The acquisition of modality

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

6

(Palmer, 1990; Perkins, 1983): dynamic, deontic, and epistemic. Dynamic modality is concerned with the ability, volition, and willingness of the subject (Palmer, 1990).

Deontic modality has to do with the necessity of “acts performed by morally responsible agents” (Lyons, 1977, p. 823), which is involved with social functions of obligation and permission. Dynamic and deontic modals are classified as agent-oriented modalities which refer to “all modal meanings that predicate conditions on the agent with regard to the completion of an action referred to by the main predicate” (Bybee & Fleishman, 1995, p. 6). When dynamic and deontic modalities lose the speaker’s agency or control, they become epistemic modality and begin to get involved with the speaker’s beliefs (Gerhardt, 1991). Epistemic modality expresses the speaker’s degree of certainty about the truth of a proposition, and is thus associated with the speaker’s knowledge and beliefs (Lyon, 1977), which is speaker-oriented.

2.2 The acquisition of modality

A number of studies have examined the acquisition of modality. Studies on the acquisition of modalities have indicated a tendency that agent-oriented modality is acquired before epistemic modality (e.g. Guo, 1995; Papafragou, 1998; Shatz &

Wilcox, 1991; Stephany, 1986; Wells, 1985). Stephany (1993) pointed out that

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

7

children start out predominantly with non-epistemic acts (i.e. requestive or imperative acts) in social interaction and then begin epistemic acts (i.e. declarative acts). A number of studies have examined the development of modality in English. Shatz and Wilcox (1991) found that English modal verbs begin with agent-oriented expressions, such as expressions of ability and volition. Wells’ studies (1979, 1985) agreed with this general development pattern, that is, children begin expressing agent-oriented modality before epistemic modality. He found that children used the expression of ability/ inability (using can/can’t) at around the age of two, and then children started to slowly produce may and might to express possibility at around the age of three. He reported that inferential uses of modals (e.g. will) appeared even later in development.

However, O’Neill and Atance (2000) argued that children began to use epistemic modals earlier than what previous studies had claimed. They found that children used epistemic modal terms productively to express uncertainty from around two years of age. Guo (1995) argued that the semantic development of modal verbs started from the three year olds’ physical abilitative meaning, continued through to the five year olds’ deontic meaning, then to the seven year olds’ epistemic-like meaning. Besides, some studies have shown that the frequency data in the use of modalities is consistent among children (Bassano, 1996; Torr, 1998; Wells, 1979). They all drew a conclusion that the expressions of agent-oriented modality are consistently more frequent than

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

8

epistemic modality in the child’s speech.

There seems to be a specific relationship between specific type of modality and subjecthood. Studies have found that children have the tendency to produce agent-oriented modal verbs with the first person subject. For instance, Pea and Mawby (1981) found that children used agent-oriented modals with the first person to express volition. In the early period of development, O’Neill and Atance (2000) reported that children used the first person subject to express future intention. In contrast, children use epistemic terms (e.g. might) mainly with the third person subject to express events and states of third-party entities (Bassano, 1996; O’Neill &

Atance, 2000).

There also seems to be a relation between specific type of modality and sentence types. Pea and Mawby (1981) found that most of the modals were used in the affirmative (e.g. gonna). Negative modals were used only to express constraints on action or an unwillingness to act (e.g. can’t, won’t) at the speech time. It has been noted that modal terms in grammatically appropriate utterances appear during the period from when a child is two to three years old (Bliss, 1988). Children tend to be conservative in the use of modals in syntactic structures compared with semantic development of modalities. Shatz and Wilcox (1991, p. 331) pointed out that “while modal vocabulary growth proceeds fairly rapidly during this early period (e.g.

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

9

can/can’t, wanna, won’t), the range of syntactic constructions in which the modals

appear changes somewhat more slowly.” For instance, early modals (e.g. can, will) are all produced in limited syntactic environments (i.e. mainly in simple declaratives) (Shatz & Wilcox, 1991). O’Neill and Atance (2000) reported that the productive construction type with modal forms started at around 3;6 and children’s syntax became more complex (e.g. modals with complement clauses) at this age.

In the studies on the development of modality in Mandarin-speaking children, Hsu (2011) investigated the functions, grammatical patterns, and input of yao. The data were spontaneous speech of three children aged from 1;5 to 3;5. The result showed that the emerging order of the three functions of yao was: dynamic > deontic >

epistemic. It was found that children first acquired the function of expressing request or volition, the function of denoting obligation or necessity was the second acquired and the function of making inference of an event was the third. As for grammatical patterns used in expressing different functions of yao, the function of expressing request or volition was expressed with various types of grammatical patterns, whereas the other two functions were expressed in a limited range of grammatical patterns.

Children went through an “old form and new function mapping” process in expanding the use of yao from one function to another. In addition, the use of yao in the input provided scaffolding for children.

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

10

Guo (1994) claimed that the meanings of modals are rooted in the social, interactional functions of language. The use of modal expressions is part of acquiring competency in discourse interaction. He examined the sentence forms, semantic meanings, and discourse functions of Mandarin Chinese modal auxiliaries among children aged 3, 5, and 7. He investigated the different uses of six modal auxiliaries:

yao “want”, xiang “desire”, yong “need”, neng “can”, hui “know-how-to”, and dei

“hafta”. Forty specific discourse functions were analyzed and categorized into two major categories: interpersonal-oriented functions and informational-oriented functions. For example, the study of the modal neng showed the acquisition order of the different meanings from physical ability to permission to epistemic uses. All three meanings serve a common interpersonal function in discourse: a challenge to the addressee. The change in the meaning of neng is primarily motivated by this interpersonal function of challenge (Guo, 1995, p. 207).

He used the term “interpersonal-oriented function” in the sense of interpersonal function, as proposed by Lyons (1977), namely, how the speaker intends to influence

the mental state or behavior of the addressee. The utterance is mainly used to modify or regulate the actional behavior of the participants of the interaction, either the addressee, the speaker, or the other parties present. This can be the initiation or prevention of others’ action (e.g. request and prohibition), keeping oneself from being

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

11

affected by others (e.g. refusal), provision for others (e.g. offer) or justification of or comment on one’s own action (e.g. self action). At the “informational-oriented function” level, the main purpose of making the utterance is to impart information rather than modifying behavior, such as reporting. Guo (1994) proposed that interpersonal-oriented functions were developed earlier than informational-oriented functions, which was related to the cognitive development of children.