The uses and utility of intelligence: the case of the British Government during the War of the Spanish Succession

ABSTRACT It is usually taken for granted that intelligence organisations provide information for decision-making and that the knowledge produced in the process is therefore deeply utilitarian. Drawing on organisational sociology, this article draws on a case study of English intelligence efforts during the War of the Spanish Succession (1701–1714) to reflect critically on the assumed direct relationship between intelligence-gathering and political decision-making. In eighteenth-century England, intelligence frequently fulfilled other, often more symbolic functions, for example when access to intelligence was employed to legitimise individual actors. In this sense, intelligence was doubtlessly useful, albeit in other ways than generally postulated by intelligence theory. These observations strongly suggest a ‘missing dimension’ in the history of intelligence in other periods as well as intelligence theory more generally.


Introduction
It appears self-evident that intelligence serves a purpose. Why else would political actors and governments invest so much time, money, and energy into both the clandestine and open acquisition of information? But what exactly is this purpose? The most common answer is that intelligence and espionage produce 'actionable knowledge or decisionmaking information'. In other words, they serve the preparation of political and military action or, to be more precise, the decision-making underlying such action. In this logic, which underpins much of the writing on the history of intelligence, information is used to prepare, safeguard, and implement decisions. 1 These assumptions rest on the essential premise that information gathered by intelligence and espionage is actually used -and used for a specific purpose, at that. Looking closely at historical as well as present instances of espionage one often cannot help wondering whether this is indeed the case. First impressions not infrequently suggest the exact opposite: intelligence services appear to amass infinite amounts of data which they do not use and often even lack the ability to exploit. Against this background, it is worthwhile asking why political actors so tirelessly engaged in information-gathering and whether the link between espionage and decision-making is the only and indeed the most plausible reason for such activity.
Moreover, it is helpful to take another step back. Rather than beginning by asking about the utility and functions of espionage, it is necessary to investigate the concrete practices connected to it. After all, decision-making is preceded by other activities, namely the collection, analysis, and interpretation of intelligence, while the particular decision-making process which they are intended to support may not even materialise itself. Whether and in what manner these various activities are practised has a profound impact on the connection between decision-making and information obtained by espionage.
Building on my recent monograph on English and (after the Union of England and Scotland in 1707) British espionage during the War of the Spanish Succession (1701-1714), which was the last and most extensive of the wars waged by King Louis XIV of France, the present article investigates the wider theoretical issues of the utility, usefulness, and uses of intelligence in the early modern period as well as the knowledge it produced. 2 During the fighting in such scattered theatres as Spain, Italy, the southern Netherlands, Bavaria, and even overseas, the French king faced a broad coalition of different European powers led by England, the Netherlands, and the Habsburg emperor. 3 In this conflict, as I have shown, the English government, above all the secretaries of state, and the commander of the English troops on the continent, John Churchill, first duke of Marlborough (1650-1722), made significant efforts to ensure effective intelligence operations. Abstracting from the findings discussed in my book, I argue that even though contemporaries were convinced that they required and made use of such information in a purely instrumental fashion for the purposes of decision-making, the historical record reveals a number of other usages and functions. Rather than presenting an empirically satiated case study, it is my aim here to focus on the general characteristics of the English government's use of intelligence derived from the close reading of a large number of sources.

Contemporary views on the utility of intelligence
That intelligence was useful for action was considered axiomatic (not only) by the English government at the turn of the eighteenth century, even if it was rarely made explicit, probably precisely because it seemed trivial to the actors in question. Daniel Defoe, the English novelist, journalist, author of Robinson Crusoe (c. 1660-1731), and a spy himself, was unusually direct when he wrote that espionage was particularly useful, if the enemy was at the same time prevented from spying: As intelligence abroad is so considerable, it follows in proporcion that the most usefull thing at home is secrecy; for, as intelligence is the most usefull to us, so keeping our enemyes from intelligence among us is as valluable a head. 4 On the whole, reflections about how and why intelligence was useful were undertaken by those who organised espionage rather than by politicians. One important spymaster working for the English government, for example, wrote that it was 'better to send superfluous news than to neglect necessary information'. 5 Another voiced a similar opinion. Although there was 'news which is regarded as fanciful and to which no attention is paid', he concluded that ignoring it was wrong. For 'in good politics, one must not neglect anything but instead must employ all means available'. 6 Of course, contemporaries would have been well aware that it was impossible to make such usage de tout and that the English government had no intention of doing so.
While letters written by members of the English government frequently draw attention to the usefulness of specific pieces of information obtained, they usually employ the subjunctive, as when Marlborough's secretary instructed a spymaster to report to Marlborough everything 'which you believe could be of any use to him' and to continue intercepting particularly revealing correspondence by the enemy 'which may be of great use'. In its ambivalence, this use of the subjunctive is symptomatic, firstly, of the belief in the utility of information and, secondly, of the generic and unspecific nature of this belief. Expressing as yet unfulfilled expectations, the phrase refers to the potential usefulness of information rather than to any concrete contexts in which that utility manifests itself. In similar fashion, a prospective spy emphasised the usefulness of the information which he promised to deliver to the government and a payment made to another spy was justified with the usefulness of his service in this capacity. 7 It is clear that, in all these cases, actors employed a notion of instrumental utility. Whenever a claim to utilité is made and whenever a piece of information is characterised as 'useful', they meant to say that the intelligence in question possessed strategic or tactical relevance. 8 Retrospective appraisals of the actual utility of individual pieces of information in these contexts, however, are much rarer. 9 More often than not, the claim that espionage was useful remained vague and general.

Analysis and consumption of intelligence in early Eighteenth-Century England
In order to better understand how English politicians dealt with intelligence, we need to take a closer look at how the English government organised its analysis and dissemination. For this purpose, it is necessary to analytically separate the different steps or stages of which a complex phenomenon like 'intelligence use' is composed and discuss the constitutive elements in turn. From the point of view of a history of knowledge of espionage, Michael Kempe has identified four epistemic operations which closely resemble the model of the intelligence cycle while placing greater emphasis on the epistemological aspects of intelligence: the collection of information, analysis (corresponding to the stage of processing and exploitation in the intelligence cycle), evaluation (analysis and production), and, finally, consumption by decision-makers (dissemination and consumption). 10 Although an examination of intelligence collection on behalf of the English government would exceed the scope of this article, some remarks about analysis (that is, the process of assembling factual knowledge from a variety of sources of information), evaluation, and consumption are in order. 11 These are surprisingly sobering when one is interested in the uses and utility of intelligence. Most importantly, such an investigation makes clear that the operations outlined above often cannot be clearly distinguished from each other and are conflated in practice. Before information could be analysed and evaluated, someone obviously had to read the spy reports. But were they actually read? The reports submitted to the secretaries of state are surprisingly devoid of traces of active readership. Even though one can find the occasional underlining or note in the margins, the absence of marginalia from the majority of such reports and letters suggests that they were ignored as often as they were read.
There was one subject, however, which received intensive interest from members of the English government: the danger of a joint French-Jacobite invasion and, related to this, the size, provisions, and armament of the French navy. The Jacobites were a political movement in England and Scotland who continued to be loyal to the Stuart dynasty which had been displaced from the throne during the Glorious Revolution of 1688. Having established themselves in France in 1689, the Stuarts several times attempted their restoration with support from the French crown, which was less interested in the fate of the old dynasty than in instrumentalising the Stuarts' claims to pursue its own objectives. 12 Both the exile court and the latter's recognition by Louis XIV during the War of the Spanish Succession were a provocation for the English crown and Jacobitism became the 'major issue in British politics in the early eighteenth century'. 13 The Act of Union which created the United Kingdom in 1707 was at least partially meant as a response to the Jacobite challenge in an attempt to tie Scotland, which was potentially more supportive of the Stuarts, closer to England. 14 More than once, Stuart supporters backed by France tried to invade Britain. During the War of the Spanish Succession, the most notable invasion attempt aimed at Scotland and was undertaken in the spring of 1708. 15 English fears of a Jacobite invasion prompted the surveillance of French ports along the Channel as potential bases for such an endeavour. 16 Spy reports relating to this issue were annotated most heavily, revealing that their readers attached the greatest importance to them. 17 On the eve of the Jacobite invasion attempt of 1708, one spymaster himself sought to direct the readers' attention to core statements and wrote short summaries of the reports for this purpose -which is quite unusual. 18 In fact, it generally was the spymasters rather than the government in London who added comments such as 'this deserves attention' (ce sy merite attention) to spy reports in order to influence their reception by the English government. 19 Consequently, the observation of French ports and the activities of the French navy is the only context in which the secretaries forwarded intelligence reports to the Commission of the Admiralty, even if they did so at irregular intervals. 20 The latter reciprocated, providing us with an example of a routine exchange of information between different branches of government which, nevertheless, was far from normal procedure. 21 This examination of the evidence shows that, most of the time, the English government's active reading interest fell short of the urgency with which it demanded its agents to send as much information as possible.
The second step, the analysis and evaluation of the information gathered, is likewise complicated. For how does one evaluate information, especially when it comes to the most fundamental of all questions: is the information correct?
In the early modern period, politicians had very few means for generating reliable knowledge from the multitude of incoming information. They were furthermore confronted with rumours as well as half-knowledge and were perfectly aware of it. 22 Especially in times of war there was usually no time to carefully evaluate intelligencefor instance by requesting second, third, or fourth eyewitness reports. Due to the war, the English government's information-gathering activities were carried out with great hurry. Everything was considered very important and urgent all the time. This is why historical actors at once were in possession of both too much information, which they could not adequately process, and too little, because they lacked the information needed to effectively evaluate other pieces of information. Seemingly paradoxically, therefore, the great amount of intelligence did not prevent actors from constantly complaining that they had too little of it.
In addition, there were no formal, institutionalised mechanisms of evaluation and analysis, let alone an organisation which would have centralised these activities. Instead, the government typically acted in an ad-hoc fashion, resulting in a low degree of systematisation in the realm of processing and evaluating information. The latter remained haphazard and vested in individual actors who often evaluated incoming intelligence individually, more rarely collaboratively, on the basis of their own sources. 23 This lack of institutionalization of intelligence in the early eighteenth century is surprising, given that actors within the English governments of the sixteenth and especially the seventeenth centuries paid particular attention to organising intelligence and espionage.-24 However, historians have also emphasized the discontinuities between the reigns of different monarchs until at least the 1660s. 25 Such discontinuities as well as the low degree of institutionalisation of English intelligence seem to be the main reason why studies have generally focused on individual actors who were particularly active in this field. 26 Even though a tendency towards greater professionalisation emerged in the second half of the seventeenth century, as Alan Marshall has shown, around 1700 the activities of the English government were still relatively unsystematic, especially as far as the evaluation of information was concerned.
Such basic evaluation rested mainly on, firstly, corroborating information against other sources and, secondly, on an appraisal of the informant in question. 27 This was supplemented, whenever possible, with letters intercepted from the enemy. The latter were valued highly because they appeared to be, and often were, authentic and accordingly had the advantage that, in a situation in which the ability to verify information was limited, they did not require further verification as far as their content was concerned. Precisely because of this assumed epistemic status, intercepted letters were used to corroborate and confirm information obtained from other sources. 28 Printed newspapers were employed in the same way and for the same end. 29 Corroboration of information from different sources was particularly important whenever intelligence was required from more distant theatres of war, especially the Iberian Peninsula. For the latter, the English government relied not only on spies, but also drew on correspondents based in France. 30 Comparable procedures are evident in the context of the Italian theatre as well. 31 Given the difficulties inherent in verifying a particular piece of information as well as the subjective impression of being pressed for time, actors relied on judging the reliability of informants instead of evaluating the reliability of the individual pieces of information supplied by them. When new informants were recruited, it was considered important, therefore, that they were embedded in existing structures such as spy networks which were already known to the government. A given piece of information was most likely to be trusted, if the informant who had provided it was part of a structure which had proven trustworthy in the past. Thus, if an informant had been successful once, his services were usually sought again. 32 Against the background of these general patterns, occasional instances of a more thorough and methodological evaluation of information are particularly striking. For the purpose of this article, an episode from the context of the 1708 Jacobite invasion attempt is particularly interesting, since the subsequently successful prevention of the invasion suggests that information-gathering and the decision-making based on it were efficient. It is hardly surprising, of course, that the Jacobites, as the key issue in English politics in the first decade of the eighteenth century, prompted initiatives to establish systematic procedures for the evaluation of information, even if these were highly charged with party political considerations. It is characteristic, however, that such organised evaluation took place after the fact -that is, not during or soon after information had become available, but considerably later in the context of a Parliamentary inquiry, when those involved had a much better overview of the matter and were able to draw on a much wider range of sources. Even then, the central question was not whether decision-makers had taken the correct course of action to prevent the invasion, but whether they had done so early enough. For the participants in this debate, the question of when certain pieces of information had been available, took centre stage.
The inquiry opened in December 1708, when the House of Lords -and simultaneously the House of Commons -requested detailed information about the exact time in the spring of that same year at which the government had learned about the arms buildup at Dunkirk. The purpose of this investigation was not only to examine the government's conduct, but, more importantly, to ensure that Great Britain would be better 28 34 The notes on the cabinet meeting of 6 January 1708/9 taken by the Secretary of State for the Southern Department Charles Spencer, third earl of Sunderland (1675-1722), reveal that he was instructed to submit his papers. In addition, the admiralty was asked to compile reports of 'all the intelligences that were given of the design'd invasion' as well as 'copies of all the orders that were given from the first news of the design'd invasion'. 35 The extensive papers provided by several departments were submitted to the scrutiny of both Houses in February 1709, which, at the beginning of March, after considerable debate, concluded that the government had acted correctly. 36 This conclusion, however, was less the result of a neutral investigation than of party politics.
In the first decade of the eighteenth century, Parliament repeatedly attempted to examine governmental decisions, particularly in situations of crisis. 37 However, only during the inquiry into the Jacobite invasion did the question of when which pieces of information had been available play a central role. In this particular case, evaluating decisions meant evaluating information. In spite of this, the inquiry did not give rise to debates about the necessity for a general reform of the mechanisms for gathering and exploiting intelligence.
Evaluating information, as a precondition of its use for making concrete political decisions, was difficult and therefore only rarely undertaken properly. However, such conclusions about a historical deficit, drawn as they are from an extreme example, are based on the assumption that all incoming intelligence was invariably meant to inform decision-making. But was information really always about making decisions? And were all decisions made on the basis of intelligence? In fact, the performance of intelligence evaluation appears in a radically different light, if we abandon such misconceptions.

English decision-making in the War of the Spanish Succession
Especially for English foreign policy until 1750, the nature of the sources makes it virtually impossible to do more than speculate about how exactly decisions were made. 38 The same is true of key military decisions taken by Marlborough such as his resolution to march to the Danube in the spring of 1704 without consulting his Dutch allies. In a 33  It is impossible to chart the process of decisionmaking which led to this extraordinary campaign, as there simply are no sources to do more than, at best, reconstruct scattered plans and agreements. 39 Even if it is occasionally possible to see the English government or some of its members attempting to determine the strategy for an entire year of campaigning, it is far more rare to find evidence of how these plans were implemented at an operational level. 40 Despite the cabinet's role as the locus of foreign policy-making, actual decision-making in the early eighteenth century took place at various other levels of the English government, which was characterized by a relatively decentralised organisational structure. 41 Although foreign policy remained a royal prerogative even after the Revolution of 1688, Parliamentary debates during the reign of Queen Anne (1702-1714) increasingly concerned questions of foreign policy. During this period, party rivalries became a hallmark of English politics, especially when compared to England's neighbours on the continent. The royal court, in contrast, lost much of its earlier importance. 42 At the same time, the early eighteenth century witnessed the emergence of a de facto ministerial hierarchy. Although the term prime minister was regarded as dubious, Sidney Godolphin and his successor Robert Harley were the first 'first ministers'. 43 The fact that the Lord Treasurer took pride of place in this way unequivocally attests to the enormous importance of finance for state formation and warfare. Moreover, with the exception of the secretaries of state, the treasurer was the only minister to officially have direct access to the queen. 44 Over the course of Anne's reign, the treasurer thus became the leading figure in domestic politics, while the secretaries of state dominated foreign relations. The latter were at the centre of English information-gathering and organised intelligence.
In Anne's reign, the power of government lay with the cabinet council, a body which had emerged from the increasingly less significant privy council. 45 Communicated as the queen's decisions, the cabinet resolutions were regarded as binding. Its main business concerned foreign policy. Here the correspondence with the queen's ambassadors was read out and analysed, foreign diplomats were interviewed, and committees on specific issues were formed. 46 At first glance, the notes of cabinet meetings taken by the secretaries of state Robert Harley (1661-1724) and Sunderland appear to provide a possible avenue for addressing the question of how intelligence played into decision-making in the cabinet. However, these notes are not official minutes but incomplete, irregular, and unofficial records, serving as memory aids for the authors' own purposes, rather than as protocols of the cabinet's meetings and their results. 47 In spite of these obvious shortcomings, they are our only means of catching a glimpse of how the cabinet actually worked. While Sunderland's notes mention intelligence only rarely, 48 Harley's more extensive records frequently show, albeit vaguely, that 'intelligence' was read in the cabinet. 49 The fact that Harley usually speaks of 'French intelligence' and 'French letters' (most likely spy reports such as those produced by Etienne Caillaud's spy company in Rotterdam) suggests that members of cabinet told each other what they had been able to learn. 50 However, there are no records of the discussions which ensued. We do not even know who provided the respective pieces of information. Intelligence, then, is indistinguishable from the large number of letters presented to the cabinet, written, for example, by English envoys and Marlborough or intercepted from French diplomats and officers. 51 These sources therefore show what business was dealt with in cabinet and how positions on certain issues were gradually formed. Yet they do not reveal the links between specific pieces of information and specific decisions. This applies more generally to decision-making outside of cabinet as well. For this reason, it remains an open question how relevant intelligence was in this process. It is generally possible to observe information, to observe actions, and (if much more rarely) to observe decisions. How these were connected, however, usually remains in the dark.
Only in exceptional cases is it possible to trace the path from an individual intelligence report to a decision. The defence against the Jacobite invasion attempt of spring 1708 is one such example. Detailed reconstruction in this case is possible for two reasons: Firstly, the parliamentary inquiry of 1709 mentioned above collected all available information in one place. 52 Secondly and more importantly, unequivocal connections between intelligence and decision-making are evident because once suspicions about the potential threat had arisen, the government intensified its efforts to collect further information on it. 53 Such a targeted search for information, however, was the exception, not the rule.
On the other hand, it is possible to discern plausible connections between specific stocks of information (hardly ever specific pieces of information) and specific decisions. In April 1707, for instance, Marlborough travelled to the camp of the Swedish king Charles XII in Altranstädt in Saxony. There he managed to dissuade Charles from his suspected plan to declare war on the Holy Roman Emperor, as doing so would have coupled the War of the Spanish Succession with the Great Northern War (1700-1721). 54 The English government sought to prevent this at all costs since it would have increased the danger of losing the war against France and its northern European ally Sweden. Although it is still unclear what exact agreement was reached between Marlborough and Charles in Altranstädt, the English government considered the trip as a success 'of as great Consequence to the common cause, as a Battle won.' 55 But why did Marlborough take this unusual step of travelling to Saxony without enjoying official status as a plenipotentiary, even if he did so on the queen's orders? 56 It is very likely that the decisive trigger was provided by the increasingly alarming reports sent by Jean de Robethon in spring 1707. 57 As former private secretary to King William III of England (r. 1689-1702) and later the private secretary as well as spymaster of the Elector of Hanover, the later British king George I (r. 1714-1727), Robethon knew English politics well. Upon the Hanoverian succession, he eventually once again returned to London. 58 Robethon knew Marlborough from his first stay in England and subsequently became one of the duke's most important informants about the Swedish king's activities during the Great Northern War. In other words, had it not been for Robethon's reports and intercepts, Marlborough would most likely have considered the situation much less dramatic than he did. Still, this conclusion does not mean that there is an unambiguous and linear connection between the intelligence provided by Robethon and Marlborough's decision to travel to Charles's camp.
In the majority of cases it is impossible to establish what exactly the English government did with the information it received -or even whether it did anything at all. Whether and, if so, how decisions were based on intelligence is usually impossible to reconstruct from the surviving documentation which, more often than not, remains silent on this issue. In fact, the postulate of the nexus of information and decision like the general idea of an unproblematic 'use' of information lack solid theoretical as well as empirical foundations. 59 Why such lack of clarity about the functions of information and espionage? Is this merely a result of a lack of sources? An alternative explanation may lie in the relatively weak formal organisation of the English government. The latter may have enabled the amassing of huge amounts of information but may not (yet) have made it possible to exploit it adequately. Both explanations, however, are insufficient. It is simply too obvious from the documentary record, e.g. the government papers preserved at the National Archives and the Blenheim Papers in the British library, that often nothing further was done with intelligence after its collection and that it accordingly did not prove relevant for either action or decision.
The commonly assumed link between information and decision-making is clearly not wrong as such, but it is too simple. This means, firstly, that it is necessary to more thoroughly reflect this link from a theoretical as well as an empirical point of view. Perhaps the relationship is more complicated. And secondly, it is necessary to take seriously the possibility that intelligence served purposes other than decision-making. In what follows, I will briefly introduce both arguments.

The nexus of information and decision-making
To understand the nexus of information and decision-making on a theoretical level, one can turn, for example, to neo-institutional theory. The organisational sociologists Martha Feldman and James March have explored the question why the handling of information in modern organisations is not necessarily goal-oriented, instrumental, and efficient. The customary view of the relationship between information and decisions observed by sociologists closely resembles the relationship postulated by intelligence theory: relevant information will be gathered and analyzed prior to decision making; information gathered for use in a decision will be used in making that decision; available information will be examined before more information is requested or gathered; needs for information will be determined prior to requesting information; information that is irrelevant to a decision will not be gathered. 60 As plausible as this model is, the connection between information and decisionmaking empirically tends to be very weak. Feldman and March point out that the information gathered often is unrelated to the issue about which a decision must be made. Moreover, information-gathering continues even when a decision has already been made. And finally, much of the information actually available is not even consulted while decision-makers nevertheless complain about a lack of information.
Feldman and March offer a number of explanations for these observations. To start with, the actors involved in information-gathering are generally not the same as those who evaluate and use the information collected. Moreover, much of the information collected by organisations is acquired not in 'decision mode', but in 'surveillance mode'. 61 This effectively means that organisations do not restrict themselves to acquiring information of immediate relevance for the decision being taken, but also collect information which, in the current context, is irrelevant but assumed to possess potential relevance in an unspecified future. It is obvious, then, that not all information which is gathered serves (or is meant to serve) decision-making. Consequently, much more information than necessary is being amassed in the belief that such information will enable solving as yet unknown problems in an equally unknown future. The purposes of such activities, then, are the attempt to gain an overview, precaution, and the minimisation of uncertainty. Precisely because the future is uncertain, this attempt at foresight in 'surveillance mode' can easily lead to the fetishization of information and thus the danger that the collection of information transforms into a goal in its own right which becomes increasingly independent of the question whether the information thus acquired will ever be of use to anyone.
This brings us to the symbolical function of information. Since a cornerstone of the self-fashioning of organisational actors, according to Feldman and March, is the assumption of the role of the decision-maker who is in possession of the best sources of information and able to use these intelligently and efficiently in the pursuit of the organisation's goals, information consequently does not so much supply a basis for action as it functions as a 'representation of competence and a reaffirmation of social virtue'. 62 Feldman and March thus show that, within organisations, information possesses other functions alongside its instrumental value for decision-making: it provides insurance as well as security but also has a symbolic function in the context of negotiating status.
Of course, the early eighteenth-century English government hardly fits the criteria of a modern organisation. To what extent, then, do the findings collected by the neo-institutionalists from the study of modern organisations apply to the constellations examined in this article? As I have shown above, espionage and the information it produced at times did indeed impact decision-making. Nevertheless, it is fair to assume that the English government was less concerned with specifics than with a general sense of security and access to information.
In fact, contemporaries already described intelligence as the 'mother of prevention'. 63 Consequently, the information acquired by the English government primarily contributed to routine efforts to minimise uncertainty and when it urged the extensive collection of intelligence, it operated in 'surveillance' rather than 'decision mode'. Information was required in order to obtain an overview of what was going on and, in the process, to reduce the uncertainty about what enemy powers were doing or planning to do. It was important to stay up to date. 64 In this context, information-gathering had no immediate end but was carried out tentatively and with a view towards future possibilities. 65 What was sought, ultimately, was 'knowledge which may not have been of any immediate value but may have been useful for decision-making at some future point.' 66 As a rule, incoming information triggered no actions; it was simply registered. Nor was it supposed to give rise to activity. Information thus was of the greatest use if there was no need to use it. Information nonetheless had considerable usefulness -for those who used it to present themselves as competent and legitimate actors.

Information fetishism and the representation of competence
tpIt is often difficult to escape the impression that the English government developed a sort of information fetishism which turned information-gathering into an end in itself and lost sight of the potential exploitability of the information collected. As discussed above, organisational sociology has tried to explain the phenomenon of quasi-ostentatious information collection differently by highlighting the secondary functions of information: information has immediate relevance not merely for decision-making and action, but, within organisational structures, is also 'a representation of competence'. 67 In this sense, information is the most important resource for distinguishing decisions as well as decision-makers as competent, rational, and efficient. Information thus becomes an argument for legitimisation which is largely independent of its usefulness for decision-making and its concrete uses for this end.
Such arguments, however, were rarely made explicit in the context of the eighteenth-century English government. Like the attempts to systematically evaluate information discussed above, offensive claims to legitimacy on the basis of access to information were rare. This is at least partly explained by the fact that the aristocratic and administrative self-image of the actors in question with its emphasis on questions of personal status rather than bureaucratic efficiency -a complex much better studied for the seventeenth century -evidently did not suggest using information in this way as an obvious option. 68 Perhaps this observation supports the conclusion that such legitimising strategies are more prevalent in precisely those institutions with a high degree of organisational sophistication studied by organisational sociologists in our own time, while the English government in this period, on the whole, was characterised by a relatively loose structure. But perhaps the predominant values of the day simply meant that such arguments remained implicit. After all, the common practice of keeping one's sources secret even from close colleagues not only protected informants from discovery by the enemy, it also kept political rivals as well as supporters at home in the dark about just how extensive and capable -or limited and incompetent -one's intelligence network actually was.
What evidence there is for the strategic use of information to demonstrate competence is usually indirect. In fact, Robert Harley who, as secretary of state had a particularly noticeable interest in information-gathering and therefore co-operated especially closely with Marlborough for this purpose, downplayed his involvement in intelligence. When he was dismissed as secretary of state in 1708, he wrote to his friend and member of the privy council John Holles, the duke of Newcastle upon Tyne (1662-1711): 'I am master of no news or intelligence and take no more pleasure in the schemes and projects which are every day new, than in hearing the dreams of a sick man.' 69 Yet Harley enjoyed a reputation as a man who loved information for its own sake and whose thirst for it was insatiable. 70 One of the most important statements concerning the use of information as a legitimising representation of competence therefore flowed not from Harley's pen, but from that of one of his opponents. At the end of 1710, Francis Hare (1671-1740), Marlborough's military chaplain and publicist, wrote a lengthy letter to Marlborough's wife Sarah in which he commented on the present state of English politics and its most important actors. In Hare's view, Harley had been a poor head of department during his tenure as secretary of state because he had wasted too much energy on building up his own spy networks in France. 71 Hare remarked: And what appeared of his intelligence with France is much to his disadvantage; it shewing an ill-managed and expensive affectation to appear considerable in his office, with very little or no advantage to the service. 72 In other words, Harley's intelligence activities had been useless, expensive, and ineffective. Yet they had been intended to make him appear more 'considerable', a remark which demonstrates that contemporaries indeed had access to an interpretative model which linked information and competence, even if they rarely made use of it.
Only once did Marlborough himself take recourse to intelligence in order to legitimise his actions and present himself as competent. But that he did so in the context of his downfall makes it all the more noteworthy. At no other time in Marlborough's career during the War of the Spanish Succession did information play as important a role in presenting him as an efficient actor. At the same time, the duke's self-projection was intended to justify his behaviour. What was at stake, however, were not specific pieces of information, but rather the use of espionage as an essential element of warfare in general.
In the autumn of 1711, Marlborough was accused of having embezzled public funds and on 30 December 1711 he was dismissed from all offices by Queen Anne. 73 From the outset, Marlborough's defence focused on the claim that he had used the money first and foremost for information-gathering, that is, for 'keeping secret correspondence, and getting intelligence of the enemies motions and designs'. 74 He accordingly emphasised the central importance of espionage for the conduct of war: Of course, Marlborough did not reveal any details about at what moment and in which situation exactly espionage had proven useful. Instead, he pointed to his military successes more generally and cited them as proof of effective, and thus also expensive, espionage on a grand scale: the many Successful Actions, such as have surpassed our own Hopes, or the Apprehensions of the Enemy in this present War in Flanders, to which our constant good Intelligence has greatly contributed, must convince every Gentleman, that such Advices have been obtained, and consequently that this Money has been rightly applied. 76 An open letter written in support of Marlborough by several European monarchs (who sought to prevent England from pulling out of the war) even went so far to claim that 'the prudent and wise management of these funds, in addition to God's blessing, has chiefly contributed to winning so many glorious victories.' 77 A treatise printed in London in Marlborough's defence meanwhile painted a broad picture of the usefulness of espionage: Can it be imagin'd he has not had his Spies in every Place, a whole Army of them in Constant Pay to shew him where to find the Enemy unguarded, where to guard himself against him? Must he not have often drawn his Intelligence from the Fountain's Head, from the first Counsells of the Enemy to be able to countermine all his Projects? And cou'd all this be done without Expence? 78 In the eyes of Marlborough and his defenders, the fact that there were no receipts and accounts of expenditure for the funds in question was evidence that they must have been spent on espionage. After all, as Marlborough himself declared, intelligence does not permit transparent book-keeping: 'The Nature of the Thing [is] not admitting of a particular Account'. 79 And the publications of his defenders further explained that the practices of secret information-gathering 'never were or can be accounted for, without destroying the End and Use of secret Service'. 80 Marlborough and his supporters thus effectively responded to a hopeless situation by emphasising the value of espionage. They attributed the duke's military successes to effective information-gathering and supported this claim with reference to the secrecy of espionage which, by its very nature, had to remain undocumented. Why did Marlborough's defence draw so heavily on the subject of intelligence? Citing espionage in this way offered itself as a useful strategy, firstly, because such activities could be used to illustrate one's prudent, efficient, and successful actions and, secondly, because activities in this sphere could plausibly be presented as inherently non-transparent. In all probability, the claim that the funds of whose embezzlement Marlborough was accused had been spent on intelligence was simply the line of argument which was most difficult to disprove -even if this defence was ultimately unable to save Marlborough. 76 Ibid., 9. 77 As an example of a multitude of similarly worded letters, see the letter of the Elector Palatine Johann Wilhelm II, BL Add. 61,326, fol. 183 r: 'la prudente et sage administration de ces deniers a principalement contribué, après la benediction de Dieu au gain de tant des Victoires glorieuses'. 78

Conclusion
In the preceding pages, I have shown that although contemporary actors were often convinced that espionage was of fundamental importance and useful for political and military decision-making, in the majority of cases they made use of it in a much less specific manner in order to obtain a general overview of the political situation. In the first instance, intelligence served the aim of avoiding error as a 'mother of prevention'. Moreover, actors used intelligence in certain cases to present themselves as competent and legitimate actors. This symbolic and legitimising function of espionage is clearly different from its immediate use in decision-making. And yet, even in this context, intelligence at times proved useful -even if it did so in a way which is very different from what contemporaries as well as historians of intelligence usually expect.
What are the implications for intelligence studies and the history of intelligence? In my opinion, we must firstly take seriously the possibility that intelligence is not alwaysand perhaps even only rarely -used for decision-making, even in our own time. Secondly, theories of intelligence would do well to account for the fact that the knowledge it produces is only partially intended to support immediate decision-making needs. Instead, it most often serves the purposes of minimising risk and uncertainty and gaining an overview of the situation in question. 81 This also implies that the stock model of the intelligence cycle with its strong focus on the action-oriented utilisation of intelligence is insufficient to encompass the empirically observed complexities of information use. 82 Finally, historical findings such as those presented here encourage us to pay closer attention than before to the similarities as well as the differences between premodern and modern intelligence and its uses.