Chapter 2 Literature Review
2.2 Requests as a Speech Act
2.2.3 Directness of request forms
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
2.2.3 Directness of request forms
As discussed previously (Section 2.1), Austin (1975) proposed that a
performative utterance convey three acts simultaneously. These three acts altogether comprise a speech act. They are locutionary act, illocutionary act, and perlocutionary act. Focusing on Austin’s illocutionary act, Searle (1975; 1976) classified speech acts into five basic categories according to the particular illocutionary act a loctionary act, or the utterance, conveys. In his classification, Searle considered the illocutionary acts that the speaker gets the hearer to do something as directives and requesting as a paradigmatic case of directives (See also Levinson, 1983). However, studies on child speech acts revealed that young children might not yet be able to differentiate requests from directives, as suggested by Garvey (1975). In light of this, requests in this study were thus used as a cover term to refer to the illocutionary act with which children’s intention is fulfilled by asking their interlocutor to perform the intended act. On the basis of Austin’s three aspects of a speech act, in this study an utterance was identified as one with the request illocutionary force by both its illocutionary and perlocutionary acts. Therefore, an utterance was considered as a request by the
addressee’s reaction to the utterance, in addition to the intended intents encoded in the utterance.
In addition to the general classification of illocutionary forces, a speech act can be further categorized in terms of directness of the illocutionary act(s) that an utterance conveys (Searle, 1975). Searle proposed that indirect speech acts, distinctive from a direct one, are ‘cases in which one illocutionary act is performed indirectly by way of performing another (in Davis, 1991: p. 266).’ The execution of an indirect speech act,
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
according to Searle, could be relied on the exploitation of the felicity conditions under which the particular speech act was felicitous. For example, to convey an indirect request, one can convey the intended illocutionary act via questioning or stating the addressee’s ability to perform the action as in Can you pass the salt (by questioning the preparatory condition). In addition, Searle also argued that some of the indirect requests could be conventionally encoded with particular linguistic forms, such as Can you…, I want…, It is possible that you would/could…(Searle, 1975). It seems that Searle’s proposal of the distinction between direct and indirect speech acts is
straightforward and clear-cut, but there may actually leave a lot to be desired.
Studies on children’s production and comprehension of direct and indirect speech acts have revealed the disadvantages of Searle’s proposal (e.g. Axia, 1996;
Garton & Pratt, 1990; Gordon & Ervin-Tripp, 1984). The findings in these studies have pointed out that at least two aspects of the distinction between direct and indirect speech acts proposed by Searle may not suffice. Gordon and Ervin-Tripp (1984) pointed out that Searle’s distinction could at best pertain to ‘formal’ differentiation between direct and indirect speech acts. What seemed formally indirect might be functionally direct in child discourse. For example, expressing one’s desire or needs to request the other party of interaction to perform an act may be formally indirect, but functionally direct. Gordon and Ervin-Tripp found that children tend to use I want… construction as a direct means to direct their mother’s action, and while requesting with this construction, they tend not to provide any justifying reasons.
However, this I want… construction, according to Searle, was actually considered as an indirect means to convey the intended directive or request, since the request was
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
conveyed via talking about the sincerity condition of a request. In addition, the formally indirect requests, such as The macaroni’s boiling over (as a directive to ask the hearer to do something about the macaroni), were found to be easily exploited by children as young as four to direct their mother’s action. Indirect as the case The macaroni’s boiling, children were found to have no difficulty in interpreting such utterance It’s noon as an indirect speech act (Gordon & Ervin-Tripp, 1984). These two cases, by Searle’s distinctive criterion, were apparent indirect directives, but direct ones for children, since they were cognitively and interactionally direct (Gordon & Ervin-Tripp, 1984).
In addition, Babelot and Marcos (1999) pointed out that children rely heavily on contextual clues to determine a particular speech act being direct or indirect. They suggested that for children there seemed no fixed association between a particular linguistic form and its illocutionary force. Therefore, this may imply that children have yet to develop the default association between a linguistic form and the illocutionary force. They develop the association through the process of pragmatic development.
Indirect speech acts can also be further classified. Searle (1975) mentioned that some indirect speech acts were conventional while some are not. For example, the request Can you pass the salt was conventional while The salt is close to you was not.
In addition, Morgan (1978) contended that conventionality of indirect speech acts were composed of a construction carrying a conventionalized meaning encoded with a conventionalized forms, which in turn was used in a conventionalized way with a conventionalized purpose. For example, when the utterance Can you pass me the salt
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
is used (conventionally) in the restaurant to the addressee with whom one is dining together, it thus (conventionally) means a request for the addressee to pass the salt over. Nevertheless, Searle and Morgan did not provide any criteria for the judgment of the conventionality of a particular indirect speech act. The distinction of
conventionalized and non-conventionalized indirect speech acts thus seems arbitrary, though factual.
With the findings in their study, Gordon and Ervin-Tripp (1984) implied a potential and reasonable account for how indirect speech acts become
conventionalized. They found that in routine context children gradually form a
cognitively or logically direct link between the formal indirect utterance, a cause, with the intended action, an effect, and this causal link (between the cause and the effect) enables children to produce and comprehend indirect directives effortlessly. This finding can thus reflect that when an utterance is regularly associated with a particular illocutionary act (as in routines), such a particular linguistic form thus becomes routinely associated with the intended action in a particular context, and an indirect speech act in turn becomes conventionalized. Hence, the conventionalization of a non-conventionalized speech act is not only theoretically argued, but also empirically testified. Also, Gordon and Ervin-Tripp’s finding lends support to Strawson’s (1974) argument that the conventionality of a speech act is fluid; a non-conventionalized speech act may become conventionalized because of frequent or regular use or mutual knowledge shared by interlocutors. Therefore, Gordon and Ervin-Tripp argued that while direct and indirect directives can be ‘formally’ differentiated, the
conventionality of them can be contextually sensitive. In other words, an indirect
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
directive can be conventional in a particular context, but the same directive can be non-conventional in other contexts.
In light of the insufficiency in Searle’s proposal of the distinction between direct and indirect speech acts, it may not be a good idea to directly apply this formal distinction to children’s use of speech acts. As discussed above, findings in previous studies (e.g., Ervin-Tripp, 1977; Garton & Pratt, 1990; Garvey, 1975; Gordon and Ervin-Tripp, 1984) reflected that the directness or indirectness of a speech act, particularly directive, is not absolute; it is relatively determined by both its formal manifestation and its logical inference. On the basis of empirical evidence obtained in child discourse and theoretical arguments, a more refined distinction between direct and indirect speech acts should be as seen in the following figure (Figure 2).5
direct indirect
imperative Formally
simple with mitigators non-imperative
Logically intention explicitly specified in locutionary acts intention inferred by the addressee Conventionalized Non-conventionalized
Figure 2. Directness of requests
As seen in the above figure, the scale on the top shows that the distinction of direct and indirect directives is not a clear or distinct dichotomy, but instead a fluid continuum. In this continuum, a directive is either more direct or more indirect, judged by the following two axes: formal representation and logical inference. When a directive is uttered with an imperative form, it is comparatively more direct and
5 Please note that the separating lines are only for clear representation. The boundaries are fluid and vague, given that each axis is a continuum.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
forceful. On the other hand, when it is uttered with a non-imperative form, such as interrogative and declarative, it is comparatively more indirect and less forceful.
Directives encoded with imperatives can be further divided into two subtypes. One type includes directives encoded with simple imperatives, and the other one includes those imperatives with mitigators, such as sentence-final particles and tag questions.6 Such imperatives usually encode the intended act explicitly in their locutionary acts and hence they are also logically direct.
As to those directives expressed with non-imperative forms, they can be further considered logically. Some of the non-imperatives may explicitly indicate the intended action (e.g., Could you pass the salt?), while others may provide no information with regard to the intended action (e.g., It’s cold in here.). The former requires the addressee’s little effort (to obtain the intended act) in inferential processing, since the intended act is explicitly coded in the linguistic form. On the other hand, the latter requires the addressee to undergo certain inferential processes so as to procure the intended action, since the proposition conveyed by the linguistic form does not specify the intended act (cf. Gordon & Ervin-Tripp, 1984). Hence, directives expressed with non-imperative forms are considered even more indirect than non-imperatives with indication of the intended speech acts, since in the former no information about the intended action is provided.
Lastly, the scale at the bottom of the figure shows that all the requests can be considered as either conventionalized or non-conventionalized ones. The distinction again is a continuum, ranging between the point of conventionalized and
6 Whether a sentence-final particle is a mitigator depends on the context. Some particles may serve the function of aggravation in some contexts (Hsu, 1996).
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
non-conventionalized requests; there is no absolute value as to whether a request is surely conventionalized or not. Basically, the figure shows that when a request is issued with a particular linguistic form that explicitly expresses the intended act, such request is usually a more conventionalized one. In addition, some indirect requests are relatively more conventionalized then others because they require the least effort of logical inference or because they are frequently used in particular situations. Other indirect requests are comparatively more non-conventionalized because these requests may be with neither linguistic nor logical explicitness of the intended act, and thus require the address to make much more effort to infer the intended act or they are simply newly innovated request forms that are infrequently used in a particular situation.
The distinction provided here has at least two advantages. First, the distinction can account for the development of non-conventionalized indirect directives into conventionalized ones. As mentioned above, the distinction is a continuum and it is subject to other interactional factors. The conventionality is determined not only by conventional linguistic forms or logical inference, but also by interlocutors’ mutual knowledge. When a non-conventionalized indirect directive is used routinely, i.e., regularly in a particular context with a particular form, it can thus become
conventionalized (Strawson, 1974; also in Davis, 1991). In other words, the degree of an indirect speech act’s conventionality is thus defined by the regular occurrence of a particular form conveying a particular illocutionary act in a particular context, and such a regular occurrence gradually becomes a part of interlocutors’ mutual
knowledge and further an element of their pragmatic competence. With this definition,
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the conventionality of indirect speech acts may no longer be arbitrary, but mutually defined by interlocutors with respect to the interactional context.
In addition, this distinction allows the addition of novel indirect speech acts. It is possible for a speaker to create a brand new form to convey his/her intended act, provided that the addressee can correctly infer the intended action with appropriate amount of inferential effort. When the speaker regularly uses this novel form in a particular context with a particular interlocutor, this form may become a shared indirect speech act, and in turn may develop into a conventionalized speech act. If, on the other hand, the addressee can never infer the intended action, no matter how many times the speaker uses this form and how much effort the addressee makes to infer, the novel form will not be used as an indirect speech act at all.
In summary, the distinction between direct and indirect speech acts provided here incorporates both linguistic and cognitive concepts. The distinction between direct and indirect speech acts is considered fluid as a continuum. The distinction between direct and indirect speech acts is not dichotomy. In addition, the
conventionality of an indirect speech act is also interactionally or collaboratively determined by interlocutors, but not predetermined by linguistic forms or propositions of utterances.
The above discussion only pertains to the directness of a speech act. It does not concern politeness nor does it consider the correspondence between politeness and directness of a speech act. As reviewed in the previous section (Section 2.1), Brown and Levinson’s (1987) politeness model considered politeness in terms of whether an illocutionary force is an FTA (face-threatening act), whether redress is directed
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
towards the interlocutors’ positive or negative face. Leech (1983), on the other hand, proposed six maxims on the basis of his Politeness Principle. In Leech’s model, each maxim is associated with a particular type of illocutionary act. These two models on politeness concurred that indirectness is a way to maintain politeness. The more politeness is needed, the more indirectness is required, and vice versa. However, the correspondence is not as straightforward as these two models have suggested.
Ideally, there should be a consistent correspondence between politeness and indirectness, as suggested by Brown and Levinson (1987) and Leech (1983). Some empirical findings indicated that the correspondence between politeness and indirectness is not fixed and not cross-culturally applicable. The correspondence is susceptible to social and cultural factors. Ervin-Tripp et al. (1990) found that in mother-child conversations, especially in the conversations between young children and their mothers, children’s indirectness in their control acts seemed not to ensure their mothers’ compliance. Nor did the indirectness considered polite by their mothers.
In other words, the mothers seemed not to expect politeness from their children in those conversational contexts. In addition, Leech (1983) also mentioned that
indirectness should be exploited with respect to the relationship between interlocutors and the conversational situations. When indirectness is used when it is least expected, the interlocutor who seems to be polite may be considered to be flattering or ironic.
In addition, interlocutors should also take contexts into account to consider how indirectness is needed so as to conform to politeness. For example, in a task-based discourse in which both participants in the task are designated a task to accomplish with a time limit, no indirectness is expected, since the main and only aim of this
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
interaction is to finish the task at hand as required. Being indirect would instead make the task even more complicated and difficult to accomplish. In addition, such
discourse type usually imposes a highly cooperative relationship on the participants.
Being indirect would instead imply to the other participant that one is not being cooperative, which will in turn be judged as impolite (Kasper, 1990; Leech, 1983).
Moreover, different cultures consider indirectness from different perspectives, although in most Western languages, indirectness is viewed as a strategy to avoid intrusion on the addressee. Clancy (1986) indicated that indirectness in Japanese culture expresses empathy between the participants in a conversation, instead of emphasizing distance. In addition, Blum-Kulka (1987) found that for Israeli people, indirectness is considered less polite and insincere. People performing indirectness are seen to be non-committing, lacking of sincerity, and imposing, since the addressee is placed a burden to infer the speaker’s intended meaning.
From the above discussion, it is clear that the range of politeness is
collaboratively determined by the interlocutors. Different social relationships between the interlocutors and different discourse types the interlocutors are involved in may change the range of politeness. Although indirectness may ideally and inherently convey a certain degree of politeness, the correspondence between indirectness and politeness is not at all straightforward and fixed. The correspondence is obviously susceptible to social and cultural factors. Therefore, for interlocutors to know how politely and indirectly they should interact with each other, they have to first conform to their cultural norms, examine the current discourse type, and then determine how much politeness they should adhere to so as to respect the social relationship between
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
them.