In chapter 1, a distinction was made between semantic gender and lexical gender. Lexical gender was assumed to be a salient meaning component associated with a noun, for instance ‘male’ or ‘female’ in some nouns, for example king and queen. Semantic gender was described as a category of thought, conveyed by pronominal resources. Consequently, a referent in the world of discourse is assigned a semantic gender by virtue of a speaker’s choice of pronoun. However, there is a close relation between, for example, the pronoun hon ‘she’ and the salient meaning component/lexical gender ‘female’. The question is how to describe this relation.
A noun is associated with a number of meaning components, which probably are best described as a structured system (see for instance Pustejovsky’s notion of qualia structure, Pustejovsky 1995). However, meaning components associated with nouns, for instance ‘male’, in association with the noun king could be thought of as weak meaning, in the sense that such components are usually negotiable and can be demoted to a higher or lower degree. For instance, a meaning component can be subject to metaphoric extension, which explains the meaning of a word such as kungs+vatten (king+water) ‘hydrochloric acid’ – this acid is the most powerful one since it can dissolve gold. A word such as kungs+sång (king+song) ‘anthem to the king’ could presumably be used also in cases where the regent is a queen. Similar examples are easily found. The way weak meanings can be demoted is especially clear in animal names, such as domherre (lit. judgement man) ‘bullfinch’, talgoxe (lit tallow oxen) ‘great tits’, and rapphöna (RAPP+hen) ‘partridge’. The right-hand segments of the compounds denote males or female animals when used in isolation, but the compound words are used for individuals of both sexes; there is thus no *domfru (intended reading “judgement lady”), *talgko (intended meaning “tallow cow”) or *rapptupp (intended meaning “RAPP+rooster”) (see also Tegnér 1925 [1964]:15f). The meaning of a pronoun such as hon ‘she’, on the other hand, does not have any elaborate qualia structure; it expresses only a few meaning components, but these are, on the other hand, not negotiable; pronouns have a strong
meaning. This is, in effect, why (4.33a) below (corresponding to example (3.29) in chapter 3)
(4.33) a Jag träffade kung-en. #Hon var glad. Sw I meet.PST king-C.DEF. she be.PST happy.
b Jag träffade honom. *Hon var glad. Sw I meet.PST him. She be.PST happy.
The idea that nouns such as kung ‘king’ and drottning ‘queen’ would belong to the genders feminine and masculine is derived from the fact that the meaning components ‘male’ and ‘female’ are very salient, hence that a referent that is referred to by the noun kung ‘king’ is also naturally also referred to by the pronoun han ‘he’. However, there are many problems associated with the idea that kung ‘king’ and drottning ‘queen’ are morphosyntactically specified as feminine and masculine. One is that very few nouns strictly denote men or women. For some nouns we would have to look out into the world and make frequency calculations before we would determine their gender. As pointed out above, is not reasonable to assume that smed ‘blacksmith’ would be a masculine noun because most (but not all) blacksmiths are men and barnmorska feminine because most midwives (but not all) are women. Sentences, such as Volvo är kung på vägen. Den flyger fram. (Volvo be.PRS king on road.C.DEF. 3C fly.PRS ahead.) ‘Volvo is the king of the road. It flies.’, would presumably have to contain a different kung ‘king’, as compared to Carl XIV Gustaf är kung ‘Carl XIV Gustaf is king’. The nouns kung2 ‘king2’ would have different genders, because of the
pronoun den in the second sentence. It would also mean that all nouns that could ever be used for humans, for instance as invectives: skåp ‘cupboard, hus ‘house’, rivjärn ‘grater’, brevlåda ‘mailbox’, lexikon ‘encyclopedia’, dator ‘computer’ …, would have a male and a female variant in the lexicon. It does not take much to realize that such a view does not provide any explanatory value. Hence, a more appropriate alternative is to view lexical gender as salient meaning components with a weak meaning, in the sense described above. It is often the case that a weak meaning component motivates the choice of a certain anaphoric pronoun, but this is not always the case. The set of meaning components associated with a noun may gradually change over time, due to linguistic and non-linguistic factors. The meaning of pronouns might of course change over time too, but much more slowly, and, from a synchronic perspective, we may safely conclude that their meaning is not negotiable.
As pointed out above, lexical gender is a salient meaning component that can also be present in the strong meaning of a pronoun. The term lexical gender usually refers to the meaning components ‘male’, ‘female’; and ‘inanimate’, but, if these meaning components correspond to genders, there is no reason why we should not extend the notion to other components of
meaning that could be salient too, such as ‘solid’, ‘round’, ‘black’ etc. The reason why such meaning components are not thought of as lexical gender features is probably that they do not correspond to features that are expressed by pronouns in Swedish and cognate languages. Having concluded that there is no SexP in the syntactic derivation of DPs, such as
vaktmästare-n (janitor-C.DEF) ‘the janitor’ or professor-n (professor-C.DEF) ‘the professor’, let
us take a look at a construction where such a functional projection is added. Josefsson (1999) discusses a construction that is quite common in spoken Swedish and Danish, in Josefsson’s work termed prenominal apposition, where information about the sex or natural gender is supplied, in a separate layer of the DP. Consider (4.34).34
(4.34) a han vaktmästare-n Sw
he janitor-C.DEF
‘he’/’the janitor’
b hon den nya professor-n Sw she DEF.C new professor-C.DEF
‘she’/’the new professor’
There are certain pragmatic functions associated with the construction in question, which are not relevant for the discussion here. What is important is that the proposed analysis suggests is that there can be a SexP merged on top of the DP in (4.34), giving rise to (4.35):
34
It should be pointed out that the sequence han vaktmästaren in (4.34) is pronounced without a prosodic break between the pronoun and the rest of the noun phrase. This makes it different from a similar construction, where a prosodic break (in writing usually indicated by commas) shows that the latter part is appositional: han, vaktmästar-en (he, janitor-C.DEF) ‘he,
(4.35) DP SexP D Sex NbP F SG DP D FP F AP Spec A’ A Complement hon den nya professorn nya professorn
What is a bit odd about (4.35) is that there are two DPs in the structure – ideally one should be enough. (The reason for assuming two DPs is that the pronoun hon ‘she’ and the DP den nya
professorn ‘the new professor’, seem to have independent definiteness, judging from the fact
that they can be used as DPs in isolation.) I have no explanation for this type of “triple definiteness”, but we may conclude that this does not seem necessary in Norwegian, where (4.36a) and (4.36b) are grammatical:35
(4.36) a ho nye professor-en N-No she new professor-C.DEF
‘she’/’the new professor’
b ho professor-en N-No
she professor-C.DEF
’she’/’the professor’
I have no explanation as to the difference between Swedish and Norwegian when it comes to the need for a determiner, such as den (DEF.C) in (4.34b), in Swedish.
35
Thanks to Marit Julien for bringing my attention to this, and for providing me with examples.