2013; van de Weijer et al. 2020; Whissell 2001) therefore seems to be justifiable from a crosslinguistic perspective in this particular case. However, synesthetic sound symbolism has only an indirect influence (see also van de Weijer et al. 2020: 16). A matching gender stereotype is also necessary to establish an association of the phonological (e.g., word medial [i]) and the semantic property (e.g., female name bearer).[34]Figure 2 depicts this mechanism.

Figure 2: Phonological gender marking via sound symbolism and gender stereotypes.

Figure 2:

Phonological gender marking via sound symbolism and gender stereotypes.

In contrast, conventional sound symbolism is based on the frequent co-occurrence of a phonological (e.g., word final [a]) and a semantic property (e.g., female name bearer) and does not involve a second semantic concept. This is depicted in Figure 3.

Figure 3: Phonological gender marking via conventional sound symbolism.

Figure 3:

Phonological gender marking via conventional sound symbolism.

Interestingly, synesthetic and conventional sound symbolism conflict in the Indo-European languages. This holds particularly for the sound [a]. On the one hand sonorous word final sounds (such as [a]) are associated with female names (due to word formation patterns; conventional sound symbolism) while on the other hand non-palatal vowels (such as [a]) are associated with male names (via association with bigger size; synesthetic sound symbolism). Hence, the connotation this sound triggers depends on its position in the name: word final position triggers a female association; word initial or word-medial position triggers a male association. In word final position the strong association of sonorous sound and female gender resulting from the specific word formation pattern overrides the general tendency to associate non-palatal vowels with bigger size and masculinity.

Both types of phonological gender marking seem to be important for name choices. Certain names are obviously perceived as more appropriate with regard to the gender of a child due to phonological properties. This reinforces correlations of gender and the phonology of the most popular names (which we analyzed here).

With regard to synesthetic sound symbolism the attractiveness of the associated meaning (e.g., smallness) also plays a role. While it might be the case that smaller sounding names are perceived as appropriate for women due to sexual size dimorphism, this does not necessarily mean that smallness is considered a desirable characteristic for women. This would diminish the attractiveness of the particular structure and would run contrary to a corresponding correlation of phonology and gender. As there are certainly cultures where this is the case it is likely that exceptions to the correlation of palatal vowels and feminine name bearers can be found (Pitcher et al. 2013: 5).

5 Conclusion

In our study, we addressed the correlation of name phonology and gender from a crosslinguistic perspective. Considering name giving practices in several (linguistically disparate) countries allows us to re-evaluate claims on how this topic relates to sound symbolism. Our empirical results yielded that the majority of the phonological patterns that have been shown to correlate with the gender of the name bearer in German and English (cf. Nübling 2009, 2012, 2018; Barry and Harper 1995; Cutler et al. 1990; Lieberson and Bell 1992; Oelkers 2003, 2004; Slater and Feinman 1985; Whissell 2001) are not associated with a certain gender universally. Instead, such correlations are limited to certain regions / language families. This substantiates Nübling’s (2018) claim that no sound as such is male or female and that specific correlations can be explained with reference to the diachrony of individual languages or language families. Also, cultural aspects (which lead to partly shared / related onomasticons) play a role here.

Even so, we also detected a phonological variable that correlates with gender in all countries studied: the number of non-palatal vowels (excluding the final sound). Here, an explanation based on synesthetic sound symbolism seems appropriate which supports Oelkers’ claim that correlations of name phonology and gender are not necessarily arbitrary. Thus, there is evidence for different kinds of sound symbolism in the domain of onymic gender marking. However, the instances of conventional sound symbolism outnumber synesthetic sound symbolism.

In addition to these results, our study has also revealed some directions for future research. For example, it would be particularly interesting to examine the relevance of non-palatal vowels experimentally. In such a study, one could ask participants with disparate linguistic backgrounds to assign “pseudonames” to female or male name bearers (cf. Cassidy at al. 1999 and Oelkers 2003 for similar studies with English and German speaking participants). Furthermore, it would be fruitful to analyze language families other than Indo-European more closely. Certainly, there are means of onymic gender marking that have not been considered here because it is obvious that they are language (family) specific. Another interesting topic would be the question how language and culture contact influence name structures and onymic gender marking. For example, it seems to be the case that Japanese names approach European patterns also in terms of phonological gender marking on names due to an increase of female names ending in a, which has been quite uncommon for Japanese female names in the past. In conclusion, there are still a number of open questions and issues that are worth addressing in the field of phonological gender marking on names. However, by adopting a crosslinguistic perspective, we hope to have shed some new light on this topic.

Corresponding authors: Tanja AckermannandChristian Zimmer, Institut für Deutsche und Niederländische Philologie, Freie Universität Berlin, Habelschwerdter Allee 45, 14195Berlin, Germany, E-mail: (T. Ackermann), (C. Zimmer)


Many thanks are due to Beijia Chen, Semra Kızılkaya, Konrad Mazur, Malka Muchnik, Eva Meier, Takuma and Vera Melber, Nina Nikulova, Milena Osterloh, and Britta Stuhl for their help with the transcription of the names.

  1. Supplement: The data samples are available for viewing at

Appendix A

Table A1 contains information on the sources of our name samples for all countries considered here. If available, a URL is given (including the date the material was accessed). The year from which the statistics originate is also given.

Table A1:

Sources of our name samples

China2015< on a representative sample collected by the company QiMingTong in cooperation with the Tsinghua University Beijing
England2017<>03/11/2019only data from England were considered for our study
Germany2017<>07/09/2018based on statistics provided by selected regional register offices
Israel2016No published data available. Data were provided by the Central Israeli Bureau of Statistics on request via e-mail.
Japan2017<>07/06/2018based on statistics provided by an insurance company
Romania2009<>02/14/2019Data provided by the Romanian Ministry of Administration and Interior
Turkey2014Türkiye İstatistik Kurumu Matbaası [Turkish Statistical Institute]. 2014. İstatistiklerle Çocuk / Statistics On Child 2014. Ankara: Türkiye İstatistik Kurumu Matbaası. ISBN: 978-975-19-6341-3.

Appendix B

01 Viktoriya11 Elena21 Stefani31 Aleksandar41 Kristiyan51 Samuil
02 Mariya12 Michaela22 Magdalena32 Georgi42 Viktor52 Aleks
03 Nikol13 Teodora23 Karina33 Martin43 Teodor53 Hristo
04 Raya14 Bozhidara24 Anna34 Iwan44 Bozhidar54 Atanas
05 Aleksandra15 Gergana25 Niya35 Dimitar45 Simeon55 Iordan
06 Sofiya16 Ema26 Vanessa36 Nikola46 Stefan56 David
07 Dariya17 Siyana27 Tzwetelina37 Daniel47 Petar57 Ioan
08 Simona18 Iwaila28 Ewa38 Boris48 Iwailo58 Wassil
09 Gebriela19 Kalina29 Darina39 Kaloyan49 Angel59 Kristian
10 Ioana20 Monika30 Plamena40 Nikolai50 Michail60 Borislav
61 Zǐ·xuān71 Jǐn·xuān81 Mèng·qí91 Hào·rán101 Zǐ·muò111 Jùn·jié
62 Zǐ·hán72 Sī·hán82 Yǜ·hán92 Zǐ·xuān102 Bó·wén112 Zhì·yuǎn
63 Shī·hán73 Kě·xīn83 Zǐ·méng93 Hào·xuān103 Yī·nuò113 Tiān·yòu
64 Kě·xīn74 Zǐ·xuān84 Zhǐ·xuān94 Yǚ·xuān104 Zǐ·hán114 Míng·xuān
65 Yī·nuò75 Ruò·xī85 Yì·hán95 Hào·yǚ105 Zǐ·ruì115 Zǐ·hán
66 Yǚ·xuān76 Zǐ·xuān86 Ruò·xuān96 Zǐ·ruì106 Ruì116 Jùn·háo
67 Xīn·yí77 Yǚ·hán87 Yī·yī97 Zǐ·xuān107 Yǚ·zé117 Hào·rán
68 Zǐ·hán78 Yǚ·tóng88 Ruò·xī98 Hào·xuān108 Míng·xuān118 Yī·míng
69 Chén·xī79 Xīn·yán89 Yì·xīn99 Jùn·xī109 Yǚ·háng119 Hào·yǚ
70 Zǐ·xuān80 Ruò·xī90 Shī·qí100 Zǐ·háo110 Zǐ·háo120 Zǐ·chén
121 Ida131 Laura141 Aya151 William161 Valdemar171 Alexander
122 Emma132 Nora142 Sofie152 Noah162 Elias172 Villads
123 Sofia133 Clara143 Ellen153 Oscar163 Magnus173 Christian
124 Ella134 Karla144 Lily154 Lucas164 Aksel174 Johan
125 Freja135 Isabella145 Mathilde155 Carl165 Frederik175 Adam
126 Josefine136 Olivia146 Maja156 Victor166 Felix176 Arthur
127 Alma137 Lærke147 Frida157 Oliver167 Elliot177 Liam
128 Alberte138 Victoria148 Emilie158 Alfred168 August178 Theo
129 Anna139 Mille149 Marie159 Malthe169 Anton179 Albert
130 Agnes140 Luna150 Esther160 Emil170 Nohr180 Mikkel
181 Olivia191 Lily201 Phoebe211 Oliver221 William231 Alexander
182 Amelia192 Sophia202 Sienna212 Harry222 Henry232 Edward
183 Isla193 Grace203 Evelyn213 George223 Alfie233 Theo
184 Ava194 Evie204 Isabelle214 Noah224 Thomas234 Isaac
185 Emily195 Jessica205 Ivy215 Jack225 Joshua235 Lucas
186 Isabella196 Sophie206 Matilda216 Jacob226 Freddie236 Ethan
187 Mia197 Alice207 Willow217 Muhammad227 James237 Max
188 Poppy198 Florence208 Elsie218 Leo228 Arthur238 Joseph
189 Ella199 Daisy209 Chloe219 Oscar229 Archie239 Samuel
190 Charlotte200 Freya210 Scarlett220 Charlie230 Logan240 Mohammed
241 Emma251 Camille261 Julia271 Gabriel281 Nathan291 Timéo
242 Louise252 Léna262 Romane272 Jules282 Arthur292 Théo
243 Jade253 Rose263 Jeanne273 Adam283 Paul293 Mohamed
244 Alice254 Inès264 Eva274 Lucas284 Nolan294 Aaron
245 Chloé255 Anna265 Lou275 Louis285 Maël295 Mathis
246 Mila256 Sarah266 Charlotte276 Raphaël286 Sacha296 Axel
247 Léa257 Zoé267 Louna277 Hugo287 Tom297 Antoine
248 Lina258 Juliette268 Mia278 Léo288 Noah298 Victor
249 Manon259 Ambre269 Nina279 Ethan289 Enzo299 Maxime
250 Lola260 Lucie270 Clara280 Liam290 Gabin300 Clément
301 Emma311 Leni321 Sophie331 Ben341 Henry351 Liam
302 Hannah312 Clara322 Charlotte332 Jonas342 Maximilian352 Moritz
303 Mia313 Lena323 Ida333 Leon343 Luca353 Julian
304 Sofia314 Luisa324 Lilly334 Paul344 Oskar354 Leo
305 Emilia315 Leonie325 Laura335 Finn345 Emil355 David
306 Lina316 Amelie326 Maja336 Noah346 Anton356 Alexander
307 Anna317 Emily327 Mathilda337 Elias347 Max357 Milan
308 Marie318 Johanna328 Lara338 Luis348 Theo358 Philipp
309 Mila319 Ella329 Frieda339 Felix349 Jakob359 Niklas
310 Lea320 Nele330 Lia340 Lukas350 Matteo360 Carl
361 Hanna371 Nóra381 Dóra391 Bence401 Zalán411 Gergő
362 Anna372 Maja382 Sára392 Máté402 Áron412 Benett
363 Jázmin373 Fanni383 Csenge393 Levente403 Balázs413 Bálint
364 Zsófia374 Laura384 Petra394 Dominik404 Kristóf414 Márk
365 Zoé375 Dorina385 Noémi395 Marcell405 Péter415 Zoltán
366 Lili376 Lilla386 Eszter396 Dávid406 Botond416 András
367 Boglárka377 Gréta387 Mira397 Ádám407 Olivér417Attila
368 Luca378 Izabella388 Flóra398 Noel408 László418 Márton
369 Emma379 Viktória389 Zselyke399 Dániel409 Zsombor419 Benedek
370 Léna380 Réka390 Liza400 Milán410 Tamás420 Gábor
421 Tamar431 Esther441 Rachel451 Noam461 Yehuda471 Raphael
422 Noa432 Lia442 Noya452 David462 Abraham472 Shmuel
423 Avigail433 Hanna443 Miriam453 Uri463 Lavi473 Ido
424 Maya434 Rivka444 Ema454 Ariel464 Israel474 Michael
425 Yael435 Roni445 Ruth455 Yosef465 Itamar475 Yitzhak
426 Adel436 Romi446 Yuval456 Eitan466 Omer476 Chaim
427 Shira437 Hodaya447Alma457 Daniel467 Yonatan477 Harel
428 Sarah438 Ela448 Halel458 Itai468 Yair478 Shimon
429 Ayala439 Michal449 Ariel459 Yehonatan469 Yaakov479 Amit
430 Talya440 Chaya450 Tahal460 Moshe470 Eliah480 Alon
481 Sakura491 Honoka501 Sara511 Haruto521 Yuusei531 Yuuki
482 Yui492 Riko502 Hina512 Souta522 Kanata532 Sora
483 Akari493 Mio503 Iroha513 Yuuto523 Souma533 Yuito
484 Mei494 Saki504 Kaho514 Haruki524 Aoi534 Itsuki
485 Hana495 Miyu505 Hinata515 Riku525 Kaito535 Kouta
486 Sana496 Ichika506 Hikari516 Sousuke526 Hayato536 Haruma
487 Rio497 Riso507 Haruka517 Minato527 Haru537 Hiroto
488 Himari498 Ema508 Kanna518 Aoto528 Asaki538 Ayato
489 Koaru499 Tsumugi509 Momoka519 Hinata529 Yuuma539 Ryou
490 Aoi500 Yuina510 Yua520 Kouki530 Rikuto540 Eito
541 Julia551 Oliwia561 Nadia571 Antoni581 Adam591 Bartosz
542 Zuzanna552 Natalia562 Marcelina572 Jakub582 Michał592 Maksymilian
543 Zofia553 Wiktoria563 Gabriela573 Jan583 Marcel593 Miłosz
544 Lena554 Emilia564 Michalina574 Szymon584 Stanisław594 Tymon
545 Maja555 Antonina565 Kornelia574 Franciszek585 Wiktor595 Oliwier
546 Hanna556 Laura566 Nikola575 Filip586 Piotr596 Alan
547 Amelia557 Pola567 Helena576 Aleksander587 Igor597 Ignacy
548 Alicja558 Iga568 Milena578 Mikołaj588 Leon598 Tymoteusz
549 Maria559 Anna569 Martyna579 Wojciech589 Nikodem599 Oskar
550 Aleksandra560 Liliana570 Jagoda580 Kacper590 Mateusz600 Dawid
601 Maria611 Mihaela621 Larisa631 Andrei641 Florin651 Eduard
602 Andreea612 Bianca622 Sara632 Alexandru642 Darius652 Mario
603 Elena613 Georgiana623 Daniela633 Ionuț643 Denis653 Matei
604 Ioana614 Nicoleta624 Miruna634 Gabriel644 Constantin654 Bogdan
605 Alexandra615 Cristina625 Roxana635 Ștefan645 Robert655 Răzvan
606 Ana616 Teodora626 Valentina636 Cristian646 Adrian656 Rareș
607 Denisa617 Diana627 Rebeca637 Mihai647 Ioan657 Valentin
608 Gabriela618 Alexia628 Raluca638 David648 Sebastian658 Cosmin
609 Ștefania619 Ionela629 Sofia639 Daniel649 Luca659 Marius
610 Daria620 Florentina630 Mădălina640 Marian650 George660 Nicolae
661 Lucia671 Carla681 Olivia691 Lucas701 Leo711 Carlos
662 Sofia672 Sara682 Elena692 Hugo702 David712 Sergio
663 Maria673 Noa683 Adriana693 Martin703 Mario713 Marc
664 Martina674 Carmen684 Laia694 Daniel704 Diego714 Antonio
665 Paula675 Claudia685 Vega695 Pablo705 Javier715 Bruno
666 Julia676 Valentina686 Vera696 Alejandro706 Enzo716 Miguel
667 Daniela677 Alma687 Lola697 Mateo707 Izan717 Gonzalo
668 Valeria678 Ana688 Irene698 Adrian708 Marcos718 Jorge
669 Alba679 Chloe689 Jimena699 Alvaro709 Marco719 Juan
670 Emma680 Marta690 Alejandra700 Manuel710 Alex720 Angel
721 Zeynep731 Meryem741 Beren751 Yusuf761 Kerem771 İbrahim
722 Elif732 Nisanur742 Ayşe752 Berat762 Ayaz772 Ömer Faruk
723 Hiranur733 Hira743 Merve753 Mustafa763 Çınar773 Umut
724 Yağmur734 Belinay744 Ada754 Ömer764 Ali774 Furkan
725 Ecrin735 Hira Nur745 Defne755 Ahmet765 Enes775 Mert
726 Zehra736 Rabia746 Fatma756 Eymen766 Muhammed

776 Rüzgar
727 Azra737 Ela747 Asya757 Muhammed767 Yunus Emre777 Hasan
728 Miray738 Esma748 Melek758 Miraç768 Hamza778 Muhammed Emin
729 Nehir739 Miray749 Esmanur759 Mehmet769 Emirhan779 Burak
730 Eylül740 Sümeyye750 Hatice760 Emir770 Hüseyin780 Yiğit


Ackermann, Tanja. 2011. Aloe vera vs. Click – Zur phonologischen Kodierung von Geschlecht bei Warennamen (Deodorants). Beiträge zur Namenforschung 46. 1–50. Search in Google Scholar

Alford, Richard D. 1988. Naming and identity: A cross-cultural study of personal naming practices. New Haven, CT: HRAF Press. Search in Google Scholar

Auer, Peter. 1993. Is a rhythm-based typology possible? A study of the role of prosody in phonological typology. KontRI Working Paper 21. Universität Konstanz. Available at: Search in Google Scholar

Barry, Herbert & Aylene S. Harper. 1995. Increased choice of female phonetic attributes in first names. Sex Roles 32. 809–819. Search in Google Scholar

Cai, Zhenguang G. & Nan Zhao. 2019. The sound of gender: Inferring the gender of names in a foreign language. Journal of Cultural Cognitive Science 3. 63–73. Search in Google Scholar

Caro Reina, Javier & Renata Szczepaniak. 2014. Introduction: Syllable and word languages. In Javier Caro Reina & Renata Szczepaniak (eds.), Syllable and word languages (Linguae & litterae 40), 8–39. Berlin & Boston: De Gruyter. Search in Google Scholar

Cassidy, Kimberly Wright, Michael H. Kelly & Lee’at J. Sharoni. 1999. Inferring gender from name phonology. Journal of Experimental Psychology 128. 362–381. Search in Google Scholar

Cutler, Anne, James McQueen & Ken Robinson. 1990. Elizabeth and John: Sound patterns of men’s and women’s names. Journal of Linguistics 26. 471–482. Search in Google Scholar

Elsen, Hilke. 2016. Einführung in die Lautsymbolik. Berlin: Schmidt. Search in Google Scholar

Elsen, Hilke. 2017. The two meanings of sound symbolism. Open Linguistics 3. 491–499. Search in Google Scholar

Fredrickson, Annie. 2007. Phonological cues to gender in sex-typed and unisex names. Available at: www. Search in Google Scholar

Gerhards, Jürgen. 2010. Die Moderne und ihre Vornamen: Eine Einladung in die Kultursoziologie, 2nd edn. Wiesbaden: VS Verlag für Sozialwissenschaften. Search in Google Scholar

Gries, Stefan Th. 2013. Statistics for linguistics with R: A practical introduction, 2nd edn. Berlin & Boston: De Gruyter Mouton. Search in Google Scholar

Handschuh, Corinna. 2019. The classification of names: A crosslinguistic study of sex-specific forms, classifiers, and gender marking on personal names. STUF – Language Typology and Universals 72. 539–572. Search in Google Scholar

Harrell Jr, E. Frank. 2019. rms: Regression modeling strategies. R package version 5.1-3. Available at: Search in Google Scholar

Hinton, Leanne, Johanna Nichols & John J. Ohala. 1994. Introduction: Sound-symbolic processes. In Leanne Hinton, Johanna Nichols & John J. Ohala (eds.), Sound symbolism, 1–12. Cambridge: Cambridge University Press. Search in Google Scholar

Hosmer, David W. & Lemeshow. Stanley. 2000. Applied logistic regression: Hosmer/applied logistic regression. Hoboken, NJ: Wiley. Search in Google Scholar

Hough, Carole. 2000. Towards an explanation of phonetic differentiation in masculine and feminine personal names. Journal of Linguistics 36. 1–11. Search in Google Scholar

Klink, Richard R. 2000. Creating brand names with meaning: The use of sound symbolism. Marketing Letters 11. 5–20. Search in Google Scholar

Knoeferle, Klemens, Jixing Li, Emanuela Maggioni & Charles Spence. 2017. What drives sound symbolism? Different acoustic cues underlie sound-size and sound-shape mappings. Scientific Reports 7. 1–11. Search in Google Scholar

Kürschner, Sebastian. 2018. Von Helga und Björn zu Sara und Aron: Der neueste Wandel isländischer Rufnamen im Vergleich mit dem Deutschen. In Seelow Hubert, Lena Rohrbach & Sebastian Kürschner (eds.), Deutsch-isländische Beziehungen: Festschrift für Hubert Seelow zum 70. Geburtstag (Berliner Beiträge zur Skandinavistik 24), 301–312. Berlin: Nordeuropa-Institut der Humboldt-Universität. Search in Google Scholar

Lieberson, Stanley & Eleanor O. Bell. 1992. Children’s first names: An empirical study of social taste. American Journal of Sociology 98. 511–554. Search in Google Scholar

Lieberson, Stanley & Kelly S. Mikelson. 1995. Distinctive African American names: An experimental, historical, and linguistic analysis of innovation. American Sociological Review 60. 928–946. Search in Google Scholar

Muchnik, Malka. 2017. Personal names in Modern Hebrew: A morphosyntactic and gender analysis. Folia Linguistica 51. 369–390. Search in Google Scholar

Neef, Martin. 2002. Das Maß aller Dinge: Sonorität. In Michael Bommes, Christina Noack & Doris Tophinke (eds.), Sprache als Form: Festschrift für Utz Maas: Zum 60. Geburtstag, 32–48. Wiesbaden: Westdeutscher Verlag. Search in Google Scholar

Nübling, Damaris. 2009. Von Monika zu Mia, von Norbert zu Noah: Zur Androgynisierung der Rufnamen seit 1945 auf prosodisch-phonologischer Ebene. Beiträge zur Namenforschung 44. 67–110. Search in Google Scholar

Nübling, Damaris. 2012. Von Elisabeth zu Lilly, von Klaus zu Nico: Zur Androgynisierung und Infantilisierung der Rufnamen von 1945 bis 2008. In Susanne Günthner, Dagmar Hüpper & Constanze Spieß (eds.), Genderlinguistik: Sprachliche Konstruktionen von Geschlechtsidentität (Linguistik – Impulse und Tendenzen 45), 319–357. Berlin & Boston: de Gruyter. Search in Google Scholar

Nübling, Damaris. 2017. Beziehung überschreibt Geschlecht. Zum Genderindex von Ruf- und Kosenamen. In Angelika Linke & Juliane Schröter (eds.), Sprache und Beziehung (Linguistik − Impulse & Tendenzen 69), 99–118. Berlin & Boston: de Gruyter. Search in Google Scholar

Nübling, Damaris. 2018. Luca und Noah – Das phonologische Degendering von Jungennamen seit der Jahrtausendwende. In Damaris Nübling & Stefan Hirschauer (eds.), Namen und Geschlechter: Studien zum onymischen un/doing Gender (Linguistik − Impulse & Tendenzen 76), 239–270. Berlin & Boston: de Gruyter. Search in Google Scholar

Nübling, Damaris, Fabian Fahlbusch & Rita Heuser. 2015. Namen: Eine Einführung in die Onomastik, 2nd edn. Tübingen: Narr. Search in Google Scholar

Oelkers, Susanne. 2003. Naming Gender. Empirische Untersuchungen zur phonologischen Struktur von Vornamen im Deutschen. Frankfurt am Main: Lang. Search in Google Scholar

Oelkers, Susanne. 2004. Der Fall Luca: Zur Männlichkeit und Weiblichkeit von Vornamen. Bulletin suisse de linguistique appliquée 80. 155–170. Search in Google Scholar

Ohala, John J. 1983. Cross-language use of pitch: An ethological view. Phonetica 40. 1–18. Search in Google Scholar

Ohala, John J. 1984. An ethological perspective on common cross-language utilization of F0 of voice. Phonetica 41. 1–16. Search in Google Scholar

Pitcher, Benjamin J., Alex Mesoudi & Alan G. McElligott. 2013. Sex-biased sound symbolism in English-language first names. PLoS ONE 6. 1–6. Search in Google Scholar

R Core Team. 2019. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. Available at: Search in Google Scholar

Sapir, Edward. 1929. A study in phonetic sound symbolism. Journal of Experimental Psychology 12. 225–239. Search in Google Scholar

Shrum, L. J. & Tina M. Lowrey. 2007. Sounds convey meaning: The implications of phonetic symbolism for brand name construction. In Tina M. Lowrey (ed.), Psycho-linguistic phenomena in marketing communications, 39–58. Mahwah, NJ: Lawrence Erlbaum. Search in Google Scholar

Slater, Anne Saxon & Saul Feinman. 1985. Gender and the phonology of North American first names. Sex Roles 13. 429–440. Search in Google Scholar

Slepian, Michael L. & Adam D. Galinsky. 2016. The voiced pronunciation of initial phonemes predicts the gender of names. Journal of Personality and Social Psychology 110. 509–527. Search in Google Scholar

Suire, Alexandre, Alba Bossoms Mesa, Michel Raymond & Melissa Barkat-Defradas. 2019. Sex-biased sound symbolism in French first names. Evolutionary Human Sciences 1. 1–17. Search in Google Scholar

Vennemann, Theo. 1982. Zur Silbenstruktur der deutschen Standardsprache. In Theo Vennemann (ed.), Silben, Segmente, Akzente (Linguistische Arbeiten 126), 261–305. Tübingen: Niemeyer. Search in Google Scholar

van de Weijer, Jeroen, Ryan G.-Y. Ren, Joost van de Weijer, Weiyun Wie & Yumeng Wang. 2020. Gender identification in Chinese names. Lingua 234. 1−18. Search in Google Scholar

Whissell, Cynthia. 2001. Sound and emotion in given names. Names 49. 97–120. Search in Google Scholar

Wright, Saundra K., Jennifer Hay & Tessa Bent. 2005. Ladies first? Phonology, frequency, and the naming conspiracy. Linguistics 43. 531−561. Search in Google Scholar

Yorkston, Eric & Geeta Menon. 2004. A sound idea: Phonetic effects of brand names on consumer judgments. Journal of Consumer Research 31. 43–51. Search in Google Scholar

Received: 2020-02-19
Accepted: 2020-10-03
Published Online: 2021-06-30
Published in Print: 2021-07-27

© 2021 Tanja Ackermann and Christian Zimmer, published by De Gruyter, Berlin/Boston

This work is licensed under the Creative Commons Attribution 4.0 International License.