ARENA Working Papers
WP 99/13

Talking Oil: A Study of Speech Acts Performed by Saudi Arabian and Norwegian Oil Ministers*

Dag Harald Claes**
ARENA, University of Oslo

Introduction

When a diplomat says `yes', he means `perhaps'; when he says `perhaps' he means `no'; and when he says `no', he is no diplomat. When a lady says `no', she means `perhaps'; when she says `perhaps', she means `yes'; and when she says `yes', she is no lady. Paraphrasing this quote from Voltaire one could suggest that when a Saudi Arabian oil minister says the price will increase he means perhaps it will, when he says the price perhaps will increase he means it will decrease, and when he says it will decrease, he is no Saudi Arabian oil minister; at least not for long.

Although this paper primarily focuses empirical cases of speech acts, some theoretical remarks are appropriate. The study of how we speak and write about the world is an important part of social science. There are, however, different approaches to this topic. The quote from Voltaire distinguishes between what is said and what is meant. Most social scientists would also distinguish between what is said and what there is, while some post-modernists would argue that what is said is all there is. The following discussion of speech acts is primarily based on the works of John Searle. In particular his book from 1969 simply called Speech Acts. Searle partly bases his argument on John Austin's pioneering book from 1962 How to Do Things with Words. Thus the discussion also contains some references to Austin's work. Regarding the relationship to postmodernism, I think the arguments to be made in this paper will speak for themselves, when it comes to Searle let me only note that he in his book The Construction of Social Reality from 1995 takes a strong realist position against postmodernism.

Searle's work is part of the study of language, especially of what can be called the philosophy of language. It is thus no wonder that the focus has been on the speech element of the speech act. Investigating speech acts from the starting point of social science concerned with explaining social behavior, I will emphasize the act element of the speech act. Acting by speaking or acting by other means have important aspects in common. In the words of John Searle: ?a theory of language is part of a theory of action, simply because speaking is a rule-governed form of behavior? (Searle 1969:17).

John Austin first distinguished between constantives and performatives, where the first category is utterances describing something and the second is utterances doing something. He also distinguished between the performance of an act in saying something and the performance of an act of saying something (Austin 1962:99-100). This established the difference between the content of an utterance and the kind of force the utterance has. Although the substantial content of an utterance might be the same, its force can differ according to the context in which the utterance is made. Saying `I do' has different implications standing at the alter in front of a priest during a wedding ceremony, or standing in the shower humming the old ABBA song `I do I do I do I do I do'. Furthermore, the context has implications for the effect of the utterance. The `I do' uttered at the alter, in front of God, is unlikely to be taken as a joke, while the same utterance to the wife to be a year earlier at a late night party was just laughed at.

The effect of utterances also invoke the problem of connecting the correct meaning to the words uttered. Keeping with the weeding couple it is obvious that the kiss at the end of the ceremony has an independent meaning and is an important part of being wed. One can hardly imagine, and only dear think of the possible consequences for the groom who does not kiss the bride, after the priest has said, ?you may kiss your bride?. Note also that the priest say `you may kiss your bride', what most of us would agree that he means is that `you are obliged to kiss your bride'. So, in addition to study the speech acts as such, we need to consider both the context and the effect of the speech acts in question.

This paper has four parts: First, I briefly describe the types of speech acts as Searle presents them. Secondly, I discuss the structure of speech acts, emphasizing the speech act of promising, and the problem of sincerity when promising. Thirdly, I devote some attention to contextual factors and their influence on the speech acts; and finally, I will end the paper with some reflections concerning the implications or effects of speech acts.

The empirical focus of this paper is speech acts performed by former Saudi Arabian oil minister Sheik Ahmad Zaki Yamani, and some of his Norwegian colleagues. The data consist of 46 interviews with Sheik Yamani, from July 1977 to June 1986. The basis for selecting the interviews has been the journal Middle East Economic Survey, as this journal has presented edited transcripts, as question and answers, of different interviews and press conferences with Sheik Yamani. In addition I will present three cases in which different Norwegian oil ministers have performed or tried to perform, speech acts influencing the international oil market. Let us first agree on what speech acts are.

Searle's taxonomy of speech acts

In his article `A taxonomy of illocutionary acts' (Searle 1979:1-29) Searle distinguish between five different categories of what he calls illocutionary acts. That is utterances that has a certain force:

• Assertives. The difference established by Austin between constantives and performatives mentioned earlier was soon found unsatisfactory, also by Austin. The distinction collapses as a speaker, when making statements about the world, commits himself to the truth of the expressed proposition. As Searle puts it: ?The point or purpose of the members of the assertive class is to commit the speaker to something's being the case? (Searle 1979:12). Assertives can be found true or false, but are not constantives. They are speech acts because the speaker wants the hearer to believe that the speaker believes what he says about the world, and thus the speaker performs a speech act. Examples are stating, complaining, and claiming.
• Directives, which are attempts by the speaker to get the hearer to do something. They can be modest, such as ask or strong such as order.
• Commissives. The point of these speech acts is to commit the speaker to some future course of action. Examples are promising, and offering. I will return to the promise in a little while.
• Expressives. They express the speaker's psychological attitude towards a state of affairs that the illocution presuppose. Examples are thank, congratulate, and apologize.
• Declarations. When successfully performed these speech acts brings about the correspondence between the propositional content and reality. The saying is the doing, like `You are fired' or `I resign'. Such utterances often include `hereby', to emphasize this effect.

Table : Categories of speech acts

Source: Searle 1979:12-20, Leech 1983:105-107.

In the 46 interviews with Yamani, covering more than 1200 questions and answers, it comes as no surprise that all these types of speech acts were performed, with the possible exception of the last category. In the following empirical discussion I will illustrate how Yamani has used some of these types of speech acts, but first, I would like to go a bit deeper into the structure and contents of one particular type of speech acts: the promise.

How to perform a speech act: empirical cases of promising

In his book Speech Acts John Searle devotes a chapter to the structure of illocutionary acts. In this chapter he concentrate his presentation around the illocutionary act of promising. By way of the discussion of promising he develops a set of necessary conditions for successful and non-defective performance of speech acts. I will start out with an empirical example of a promise and discuss Searle's conditions from this. The focus in on how to perform successful and non-defective speech acts. I will not be able to go into all details of Searle's discussion of the promise and have excluded aspects that seemed most of linguistic or philosophical interest. Characteristically Searle calls his presentation: ?How to promise: a complicated way? (Searle 1969:57-61). This empirical exposition can thus be called: Did they promise, a complicated answer.

The empirical case in question I have named the �ien case, as Arne �ien became Minster for Oil and Energy when Gro Harlem Brundtland formed a new government in Norway in May 1986. In the government's inaugural address, it was stated that:

?if the opec countries agree on measures capable of stabilizing the oil prices at a reasonable level, the Government will contribute to such stabilization.?

Searle's definition of a promise can be set forth as follows:

?Given that a speaker utters a sentence in the presence of a hearer, then, in the literal utterance of this sentence, the speaker sincerely and non-defectively promises something to the hearer if and only if the following conditions obtain? (Searle 1969:57):

1. Normal input and output conditions obtain.
Searle uses the terms " input " and " output " to cover the large and indefinite range of conditions under which any kind of serious and literal linguistic communication is possible. "Output" covers the conditions for intelligible speaking and "input" covers the conditions of understanding. Together they include such things as that the speaker and hearer both know how to speak the language; both are conscious of what they are doing; they have no physical impediments to communication, such as deafness, aphasia, or laryngitis; and they are not acting in a play or telling jokes, etc.
I think we can safely assume these conditions to have been met in the �ien case.

2. In expressing the promise, the speaker predicates a future act by himself.
A promise must include a prediction of a future act, not a past act. One cannot promise to have done something, and one cannot promise that someone else will do something. We should note that Searle includes in this definition the promise not to do something.
The Norwegian promise was a promise of a future act by the Norwegian authorities.

3. The hearer would prefer the speaker doing what he promise to his not doing what he promise, and the speaker believes the hearer would prefer his doing what he promise to his not doing what he promise.
This condition points to the distinction between promises and threats. ?A promise is a pledge to do something for you, not to you; but a threat is a pledge to do something to you, not for you? (Searle 1969:58). A promise is defective if the thing promised is something the promisee does not want done; and it is further defective if the promisor does not believe the promisee wants it done. A non-defective promise must be intended as a promise and not merely conceal a threat or warning. Searle illustrates this point with an example, that goes as follows:

Suppose I say to a lazy student, " If you don't hand in your paper on time I promise you I will give you a failing grade in the course ". Is this utterance a promise? I am inclined to think not; we would more naturally describe it as a warning or possibly even a threat. But why, then, is it possible to use the locution " I promise " in such a case? I think we use it here because " I promise " and " I hereby promise " are among the strongest illocutionary force indicating devices for commitment provided by the English language. For that reason we often use these expressions in the performance of speech acts which are not strictly speaking promises, but in which we wish to emphasize the degree of our commitment. (Searle 1969: 58)

Regarding the empirical case, Nils Morten Udgaard, state secretary in the outgoing conservative government in 1986, has argued that a Norwegian contribution to the stabilization efforts of Opec was not desired by Saudi Arabia, as it would jeopardize an ongoing Saudi Arabian attempt to punish Opec members that were producing above agreed quotas (Udgaard 1989). This argument assumes that the Norwegian contribution could increase the oil price above the Opec target, or at least so that the price effect of others contributions would be lost due to the Norwegian contribution. It seems more reasonable to argue that Opec, and Saudi Arabia, would prefer Norway and others to contribute to price stabilization. The new Norwegian government seemed to believe this to be the case.

4.It is not obvious to both the speaker and the hearer that the speaker will do what he promise in the normal course of events.
If I promised to do something that it was obvious to all concerned that I was going to do anyway, my audience would most likely assume that I believed that it was not obvious that I was going to do the thing promised. A happily married man who promises his wife he will not desert her in the next week is likely to provide more anxiety than comfort. This could imply a conditional aspect of promises, since it is the conditions specified in the promise that brings the course of events out of its normal path. In the �ien case this would be if the Opec countries agreed on stabilizing the oil price at a reasonable level. However, as far as my reading goes, this has not been explicitly discussed by Searle. I suspect that this is a point where the interests of philosophers of language and social scientist diverge. We see this in the work of Knut Midgaard (1993) who bases some of his arguments on Austin and Searle, but focuses his discussions on the concept of interlocution rather than illocution.
This condition of Searle does seem to have been met in the �ien case. There was no reason to believe that Norway would reduce its oil production if no one else did contribute. The condition was also clearly stated in the promise. The Norwegian measures depended on opec itself enforcing measures inclined to stabilize prices, it was thus conditioned on a future act of opec.

5. The speaker intends that the utterance will place him under an obligation to do what he promises.
The point here is that the sentence uttered is intended as a promise, not as something else. Searle explicitly reserve this condition to the utterance, thus it does not include conditions related to what extent the speaker's intentions are relaized.
As the statement by the Norwegian government was presented in a formal way in the inaugural speech in parliament it increases the perceived seriousness of the promise. It can obviously have been other reasons for making this statement than making a promise to Opec. Even if the government did have several intentions with the statement, it did make a promise to Opec.

6. The speaker intends to produce in the hearer the knowledge that the utterance is to count as placing the speaker under an obligation to do what is promised.
Searle describes this condition as follows: ?The speaker intends to produce a certain illocutionary effect by means of getting the hearer to recognize his intention to produce that effect. He also intends this recognition to be achieved in virtue of the fact that the meaning of the item he utters conventionally associates it with producing that effect?.
After the Norwegian promise diplomatic relationship with Opec was activated to ensure that the message was conceived correctly.

7. The speaker intends to do what he promises.
This reflects the distinction between sincere promises where the speaker intends to do the act and insincere promises where the speaker does not intend to do the act. Searle call this condition the sincerity condition. This condition cannot have the same status as the others, because Searle in the following section of his article argues that also insincere promises are promises. This leads to the discussion of how to make promises credible, and to what extent it is possible to make insincere promises as credible as sincere promises. Following that path of reasoning would invoke the literature and debate of bargaining and game theory. Knut Midgaard (1993) has shown the substantial benefits from exchange between game theories and theories of the use of language. However, I do not find it possible to develop this tradition also in a short paper like this.

Without going into details of the analysis of the Norwegian petroleum policy, the statement of May 1986 was sincere. This was revealed when the Ministry of Oil and Energy in September 1986 reduced the net export of crude by around ten percent in.? ^[1] Thus removing some 80,000 barrels per day from the international market. The promise was performed successfully and non-defectively.

Let me now add two other examples of market related utterances by Norwegian oil ministers that might illustrate how a speech act can go wrong, or in the words of Searle: be unsuccessful or defective.

First, the case I have called the Kristiansen case. On 23 January 1986 a press release was issued from the Norwegian Ministry of Oil and Energy (I quote): ?As far as opec is concerned, measures taken by Norway are only conceivable in a situation where opec is willing and able to regulate production in an effective manner.? ^[2] Platt's Oilgram News, a daily oil industry news sheet, wrote as follows: ?A flurry of fact and rumor surrounding potential moves toward a united opec/non-opec producer front kept oil and financial markets on edge today. Norwegian Petroleum Minister Kaare Kristiansen made the biggest splash by suggesting that his government would go along if Britain and other non-opec producers reach agreement with opec to cut output. ?There can be a stable oil price only if there's a high degree of agreement between oil producing countries inside and outside opec,? Kristiansen's spokesman quoted him as telling a conference in Sandefjord, Norway.? ^[3]

The following day Platt's reported: ?Yamani praised Norwegian Oil Minister Kaare Kristiansen's statement yesterday indicating that his country would be willing to cut crude production.? ^[4] A few paragraphs further down another news story reads as follows: ?Norwegian Oil Minister Kristiansen, who won Yamani's praise for his earlier comments, took back much of the support he implied yesterday for such a pact. Yamani had no sooner set forth his statement than Kristiansen issued a counter statement largely rescinding his remarks of the day before. Today's version of the Norwegian position defined it as ?up to Saudi Arabia? to do something about prices and production. Kristiansen added, ?I can't see any reason for the Norwegian government to impose on companies a reduction in production?. ^[5]

A year later Kristiansen described the situation thus: ?I recall an odd episode. Now, it was not my mistake, but a press release went out from the Ministry concerning an address I was to hold at a seminar in Sandefjord just after opec had caused the dramatic oil price drop in the winter of 1986. A secretary wrote in the press release something quite different from what I said ? namely, that Norway was willing to cooperate with opec on certain conditions. But that was not what I said in Sandefjord? (Lind�e 1987:102?103, my translation). On the question that he had gotten cold feet and deleted the paragraph before his speech Kristiansen adds: ?No, the paragraph was part of an earlier draft, but we had discussed the speech in the Ministry and decided to deleted it. However, it became part of the press release? (Lind�e 1987:102?103, my translation). We thus have an interesting case where the speech act was not actually uttered by the speaker: Oil Minister Kristiansen, but still reacted upon as if it had been uttered.

The final Norwegian case I have called the Arnstad case. On the 26^th of March 1998 the Norwegian oil minister, Marit Arnstad announced that a press conference was to be held later in the evening. This increased the Brent oil price with more than half a dollar during a couple of hours before the press conference. This was due to expectations among market actors that the press conference would include an announcement of Norwegian production cuts. At the press conference it became clear that Arnstad did not have the necessary support in the parliament's enlarged foreign affairs committee. This caused the price to fall back to its previous level.

The problem was that Arnstad had not consulted the opposition parties in beforehand. To the opposition the issue was thus perceived as a fait accompli when they understood that a press conference was announced before they had been given an opportunity to voice their opinions. Although Arnstad at the press conference denied it had been a problematic meeting, and called it a ?first consultation with the parliament? she would hardly have summoned a press conference to tell only this. She said to Dagsavisen the next day that she had not anticipated the reactions in the Parliament. ^[6] Arnstad had obviously planned to announce a cut in the Norwegian oil production at the following press conference, but did not get the support of the parliament's enlarged foreign affairs committee, and had to smooth the situation, by calling it a first consultation with the parliament. It should be noted that a compromise, cutting the production, was agreed upon between the Government and the Labor party the very next day.

If we compare these Norwegian cases (table 2) we can illustrate how the relationship between what is said and what is meant ca be distorted. Kristiansen did not intend to perform any speech act, he did, or his secretary did it for him. Arnstad did intend to perform a speech act, but was prohibited by the Parliament from doing so.

Table : Comparing the Norwegian cases

Evaluating the promises, �ien seems to have been successful and non-defective. Kristiansen defected as he performed a promise that he did not intend to perform, while Arnstad was unsuccessful in performing the act she intended to perform.

Our first conclusion is that performing successful and non-defective speech acts is no easy task in itself. It becomes even more difficult when we now turn to the possible distorting role of the context surrounding the speech act.

Context

John Austin (1962:14-15) formulates the contextual conditions for successful speech acts as follows: ?There must exist an accepted conventional procedure having a certain conventional effect, the procedure to include the uttering of certain word by certain persons in certain circumstances.?

Our empirical investigations of speech acts must therefore include a discussion of these ?certain circumstances? and their likely implications for the speech acts performed. The number of circumstances likely to influence a particular speech act is indefinite. One should thus be cautious when evaluating the explanatory power of the empirical factors I am going to discuss, as there is a large residual category of possible important contextual conditions. I will distinguish between background factors and factors relating to the situation in which the speech acts takes place.

In the background category we find factors like the present oil price, production, political conditions and personal background factors. In the case of Yamani one should not rule out his family life as an important factor: in a part of the period discussed, more precise during the seven years from 1976 to 1983, his wife Tammam gave birth to five children. I guess most parents would agree that having children does something to you, although it is harder to say exactly what.

In the situational category we find factors related to the particular situation when the speech act is performed. This could be factors like the air conditions in the room of the interview, the interviewers appearance, whether he smile or express anger, the language used at the press conference or the physical and mental condition of the person at the moment of the utterance. I have collected some empirical information regarding one background factor: the market situation, and one situational factor that I have called the cultural setting. Let us first consider the market situation.

Figure : Oil price and Saudi Arabian production

Source: OPEC Bulletin & Petroleum Economist

Figure 1 shows the development of Saudi Arabian oil production and the oil price in the period covered by the Yamani interviews. As the figure clearly shows this is a turbulent period.

Let us now compare the substantial content of some of the statements Yamani made during this period, based on the interviews reported in Middle East Economic Survey:

1. In April 1981 Yamani was asked when the world could expect Opec to be the old unified Opec that operates so effectively as a cartel? He replied: ?Today, if we want. The Saudis can reduce their production to 6 mbd and live happily with that ... then immediately you will have a shortage and the price of oil will go up ... but we belive in a lower oil price?. When it comes to performing for instance directives, that was speech acts trying to get other to do something, he is for instance rather conciliatory about how to get other producers to perform preferable future action. He says: ?I think that the others will realize the fact and come and sit down with us and reduce their prices.? ^[7]
2. In March 1982 he declares the Saudi Arabian swing producer role and states that the production will be 7 mbd for the next month. He also stresses that a lower production ceiling can be imposed, ?The possibility of going down is there ? I just want this to be well known.? ^[8]
3. In December 1982 Yamani performs another directive, addressed to the other Opec members that exceeded the production quotas and gave price discounts, he simply stated: ?These activities will have to stop.? ^[9]
4. In January 1983 he is asked ?are you still the swing producer??, he answers: So far, yes. But I'm reaching the bottom. There is no way to go below.? ^[10]
5. By December 1985 the swing producer role was definitely abandoned when Opec changed its strategy from defending the price to defending a fair market share. Yamani stated that he hoped Saudi Arabia did not have to increase production, but he indirectly admitted that he did no longer have control when he said ?we are heading towards something unknown ... anything can happen ? I don't know.? ^[11]
6. In April 1986 he looks back and explain the change in the market with as follows: ?The whole matter happened when Saudi Arabia was not able to carry the burden.? ^[12] Later in the same interview the following exchange occurs: Q: Did anyone ask you to resume your swing role? A: Of course. Q: And what was your answer? A: Silence. I was silent most of the time this time. I enjoyed that.

This last remark clearly expresses Yamani's psychological attitude, and is thus an example of the category Searle calls expressives.

If one should dare a general interpretation of these statements one could claim that Yamani, in the interviews from December 1985 and April 1986 expresses a resignation regarding the Saudi Arabian role in the oil market. By statements like: ?anything can happen, I enjoyed being silent, we failed to carry the burden? and so on. This is quite different from the statement of 1981, when he argues that Saudi Arabia could turn Opec into a genuine cartel at once. As the potential for Saudi Arabian governance of the market disappears so does the potential for effectual speech acts. The possibility to influence the market through statements in interviews was gone. In the interview from December 1985 Yamani is also remarkably less wordy. Almost all the questions are answered by one or two sentences, often with single words, like `yes' or `no'. The instrumental role of speaking to the press was gone. It seems reasonable to argue that changes in the market conditions had influenced the contents and effects of the speech acts performed by Yamani.

Let me briefly comment the possible influence of what I have called personal context. Yamani was fired from the post as Saudi Arabian oil minister in October 1986. In an unauthorized biography of Yamani, Jeffrey Robinson argues that Yamani already in July 1985 began removing his personal papers from the ministry and set up offices both in Geneva and London, actually planning his exit as Saudi Arabian oil minister (Robinson 1988:268). If this holds true the more fatalistic statements can also have been influenced by Yamani foreseeing his lack of confidence at the house of Saud.

Let's turn to some data on one aspect of the situational context of the Yamani interviews namely the cultural location of the interviews. The interviews included in the empirical study can be categorized as follows:

1. Western newspaper or broadcasting
2. Lecture in the West
3. Press conference after Opec meetings
4. Exclusive interview with MEES
5. Lecture in Arabic countries
6. Interview in Arabic newspapers or broadcasting

This variable has been re-coded into a dummy variable called culture:

0. Western media or lecture (Interview 1 thru 4)

1. Arabic media or lecture (Interview 5 thru 6)

The frequencies of these variables are reported in table 3.

Table : variable of the situational context

This culture variable has then been correlated with a variable that express to what extent Yamani argues in favor of oil market power as a political weapon, or at least that oil is a political issue (table 4).

Table : Situational context and politicization

Correlations

** Correlation is significant at the 0.01 level (2-tailed)

This correlation is one to one. Yamani never argues in favor of the use of oil as a political instrument or weapon in Western media, all arguments in favor of political use of oil is done in Arabic media. Although the number of observations are low the correlation is as strong as possible. This can be substantiated by a more qualitative methodological approach. The following quote is taken from an interview with Yamani on the NBC television program ?Meet the Press? on 19 April 1981. The reporter states that there is a glut on the international oil market. Yamani's response is as follows: ?Well, as a matter of fact, this glut was anticipated by Saudi Arabia and almost done by Saudi Arabia. ... We engineered the glut and want to see it in order to stabilize the price of oil.? ^[13] Five months later Yamani is interviewed by the Arabic-language newspaper al-Sharg al-Awsat where he states the following: ?All economists agree that the accusations ... that Saudi Arabia is the cause of the oil surplus or of the increase in stockpiles is inadmissible, if not ridiculous.? ^[14]

While the previous section identified insincerity and defective speech acts, this section has at least shown how Yamani spoke with two tongues. So, oil ministers deliberately lie and perform incomplete or defective speech acts. The question then arises what are the implications of this? Does it matter whether they lie or tell the truth? This lead to the final topic of this paper, the effect of speech acts.

Effects

Let me start by extracting a bit more empirical information from the Yamani interviews. In the interviews with Yamani the amount of mere assertive statements is striking. As described above, assertives was according to Searle (1979:12-13), statements about the world which the speaker claimed to be true.

Several questions to Yamani are simply of the following kind: ?How much is Saudi Arabian oil production at the moment? When did your government take this decision? What are the quotas?? He is also asked about his predictions about the future: ?What is your fourth quarter production likely to be? What do you think the oil price will be a year from now?? Some of the answers to these questions are obviously assertives that can be proven true or false, even of course his prediction about the future. shows Yamani's predictions correlated with the actual development of Saudi Arabian oil production and the oil price.

Table : Yamani's predictive power

Future Price

Correlations

* Correlation is significant at the 0.05 level (2-tailed)

Future Production

Correlations

The two tables of correlations suggest a fairly high predictive power. Although a bit surprising, Yamani is better in predicting the price than the future production of Saudi Arabia. It should be noted that it is only the price correlation that is significant, and that the variables only express the direction of the price, not its actual level. The point here is, however, that Yamani could have had many motives for presenting predictions he knew to be wrong. Still these questions are asked over and over again. Why? And why should these unreliable statements influence the oil market?

In order to answer this question we need to consider the formation of expectations in future markets.

The trade system of the international oil market has changed during the last two decades from a market with few large actors, long-term contracts, and low price transparency; into a market characterised by a large number of traders, spot, forward and future markets and an enormous flow of market information. A shipment of oil is sold several times from port to port. The development of future markets makes it possible for traders to trade on oil physically being delivered several months into the future, and has made it possible for the traders to hedge their portfolios, and for refiners to secure profits through netback arrangements.

When market actors make simultaneous and instant moves in the marketplace, the behavior of all other actors could be regarded parametrically, in other words they could be taken as given. Or if the actor should make assumptions about other actors behavior it would have to be exactly that ? assumptions, as no observation about others behavior is possible.

However, when the market transactions are extended in time a new kind of uncertainty enters the calculations of market actors. In such a market, actors are concerned not only with the price of the traded good today, but also the price some times in the future, the price of the good today will be influenced by the traders expectations about the price tomorrow. This price will in turn be influenced by the expectations about the price the day after tomorrow and so on. This does not necessarily change the formation of expectations, but it creates room for more and extended strategic interaction among the market actors.

As Stephen Burnell (1989:394) argues:

a rational agent will recognize that the price tomorrow will depend upon the actions of all other agents tomorrow, and hence upon their theories of price formation ... The potential then exists for a rational agent to believe that certain events will affect the price ... simply because (he believes) other agents believe these events matter.

This open up for the possible influence of factors that cannot reasonably be claimed to affect the market fundamentals, such as endowments, preferences or production capacities, but that can be claimed to affect other actors believes.

Factors that have an effect on the fundamentals are intrinsic to the economy, while other factors are extrinsic. The activity of the sun, creating sunspots is one such extrinsic factor. `Sunspots' have thus become the name for such factors in the economic literature. The economic literature on sunspots has primarily been focused on how equilibrium emerges in such markets. The empirical cases have mostly been the stock market and to some extent the foreign exchange market. But commodity future markets, including the oil market, are also regarded as having the characteristics making the market vulnerable to sunspots.

In such markets price formation can be thought of as a social convention. This idea of price formation as a social convention was developed by Keynes, in his book The General Theory of Employment, Interest and Money from 1936:

The essence of this convention lies in assuming that the existing state of affairs will continue indefinitely, except in so far as we have specific reasons to expect a change. ... We are assuming, in effect, that the existing market valuation, however arrived at, is uniquely correct in relation to our existing knowledge ... and that it will only change in proportion to changes in this knowledge. (Keynes 1936:152)

Keynes then goes on to discuss different weaknesses in this convention. For the purpose of this paper one such weakness is particularly important. Keynes argues that a

conventional valuation which is established as the outcome of the mass psychology of a large number of ignorant individuals is liable to change violently as the result of a sudden fluctuation of opinion due to factors which do not really make much difference. ... In abnormal times the market will be subject to waves of optimistic and pessimistic sentiment, which are unreasoning and yet in a sense legitimate where no solid basis exists for a reasonable calculation. (ibid.:154)

The speech acts discussed above can be regarded as an extrinsic factor influencing the market convention. The utterances of Yamani or anybody else do not in themselves alter any fundamental parameters of the oil market. They do however, influence the assumptions held by market actors, in particular the traders, about the future conditions of these parameters. By influencing the traders' assumptions about the future, the sunspots (or speech acts) influence the price of oil in the forward and future markets. As mentioned earlier, changed believes about the future price of oil, subsequently changes believes about what is the ?right? price of oil today. The point is thus not if one believes what Yamani is saying, but the probability that other actors believe it. This logic can move markets far away from what the ?fundamentals? of the market indicates.

If we assume that market actors, in particular the oil traders, behave according to this model, basing their own actions as much on believes about the future action of other actors than on the market fundamentals, the speech acts becomes important whether they are sincere, non-defective, consistent or not. The point for the speaker is then not to be sincere, but to create the desired believes among market actors.

A strategic use of speech act in order to talk the price up does however seem difficult. It might also be a risky business, because an overexploitation of the sunspot phenomenon might weaken the force of ones statements over time. If we apply the argument of Keynes oil traders might fall into waves of optimistic and pessimistic market behavior, increasing price fluctuations. It is hard to prove the effects of speech acts in such market conditions. This is clearly demonstrated if we just briefly return to the Norwegian oil ministers and their speech acts discussed earlier. What was their actual effect on the market? The following figures show the oil price around the statements of the Norwegian oil ministers: First the Kristiansen statement, then the statement by �ien and finally the aborted statement by Arnstad.

Figure : The effect in the Kristiansen case

Source: Petroleum Economist

Figure : The effect in the �ien case

Source: Petroleum Economist

Figure : The effect in the Arnstad case

Source: Petroleum Economist

In the case of Kristiansen the price fell, but he had no intention of doing anything neither with words nor deeds. The inaugural speech in May 1986 of the �ien case, was sincere, but did not have a substantial price effect. �ien might have had an effect when the promise of May 1986 was substantiated in September the same year. Arnstad did not performed the intended speech act, but the in the following months the oil price increased slightly. Obviously we don't have any hard evidence for establishing a causal connection between Norwegian oil ministers' speech acts and the price of oil.

Conclusion

This paper has identified some types of speech acts performed by oil ministers, discussed the success and defectiveness of certain promises, shown the influence of contextual factors on speech acts, and discussed the problems of identifying their effects on a market like the international oil market.

I have discussed theories and empirical illustrations on a topic that cuts across the philosophy of language, political science and economics. By studying a particular phenomenon from different angels, and with different perspectives, one can increase the empirical validity of the arguments. Empirical studies of speech acts tends to become either highly specific, concentrating on one or a few rather trivial utterances; or if they are more general their empirical significance are usually lost. The cases and correlations presented have tried to find a middle-way in the empirical analysis. As pointed out by Dagfinn F�llesdal (1982:561):

In studying man, we need a very comprehensive theory, which includes a theory of action, a theory of meaning and communication, a theory of reference, and a theory of knowledge. All of these are interconnected, in such a way that evidence is transmitted between the theories.

Footnotes

* This paper is a slightly revised version of a trial lecture over a chosen subject for the degree of doctor politicarum rerum, held at the Department of Political Science at the University of Oslo on 4th of December 1998.

** ARENA, P.O. Box 1143 Blindern, 0317 Oslo, Norway, tel.:+47 22 85 78 57, fax: +47 22 85 78 32, e-mail: dagc@arena.uio.no

^[1] Press release from the Norwegian Ministry of Oil and Energy, 10 September 1986.

^[2] Press release from the Norwegian Ministry of Oil and Energy, 21 January 1986.

^[3] Platt's Oilgram News, 23 January 1986.

^[4] Platt's Oilgram News, 24 January 1986.

^[5] Platt's Oilgram News, 24 January 1986.

^[6] Dagsavisen (Oslo) 27. March 1998 p. 10.

^[7] Middle East Economic Survey, 27 April 1981.

^[8] Middle East Economic Survey, 29 March 1982.

^[9] Middle East Economic Survey, 20 December 1982.

^[10] Middle East Economic Survey, 31 January 1983.

^[11] Middle East Economic Survey, 16 December 1985.

^[12] Middle East Economic Survey, 28 April 1986.

^[13] Middle East Economic Survey 27 April 1981.

^[14] Middle East Economic Survey 14 September 1981.

References

Austin, J.L. (1962). How to Do Things with Words. Oxford: Oxford University Press.

Burnell, Stephen (1989). ?Sunspots? in Frank Hahn (ed.) The Economics of missing markets, information, and games Oxford : Clarendon Press.

F�llesdal, Dagfinn (1982). Intentionality and behaviorism, Proceedings of the International Congress of Logic, Methodology and Philosophy of Science, Hannover 1979, s. 553569.

Keynes, John Maynard (1936). The general Theory of Employment Interest and Money London: Macmillan.

Lind�e, John Ove (1987). ?Intervjuer med fem statsr�der?, in Egil Helle (ed.), Olje og Energi i ti �r 1978?88. Oslo: Ministry of Oil and Energy.

Midgaard, Knut (1993). Strategi, politisk kommunikasjon og forhandlinger Institutt for Statsvitenskap, Universitetet i Oslo.

Robinson, Jeffrey (1988). Yamani ? the inside story. London: Simon & Schuster.

Searle, John R. (1969) Speech Acts ? an Essay in the Philosophy of Language Cambridge: Cambridge University Press.

Searle, John R. (1979). Expression and Meaning ? Studies in the Theory of Speech Acts Cambridge: Cambridge University Press.

Searle, John R. (1995). The Construction of Social Reality New York: The Free Press.

Udgaard, Nils Morten (1989). ?Norway between iea and opec: oil prices and foreign policy?, in Ole Gunnar Austvik (ed.), Norwegian Oil and Foreign Policy. Oslo: Vett & Viten.

[Date of publication in the ARENA Working Paper series: 15.04.1999]