The Nigerian professor Ozidi Bariki describes a conversation in which he said to a friend:
“I love your left hand.” (The friend had a cup of tea in his hand). The friend, in reaction to my utterance, transferred the cup to his right hand. That prompted me to say: “I love your right hand”. My friend smiled, recognized my desire for tea and told his sister, “My friend wants tea”… My friend’s utterance addressed to his sister in reaction to mine was a representative, i.e. a simple statement: “my friend wants a tea”. The girl rightly interpreted the context of the representative to mean a directive. In other words, her brother (my friend) was ordering her to prepare some tea. (Bariki 2008)
This brief dialogue contains two examples of indirect speech acts. In both cases, the utterance has the form of a simple statement, but is actually intended to perform a different kind of act: request in the first case and command in the second. The second statement, “My friend wants tea,” was immediately and automatically interpreted correctly by the addressee. (In African culture, when an older brother makes such a statement to his younger sister, there is only one possible interpretation.) The first statement, however, failed to communicate. Only after the second attempt was the addressee able to work out the intended meaning, not automatically at all, but as if he was trying to solve a riddle.
Bariki uses this example to illustrate the role that context plays in enabling the hearer to identify the intended speech act. But it also shows us that context alone is not enough. In the context of the first utterance, there was a natural association between what was said (your left hand) and what was intended (a cup of tea); the addressee was holding a cup of tea in his left hand. In spite of this, the addressee was unable to figure out what the speaker meant. The contrast between this failed attempt at communication and the immediately understood statement My friend wants tea, suggests that there are certain principles and conventions which need to be followed in order to make the illocutionary force of an utterance clear to the hearer.
We might define an indirect speech act (following Searle 1975) as an utterance in which one illocutionary act (the primary act) is intentionally performed by means of the performance of another act (the literal act). In other words, it is an utterance whose form does not reflect the intended illocutionary force. My friend wants tea is a simple declarative sentence, the form which is normally used for making statements. In the context above, however, it was correctly interpreted as a command. So the literal act was a statement, but the primary act was a command.
Most if not all languages have grammatical and/or phonological means of distinguishing at least three basic types of sentences: statements, questions, and commands. The default expectation is that declarative sentences will express statements, interrogative sentences will express questions, and imperative sentences will express commands. When these expectations are met, we have a direct speech act because the grammatical form matches the intended illocutionary force. Explicit performatives are also direct speech acts.
An indirect speech act will normally be expressed as a declarative, interrogative, or imperative sentence; so the literal act will normally be a statement, question, or command. One of the best-known types of indirect speech act is the Rhetorical Question, which involves an interrogative sentence but is not intended to be a genuine request for information.
Why is the statement I love your left hand not likely to work as an indirect request for tea? Searle (1969; 1975) proposes that in order for an indirect speech act to be successful, the literal act should normally be related to the Felicity Conditions of the intended or primary act in certain specific ways. Searle restated Austin’s Felicity Conditions under four headings: preparatory conditions(background circumstances and knowledge about the speaker, hearer, and/ or situation which must be true in order for the speech act to be felicitous); sincerity conditions (necessary psychological states of speaker and/or hearer); propositional content (the kind of situation or event described by the underlying proposition); essential condition (the essence of the speech act; what the act “counts as”). These four categories are illustrated in Table 10.1 using the speech acts of promising and requesting.
Generally speaking, speakers perform an indirect speech act by stating or asking about one of the Felicity Conditions (apart from the essential condition). The examples in (7) show some sentences that could be used as indirect requests for tea. Sentences (7a–b) ask about the preparatory condition for a request, namely the hearer’s ability to perform the action. Sentences (7c–d) state the sincerity condition for a request, namely that the speaker wants the hearer to perform the action. Sentences (7e–f) ask about the propositional content of the request, namely the future act by the hearer.
(7) a. Do you have any tea?
b. Could you possibly give me some tea?
c. I would like you to give me some tea.
d. I would really appreciate a cup of tea.
e. Will you give me some tea?
f. Are you going to give me some tea?
All of these sentences could be understood as requests for tea, if spoken in the right context, but they are clearly not all equivalent: (7b) is a more polite way of asking than (7a); (7d) is a polite request, whereas (7c) sounds more demanding; (7e) is a polite request, whereas (7f) sounds impatient and even rude.
Not every possible strategy is actually available for a given speech act. For example, asking about the sincerity condition for a request is generally quite unnatural: #Do I want you to give me some tea? This is because speakers do not normally ask other people about their own mental or emotional states. So that specific strategy cannot be used to form an indirect request.
We almost automatically interpret examples like (7b) and (7e) as requests. This tendency is so strong that it may be hard to recognize them as indirect speech acts. The crucial point is that their grammatical form is that of a question, not a request. However, some very close paraphrases of these sentences, such as those in (8), would probably not be understood as requests in most contexts.
(8) a. Do you currently have the ability to provide me with tea?
b. Do you anticipate giving me a cup of tea in the near future?
We can see the difference quite clearly if we try to add the word please to each sentence. As we noted in Chapter 1, please is a marker of politeness which is restricted to occurring only in requests; it does not occur naturally in other kinds of speech acts. It is possible, and in most cases fairly natural, to add please to any of the sentences in (7), even to those which do not sound very polite on their own. However, this is not possible for the sentences in (8). This difference provides good evidence for saying that the sentences in (8) are not naturally interpretable as indirect requests.
(9) a. Could you possibly give me some tea, please?
b. Will you give me some tea, please?
c. I would like you to give me some tea, please.
d. Are you going to give me some tea (?please)?
e. Do you currently have the ability to provide me with tea (#please)?
f. Do you anticipate giving me a cup of tea in the near future (#please)?
The contrast between the acceptability of (7b) and (7e) as requests vs. the unacceptability of their close paraphrases in (8) suggests that the form of the sentence, as well as its semantic content, helps to determine whether an indirect speech act will be successful or not. We will return to this issue below, but first we need to think about a more fundamental question: How does the hearer recognize an indirect speech act? In other words, how does he know that the primary (intended) illocutionary force of the utterance is not the same as the literal force suggested by the form of the sentence?
Searle suggests that the key to solving this problem comes from Grice’s Cooperative Principle. If someone asks the person sitting next to him at a dinner Can you pass me the salt?, we might expect the addressee to be puzzled. Only under the most unusual circumstances would this question be relevant to the current topic of conversation. Only under the most unusual circumstances would the answer to this question be informative, since few people who can sit up at a dinner table are physically unable to lift a salt shaker. In most contexts, the addressee could only believe the speaker to be obeying the Co-operative Principle if the question is not meant as a simple request for information, i.e., if the intended illocutionary force is something other than a question.
Having recognized this question as an indirect speech act, how does the addressee figure out what the intended illocutionary force is? Searle’s solution is essentially the Gricean method of calculating implicatures, enriched by an understanding of the Felicity Conditions for the intended speech act. Searle (1975) suggests that the addressee might reason as follows: “This question is not relevant to the current topic of conversation, and the speaker cannot be in doubt about my ability to pass the salt. I believe him to be cooperating in the conversation, so there must be another point to the question. I know that a preparatory condition for making a request is the belief that the addressee is able to perform the requested action. I know that people often use salt at dinner, sharing a common salt shaker which they pass back and forth as requested. Since he has mentioned a preparatory condition for requesting me to perform this action, I conclude that this request is what he means to communicate.”
So it is important that we understand indirect speech acts as a kind of conversational implicature. However, they are different in certain respects from the implicatures that Grice discussed. For example, Grice stated that implicatures are “non-detachable”, meaning that semantically equivalent sentences should trigger the same implicatures in the same context. However, as we noted above, this is not always true with indirect speech acts. In the current example, Searle points out that the question Are you able to pass me the salt?, although a close paraphrase of Can you pass me the salt?, is much less likely to be interpreted as a request (#Are you able to please pass me the salt?). How can we account for this?
Searle argues that, while the meaning of the indirect speech act is calculable or explainable in Gricean terms, the forms of indirect speech acts are partly conventionalized. Searle refers to these as “conventions of usage”, in contrast to normal idioms like kick the bucket (for ‘die’) which we might call conventions of meaning or sense.
Conventionalized speech acts are different from normal idioms in several important ways. First, the meanings of normal idioms are not calculable or predictable from their literal meanings. The phrase kick the bucket contains no words which have any component of meaning relating to death.
Second, when an indirect speech act is performed, both the literal and primary acts are understood to be part of what is meant. In Searle’s terms, the primary act is performed “by way of” performing the literal act. We can see this because, as illustrated in (10), the hearer could appropriately reply to the primary act alone (A1), the literal act alone (A2), or to both acts together (A3). Moreover, in reporting indirect speech acts, it is possible (and in fact quite common) to use matrix verbs which refer to the literal act rather than the primary act, as illustrated in (11–12).
(10) Q: Can you (please) tell me the time?
A1: It’s almost 5:30.
A2: No, I’m sorry, I can’t; my watch has stopped.
A3: Yes, it’s 5:30.
(11) a. Will you (please) pass me the salt?
b. He asked me whether I would pass him the salt.
(12) a. I want you to leave now (please).
b. He told me that he wanted me to leave.
In this way indirect speech acts are quite similar to other conversational implicatures, in that both the sentence meaning and the pragmatic inference are part of what is communicated. They are very different from normal idioms, which allow either the idiomatic meaning (the normal interpretation), or the literal meaning (under unusual circumstances), but never both together. The two senses of a normal idiom are antagonistic, as we can see by the fact that some people use them to form (admittedly bad) puns:
(13) Old milkmaids never die — they just kick the bucket.7
Birner (2012/2013: 196) points out that under Searle’s view, indirect speech acts are similar to generalized conversational implicatures. In both cases the implicature is part of the default interpretation of the utterance; it will arise unless it is blocked by specific features in the context, or is explicitly negated, etc. We have to work pretty hard to create a context in which the question Can you pass the salt? would not be interpreted as a request, but it can be done.8
Searle states that politeness is one of the primary reasons for using an indirect speech act. Notice that all of the sentences in (7), except perhaps (7f), sound more polite than the simple imperative: Give me some tea! He suggests that this motivation may help to explain why certain forms tend to be conventionalized for particular purposes.
7 Richard Lederer (1988) Get Thee to a Punnery. Wyrick & Company.
8 Searle (1975: 69) suggests that a doctor might ask such a question to check on the progress of a patient with an injured arm.