Transcript LOGIC FOR PRAGMATICS

On the Meaning of
the Novel Use of Bei-construction
Yan JIANG 蒋严
1. Chinese passives are non-canonical
passives in typology of world lgs
It does not have passive morphology for verbs
It is lexically marked by a set of passive words such as
bei被, gei给, rang让, jiao叫, with bei as the more standard
use and the others as the more colloquial/regional uses.
It is sometimes possible to have two passive markers cooccurring in one simplex sentence, i.e.
让他给骗了， 又被他给幽了一默， 叫他给拖累了
It typically has malficiary meaning
2
2. The Syntactic status of bei
1. Taken as a preposition, like the English “by”. More agreeable to
intuition, as bei’s meaning in modern Chinese is rather empty, with
no more than a weak suggestion of a patient thematic role of the
subject, although it can actually head an optional NP with an agent
role. But unlike the English “by”, bei can be dangling, without a
following NP.
2. Taken as a light verb: either as a raising verb or as a control verb.
This partially reflects the historical development of bei, which was
originally a verb [meaning ‘to cover’]. But the main motivations
come from formal syntax
3. Long and short passives as two distinct structures, not as complete
or elliptical structures.
4. The dual-bei analysis, with one bei taken as the light
verb, and the other as the preposition. The two bei
usually merge into one when realized as a3 sentence.
3. The novel use of bei sentences
 Starting from 2008, the novel use of bei has emerged as
a new grammatical construction.
 It only involves bei, but not the other passive markers
 It resembles the short passives in that bei does not head
an agent NP.
 It differs from the short passives in that bei is followed by
an intransitive verb, a predicate NP, or a predicate ADJP.
Such predicates are normally used in active voice
equivalents, with the subjects as agents/experiencers,
but there are some such NPs or ADJPs whose usage as
a predicate is novel in itself. That is, it is only
accommodated by the novel bei construction. [the
4
examples in red in the next page]
Some examples
 被自杀、被代表、被捐款、被失踪、被自愿、被失业、被
就业、被开心、被小康、被寂寞、被增长、被全勤、被幸
福、被剩男、被缓缴、被盖楼、被动车、被高铁、被房奴
、被出线
5
4. The meaning of new bei sentences
1. Obviously, it goes beyond passive meaning, but the malficiary
meaning is retained, or even strengthened.
2. Not just about the experience or the reception of an action on the
part of the patient/experiencer, but also involves an action/event/
state “being forced on” , and the “involuntarity”, “unawareness” and
“unwillingness” of the patient/experiencer, even though the verb
only accepts as its theta role an agent NP that can initiate an action
on its own. This is a conflict that can be perceived both
semantically and syntactically.
3. The subject, which should normally be the agent of the predicates,
now enters into a passive frame, where the subject position
markedly accommodates a non-agentive NP (as its grammatical
subject), which is also the only prominent NP as an argument in the
sentence.
6
4. Since the predicate following bei cannot take on another
NP as its logical subject, i.e., one that immediately follows
bei, it acquires a non-agentive nature and can only
describe a resulting state.
5. “bei xx” reports an event, but also partially negates the
veracity of the event, conveying the negation of and
suspicion on the event reported by xx, admitting only a
resulting state that seems to have been caused by xx.
6. The real agent/causer is not given or cannot be given
and can only be contextually accommodated.
7
5. Issues to be addressed
1. How to resolve the conflict as stated in point 2 of page
4, thereby giving justification to the validity of the novel
use of the “bei” construction?
2. How to give a unified explanation to the relationship
between the more normal use of bei constructions and
the novel use?
3. How to give an overall semantic representation of the
bei construction?
4. How to derive the pragmatic implicatures that can be
perceived in the novel use of bei construction that
reflect the intuition of the language users?
8
6. Existing syntactic treatments
1. Treating bei as a verb
2. bei heads either a clause or a VP.
3. bei heading a clause forms the long passive structure:
bei + NP + V…
4. bei heading a VP forms the short passive structure: bei
+ e + V, where the e is either an operator OP that
relates itself to the raised subject or is a PRO whose
reference is to be resolved by an overt or arbitrary NP
that controls it.
9
7. Possible semantic representations
All formulae adapted from Kate Kearns (2011) Semantics
I. Long passives
a. DO (x, [...])
Jones walked. => DO(j,[walk(j)])
Jones pushed Bill. => DO(j, [push (b)(j)])
Bill was pushed by Jones. Same as above, but contains a
derived formula: j CAUSE [BECOME [pushed (b)]]
II. Short passives
b.  x [DO (x, […])].   x [DO (x, [push (b)(x)])].
Bill was pushed.
Derived formula: x [x CAUSE [BECOME[Pred (b)]]]
10
The novel passive
x V active. Example: x自杀：DO(x, V(x))
x 被V-intransitive. Example: x 被自杀：
The light verb 被 initiates type coercing to produce the
following formula: x [x CAUSE [BECOME[Pred (b)]]]
Or y[x [x CAUSE [BECOME[Pred (y)]]] ]
Its difference from the normal short passives is that the
novel passive has the above formula directly projected as
its argument structure. Its causative formula is not a
derived one. Crucially, it does not have a non-causative
equivalent, unlike a standard short passive:
*  x [DO (x, […])].
* y[ x [DO (x, [自杀（y）(x)])]]. (this formula11contains a
type conflict or type mismatch)
Pragmatic implicatures
 Deriving the pragmatic implicatures from the
recoverability or non-recoverability of the missing x
argument as well as the process of type coercing.
 Conclusion: Chinese passives are indeed chameleonic,
both syntactically and semantically, but we should
endeavour to provide a unified mechanism.
12