Map 117 shows the strategies in contact languages that specify whether animals are female or male. Only morphologically productive, non-lexicalized patterns like lion-lioness are considered here. Lexicalized pairs such as English bull-cow are disregarded, as are unproductive patterns like fox-vixen, even if they are etymologically related. Thus, Guadeloupean Creole poul ‘hen’ and kòk ‘cock’ (Colot & Ludwig 2013a) do not count as words with sex-denoting elements for the present purposes.
The question whether the base word, apart from the generic meaning, can also denote only the female or only the male animal (like English tiger meaning ‘male and female tiger’ or only ’male tiger’) is irrelevant for this feature. Thus, Seychelles Creole—where bourik ‘donkey’ “refers to both sexes, but one can distinguish the two sexes by adding femel and mal” (Michaelis & Rosalie 2013)—is classified as a language with sex-denoting elements. For this reason, only when a language does not have a single sex-denoting element is it classified as not sex-denoting (value 5 “No sex-denoting element”).
Also irrelevant for this feature is the token frequency of the element, that is, sex-denoting elements for animals that are rarely referred to count the same as those for animals that speakers talk about more frequently. As to the type frequency, as long as there is at least one word among all the words for animals that can receive a sex-denoting element, this suffices for a value to be “true”, the actual frequency of the particular pattern in comparison to other patterns attested in the same language is indicated in the relative importance field in the APiCS database.
We distinguish between sex-denoting words (values 1 and 2) and affixes (values 3 and 4), which can either be preposed (values 1 and 3) or postposed (values 2 and 4). Sex-denoting words can include 3sg pronouns (‘he’, ‘she’), nouns (e.g. ‘man’, ‘woman’) or adjectives (e.g. ‘male’, ‘female’).
|Preposed sex-denoting word||31||7||38|
|Postposed sex-denoting word||23||7||30|
Value 1. A preposed sex-denoting word is the most frequent value in our sample, 31 languages (44%) relying exclusively on this strategy, and another 7 (10%) along with other strategies. Many APiCS languages have a preposed word for both male and female animals:
Contact languages can have different preposed elements to denote the same sex. These can be nouns that are inherently male or female, as shown in (5), or items from different word classes, e.g. a pronoun and a noun, as in (6), or a pronoun and an adjective, as in (7):
The elements can also derive from different lexifiers, as seen in Chinuk Wawa, where the male-denoting word derives from English man and the female-denoting word from Nootka/Nuuchahnulth ƛoːcsma ‘woman’. The latter was remodelled in Chinuk Wawa by analogy with English wo-man (Grant 2013):
Value 2. 23 APiCS languages (32%) exclusively rely on postposed sex-denoting words, and another 7 (10%) have this along with other values. Again, many languages have words for both female and male animals:
Interestingly, in Sango, which in many respects patterns like Kikongo-Kituba and Lingala, “Ti ‘of’ is never used in [such] constructions” (Samarin 2013). Rather, Sango uses preposed sex-denoting words (value 1).
Value 3. Sex-denoting prefixes are found in only one language, Sranan, where it is the exclusive strategy:
Value 4. Only three languages (4%) have sex-denoting suffixes and another one (Afrikaans) shares this with other strategies:
For the majority of European-lexified languages in APiCS, the syntax of the lexifier seems to have provided the model for the order of the animal term and its sex-denoting element. 20 (or 21, including Sranan) of the 27 English-lexified contact languages have preposed sex-denoting words, which are either adjectives or nouns. These languages follow the English adj n or right-headed compound order. The Portuguese-lexified languages also appear to have adopted the Portuguese model: 12 of 13 have postposed sex-denoting words, following the predominant postnominal position of adjectives and left-headedness of compounds in Portuguese. Interestingly, although French compounds are also predominantly left-headed and most adjectives follow the noun, 8 of 10 French-lexified APiCS languages have preposed sex-denoting words. The explanation could be that this reflects the more common prenominal adjective position in earlier French (Boucher 2004: 53-54). Three of the five contact languages with African lexifiers have no sex-denoting strategy (Kikongo-Kituba, Lingala, Mixed Ma’a/Mbugu). Fanakalo is rather like these, as its suffix is rarely used, and it has no other strategy. Only Sango has a common preposed sex-denoting word.