國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
lexical meaning diachronically due to the process of grammaticalization so as to have
acquired many semantic/pragmatic functions, namely they are polysemous. So it
seems plausible that we can acquire the accurate modal interpretation by examining
the insight of polysemous modal verbs. However, when we encounter multiple-modal
construction (discussed in a later section), which can allow two or three modal verbs
to co-occur in one sentence in Mandarin Chinese, we cannot solely depend upon the
particular modal verb used but take the aspects beyond word level into consideration.
Thus, in order to account for the fact that modal meaning is readily recognized as such
we should consider other factors that may control the selection of appropriate modal
interpretation, such as the syntactic environment or the semantic contribution of the
different parts of the utterance.
1.1 Motivation and Purpose
In recent decades the modal construction also has raised an extensive attention.
Much of the literature (Coates, 1983; Hofmann, 1993; Bybee et al., 1994; Heine, 1995)
takes semantic or syntactic approaches to analyze the modal construction and
observes that there are many similarities cross-linguistically. In terms of Chinese,
various proposals have been made to account for how various modal interpretations
are realized in syntactic structure (Chao, 1968; Li and Thompson, 1981; Lin and Tang,
1995; Tang and Tang, 1997; Huang, 1999; Huang, 2009). Among the several syntactic
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
properties of Chinese modals, one is worth to be addressed: the multiple-modal
construction. Unlike English, Mandarin Chinese allows the possibility of
co-occurrence of modal verbs, as in (1):
(1) a. *He can will come.
b. 他 應該 會 答應跟我們去旅行
ta yinggai hui daying gen women qu luxing he should would promise with us go travel
‘It is supposed that he would be likely to agree to go traveling with us.’
However, if we examine the multiple-modal construction closely, we can see some
ordering restrictions are at work, as in (2):
(2) *他 會 應該 去上學
ta hui yinggai qu shangxue he will should go school
In addition, there exist some polysemous modal verbs in Chinese modal system, such
as hui 會, neng 能, yao 要, yinggai 應該, keyi 可以. Because these polysemous modal
verbs seem to encode multiple-layer of modality interpretation, it is plausible that
there exist rich multiple-modal constructions in Mandarin Chinese. However, the data
of Chinese modal adjacency shown in Huang’s (1999) study are very restricted and
these polysemous modal verbs cannot be stacked in one sentence randomly.
In early time, Chao (1968:609-610) argues that modal verbs can co-occur in one
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
sentence. Such as hui yao會要, hui neng會能, hui ken會肯, keyi hui可以會, gai yao 該要, yinggai keyi應該可以, but he didn’t explain the ordering restriction of modal
co-occurrence. Her (2008) assumes Chinese modal system has two dimensions:
speaker-oriented and subject-oriented. He observes that speaker-oriented modals
always precede subject-oriented modals. See the following examples:
(3) *他 能 可能 唱歌
ta neng keneng changge he can possible sing ‘It is possible that he can sing.’
(4) *他 願意 必須 唱歌
ta yuanyi bixu changge he willing to must sing
‘It is obligatory that he is willing to sing.’
We can see the dynamic modals neng ‘can’ and yuanyi ‘willing to’ cannot occur
before either epistemic modal keneng ‘possible’ or deontic modal bixu ‘must’. He also
further divided speaker-oriented into possibility and obligation while subject-oriented
into volition and capacity. Tsao (1993) also generates some possible patterns of
Chinese modal sequence:
(5) 1 2 3A/B epistemic(yinggai, keneng) + hui + deontic/dynamic
(a) keneng+hui(1+2 or 1+3B)
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
(b) yinggai+keyi(1+3A or 1+3B) (c) hui+yao(2+3B)
(d) yinggai+hui+ken(1+2+3B)
Following Tsao’s (1993) generalization, Zheng (2001) adopts threefold classification
of Chinese modal system and also has similar findings on Chinese modal sequence
based on corpus-based approach. She observes the real usage of Chinese modal
sequence can be generated into three basic patterns: epistemic+deontic,
epistemic+dynamic and deontic+dynamic. Guo (1994) also claims that the order of
multiple-modal construction in Chinese is not arbitrary. The order should be
epistemic>deontic>dynamic and Huang (1999) further argues the multiple-modal
construction logically has four variants of multiple-modal expressions as follows:
I. epistemic > deontic II. deontic > dynamic III. epistemic > dynamic
IV. epistemic > deontic > dynamic
In sum, these studies assure that there exists restriction of multiple-modal
co-occurrence but what mechanism governs this ordering constraint is still needs to be
discovered. Also currently we are not sure whether any modal verbs can randomly
co-occur in one sentence.
Therefore, what interests us is how one specific modal interpretation of
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
polysemous modal verb is selected when filling in this construction. Though these
previous studies have reached agreement that epistemic modal verbs and
non-epistemic modal verbs are merged at different level, in which can provide a
fundamental explication to multiple-modal construction in Mandarin Chinese, there is
still very little consensus among them as to the best way to formulate a general
representation to predict such modal adjacency restriction. As a result, in the present
study we propose that such association is not arbitrary but may be restricted by some
mechanisms, e.g. the interaction between linguistic semantics of each polysemous
modal verb and sentence pragmatics in this construction. So the goal of this thesis
aims to explore this correspondence between modal structure and modal interpretation
in Mandarin Chinese so that we can find out a better solution to account for every
possible distribution of the multiple-construction of Chinese polysemous modal verbs.