differently. For example, a Greek General or a Turkish ambassador may honestly have different ideas of what is crisis management and how to ‘save face’ during a crisis negotiation. There exists the danger of mirror imaging in the interpretation and application of these principles and ultimately, the Turks may understand the crisis management principles differently to the Greeks.
Scholarly research in leadership studies has identified two main types of intelligence-policy relations, the mono-centrist and the poly-centrist. A mono-centrist approach involves a lack of consultation between decision-makers and intelligence staffs and, the formulation of an elite group. This closed group is often isolated from the intelligence bureaucracy and employs it to deliver, rather than analyse, any information. Gradually, this elite team is affected by what is termed ‘groupthink’ which serves to constrain creative thinking and rigorous intelligence analysis in a crisis. Groupthink refers to the stereotyped views of enemy leaders and opponents’ policies and of the inherent morality with which the decision-makers feel for themselves. It also incorporates the exertion of direct pressure on other officials to agree with the prevailing policy options and assessments.37
In most cases, intelligence personnel are kept out of this policy making group. The groupthink phenomenon refers to the situation when political, military and intelligence officials follow the same patterns of thought, interpretation and analysis. This can lead to a disregard for parameters and hypotheses that could aid the understanding of the crisis situation as a ‘dynamic intelligence reality’. CIA analyst Frank Watanabe warns that total agreement among decision-makers during the analysis of the opponent’s intentions, especially during a crisis, ‘may be an indicator that something in the estimate is probably wrong’.38 In contrast, the poly-centrist approach seeks to prevent the emergence of groupthink and favours rigorous cross-bureaucratic communication and consultations among various ranks, including direct competition for the most accurate and precise intelligence assessment.39
Chapter II identifies that the patterns of consultation within the Greek administration in the 1980s, were mono-centrist and subject to high degree of groupthink under the influence of Prime Minister Andreas Papandreou. Moreover, the Greek legal framework of the intelligence services in the 1970s and 1980s contributed significantly to the mono-centrist approach. A close study of published memoirs and interviews with key officials willing to speak can reveal the consultation patterns within the top echelons of the Greek government in the 1980s. Regarding the mid 1990s, Chapter V presents the consultations of Prime Minister Costas Simitis with his secretaries and the chief of the National Defence General Staff during the January 1996 Greek-Turkish crisis.
Intelligence Relating to Intentions during a Crisis
During the period preceding a crisis, intelligence agencies watch for indicators that may show a change in the behaviour of their perceived opponent. A change in attitude may indicate to the intelligence community that national security and the interests of the state, as they are perceived and defined, may be jeopardized by the actions or decisions of the opponent state. The focus should remain on identifying the significance of every indicator, or of the accumulation of these indicators and their impact on short-term and medium-term bilateral relations. Professor Kam argues that the enemy produces indicators through his behaviour. Intelligence analysts seek to identify the most useful threat assessment indicators (e.g. demarches of diplomats, troop mobilisation and deployments, new high ranking military appointments and ultra-nationalist remarks by politicians.) The challenging issue that arises is whether the analysts, in this case the Greek intelligence community, are focusing on the correct indicators for the situation at hand.40
Information on the real political and military intentions of the opponent in a rapidly changing situation constitutes the core requirement of intelligence in a crisis. Ben-Zvi makes a distinction between ‘basic’ and ‘immediate’ intentions. Basic intentions have a long-term, strategic essence and are assessed regularly by various sources (i.e. diplomatic, military). There is an anticipation of the opponent’s style of behaviour in international organisations, in military exercises and in diplomatic relations with allies and adversaries. However ‘basic’ intentions may be difficult to define because deception and secrecy as well as flaws in analysis (i.e. mirror imaging, ethnic stereotypes, trend analysis on historical events and the prediction of an opponent’s behaviour in the future), may interfere. In contrast, ‘immediate’ intentions refer to the perception of how the enemy may behave and how he might implement his policies in the tactical environment of a crisis. Analysts formulate short-term military and diplomatic-political expectations about the opponent’s decisions and observe the opponent’s deployment of capabilities and the armed forces operational status. However uncertainty remains a factor to be counted in all assessments.41 As discussed in Chapter V, during the Greek-Turkish crisis in 1996, Ankara challenged the sovereignty of two islets within Greek territorial waters. In Athens, the Ministry of Foreign Affairs and the MoD considered this to be the first time Turkey had undertaken such an initiative since the signing of the Lausanne Treaty in 1923. As a result, the realisation that ‘something changed’ in the behaviour of a ‘revisionist’ Turkey considerably influenced the contemporary Greek assessment of the crisis.
In addition, warning is divided into three main categories: political, strategic and tactical. Political warning refers to the history of bilateral tensions that officials assume could lead to a crisis at any moment. A crisis may be the pretext or the cause of resorting to arms. As Professor Richard Betts argues, ‘strategic warning covers issues of developing military assets according to a preconceived attack plan of the opponent state’.42 A warning requires a prompt analysis and a decision that ultimately leads to action. This action creates indicators that are picked up by the opposite side, thus initiating the ‘cycle of warning’ process (i.e. warning-analysis-decision-action). The opponent’s reaction creates another warning which begins the process again. During international crises, this two-party process takes place within days, hours or minutes. The interaction of the cycles of warning directly affects the assessments of both opponents. Intelligence agencies must establish if the opponent has optioned for a ‘decision stairway’. This process shows that the other party is seriously considering ‘all options’ (i.e. resorting to the use of force) and making operational and political contingency plans. Thomas Belden, a CIA analyst, talks of the decision stairway being a military-political-intelligence process that is affected by the other party’s initiatives and decisions.43
As shown through the analysis of the research undertaken, the concept of cycles of warning is directly relevant to the way Greek reactions were shaped during the crises in 1987 and 1996. With regards to Turkish intentions, as examined in Chapter IV, Ankara did not intend to initiate military action against Greece during the crisis of 1987. The deployment of Turkish forces was limited and not positioned offensively. Moreover, during the crisis in January 1996, Ankara only mobilised naval units around the disputed area and decided against involving the Turkish land forces (with the exception of a small commando team), thus indicating an intention to manage the crisis rather than implement plans for war.
Professor Betts argues that intelligence failures in a crisis are due to a total absence of warning, very late warning, or a lack of appreciation for the information provided to the decision-makers. The warning is either ‘factual-technical’ or ‘contingent-political’. The factual-technical warning is based on the intelligence gathered regarding the enemy’s capabilities in a given area (e.g. military movements on mainland Turkey, naval movements in the Aegean international waters and Greek-Turkish air dogfights). The contingent-political warning depends on the analysis of the information gathered and on the hypothetical construction of the opponent’s intentions.44
In his study of the role of intelligence in a crisis, Thomas Belden offers four general principles of what he calls ‘probability statements’. Firstly, one should bear in mind that very precise predictions (of dates and hours) have a low probability. One cannot be sure that at X time, on Z date an offensive/provocative action will take place. Secondly, the greater the number of estimated data/information elements (e.g. movement of ships, troops, politicians’ statements within a short period of time and the influence of international
