This is the third chapter on experiencer constructions. After ‘headache’ and ‘like’ in Chapters 66-67, we now consider ‘fear’, i.e. clauses like English The child is afraid. Such ‘fear’ constructions may involve three semantic entities: the experiencer (‘the child’), the emotion (‘fear’), and potentially a body-part (‘my heart/my soul’) that is thought of as the place where the emotion resides.
As was true for the previous two chapters, there is a lot of interesting variation in such ‘fear’ constructions, too. However, here we only concentrate on the question which semantic entity is expressed as subject, and if the subject is the experiencer, whether the concept of fear is expressed verbally or non-verbally. Subject is defined here as an argument that is coded like the typical transitive agent or the single argument in a monotransitive clause.
In addition to experiencer and emotion, ‘fear’ constructions may of course include a stimulus, the entity that the experiencer is afraid of, but this is typically optional and we do not consider its expression. Note that in some languages the expression of the stimulus may entail different constructions, e.g. English I am afraid (of the storm) vs. I fear the storm. If this is the case for a given language, we only consider constructions in which the stimulus is not overtly expressed, e.g. English The child is afraid, but not sentences like The child fears the storm because such constructions may be different from the ones we are studying in this feature.
A language may have several different ‘fear’ constructions.
Thus, in this feature we distinguish seven values:
|Experiencer is subject, 'fear' is verbal||35||20||55|
|Experiencer is subject, 'fear' is non-verbal||11||17||28|
|'Fear' is subject, experiencer is object||0||10||10|
|Experiencer is dative||1||2||3|
|'Fear' is subject, experiencer is oblique||0||1||1|
|Body-part is subject||0||6||6|
In constructions of value 1, the experiencer is treated like the subject of a typical transitive verb like ‘break’ or ‘kill’, and the semantics of fear is expressed verbally, as in examples (1)-(5):
(1) Ambon Malay (Paauw 2013)
‘I’m afraid.’See example 19-129
In the European-based creoles, reflexes of adjectival or nominal source lexemes meaning ‘afraid, fear (n.)’ often make up for the ‘fear’ verb in the APiCS languages, e.g., Spanish/Portuguese miedo/medo, French peur, Dutch bang and English afraid.
(2) Ternate Chabacano (Sippola 2013b)
Ta myédu mótro.
IPFV fear 1PL
‘We were afraid.’See example 44-123
(3) Martinican Creole (Colot & Ludwig 2013b)
‘I am afraid.’
(5) Gullah (Klein 2013)
I don fraid.
I NEG fear
‘I am not afraid.’
In Gullah, the negator don is used to negate verbs and not adjectives. Therefore fraid is verbal in contrast to its English source word (see Klein 2013).
In the second construction type (value 2), the experiencer is still treated like the subject, but this time the concept of fear is expressed in a non-verbal noun-like or adjective-like way. There are different verbal elements which link the subject to the noun-like element ‘fear’. The copula ‘be’ is only one type here, as can be seen from examples (6)-(9). Other linking verbs are ‘stay’, ‘have’, and ‘feel’ (see 7 and 8). What is important in this value is the fact that the lexeme carrying the semantics of ‘fear’ is not verbally encoded:
In the next construction type (value 3), ‘fear’ is treated as the subject, and the experiencer is coded like the object of a transitive verb. The relationship between the emotion and the experiencer is expressed by a general affect verb (‘do’). Examples come from Sango and Nigerian Pidgin:
(9) Sango (Samarin 2013)
mbito a-sara ala
fear PM-do 3PL
‘They’re afraid.’ (Lit. ‘Fear does them.’)
Constructions of value 4 again treat ‘fear’ as subject, but the experiencer is coded as dative, i.e. like the recipient of a typical ditransitive construction. The three APiCS languages which show this value are Chinuk Wawa, Korlai and Diu Indo-Portuguese:
(10) Chinuk Wawa (Grant 2013)
kwas kápa náyka
fear to 1SG
‘I’m scared.’ (Lit. ‘Fear is to me.’)
In example (10) from Chinuk Wawa, the dative experiencer follows the ‘fear’ subject argument, but in the two cases from South Asian languages, Diu Indo-Portuguese and Korlai, the dative experiencers are in preverbal position and precede the ‘fear’ argument, which is sentence-final (see 11). We already saw this construction type in the two other chapters on experiencer constructions with ‘headache’ and ‘like’ (Chapters 66-67).
The fifth construction type (value 5) has the experiencer treated as an oblique with a ‘fear’ subject. The only APiCS language showing this value is Cape Verdean Creole of São Vicente:
(12) Cape Verdean Creole of São Vicente (Swolkien 2013)
Med dá na es.
fear give on 3PL
‘They became overcome by fear.’ (Lit. ‘Fear is giving on
them.’)See example 32-111
The last option to express fear is value 6: A body-part (‘heart’) is treated as subject, and the experiencer is coded as a possessor:
(13) Nengee (Migge 2013)
Mi ati saka.
my heart drop
‘My heart dropped.’ OR: ‘I was afraid.’
It is interesting to note that for none of the six APiCS languages with this value this is the only way to express ‘fear’. This is in sharp contrast to Chapter 66 (‘headache’) where 31 APiCS languages have the body-part construction as their only means to express ‘to have a headache’.
By far the most widespread construction type in the APiCS languages is the one with the experiencer treated as subject, and ‘fear’ is treated either verbally (value 1, 55 languages) or non-verbally (value 2, 28 languages). Many APiCS languages have one of the two values exclusively, i.e. just show one construction type. For all remaining values (3)-(7), which only cover few languages, no APiCS language has one of these values as its only option (except for Chinuk Wawa, which only shows value 4).
As for potential substrate/adstrate influence, we can confidently trace back the "dative subjects" in Diu Indo-Portuguese and Korlai to their South Asian substrates/adstrates (see also the relevant discussion in Chapters 66 and 67).
Looking at the bulk of the other European-based APiCS languages, it is noteworthy to mention that they have verbal ‘fear’ constructions as their only option (e.g. Angolar, Guinea-Bissau Kriyol, Ghanaian Pidgin English, Tayo). It is true that the European base languages mostly also have the verbal construction type, but certainly this is not the most prominent option and, moreover, the creole ‘fear’ lexemes do not go back to these verbal constructions (see also §2). For example, the Seychelles Creole verb pe ‘be.afraid’ goes back to French non-verbal j’ai peur [I=have fear] and not to verbal je crains [I fear], and the Bislama verb frait goes back to English non-verbal I am afraid, and not to verbal I fear. Moreover, Spanish and Portuguese only show non-verbal constructions with a nominal ‘fear’ lexeme: Portuguese Tenho medo [I.have fear]. The base languages thus do not offer a model for the verbal constructions in Spanish- and Portuguese-based creole languages. Therefore, it seems appealing to look for an explanation for the creole patterns in the relevant substrate/adstrate languages. But, unfortunately, there is too little data on experiencer constructions in West African languages, for example (Ameka 1990 on Ewe is the only work we know of.