Ascription of Surprise in Cantonese (EACL9 500words)

Language of presentation: English
Surprise Ascription of Cantonese Mirative Adverbs
Lawrence Y.-L. Cheung
Margaret N.-Y. Lam
Department of Linguistics and Modern Languages
Chinese University of Hong Kong
Email: [email protected], [email protected]
Keywords: mirativity, surprise experiencer, entailment, conversational implicature
Puzzle: Research on mirativity (DeLancy 1997, Rett and Murray 2013) typically assumes
that the surprise associated with a mirative expression is ascribed to the speaker. This study
investigates the surprise ascription of Cantonese mirative adverbs jyunloi 原來 and gingjin 竟
然. While the surprise ascription of the mirative adverb in simple sentences is always
speaker-oriented (1), we have discovered that the ascription of the mirative adverb in
complement clauses could be one of the following:
A. Higher-Subject-Oriented Ascription (HSOA) (2)
B. SPeaker-Oriented Ascription (SPOA) (3)
C. Failure of ascription (4)
What determines the choice of the surprise experiencer (i.e. A—C) of the mirative adverbs?
Analysis: Successful ascription to experiencer X must fulfill two requirements.
A. Belief Entailment: The experiencer X must be lexically specified to, semantically entailed
to or pragmatically implicated to believe p, where p is the clause the mirative adverb occurs.
HSOA is found only when the higher predicate (e.g. hou caaji ‘be surprised’) lexically
specifies that the higher subject believes p. In (2), it is John who is surprised that “the TV is
broken” (=p). In contrast, SPOA can be triggered in two ways. In (3), the higher predicate
denies that the higher subject believes p. However, (3) semantically entails that the speaker
believes p because the higher predicates are factive. The belief entailment is critical to SPOA.
If one replaces the higher predicate with a non-factive like m soengseon ‘not believe’, SPOA
becomes unavailable (4) because nobody is entailed to believe p. The contrast between (3)
and (4) is previously unexpected but receives an account. The second way of triggering
SPOA is to pragmatically implicate (via Maxim of Quality) that the speaker believes p (1),
explaining obligatory SPOA when the adverb in is the root clause.
B. Belief Revision: X must have undergone revision from previous belief ~p to current belief
p. Although both zidou ‘know’ (6a) and m zidou ‘not know’ (6b) entail p, zidou implies a
stative belief without revision, leading to ungrammaticality in (6a).
[S-Exp = surprise experiencer]
(1.) Bou dinsi {jyunloi / gingjin} waai-zo.
CL TV JYUNLOI GINGJIN be.broken-PERF
‘The TV is broken. [S-Exp: SPEAKER]’
1 (2.) John {hou caaji
/ faatgok} [ bou dinsi {jyunloi / gingjin} waai-zo ].
very surprised realized
CL TV JYUNLOI GINGJIN be.broken-PERF
‘John {was very surprised / realized} that the TV is broken. [S-Exp: John]’
(3.) John {m caaji
/ m zidou} [ bou dinsi {jyunloi / gingjin} waai-zo ].
not surprised not know CL TV JYUNLOI GINGJIN be.broken-PERF
‘John did not {realize / know} that the TV is broken. [S-Exp: SPEAKER]’
(4.) *John {m soengseon / m jingwai} [ bou dinsi {jyunloi / gingjin} waai-zo ].
not believe
not think
‘John did not believe/not think that the TV is broken. [S-Exp: ∅]’
(5.) John {a. *zidou / b. Pm zidou} [ bou dinsi {jyunloi / gingjin} waai-zo ].
know
not know
a. * ‘John knew that the TV is broken.’
b. P ‘John did not know that the TV is broken. [S-Exp: SPEAKER]’
(500 words)
References:
DeLancey, Scott. 1997. Mirativity: The grammatical marking of unexpected
information. Linguistic Typology 1: 33-52.
Rett, Jessica and Sarah Murray. 2013. A semantic account of mirative evidentials.
Proceedings of SALT 23, 453-472.
2