Penultimate version of Yukio Irie Question and Inference in,begegnungen in Vergangenheit und Gegenwa rt, Claudia Rammelt, Cornelia Schlarb, Egbert Schlarb (HG.), Lit Verlag Dr. W. Hopf Berlin, Juni, 2015, pp. 365-375. Question and Inference Yukio Irie (irie@let.osaka-u.ac.jp) The meaning of an utterance can be understood only in relation to a question to which the utterrace is an answer ( Collingwood thesis 1 ). On the other hand, there is an inferentialist theory of meaning, which seeks to understand the meaning of an utterance from its role in an inference. If these two approaches to the theory of linguistic meaning are correct, then we could reach a more profitable theory of meaning by making the relationship between a question and an inference explicit. We will argue the following two points. 1. An inference presupposes* a question. 2. How does the question arise? Section 1 An inference presupposes* a question. 1 Definitions of a theoretical question and a practical one Questions and Answers are divided into a monolog of answering one s own question and a dialog with others. (1) Definitions of two types of questions in a dialog A theoretical question is a question whose answer is an assertion, which has a truth value. A practical question is a question whose answer is an intention-expressive utterance, which has no truth value, such as an order and a promise. One and the same sentence can become an assertion with truth value and an intention-expressive utterance. For example, This is yours would be an assertion in some cases and an intention-expressive utterance of an assignation in other cases. Therefore, the following research into questions and inferences does not deal with sentences but, rather, utterances. In this article, an utterance that describes a fact and has a truth value is called an assertive utterance, and an utterance that has no truth value and addresses others, such as requests, orders, promises, attitude-expressive utterances, and declarations, is called an intention-expressive utterance. John Searle divided illocutionary acts into five types (the assertives, the directives, the commissives, the expressives, and declarations). What is called here an intention-expressive utterance includes the directives, the commissives, the expressives, and the declarations. To this extent, I owe much to J. Searle. But, I cannot agree with him regarding the 1 Cf. Yukio Irie, A Proof of Collingwood s Thesis in: Philosophia Osaka 4 (2009), 69 83. 1
utterance of a question. Searle rendered the utterance of question as a kind of request. 2 I think we should deal with it as different from an intention-expressive utterance, which includes the declaration. 3 (2) Definitions of two types of questions in a monolog A theoretical question is a question the answer to which is an utterance that has a truth value. A practical question is a question the answer to which is an utterance that expresses an intention and has no truth value. I will call an utterance that describes a fact and has a truth value an utterance of fact, as distinct from an assertive utterance, and an utterance by which we intend something in a monologue, an utterance of intention, as distinct from an intention-expressive utterance. An utterance of a question in a monologue has no truth value, but we would like to deal with it as distinct from an utterance of intention. (3) Illocutionary acts in dialogue and monologue Searle undertook detailed research into illocutionary acts in dialogue. He did not divide illocutionary acts into dialogues and monologues. However, what speech in a monologue does is different from an illocutionary act in dialogue. The utterance of fact in monologue has a truth value but does not attempt to transmit something to an addressee. The utterance of intention will not order, promise, express an attitude, or make a declaration to an addressee. This said, it performs not only a propositional act but also something else. It resembles the illocutionary act, because it has a direction of fit, which Searle noted in the illocutionary act. The utterance of fact has a direction of fit from word to world, and the utterance of intention has a direction from world to word. I call this act the monological illocutionary act. As we cannot perform a propositional act unless we perform an illocutionary act in dialogue, we cannot perform a propositional act unless we perform a monological illocutionary act in monologue. However, in dialogue the monological illocutionary act is not more fundamental or primitive than is the illocutionary act, because the language would have initially arisen (i.e., in terms of both the history of humankind and the development of infants) in dialogue. Indeed, as the monologue becomes possible only after the emergence of dialogue, the monological illocutionary act is thought to be a derivative of the illocutionary act in dialogue. 2 Definitions of a theoretical inference and a practical one (1) Inferences in a monologue We will first define two types of inference. (As I mention later, there are also other types). A theoretical inference is a combination of utterances such that one or more utterances of facts constitute premises and one utterance of fact constitutes a conclusion, and if the premises are all true, then the 2 Cf. J. R. Searle, Expression and Meaning, Cambridge U.P. 1979, 14. 3 Cf. Yukio Irie, The Illocutionary Negation and the Question (in Japanese) in: Bulletin of Osaka Shoin Women s College 29 (1992), 201 217. 2
conclusion becomes necessarily true. A practical inference is a combination such that one utterance of intention and, in many cases, utterances of facts, constitute premises, and an utterance of intention constitutes a conclusion, and (1) if we commit the premises, then we necessarily commit the conclusion, or (2) if we commit the premises, then we can commit the conclusion. (1) and (2) are inferences as follows. Practical inference (1) The only one measure of doing X is doing Y. I shall do Y. Practical inference (2) If I do Y, then I can do X. I shall do Y. (1) corresponds to what Von Wright proposed as a form of practical inference, and (2) corresponds to what Anscombe proposed. 4 (1) shows us that in order to do X, it is necessary to do Y. However, such cases are rare. In many cases, there are plural ways to do something and we would make practical inferences in such cases. Therefore, (2) is a more general form of practical inferences. There are other types of inferences as follows. (a) It is possible to theoretically construct an inference in which the premises contain utterances of intentions, and the conclusion is an utterance of fact. We can deduce the conclusion only from utterances of facts in premises, and/or we can deduce an utterance of fact from an utterance of intention, e.g., that is salt from I shall take that salt ; or we can deduce an utterance of fact from an utterance of fact that is deducible from an utterance of intention and other utterances of facts in premises. I want to call this type practical-theoretical-inference. (b) By contrast, it is also possible to construct an inference in which all premises are utterances of facts and a conclusion is an utterance of intention. If we can infer from the commitment to the premises to the commitment to the conclusion, then the inference can hold according to the above definition of practical inference (2). I will call this type theoretical-practical-inference. (2) Inferences in dialogue The theoretical inference in dialogue is envisaged to be functionally reducible to a theoretical inference in monologue. By contrast, the practical inference in dialog is difficult to reduce to the practical inference in monologue, because there is no utterance of so-called illocutionary acts such as orders, promises, and so on. These ideas are very important, but we will not deal with them here. 4 Cf. E. Anscombe, Intention, Cambridge, 1957, 2nd version 1967; E. Anscombe, Practical Inference, in R. Hursthouse, G. Lawrence, and W. Quinn (eds.), Virtues and Reasons, I 34, Oxford, 1989; von Wright, On So-Called Practical Inference, Acta Sociologica 15, 39-53; and J. Searle, Rationality in Action, Cambridge, Chap. 8. 3
3 Theoretical inference presupposes* a theoretical question In the theoretical inference, if premises are true, then a conclusion is necessarily true. Furthermore, in the theoretical inference, if premises are true, then not one but many utterances are to be logically true as follows. All penguins are birds. All birds lay eggs. All penguins lay eggs. All that do not lay eggs are not penguins. Some penguins lay eggs. There is no penguin that does not lay eggs. All that are not birds are not penguins. All that do not lay eggs are not birds. In spite of this situation, why is only one utterance selected as a conclusion? In order to answer to this question, we need to consider that the inference is made to answer some question and the conclusion of the inference becomes an answer to the question. The above inference is performed to answer the following question. Does a penguin lay eggs? All penguins are birds. All birds lay eggs. All penguins lay eggs. Therefore, the same premises would produce a different conclusion, when we infer to answer a different question. Are there penguins that do not lay eggs? All penguins are birds. All birds lay eggs. There is no penguin that does not lay eggs. So, we can claim that the theoretical inference presupposes* a theoretical question. This presupposition* is different from the usual use of presupposition in logics, so I have marked it with *. 4 Practical inference presupposes* a practical question Additionally, in practical inferences, many utterances can be deduced as a conclusion from the premises. First, we consider the practical inference (1). The only one measure of doing X is doing Y. I shall do Y. If I cannot do Y, I give up doing X. If I do not want to do Y, I must give up doing Y. If I intend to do X, I need to intend to do Y. 4
If I do not do Y, I cannot do X. The last three of these conclusions are utterances of fact that have truth values. So, one of the three becomes a conclusion and then the inference is not a practical inference but a practical-theoretical-inference. In order to explain why I shall do Y is selected as a conclusion from many candidates, we need to consider that the inference is performed in order to answer a practical question of conclusion, and the conclusion becomes an answer to the question. The above inference is performed to answer the following question. What shall I do to do X? (practical question) The only one measure of doing X is doing Y. I shall do Y. The practical inference is an inference we perform to decide what I shall do to fulfill an intention. Therefore, the inference constitutes a process to answer a practical question like What shall I do to do X? This question arises after the intention I shall do X arises. Then, the practical inference is performed to answer the question in the above way. An answer to this question of What shall I do to do X? is not The only one measure of doing X is doing Y, because this is an utterance of fact that has a truth value. So, the answer to this question is I shall do Y. We can explain why the utterance of the intention I shall do Y follows from the beginning two premises. By the way, the same premises would produce a different conclusion when we infer to answer a different question. In which case I cannot do X? (practical question) The only one measure of doing X is doing Y. I cannot do X, if I cannot do Y. Then we consider the practical inference (2) If I do Y, then I can do X. I shall do Y. If I do something other than Y, I could do X. If I cannot do X, then I cannot do Y. In order to explain why I shall do Y is selected as a conclusion from many candidates, we need to consider that the inference is made in order to answer a practical question of conclusion, and the conclusion becomes an answer to the question. The above inference is made to answer the following question. What shall I do to do X? (practical question) If I do Y, then I can do X. 5
I shall do Y. The last two of these conclusions are utterances of fact that have truth values. So, if one of two candidates becomes a conclusion, then the inference is not a practical inference but a practical-theoretical-inference. A concrete example will help to show that different practical questions lead us from the same premises to different conclusions. I shall repair a window. If I get tools in a DIY shop, I can repair a window by myself. I shall get tools in a DIY shop. I shall repair a window by myself. I shall go to a DIY shop. There are many candidate conclusions. The choice of the actual conclusion depends on the question to which a practical inference is the answer. I shall repair a window. How shall I repair a window? If I get tools in a DIY shop, I can repair a window by myself. I shall repair a window by myself. I shall repair a window. What do I need to do to repair a window? If I get tools in a DIY shop, I can repair a window by myself. I shall get tools in a DIY shop. I shall repair a window. What shall I do first to repair a window? If I get tools in a DIY shop, I can repair a window by myself. I shall go to a DIY shop. Thus, we can claim that the practical inference presupposes* a practical question. Therefore, all inferences (i.e., theoretical inferences and practical inferences), including the practical-theoretical-inferences and the theoretical-practical-inferences, presuppose some questions. When we consider inferences, we always need to attend to the fact that the inference is made only by a presupposed* question. I will call the inferences understood in this way question-answer-inference and define them as follows. A theoretical question-answer-inference is a combination of utterances such that one theoretical question and one or more utterances of facts constitute premises and one utterance of fact constitutes a conclusion, and if the premises are all true, then the conclusion becomes necessarily true. A practical question-answer-inference is a combination such that one utterance of intention and one practical question and, in many cases, utterances of facts, constitute premises and an utterance of 6
intention constitutes a conclusion, and (1) if we commit to the premises, then we necessarily commit to the conclusion, or (2) if we commit to the premises, then we can commit to the conclusion. Section 2 How does the question arise? As we have seen, the inference is performed to answer a question. Then, how does the question arise? We consider here how the theoretical question and the practical one arise. In this process, the relationship between the two types of question will become clear. 1 How does the theoretical question arise? There are five cases in which theoretical questions are settled. (1) A theoretical question is settled in order to answer another question. (2) A theoretical question is settled when we notice a contradiction between two utterances of facts. (3) A theoretical question is settled to answer a practical question. (4) A theoretical question is settled to answer a question asked by another. (We do not analyze here questions and answers in dialogue.) (1) A theoretical question is settled in order to answer another question. In some cases, a theoretical question is settled to get a premise for constructing an inference for answering another theoretical question, as follows. Does a penguin lay eggs? (theoretical question 1) A bird lays eggs. (utterance of fact 1) If a penguin is a bird, then the penguin lays eggs. (utterance of fact 2) Is a penguin a bird? (theoretical question 2) A penguin is a bird. (answer to theoretical question 2) A penguin lays eggs. (answer to theoretical question 1) The three underlined utterances consist of an inference whose conclusion is an answer to theoretical question 1. Theoretical question 2 is settled to answer theoretical question 1.We could also say here that the asking of theoretical question 2 has arisen as a conclusion of the following practical inference. I shall answer theoretical question 1. (intention 1) What shall I do to answer theoretical question 1? (practical question) If I answer theoretical question 2, then we can probably realize intention 1. (utterance of fact) I shall answer theoretical question 2. (answer to practical question) The realization of this intention is to ask theoretical question 2. To ask a theoretical question is an intentional action, and an intentional action is always a conclusion of a practical inference; therefore, to ask a theoretical question is a conclusion of a practical inference. (2) A theoretical question is settled when we notice a contradiction between two utterances of 7
facts. Even if two utterances of facts are in contradiction, it is not enough to generate a question. It is necessary that one notices the contradiction between them. There are two cases in which the contradiction of two utterances of fact may be noticed. The first case is while making an inference. For example, when we infer not q from p, we notice that p and q fall into contradiction. However, the recognition of the contradiction is still not enough to generate a question. It is necessary that we think it necessary to resolve the contradiction. The necessity comes from trying to resolve a theoretical question. Given that we have found the contradiction during the making of an inference to solve the question, we need to resolve the contradiction in order to answer the question. So, it becomes necessary to pose a theoretical question of How shall I solve this contradiction? So, (2) is a special case of (1). The second case is when we encounter an event against an expectation, such as being surprised. In such a case, questions such as Oh, what? and Uh, why? would arise almost automatically. Cognition of the contradiction between the expectation and the event against it (i.e., the contradiction between two utterances of the expectation and the event) is verbalized as an utterance of the question. Sometimes we can immediately receive an answer to the question but not in all cases. The question arises automatically by being aware of the contradiction; it does not arise in order to solve another question, so we would leave it, even if we cannot solve it. (3) A theoretical question is settled to answer a practical question. In some cases, we settle a theoretical question to answer a practical question. We explain this process in subsection 2. (4) A theoretical question is settled to answer a question asked by the other. Here, we do not deal with questions and answers in dialogue. However, we will comment on the above cases. Some cases of (2) are special cases of (1). In case (1), a theoretical question arises in order to answer to another theoretical question. The other one might also arise in order to answer another, higher question. To avoid an infinite regress, the highest theoretical question in the regress must arise in an alternative way to (1). The highest theoretical question arises in (2), (3), or in dialogue. However, we do not analyze here questions and answers in dialogue. 2 How does the practical question arise? The cases in which practical questions arise seem to be divided into two cases. The case when we decide how to put an intention into place. The case when we decide an intention about an issue. The question What shall I cook for a dinner today? is a question of the latter case. But if we the have an intention of I shall enjoy a dinner today, the question is a question of the former case. In the case of a monologue and answering one s own question, we could say, every practical question arises to decide how to put an intention into place. Regarding the case of dialogue, we need to consider it in another way. 8
By the way, intentions are divided into intentions in actions and previous intentions. In both cases an intention always entails cognition of reality, which is in conflict with the intention. However, a conflict between an intention in action and cognition of reality does not give rise to a practical question, because the intention is already in the process of its realization and, therefore, we need not ask how to put it into place. On the other hand, the conflict between a previous intention and cognition of reality gives rise to a practical question that asks how to put it into place. In short, the practical question arises from a conflict between a previous intention and cognition of reality. By the way, previous intentions are divided into two types. (a) The previous intention that is immediately performed when the conditions are fulfilled. (PIa) E.g., I shall cross at the crosswalk, when the signal turns green. (b) The previous intention that needs to be decided in the way of performing it in order to be performed. (PIb) In (a), if we cognize the fulfillment of conditions concerning the time, space, and situation, the previous intention moves into an intention in action; therefore, we ask no practical question. In (b), we ask a practical question such as What shall I do to put the intention X into place. What is the difference between (a) and (b)? In which case do we need to decide on a method of realization in order to perform an intention? Our actions can be decomposed into basic actions in the sense that Arthur Danto coined. 5 Accustomed actions, such as brushing one s teeth, making coffee, crossing the crosswalk, and so on, can be performed automatically without thinking about the way of performing them. But, actions that are not basic actions and accustomed actions have to be decided by a method of performing them, by asking a question such as What shall I do to do the intended X? We can further divide (b) into two cases. (b1) There are numerous ways to perform a previous intention, and one must select one. (b2) One does not know the way to perform it. In (b1), after having a previous intention of X, a practical question of In which way I shall put it into place? arises. (intention 1) What shall I do to do X? (practical question ) There are numerous ways to do X. (utterance of fact) Which is the best way to do X? (theoretical question) The best way to do X is to do Y. (answer to the theoretical question) I shall do Y. (answer to the practical question) What should we do when we cannot answer the above theoretical question or when we can limit the options but not to the best option? We have the intention to do X, therefore, we would select one option by making a decision. In (b2), we do not know the way to perform an action; therefore, such intentions might be rather called purposes, plans, and so on. The question of How can I do X? is also a theoretical question. 5 Cf. Arthor Danto, What We Can Do, in The Journal of Philosophy, Vol. 60, No. 15 (1963), 435 45. 9
(intention 1) What shall I do to do X? (practical question 1) I do not know the way to do X. (utterance of fact) What can I do to do X? (theoretical question) The answer to the above theoretical question will be a description like I can do Y to do X. One can make the following practical inference based on this answer. (an intention 1) What shall I do to do X? (a practical question 1) I do not know the way to do X. (an utterance of fact) What can I do to do X? (a theoretical question) I can do Y to do X. (an answer to the theoretical question) I shall do Y. When one cannot get an answer to the above theoretical question, one must give up on doing X. If the intention to do X is constructed for a higher-order intention, then he/she sets another intention to realize the higher-order intention. If there is no higher-order intention, then he/she would move onto a quite different intention. In both (b1) and (b2), one becomes aware of the theoretical question only when he/she cannot get an answer immediately. If it is given immediately, the theoretical question is not noticed. Even in such a case, we as interpreters of the circumstances can understand the above utterances of fact are answers to theoretical questions according to CT. The intention I shall do Y as a conclusion would be one of the above three kinds of previous intentions (PIa, PIb1, PIb2). There are five following cases in which the practical questions are settled. (1) A practical question is settled in order to answer another higher practical question. (2) A practical question is settled when plural intentions come into conflict. (3) A practical question is settled when we need to adjust plural intentions to realize them all. (4) A practical question is settled to answer a practical question asked by another. (We do not analyze questions and answers in dialogue here.) (1) A practical question is settled in order to answer other higher practical question. We will consider the following process. (intention1) What shall I do to do X? (practical question 1) I can do Y to do X. (utterance of fact) I shall do Y. (an answer to practical question 1) What shall I do to do Y? (practical question 2) Practical question 2 seems to be asked after getting an answer to practical question 1 and not to be asked in the process of answering practical question 1. However, the answer I shall do Y to practical question 1 is a tentative answer. Because, if Y turns out to be unable to be realized, practical question 1 10
must be asked again. If it turns out to be necessary that we must do Z to do Y, we must do Z to do X. In this case, I shall do Z emerges as an answer to the practical question. We can obtain a final answer to the practical question only when we obtain the way that is explicitly unknown as the way that is completely realizable. (2) A practical question is settled when plural intentions come into conflict. When we have intentions I shall do X and I shall do Y and notice the two intentions are incompatible, then we would pose a practical question, such as Which do I prefer, doing X or Y? In order to answer the question, we might pose a theoretical question like Which must be preferred? or By which criterion should I decide it? and get an answer, e.g., I must prefer X and decide to do X based on the answer. In some cases, we might make a decision without posing such theoretical questions or, even if it is posed, without getting any answer to such theoretical questions. By the way, in a case in which we notice incompatibility between plural intentions, if the incompatibility is easy to discern, we will be aware of the incompatibility when the latter intention that is formed later than another is formed, and we can avoid settling for such an intention, which is incompatible with the intentions we already have. However, if the incompatibility is not easy to ascertain, we become aware of the incompatibility after forming the intentions. When we seek to answer the practical question What shall I do to do X based on the intention I shall do X, we would discover the incompatibility. Then, a practical question arises: Which shall I prefer, doing X or Y? This question is posed for the higher intention of I shall do X or Y. Therefore, (2) is a special case of (1). (3) A practical question is settled when we need to adjust plural intentions to realize them all. The two intentions of I shall go to the seaside this summer and I shall go to the mountains this summer are compatible. However, in order to realize both we need to adjust them. We need to ask a question such as How do we adjust the intention to go to the seaside and the one to go to the mountains this summer? A theoretical question might also be posed in this case, and we get finally an answer to the practical question by making a decision. (4) A practical question is settled to answer a practical question asked by another. We do not deal with questions and answers in a dialogue. However, I will comment on the above cases. Some cases of (2) and (3) are special cases of (1). In case (1), a practical question is settled in order to answer another higher practical question. When we explain how the higher practical question was settled, we need to suppose another further higher practical question. Insofar as we think in the domain of monologue, we cannot avoid an infinite regress. To avoid it, we need to consider that the first practical question arises in order to answer a practical question from some other person. Conclusion In the first section, we analyzed the relationship between questions and inferences in monologue. Questions are divided into theoretical and practical questions and inferences are also divided into 11 コメント [Ed1]: コメント [Ed2R1]: Please clarify as this sentence is a bit unclear.
theoretical and practical inferences. We found that the theoretical inference and practical inference are processes for answering the theoretical question and the practical one, respectively. All inferences turned out to presuppose* the questions. (We did not inquire whether the process of answering a question always becomes an inference.) In the second section, we analyzed how questions arise. A question comes into existence not in isolation but in relation to other questions and inferences. So, we explained the cases in which a theoretical question and a practical one arise. This research is ongoing. One point requiring further work is the analysis of cases in dialogue. Another such point is that we need to inquire about the distinction between a desire and an intention. The analysis of the complicated relationships among questions and inferences concerns research regarding not only interrogative logic but also the theory of meaning. The English in this document has been checked by at least two professional editors, both native speakers of English. For a certificate, please see: http://www.textcheck.com/certificate/r6lkqq 12