Long cycle theory and the hegemonic powers' basing networks

Political Geography 18 (1999) 941–972 www.elsevier.com/locate/polgeo Long cycle theory and the hegemonic powers’ basing networks Robert E. Harkavy * Department of Political Science, The Pennsylvania State University, University Park, PA 16802, USA Abstract One recent focus of research in international relations theory is that of “long cycle theory,” associated primarily with George Modelski and William Thompson, which posits serial cycles of hegemonic dominance — Venice, Portugal, the Netherlands, Great Britain, the United States — lasting approximately for one century. These hegemonic cycles are highly correlated with, or underpinned by, maritime and commercial dominance. Some aspects of long cycle theory have been contested by the rival “world systems” theory, that has fewer cycles and a disinclination to separate the military and economic dimensions of hegemony. Heretofore, naval power, as reflected in capital ship construction and orders of battle, has been used to measure maritime dominance. This research suggests that data for rival and successive global basing access networks could be used to inform and query the basis of long cycle theory; i.e., to provide a measure of “global reach”. There are additionally, interesting conceptual questions about the basis for basing access, as it has evolved historically; specifically, from a basis in conquest to one dependent upon diplomacy and various quid pro quo. The article suggests the need for more historical data collection on bases.  1999 Elsevier Science Ltd. All rights reserved. Keywords: Long cycles; Bases; Hegemony Introduction Recent years have seen the development of long cycle theory as one of the primary foci of efforts toward “grand theory” in international relations. Stated in summary form by its most well known proponents, George Modelski and William Thompson, “the long cycle of global politics refers to the process of fluctuations in concentration * Tel.: +1-814-865-7515; fax: +1-814-863-8975. 0962-6298/99/$ - see front matter  1999 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 2 - 6 2 9 8 ( 9 9 ) 0 0 0 3 3 - 5 942 R.E. Harkavy / Political Geography 18 (1999) 941–972 of global reach capabilities which provide one foundation for world leadership” (Modelski & Thompson, 1988: 97; see also Modelski 1978, 1987). Crucially, it deals with the serial (and approximately 100 years in length) periods of maritime and commercial dominance by Venice, Portugal, the Netherlands, Great Britain (two century-long periods straddling the intervening Napoleonic Wars), and the United States. In relation to these long cycles, the theory further identifies those “phases of global war (1494–1516, 1580–1608, 1688–1713, 1792–1815, and 1914–1945) for it is these system-transforming bouts of warfare that constitute the watershed phases of the hypothesized long cycle process”. Further, according to Modelski and Thompson, “it is during the global war phase that the system’s tendency toward concentrationdeconcentration essentially switches from deconcentration to a fairly high level of concentration for some finite post-war period of time” (Modelski & Thompson, 1988: 97). As it happens, the extant corpus of long cycle theory raises a few interesting issues in relation to International Relations (IR) theory and the state of the discipline which we may only briefly touch upon here. It is closely related to, and in some crucial ways rivalrous to, “world system theory” associated with Immanuel Wallerstein, et al., which focuses on the origins and dynamics of the modern global capitalist economy and worldwide uneven development, that is, a capitalist world system that has continually contained a core, periphery and semi-periphery.1 The shifting serial hegemony, or what otherwise has been referred to by Thompson (1997) as “system leader lineage” forms the core of capitalist domination. Interestingly, too, long cycle theory, as noted by Rosecrance (1987), constitutes one example of a cyclical theory of IR, by contrast with linear ones as represented, for instance, by “realism”, or millennial constructs such as Marxism or perhaps “endism” in its recent manifestations. Also interesting, in relation to structural theories of IR, is the seeming implication of long-cycle theory that the international system tends constantly towards unipolarity, or to an alternation between unipolarity (during concentration phases of long cycles) and bipolarity, the latter during periods of hegemonic warfare. This contrasts with the widely held assertion by some IR theorists that the system has tended to alternate between bipolar and multipolar periods (Hopf, 1991). One other interesting relationship appears to have emerged with long cycle theory, and that is the extent to which it forces attention to the connection between the hitherto almost wholly separate worlds of IR theory and military history; specifically, the history of maritime rivalries and naval warfare. An analyst looking to flesh out the historical details feeding into long cycle theory will necessarily find himself immersed in the magisterial works of Boxer (1965, 1969) and Parry (1967) on the Portuguese, Spanish, and Dutch seaborne empires; in the exemplary two-volume series by Padfield (1979, 1982) on maritime and commercial rivalries in the 15th to 18th centuries, and the historical analyses of naval warfare and maritime supremacy by Mahan (1980), Graham (1965) and Rosinski (1957) and more latterly, Gray 1 Summarized for the uninitiated in Viotti and Kauppi (1993: pp. 449–470). See also Wallerstein (1979). R.E. Harkavy / Political Geography 18 (1999) 941–972 943 (1992). Those analyses are juxtaposed to those more focused on, but related to, the economic and technological roots of long cycles, particularly, the historical progression of Kondratieff waves, as analyzed by Goldstein (1985) and Thompson (1997). Though it is beyond the scope of this paper, it is worth remarking that whereas some analysts see long cycles driven primarily by domestic economic factors and technological innovation, historians such as Padfield (1979, 1982) and Boxer (1965, 1969), wholly innocent of the theoretical baggage of long cycle theory, see the successive periods of maritime supremacy driven only in part by K-waves (or some looser interpretations of economic dynamism), but also by envy, piracy, and idiosyncratic geographical factors related to maritime dominance.2 Padfield, indeed, wholly absent from the vocabulary of modern IR theory, portrays a lengthy skein of balance of power combinations and recombinations, or at the least, of the rise and fall of global maritime empires. Finally, it is to be noted that some serious general criticisms have been leveled at long cycle theory, some of them summarized in a review article by Rosecrance (1987) more than a decade ago. The almost too-coincidental nature of century-long cycles (literally corresponding to the actual centuries) provides an almost bizarre twist. The omission of the major land wars of the 17th century in Europe is also raised, as is the question of whether periods of dominance or aspirations of same by major land powers in Europe — France under Louis XIV, Germany in the twentieth century, the USSR during the Cold War — does or doesn’t chip away at long cycle theory’s thesis of continuous, serial hegemony by major seapowers. This problem is finessed somewhat by Modelski and Thompson (1988), who include these powers within their definition of “global powers”, albeit short of hegemony. The question of whether Spain was at least co-equal to Portugal as a hegemonic power in the 16th century is also raised not only by Rosecrance (1987), but by devotees of world systems theory such as Taylor (1989). Then too, these devotees of world systems theory criticize what they perceive as long cycle theory’s tendency to a “dual logic” that separates military-security and economic forces and trends (these critics naturally see such matters through neoMarxist lenses), even despite the recent efforts by Thompson and Modelski to focus on Kondratieff waves as economic underpinnings of military, mostly naval, hegemony. Indeed, world systems theory advocates perceive three cycles of hegemony, but not so congruent with actual centuries, broadly parallel to Modelski’s Dutch long cycle, the second British long cycle, and the ongoing U.S. long cycle. They largely omit Portugal despite its extensive naval reach and basing structure. But it is worth 2 Somewhat ironically, the historian Padfield (1979, 1982), while not attuned to K-waves or, generally, the economic bases of capitalist hegemons, raises throughout his works another important political and economic “variable” of sorts, namely, the comparative ability of then major power governments to “extract” money for military purposes involving bank loans, government budgets, etc. (nowadays that problem tends to be encompassed by ratios of military expenditures to GNP). Padfield (1982: 145–151) covers this for Britain and France in the 1690s. For a more generalized analysis of this problem, covering numerous nations in the 20th century, see Millett and Murray (1988). 944 R.E. Harkavy / Political Geography 18 (1999) 941–972 noting that basing structure could as easily be used as a correlative measure of world systems’ theory’s three hegemonic cycles as of Modelski’s five long cycles; indeed, it might be one basis for comparing the comparative validity of the rival theses of long cycles and “cycles of hegemony” a la world systems theory. What then has been central to definitions of (or causes of) hegemony in the long cycle context? In general, that means seapower and commercial dominance, the latter referring to the successive roles of Venice, Lisbon and Antwerp, Amsterdam, London and New York as global financial cores (the now apparent near collapse of the power of Tokyo banks in the 1990s removed for now another possible anomaly in relation to this thesis). Thompson (1997) and Goldstein (1985) stress the relation to K-waves, i.e., technological innovation and dynamism, as Cipolla (1965) did earlier in pointing out the crucial nature of the development of iron production and naval ship and armaments developments in the early phases of the rise of the West. Modelski and Thompson (1988), in their more recent work, dwell upon the central measurement of numbers of capital ships, concluding that long cycle dominance has been associated with a 50% or more share of the world’s largest fighting ships (echoes of Mahan here, and his thesis of indivisible seapower, and the traditional British naval doctrine that insisted upon a navy equal to or better than the next two ranked navies combined). Interestingly, much of the literature on comparative US–Soviet naval power during the latter phases of the Cold War dwelled not just on comparisons of numbers of ships (surface and sub-surface), but on comparative “ship-days” spent by rival fleets in the major oceans and seas, i.e., Indian Ocean, Mediterranean, etc. (see Watson, 1982). Such measures could conceivably be reconstructed for earlier rivalries, i.e., Dutch and Portuguese ship-days in the Indian Ocean year-by-year in the late 16th century. Most fundamentally, long cycle theory is concerned with seapower as one form of global reach. As noted by Modelski and Thompson (1988: 3), the questions of who exercises leadership in global politics and why are perhaps the cardinal questions of world politics, and the concept linking these two questions is seapower. According to them, “only those disposing of superior navies have, in the modern world, staked out a claim to world leadership”. They refer to an “oceanic system” in the modern world, and aver that the advent of the modern world system was at the same time also the onset of use and control of the seas on a global scale, hence the opening of an entirely new age of seapower. Indeed, in ancient Greece, there came to be known the term thalassocracy, which meant “maritime supremacy” or “rule of the sea” (Modelski and Thompson, 1988: 5). Modelski and Thompson trace the development of this concept from the ancient Greeks via Byzantium and Venice, and in connection with periods of Arab and Chinese naval supremacy in the Indian Ocean. But they also conclude that “in West European experience the first to evolve global oceanic concepts were the Portuguese”. Prior to 1500, maritime supremacy played out within the confines of narrow seas, i.e., the Mediterranean, Arabian Sea, the Baltic, the Black Sea, the North Sea, and the South China Sea. But after 1500, one could begin to speak of truly global maritime activities and predominance, so that the subsequent serial maritime hegemons could be seen as striving towards a truly global reach. R.E. Harkavy / Political Geography 18 (1999) 941–972 945 As noted, Modelski and Thompson have chosen to focus upon data for the number of capital ships as the primary criterion for global seapower, and for defining hegemonic long-cycle phases. While it was dealt with in the interstices of that analysis, there was much less attention paid to the related and important subject of the development of global — or at least extensive short of global — systems of overseas bases. In the earlier periods of Mongol, Southern Sung China, Venetian, Portuguese, Dutch and British dominance, that referred, of course, solely to naval bases. As we shall see, in the 20th century, the development of airpower, missiles, satellites and a host of new communications and intelligence functions has produced a far more diverse set of functions and purposes for external basing access. And, indeed, the critical role of basing access in defining global military and commercial dominance has not changed. That then leads into the central research question here provisionally broached pending a much larger research effort. The question deals with the extent that an analysis of the far-flung basing networks of the serial maritime hegemons can further inform and question the basic tenets of long cycle theory. Key questions, criteria involved in the relationship of basing access and longcycle hegemony This paper is intended as an initial cut — an outline — of the relationship of external basing access and long-cycle hegemonic dominance, as it developed historically from China in the Middle Ages, to Venice, to Portugal and Spain, to the Netherlands, to Britain and to the U.S. in serial phases of maritime and commercial dominance (the basing access of rival “global powers” in Modelski’s and Thompson’s terms — such as France during and after Louis XIV, Germany, Japan, and the Soviet Union — will also be considered, if only for juxtaposition to the superior basing networks of hegemonic powers). As such, we have identified a range of conceptual and historical-developmental questions, or issues, which may serve as the basis for a more comprehensive analysis and effort at data collection on bases, spanning 500 or more years of great power rivalry. O The fundamental problem of reciprocal causation involved in the question of whether maritime dominance has resulted in corresponding dominant patterns of basing access, or whether the latter has provided for maritime dominance, or both have developed simultaneously in ways perhaps too complex to allow for easy attribution of causality. O The long term evolvement of basing networks from “regional reach” (Venice in the Mediterranean) to “quasi-global reach” (Portugal around the Indian Ocean littoral and across the Atlantic in Brazil), to truly global reach in the more recent cases of Great Britain and the U.S. “Global reach” has become increasingly truly global. O The vastly increasing proliferation of types of basing access. Venice, Portugal, and the Netherlands sought and acquired naval bases only (albeit of somewhat 946 O O O O O O O O O O R.E. Harkavy / Political Geography 18 (1999) 941–972 differentiated types), i.e., basing networks were almost entirely defined by naval installations. By the late 20th century, rival U.S. and Soviet global networks comprised a diverse array of military functions under the general rubrics of air, naval, ground, missile, communications, space-related, intelligence, research, logistics bases, etc. Whereas earlier bases related almost entirely to surface naval activities around the major seas and oceans, today they relate to complex interactions between land, air, outer space, and underseas military activities as well as those on the surface of the major seas. The earlier mixing or co-mingling of military (naval) and commercial (entrepôt) functions for specific external bases; later, these military and economic functions would largely be separated out.3 The relationship of rival basing networks (often highly asymmetric) to the traditional Mahanian concept or dictum about the indivisibility of seapower or “command of the sea”. Has the co-existence of rival basing networks, albeit asymmetric, diluted this maxim or dictum? Indeed, does the co-existence of naval basing networks even amidst periods of seemingly global dominance dilute the central thesis of long cycles and perhaps provide indicators of degrees of multipolarity or power diffusion? The long-range connection of long cycle theory to geopolitical theory (Mackinder, 1904; Mahan, 1980; Spykman, 1944; Cohen, 1963), as informed by analysis of basing networks as reflective of “heartland versus rimland”, or landpower versus seapower dualisms. The relationship of basing access networks to the outcomes of hegemonic wars, but also, the gradualness of “base races” or contests over access before, after, and in-between major wars. The ability of dominant naval powers to “pick off” rival bases and colonial outposts during wars; further, the deterrent aspects of this possibility. There may be a relationship here to the recently evolved concept — during the latter part of the Cold War — of “horizontal escalation” (Epstein, 1984). The use of external bases for off-shore ship production in earlier phases of hegemonic cycles; in modern times, the crucial role of ship repair but also licensed ship and other weapons production as a contemporary analog. The long term shift in the mix of basing functions, i.e., from commerce raiding in earlier times to various aspects of interventions, arms re-supply, coercive diplomacy, etc., in modern times. The evolvement of mixed government-private basing activities earlier (trading companies, privateers) to the more or less entirely state-based nature of contemporary basing diplomacy. How bases are acquired and retained in historical context, i.e., the varying roles 3 This point is only partly modified by the fact that in the modern world, basing structures are critical to protection of key economic assets, even if not co-located with them, as witness U.S. naval and air bases in and around the Persian Gulf in proximity to major oil fields. R.E. Harkavy / Political Geography 18 (1999) 941–972 947 of colonial conquest, alliances, arms transfers as quid pro quo, etc., in different historical epochs. Stated otherwise, the historical evolution of basing diplomacy. O The changing technological requirements for bases, viz., for naval basing, the progression from galleys, to galleons, to coal-fired ships, to oil-fired ships, and nuclear-powered ships. O The correlation of basing access dominance to the phases of long cycles identified by Modelski, 1987, i.e., delegitimization, deconcentration, etc., as further correlated with statistics regarding capital ship production and deployment. O Thompson’s recent work on “system leader lineage”, the extent of congruence of the serial basing networks of Portugal, the Netherlands, Great Britain and the U.S., emanating from a core area of hegemonic dominance, and the long-term nature of the strategic importance of certain key oceanic areas and routes, naval chokepoints, islands, etc. (Thompson, 1997). Many of these questions will be addressed below, not necessarily in the above order, as they are in some cases inextricably entwined with one another. Regional reach to global reach Though on a superficial level one may talk about serial century-long periods of long-cycle hegemony based on seapower, and also about “system lineage” patterns whereby basing networks (along with other aspects of “global” dominance) partly clone each other, in fact there has been a progression from less-than-global basing systems to more truly global ones. [Perhaps also one might say the same thing about the reach of western capitalist interests in the context of “world systems analysis” and/or core-periphery relations.] China, in the early period of the Ming dynasty, under the aegis of the Muslim eunuch admiral Zheng He, had a naval reach — and some associated points of access — extending from south China to Southeast Asia and on through the Indian Ocean, at Colombo (Ceylon), the Philippines, Malacca (Sumatra), Calicut (India), Chittagong (modern Bangladesh), Hormuz, Aden, Jidda, and the area around the Bab El Mandeb, also Mogadishu (Somalia) (Swanson, 1982). Previous to that, the Mongol Empire in its latter stages, albeit primarily a landpower, had utilized naval bases along the Chinese and Vietnamese coasts to project power to Japan and the East Indies.4 Venice, perhaps to be characterized as holding, in long cycle terms, regional leadership prototypic of later cases of global leadership, had extensive basing access throughout the Mediterranean, extending to Crete, the Greek islands, the Levant (Accra), Constantinople, Corfu, Ragusa on the Adriatic, and Negroponte (in Greece). (Rival Genoa had similar but less extensive access) (Lane, 1973). But it had truly 4 The activities of the Mongol navy and associated basing problems are discussed in Benson (1995: 374), including Mongol naval operations in the 1270s and 1280s in Korea, Japan, and Vietnam, particularly the use of naval vessels for amphibious operations. The bases used by the Mongols were, apparently, on the basis of conquest rather than diplomacy. 948 R.E. Harkavy / Political Geography 18 (1999) 941–972 nothing outside the enclosed region of the Mediterranean Sea. Portugal, commonly characterized as the first global maritime power, had a basing system that encompassed much of the African littoral, west, south and east, islands offshore of Africa such as Madeira, Cape Verde, etc., and all around the Indian Ocean arc extending from East Africa to the Arabian Sea, India and what formerly was known as the East Indies and on to China (also in Brazil). Portugal also had access to Northern Europe via Antwerp and Southampton, and to the Mediterranean via Seville. That was an impressive extended reach, quasi-global in scope, but lacking in access to North America, and South and Central America other than Brazil, i.e., the entire Pacific coast of the Western Hemisphere, and also northeast Asia. Also, Spain during this period maintained extensive basing access in areas not reached by the Portugese empire, i.e., in Peru, Mexico, and throughout the Caribbean including what is now Venezuela and Colombia. The Dutch founded new forts and trading posts, and inherited others from the Portuguese. Great Britain, on the other hand, in its two long-cycle phases bracketing the Napoleonic wars, developed a more truly global basing network in accord with an empire upon which “the sun never set”. That system of access included, at its zenith, naval bases in, among other places, all along the east and west African coasts (Freetown, Capetown, Zanzibar, Mombasa), throughout the Mediterranean (Gibraltar, Malta, Cyprus, Suez, Alexandria), in the Arabian Sea/Indian Ocean area (Aden, Basra, Bombay, Colombo and Trincomalee, Singapore), in the Antipodes, in East Asia (Hong Kong), various southwest Pacific islands, in British Columbia (Esquimalt), in the Caribbean (Kingston, Trinidad), around the North Atlantic (Halifax, Bermuda) and also in the South Atlantic (the Falklands). The U.S. basing network during the Cold War was also truly global, albeit perhaps less quantitatively extensive than its predecessor. After all, it was denied access in the broad swathe of territories under communist sway, i.e., China, the USSR and its nearby satrapies. It also had little access in Latin America, as that region tried to emerge from the Monroe Doctrine and keep its distance from Cold War rivalries. As we shall note in another subsequent context, that had to do both with political factors (colonialism as a basis of imperial control versus various quid pro quo with essentially independent actors), and also technological factors; specifically, the lesser requirements for large numbers of bases as technology came to transcend factors of distance. Proliferation of types of bases and/or basing access In earlier times, indeed up to the interwar period after World War I, basing access was almost entirely to do with navies, with ships. There were, of course, no aircraft nor “technical” facilities in the modern sense, all of which required the invention of electricity. There were some predecessors to contemporary land force access, mostly pertaining to small army garrisons (co-located with naval bases and fortresses built to protect them.) Even in earlier times, the Venetian, Portugese and Dutch basing access systems revealed some diversification, subsumed beneath the generality that almost entirely, R.E. Harkavy / Political Geography 18 (1999) 941–972 949 naval basing was involved. In each of these early systems, what would today be familiar as “navy bases” actually combined, sometimes, the functions of naval bases and entrepôts, i.e., factories and commercial trading stations used for warehousing, transshipments of goods, etc. (Padfield, 1979: 4). In short, there was perhaps a less clear line then between specifically military and specifically economic functions, centuries before the U.S. overseas military presence would inspire charges of “imperialism”, that it was intended to protect corporate assets overseas. Venice and Portugal not only used bases for provisioning, ship repairs, R and R (rest and recreation in modern lingo), but also had small fleets more or less permanently stationed overseas (or in Venice’s case, at the other end of the Mediterranean). That anticipated later U.S. use of “station fleets” in the Caribbean and the Far East early in the 20th century, not to mention the homeporting of fleets during the Cold War in places such as Yokosuka (Japan), Bahrein, and Naples (Italy) (Roberts, 1977). In the Portugese case as well, whole fleets of ships were actually constructed overseas, for instance, in India, so that overseas fleets need not have been rotated back and forth between external bases and homeports in Portugal (Padfield, 1979). Even with extensive license-production of weapons systems all over the world by the U.S. in recent years, there has been no overseas or offshore production of capital ships or submarines for use by the U.S. Navy (Germany under the Versailles regime had produced submarines offshore in Turkey and Spain) (Harkavy, 1975). As it happens, and apparently beginning with Venice, the basing structures of the serial hegemonic powers have probably all had land forces associated with naval bases and fortresses that guarded harbors. The extent to which these were permanent garrisons, as opposed to armed colonists, personnel from trading companies, or temporarily stationed personnel from fleets, is a question that needs further research. For instance, the Atlas of British Overseas Expansion avers that “from the mideighteenth century Britain began stationing permanent military garrisons in her colonies, at first haphazardly and then routinely, as an adjunct to naval protection” (Porter, 1991: 118). That was seen as a qualitative shift, and these commitments constantly overstretched London’s budget “at a time when Parliament begrudged expenditure on a large standing army in years of European peace” (Porter, 1991: 118). France maintained large garrisons in places such as Algeria, Morocco and Senegal at the same time, as would the U.S. later in Germany, South Korea, Japan (to a lesser degree, Italy and elsewhere) during the Cold War, albeit on the basis of alliances (hence, limited sovereignty over bases) rather than naked colonial control. But skipping over the centuries, the 20th century was to see a vast proliferation of new basing requirements driven by the ongoing march of military technological innovation. Before World War I, Britain and Germany had underwater cable systems going to their colonies that required terminals in these overseas possessions, perhaps the first of the C3I (command, control, and communications) technical facilities (Kennedy, 1971). By the interwar period, all of the major powers had external air bases — in the cases of the major colonial powers, mirroring the locations of major naval bases. Britain even had an extensive network of mooring masts for long-range dirigibles. The 1930’s saw the extensive proliferation of requirements for communi- 950 R.E. Harkavy / Political Geography 18 (1999) 941–972 cations facilities (also, for communications intercepts) and the beginnings of external access for radar stations. But during the Cold War period, there was a truly staggering explosion of basing functions. The Stockholm International Peace Research Institute (SIPRI) has developed a typology of forward military presence functions with the following general categories: naval, air, land, missile, command and communications, intelligence, space-related, research and logistics (Harkavy, 1989). Each of these categories subsumes a host of others. Under air bases, there are forward fighter and bomber bases, tanker bases, maritime patrol aircraft bases, transport staging bases, etc. Under naval, there are surface ship and submarine bases, including a spectrum from homeporting and major drydocking facilities to places merely utilized for port visits. There are separate facilities for strategic surface-to-surface missiles and surface-to-air missiles. Under intelligence are a host of basing functions, for example, nuclear detection, SOSUS (sound surveillance systems to locate submarines), early warning radars, satellite down-links and master control stations, etc. The U.S. also made extensive use of overseas facilities for such purposes as solar flare detection, tracking of others’ satellites, weather forecasting, Voice of America transmission stations, etc. The full list was much, much longer. Whereas earlier, basing pertained only to the surface of the ocean, now its functions deal with the land, sea surface, the underseas, the atmosphere and outerspace, and increasingly, the links between all of these, so that, for example, there are land-based and space-based communications links to submarines patrolling under the seas. Changing basing functions Over the centuries, there has been somewhat of a shift in the general functions of bases, and in the “norms” associated with their use. This ramifies into some very broad issues of IR Theory and the relationship between security and economic affairs. Nowadays, in assessing the recent (U.S. and Soviet) functions of bases in the most general sense, a typology of sorts may be assayed. First, there is the big divide between the nuclear strategic functions of bases on the one hand, and conventional power projection on the other. The former area pertains both to deterrence in times of peace and (hypothetically) the use of bases for nuclear warfighting (or the use of bases in a “protracted conventional phase” that might have seen efforts by both sides to alter the nuclear balance, in part by interdicting or neutralizing the other side’s nuclear-related bases). Conventional power projection pertains variously to direct intervention (as with U.S. use of bases en route to Desert Storm); arms resupply during conflicts on behalf of client states (as with U.S. use of Portugal’s Azores bases in resupplying Israel in 1973); coercive diplomacy, and “showing the flag”, i.e., symbolic issues of “presence”. The current use of forward bases for force interposition or peacekeeping activities (see the U.S. use of bases in Hungary in relation to Bosnia) is still another modern category. In the past, from Venice to the heyday of Britain’s empire, conventional power projection by way of interventions, coercive diplomacy, and “presence” were all R.E. Harkavy / Political Geography 18 (1999) 941–972 951 omnipresent, always utilizing naval forces (the U.S. nowadays uses AWACS (Airborne Warning and Control System) or JSTARS (Joint Surveillance and Target Attack Radar System) planes in its ventures into “gunboat diplomacy”). But in addition, bases were vital to commerce raiding and the activities of privateers. This is a difficult area for analysis here. A review of the extant literature on maritime history reveals the extensive use of overseas bases (for instance, in the Caribbean) by privateers or even pirates who, while nominally acting independently and mainly for profit, were nonetheless citizens or subjects of one or another European country, or acting de facto under such national control. It is possible that a detailed analysis of Chinese naval activities during the era of Zheng He would reveal similar analytical problems. France in particular, during the long period between the wars of religion and the Napoleonic wars, during which its continental military activities usually precluded matching Britain’s navy, made extensive use of such “privateers” in a guerre de course that was conducted primarily around the English Channel and Bay of Biscay, but also far afield in the Caribbean (Padfield, 1982: chapter 5). Nowadays, of course, privateers are long passé; likewise, the practice of commerce raiding outside periods of intense warfare bracketed by formal war declarations and treaties (privateers could seemingly conduct commerce raiding outside periods of declared warfare just because they could claim to be operating independent of national control). Finally, it is to be noted that earlier, “bases” and “factories” overseas were colocated (some of the historical maps use these terms interchangeably). That is, fortresses and naval stations were co-located with economic enterprises overseas, with the military presence needed on the spot to guard the “factories” or entrepôts. The basis for basing: territorial control vs. quid pro quo What, historically, has been the political basis for basing? Why or how do hegemonic (or other) naval powers acquire such basing access? What have been the changes in practice over time? Basically, this breaks down into a spectrum with two strong contending poles, and grey areas in between. Simply stated, basing access may be taken by force, i.e., it is a function of imperial or colonial control; or, it may be acquired via various forms of quid pro quo, diplomatic trade-offs, reciprocal albeit asymmetric transactions, etc. In the latter cases, some degree or another of coercion, albeit subtle and implicit, may be involved. Much of the history of basing access derives from outright conquest or the imposition of foreign access upon unwilling “hosts”. Venice fought for and retained access in various places in the eastern Mediterranean, which then had to be protected by fortifications, troops, and galley fleets. Earlier, Ming China under Zheng He’s admiralship had obtained access in nearby Southeast Asia primarily by supporting clients in key strategic positions such as Malacca. But its “bases” in Africa and the Middle East really were just fleet visit sites with permission by local rulers eager to trade (again, the blurring of military and economic functions). The later Portuguese and 952 R.E. Harkavy / Political Geography 18 (1999) 941–972 Dutch seaborne empires relied on conquests and occupations of key strategic points d’appui, but with a mix of raw force and deals with local leaders who could be bribed, threatened into submission, or very often, needed the European power’s navy in its conflicts with nearby rival states. Portugal, for instance, did an effective job of allying with some principalities on India’s west coast and playing off one local ruler against another (Padfield, 1979: 65–69). In what would become a familiar theme, arms transfers were a particularly effective medium of exchange to secure basing rights. So were bribes and other subventions. Britain’s empire, mostly hinged on outright colonial control, also utilized subsidies to European allies in their myriad wars in exchange for naval base access; earlier after 1585, there were English bases in the Netherlands. As would become familiar to 20th century students, basing access was often granted as a way of securing support against contiguous regional foes, i.e., a trade-off of bases for security (Padfield, 1982: chapters 6 and 7). In the twentieth century, which saw a transition from British to U.S. maritime supremacy, there was a long-term trend away from access via colonial control to that based largely upon various quid pro quo, i.e., security and economic assistance. In the interwar period, basing systems were almost entirely a function of colonial control. Britain (now perhaps beyond the point of its maritime supremacy) had by far the largest basing system, correlated with the size and scope of its empire. France, the Netherlands, Spain and Portugal followed. Germany, shorn of its overseas empire by the Treaty of Versailles had nothing. Neither did the Soviet Union. The U.S. had a few external bases in the Philippines, Hawaii, Samoa, Midway Island, Guam, etc., mostly taken from Spain in 1898. In this case there was a lack of congruence between maritime power, measured by capital ships, as per Modelski/Thompson, and bases. That lack of congruence was partly restored via the Anglo-American destroyers-forbases deal in 1940 and the subsequent elaboration of the American basing network during World War II. In the 1950s and 1960s, the basis for bases shifted from colonial control to diplomatic quid pro quo. The U.S. (and later the USSR) first availed itself after 1945 of a global basing network provided by its allies’ colonial control in places such as Singapore, Mauritius, Cyprus, Aden, Capetown, Mombasa and Malta. But as those colonial dominoes fell one by one, translating into more than 100 new states, both the U.S. and USSR had to utilize formal alliances and arms transfers (often, the cement of security and economic assistance) to acquire and secure bases. Indeed, by the 1980s, one could almost read into the basic numbers on security assistance to ascertain where America’s most crucial basing points were: in Spain, Portugal (the Azores), Greece (including Crete), Turkey, the Philippines, Thailand, Morocco and Kenya (United States Congress, annual). Some were afforded by tight alliance relationships with nations not needing security assistance, i.e., Japan and Germany, which provided offsets to maintain a U.S. presence and commitment. In the military and academic literature on this subject in the 1970s and beyond, a distinction was drawn between “bases” and “facilities” (Harkavy, 1989). The former pertained to situations where the user (mostly the U.S. and USSR) had something close to complete control over external ports and airfields. As the Cold War progressed, however, what had been “bases” came often to be characterized as “facili- R.E. Harkavy / Political Geography 18 (1999) 941–972 953 ties,” whereby the host nation exercised its sovereignty and control over the use of such “bases”. Thus the U.S. was able to use facilities in Taiwan, Okinawa and the Philippines for combat missions in the Vietnam war, and British facilities for a raid on Libya in 1986. In 1998, Turkey and Saudi Arabia were reluctant to allow the U.S. Air Force to attack Iraq even from bases where U.S. aircraft were more or less permanently stationed. By the 1980s many nations hosting to U.S. intelligence gathering facilities were demanding a partnership in utilizing the “take” from such facilities. This is perhaps an important research question in longer term basing history, where alliances and related issues of sovereignty must presumably have come into question when the user wanted to conduct military operations not approved of by the host. Generally speaking, bipolar systems with tight alliances and an ideological basis for conflict (the Cold War) seem to have experienced more permissive basing access for the hegemons. Multipolar (and less durable) alliance structures seem to dictate less basing access for major maritime powers. In the 1930s, outside the realm of colonial control, only Japan’s use of Siamese naval bases and Germany’s submarine bases in Spanish-controlled islands (Canary and Balearic Islands) constituted examples of basing access provided one independent nation by another (Harkavy, 1982: chapter 3). Though a full discussion is not possible here, it is worth noting that throughout long cycle history, islands along the rimland littorals have been important to basing. Malta, Minorca, Cyprus, Socotra, Mauritius, Madagascar, Sri Lanka, various Indonesian and Pacific islands, etc., have been critical to naval basing through several long cycles. More recently, one notes the crucial importance of Central Pacific islands to support Japan’s onslaught in 1941–1942, and the critical roles of Guam, Diego Garcia and Ascension in recent U.S. and British military operations or the extent to which Ethiopia’s Dahlak Archipelago became a strategic prize for the USSR (and earlier perhaps, Israel). In recent years, some of these small islands have formed the last remnant of “colonialism” just because they may have been too small for independence but have been of critical military value. System leader lineage In some recent work, William Thompson has written about the various aspects of “system leader lineage,” conceding, however, that “strictly speaking, the previous system leader does not give birth to its successor” (Thompson, 1997: 19). Yet, he avers, “there are clear lineage patterns in the technological innovations that provide the foundations for systemic leadership”, “a long history of assistance and resource transfers from the old leader to the new that could be likened to parental nurturing and offspring learning”, and “an even longer history of interaction in the security realm that paints system leaders as a special community in international relations — not unlike a kin or clan relationship linked by some real or imagined bloodline” (Thompson, 1997: 19). He even refers loosely to a longitudinal “geopolitical community formation”, whereby despite the wars between the Portuguese and Dutch, 954 R.E. Harkavy / Political Geography 18 (1999) 941–972 the Portuguese and English, the Dutch and the British, and the British and the Americans (Genoa and Venice are left out here), there are also historical security linkages, primarily among the British, Dutch and Americans. He concludes that “the states that have become system leaders have been unusually prominent suppliers of protection and security assistance before, during, and after the recipients’ periods of systemic leadership” (Thompson, 1997: 19). In this same vein, one may point to the lineage aspects of the serial basing networks of the Chinese, Venetian, Portuguese, Dutch, British, and U.S. maritime empires. Some of these links are stronger than others, i.e., there are no links between the Chinese and Venetian networks, and none either between Venice and Portugal. Nonetheless, some obvious longitudinal linkages may be cited, on two related levels. First, in the progression from Portugal, the Netherlands, Britain and the U.S. there is a tendency for some naval bases or strongpoints or points d’appui (either precisely or in the vicinity) to be passed along from one empire to the next. In some cases that has resulted from conquests, the results of wars; in other cases, the inheritance has been more peacefully acquired. Second, it is noted that throughout this long history, certain areas, certain littorals, have been the primary foci of basing networks, i.e., the Mediterranean, the entire African littoral from Morocco around the Cape of Good Hope and on to the Bab El Mandeb; the entire arc of the Indian Ocean from southern Africa to Southeast Asia bracketing the South Asian sub-continent; and the Asian littoral running from present day Vietnam on up to China, Korea, and Japan. Access to and control over strongpoints throughout these areas has been a consistent bone of contention throughout most of this past millenium. The relation to Mahan and Spykman, i.e., sea control and control of the rimlands or (in Saul Cohen’s terminology) “shatterbelts”, is fairly apparent (Spykman 1942, 1944; Cohen, 1963). Interestingly, many of the naval access points that were established by Zheng He’s Chinese navy in the early 15th century were those later re-established by Portugal and the Netherlands; indeed, were later contested for by the U.S. and the USSR during the heyday of the Cold War in the 1970s and 1980s: bases near and around the crucial Indonesian Straits, Sri Lanka, India; Hormuz, Aden, and Jidda near and around the crucial chokepoints of the Straits of Hormuz and Bab El Mandeb. (The Japanese navy tried to take over many of these same positions in its attempt at supplanting the British Navy in the Indian Ocean in 1942.) The Dutch eventually took over many of the bases established first by Portugal: Colombo and Jaffna (Ceylon), Calicut and Cochin (India), Malacca, Macassar (Indonesia), several places in southwest Africa, Capetown, and also Recife and Bahia in Brazil (the latter have not entirely been part of a system lineage, but note the use of air bases in Northeast Brazil by the U.S. in ferrying aircraft to the British in the Middle East in 1942) (Harkavy, 1982: 111). Later, Britain was to inherit much of what had been the Portuguese and Dutch naval basing networks in the Indian Ocean/South China Sea area, establishing naval bases at Capetown (Simonstown), Mauritius, Mombasa, Aden, Muscat, Basra, Bombay (near the former Portuguese base at Diu), Colombo, Singapore (near Malacca) and Hong Kong (near Macao). During the early part of the Cold War, the U.S. used many of the British bases in this area. But as the Cold War progressed, the Soviets made competitive inroads, R.E. Harkavy / Political Geography 18 (1999) 941–972 955 establishing naval bases or lesser forms of access in Nacala and Lourenco Marques (Mozambique), Berbera and Mogadiscu (Somalia). Aden (South Yemen), Port Sudan (Sudan), Vishakapatnam (India) and Haiphong and Camranh Bay (Vietnam) even as the U.S. maintained access at Capetown, Mombasa (Kenya), Muscat and Masirah Island (Oman), Bahrein, Colombo, Singapore and Subic Bay (the Philippines), among others. There was a competitive, rival conflict over access points just as there had been in the 17th century as the Dutch only gradually took over much of what had been the Portuguese seaborne empire. The entire Mediterranean littoral has been an area of competition over basing access over a long stretch of time, even though neither the Portuguese nor Dutch seaborne empires penetrated into this inland sea area (Portugal had bases at the Atlantic entrance at Tangier and Ceuta). But it is striking to note that in the struggles between Venice and the Ottoman Empire, then later between Britain, France and Spain, and still later between the U.S. and USSR, many of the same naval strongholds were contested for. Venice, for instance had important bases on Crete (as did the U.S. throughout most of the Cold War, inheriting the former British base at Souda Bay) and at Ragusa on the Adriatic (later a Soviet point of access) and on the Turkish coast. Britain, at the height of its empire, established important bases first at Gibraltar and Minorca, later at Malta, Crete, Cyprus, Alexandria and Port Suez. During the peak of the Cold War, U.S. main naval and air bases at Rota (Spain), Sigonella (Sicily), La Maddalena (Sardinia), Naples, Piraeus, Souda Bay (Crete) and Izmir (Turkey) were balanced off by Soviet access to Annaba (Algeria), Tripoli (Libya), Alexandria and Port Said (Egypt), Tartus and Latakia (Syria), and several ports along Yugoslavia’s Adriatic coast (Dismukes & McConnell, 1979). Perhaps further afield, one might note that the rivalry among Britain, France, Spain and the Netherlands (ending up with British dominance after the Napoleonic wars) for points of access around the Caribbean Sea was somewhat followed upon by the U.S.-Soviet rivalry during the Cold War, with Moscow establishing naval base access in Cuba (and the U.S. fearing further incursions in Guatemala in 1954, the Dominican Republic in 1965, Nicaragua in the early 1980s, and acting accordingly, for good or worse). In short, there has been a form of system leader lineage as pertains to what Saul Cohen refers to as the “world that matters”, some consistently contested over strategic seas, strategic regions and strategic basing points whose importance has been remarkably enduring. It may be worth pointing out that the Pentagon has already begun to sweat over early adumbrations of China’s putative desire for long range power projection capability in the Indian Ocean area. A few years ago, reports (true or false) of Chinese bargaining for access to some islands off shore of Burma inspired scenarios bearing an eerie resemblance to Zheng He’s westward thrust in the direction of the Persian Gulf, albeit then absent the oil factor. Technological change, basing requirements and global networks Needless to say, changes in military technology have had a huge impact on basing requirements over the 500 or so years since Venice and then Portugal were (regional 956 R.E. Harkavy / Political Geography 18 (1999) 941–972 to semi-global) hegemons. This is a large subject, but several main preliminary points may be made. First, there is the important impact of the progression, as regards naval propulsion, from galleys to sail to coal to oil and then nuclear naval propulsion. Each phase has had its own requirements regarding bases. During the age of sail, for instance, particularly as applied to Portuguese naval activities all around the Indian Ocean and across the Atlantic to Brazil, the major sailing routes were greatly altered or determined by patterns of wind and currents. Hence, the famous “Carreira da India”.5 Thus some base locations in West and East Africa derived their importance. The bases were then utilized, variously, for reprovisioning, R and R (Rest and Recreation) (all the more important before antidotes for scurvy were developed), and ship repair and rebuilding. Later, during Britain’s reign in the 19th century, the development of coalpowered, steam-driven ships led to requirements for networks of coaling stations whereby coal was stockpiled even on islands in the middle of oceans (Kemp & Harkavy, 1997: chapter 2). That became an important diplomatic desideratum, as the Russians learned in trying to move a fleet from Europe to the Sea of Japan in 1905 and having to rely on British coaling stations. Oil powered ships and the development of fleet oilers to accompany fleets changed this equation again, lessening overall requirements for bases but still in some instances retaining the use of facilities for refueling. The development of aircraft added a whole new dimension to basing requirements, while the need for naval bases was retained. But over the past half century or more, the development of longer range aircraft and ships, plus the development of techniques for aerial refueling of planes and at-sea refueling of ships has had the effect of greatly decreasing the number of basing points required by major powers to maintain global access networks. In 1942, the U.S. needed an extensive basing chain (Florida–Cuba–Trinidad–British Guyana–Recife–Takoradi–Kano–Khartoum) to ferry aircraft and other supplies to beleaguered British forces in the Middle East. In 1973, the U.S. was able to resupply Israel with arms using just one transit point in the Azores. In 1991, U.S. B-52 bombers, with the aid of tankers, conducted bombing raids over Iraq all the way from a base in Louisiana. Further, the number of aircraft and ships in the inventories of all major nations has declined (more combat power per ship and plane, fewer of each), and this too has militated toward lesser basing requirements in a quantitative sense. In the modern world, of course, the development of satellites and the advent of an array of new communications and information technologies has spawned numerous new basing requirements: satellite down-links, radars for early warning, signals intelligence (SIGINT), satellite-tracking, nuclear detection, etc. This has meant a massive long-run trend, whereby major powers require fewer naval and air bases, but a proliferated array of global networks for various “technical functions”. 5 Boxer (1965: 54–55) provides a map of the “Carreira” replete with patterns of winds and currents. R.E. Harkavy / Political Geography 18 (1999) 941–972 957 Rival basing networks and “command of the sea” There is the familiar albeit arguable Mahanian dictum about “command of the sea” that posits a tendency for maritime hegemons to maintain complete mastery of the high seas. That mastery is said to be based at any time on the hegemon’s main battle fleet’s ability to defeat a rival’s main battle fleet, rather a Clausewitzian notion applied to naval warfare. For a long time, in a related vein, Britain maintained a policy of fielding a navy equal at least to the two next most powerful navies, a policy only reluctantly abandoned after World War I in the face of the growth of the U.S. Navy. But even though the serial hegemons have maintained “command of the sea” in Mahan’s terms, and even though there has been a tendency for each hegemon to inherit the previous one’s global basing network, basing access has never been the exclusive preserve of one power. Rather, there have been asymmetries in varying degrees, and also constant competition for basing access in periods of peace as well as war. At the conclusion of hegemonic wars, large shifts in the balance of basing access have occurred, but not always to the point of total exclusion of losers from such access. And, there have been some rivalries — U.S.–USSR during the Cold War, Portugal–Spain in the 1500s — when more than one maritime power has been availed of significant levels of access on a global or quasi-global basis. The history is complex and not so easily reduced to simple generalizations. Venetian and Genoese basing structures In the late 15th and early 16th centuries, just as Portugal was embarking on what would be its expansion around Africa and into the Indian Ocean, the Mediterranean (then the sub-global cockpit for naval rivalry among the extant rival major powers) saw what was perhaps a four-way competition for basing access, and which competition was at the heart of the security and economic rivalries of the period. Both Genoa and Venice lost important prior bases after the Ottoman Turks’ takeover of Constantinople in 1453 — the former lost bases on the Black Sea and then lost most of the chains of eastern Mediterranean coastal bases which formed the other leg of the Levant operation. Venice too gradually lost access in the Mediterranean, with a turning point in 1503, when she made naval base concessions for peace with Turkey in order to devote herself to the new power balance brought about by the rivalry of France and Spain in Italy. That entailed the loss to the Ottoman Empire of the greater part of a chain of seaside fortresses in Greece and the Aegean, even at the gates of the Adriatic Sea. Spain, meanwhile, by around 1505, had established bases in North Africa at Mers El Kebir, Oran, Mostaranem, Tenes, Algiers, Velez, and Bougie, and was constructing a fortress at Algiers. The Ottomans, moving westward, built fortresses on the island of Goletta, which commanded the harbor at Tunis. And, in an early example of basing diplomacy (as contrasted with outright conquest), Turkish fleets were granted the use of French ports, principally Marseilles. The Venetians, seemingly fading as a power, retook, for a while, bases in Greece and the Aegean. 958 R.E. Harkavy / Political Geography 18 (1999) 941–972 The Ottoman capture of bases in Crete caused Venice, Spain and the Papal States to create the Holy League. But Genoa retained a base at Chios in the Aegean, seemingly with the approval of the Ottoman Sultans, while Venice retained only bases on Crete and Cyprus. The early 1500s saw a complex pattern of alliances and basing access in the Mediterranean involving Spain, France, Genoa, Venice and Turkey, with Spain and Venice often but not always aligned against Turkey and France (Padfield, 1979: chapter 3). Genoa ended up firmly in the Spanish orbit in 1528 (Spain itself had become part of the Habsburg Empire in 1516). Earlier, in 1503, Venice, losing its power, had made naval base concessions for peace with Turkey, and lost to the latter the greater part of its seaside fortresses in Greece and the Aegean. Spain during this period was engaged in a struggle with France, but also with Turkey pressing up from the Balkans along the Danube, Turkey was engaged with Persia to the east, but mounted attacks on the Portuguese in the Indian Ocean from its Red Sea and South Arabian bases. According to Padfield (1979: 87), “the struggle for the Mediterranean was never a simple matter of Christianity against Islam; it was more of a struggle between the two great empires of Spain and Turkey counterpointed by the feud between Spain and France and the extraordinary alternating hostility and mutual dependence of Turkey and Venice. When the two conflicts coincided and France joined with Turkey against Spain, or Spain joined with Venice against Turkey, the alliances were never satisfactory because each partner had different objectives; it was no part of Charles V’s plan to smash Turkish naval power so decisively that Venice would become undisputed master of the eastern Mediterranean, nor was it Venice’s idea to assist Charles to gain control of all Italy by helping him against his French rivals or the Algerine corsairs who interfered with his lines of communication. In short, the two halves of the sea each supported its own system of trade rivalry and conquest, and the powers of the other half, together with small independent squadrons from the Knights of St. John, the Papal States of central Italy, the Princes of Savoy, Tuscany and Monaco, were called in by subsidy and diplomacy to tilt or redress the balance when it was threatened.” In 1538, a joint Papal-States–Venice–Spain fleet was defeated by the Barbary admiral Kheir El Din and the Turks, causing Venice to sue for a separate peace in the aftermath and leading to Turkish use of the French port of Marseilles. The battle of Lepanto in 1571 reversed this earlier Turkish victory, after which Spain and Turkey reverted to interests in their own separate spheres. The naval wars in the Med in the 1500s are characterized as having “settled into a series of amphibious assaults to gain, retake or defend the bases useful for the struggle for ‘command’ as in naval historical dogma, but for the capture of bases from which attack and defense of trade and coasts could be mounted”. Fought over in campaigns were Algiers, Tripoli, Bougie, Djerba, Malta, and Tunis, the latter defended in an epic of resistance by the Maltese Knights before they were relieved by a Spanish squadron. By the 1570s, Venice had made a separate peace with Turkey, R.E. Harkavy / Political Geography 18 (1999) 941–972 959 giving up its position on Cyprus, and Spain and Turkey maintained control of mostly the opposite ends of the Mediterranean. There was no hegemon during this transition period as the Portuguese access network grew outside of the Mediterranean cockpit. Portugal’s and Spain’s basing structures Portugal’s basing network in the Indian Ocean and in West Africa and across the Atlantic to Brazil evolved gradually from the late 15th into the 16th century. And it was uncontested by other powers, even though Portugal was defeated in its attempts to take Aden by local Arab rulers. The three main fortress bases in the Indian Ocean were Ormuz, Goa and Malacca, but with a string of “tributary ports” and “fortress factories” down the Malabar Coast of India, the Muscat coast of Arabia, and the east coast of Africa from Sofala to Mombasa and Malindi. In the 1530s bases were acquired in Brazil, Macao, and Ceylon. Portugal’s rivals were the local fieldoms whom it was trying to subdue (Padfield, 1979: chapter 2). But at the same time as the 16th century progressed, Spain acquired a string of bases along the Spanish Main in the Caribbean — at Santo Domingo on the southwest coast of Hispaniola, Nombre de Dios on the Panama Isthmus, Vera Cruz and San Juan de Ulloa in Mexico, Havana and on the Florida coast, with a fortress being built at St. Augustine in 1564 (Padfield, 1979: chapter 4). The bilateral Portugal– Spain Treaty of Tordesillas, later confirmed by a Pope, had created a virtual global condominium regarding sea control, basing access, resource control, etc., but in which the respective hegemonic domains of Spain and Portugal were sharply and fully demarcated. Hence, it was distinct from the kind of rivalrous bipolarity exemplified by the Cold War. As it happens, the Spanish basing network in the Caribbean was, around the time of the defeat of the Spanish Armada, under pressure from upstart Britain, whose navy sacked Santo Domingo and Cartagena (the British planned but did not succeed at this juncture in taking over the Azores, so as to better interdict Spanish shipping en route home from the Caribbean and South America) (Padfield, 1979: 130). Scholars argue whether Spain deserves to share with Portugal the role of global hegemon during this period. There is a contrast here with the later overlapping and more directly confrontational basing networks of the U.S. and USSR during the Cold War. The Dutch basing structure During the 17th century, the Dutch took over most of what had been the Portuguese seaborne empire (Portugal had been taken over by the Spanish monarchy in the 1580s, but continued to operate semi-autonomously abroad). Most crucially, the Dutch took over previous Portuguese bases at Tidor (Halmahera Islands), Ternate, Malacca, Amboina, Macassar, Colombo, Cochin, the Cape of Good Hope, several places along the West African coast in Angola and the Gold Coast, and Pernambuco and Bahia in Brazil (Padfield, 1979: chapter 5). But it is important to note that the 960 R.E. Harkavy / Political Geography 18 (1999) 941–972 takeover was not complete, as the Dutch failed militarily to take over bases in Macao, Timor, and Mozambique (the Dutch also established a new main base in the East Indies at Batavia, and also took the Portuguese West African fortress of El Mina (the base of the Guinea gold trade and a major slaving center). Spain meanwhile, even as its power waned, hung onto some positions in South America and the Caribbean, while the important Portuguese base at Hormuz was taken over (in 1622) by Persia, which was re-establishing some traditional trade routes between South and Southeast Asia and Europe via the Middle East. Persia was helped by the Dutch in this venture (Padfield, 1979: 70–71). Even at the peak of Dutch maritime and commercial power and its (incomplete) rout of the Portuguese basing network, the other main powers, i.e., Spain, Britain, and France, retained some overseas basing points. Spain, as noted, kept its chain of fortified bases that allowed the annual Spanish fleets and their treasure to get home. English and French companies, meanwhile, established settlements in North America which in some cases were used as bases to attack Spanish shipping. The British established bases in Bermuda, Barbados, St. Christopher, Nevis, and Antigua; France in Martinique, Guadeloupe and half of St. Christopher. The English, of course, settled Nova Scotia, New England, Maryland, Virginia, etc; France was installed in Quebec and the St. Lawrence Valley. Meanwhile, the Dutch took over Manhattan and established entrepôts at Curacao, St. Eustatius, St. Martin, and Aruba, much of which involved the trade in sugar and tobacco. During this period, that is with reference to bases and to the establishment of colonies, there was a degree of “multipolarity” even as the Dutch were for a time the wealthiest and strongest trading nation there had ever been, dominating world commerce and the carrying trade, via a singleminded war of attrition on the hitherto Spanish and Portuguese oceanic monopolies.6 The British basing structure Britain succeeded the Netherlands as the next great maritime hegemon, and it established an even more extensive global basing network, first defeating Holland at sea, then joining it via dynastic marriage. In the eighteenth century, first during the reign of Louis XIV up to the treaty of Utrecht and then later in the Seven Years’ War, it established an unrivalled basing network. But the Dutch, Portuguese, and Spanish retained some older points d’appui, as did France even after the major defeats that cost it most of its colonies and bases in North America. Even at the peak of the British Empire in the late 19th century, its basing access network, while dominant, was not exclusive. Indeed, Germany entered the fray in the latter part of the century, with access to Tanganyika, Southwest Africa, Togo, and some islands in the central Pacific. To some extent, the asymmetric rivalries for global basing networks may be encap- 6 The considerably multipolar nature of rival basing networks during this period contrasts somewhat with long cycle theory’s assumption of unipolar hegemony or bipolar rivalry (Scammell, 1989). R.E. Harkavy / Political Geography 18 (1999) 941–972 961 sulated by Padfield’s rather crude three-way typology, in which he characterizes the various aspirants for hegemonic power as “sea creatures”, “territorial animals”, and “hybrids” (Padfield, 1979: 14–16). The long-cycle hegemons — Venice, Portugal, Netherlands, Britain, and the U.S. — have all been “sea creatures” in the sense that their respective drives for maritime hegemony were not diluted or diverted by the need to maintain large armies for continental wars. The “hybrids” on the other hand — Spain, France in particular — always had to bifurcate their security strategies between issues of naval mastery and land threats from nearby powers not hindered by water barriers such as the English Channel and the Atlantic Ocean (the Dutch were threatened on land and that cut into their naval strength; Britain was threatened by invasion by the Spanish in 1588, the Dutch, French, and later, the Germans). The real “territorial animals”, i.e., Germany from the late 19th century up to 1945, the USSR during the Cold War, both tried to match British and American naval power (and to establish rival basing networks), all the while tied to the massive budgetary costs associated with the need to maintain large land armies that could dominate Central Europe. Japan in the 1930s, seemingly a “sea creature” and one that sought major naval basing access in Southeast Asia (it had navy bases in Siam before World War II) also ended up de facto a “hybrid” due to the extent its army was tied down in China beginning in 1931. Theoretically speaking, there is the interesting question of the relationship between the long cycle literature associated with Thompson and Modelski, and the by now venerable tradition of geopolitical theory running from Ratzel to Mackinder and Mahan, and on to Spykman and the more contemporary writers such as Saul Cohen, Geoffrey Parker, and Colin Gray (Cohen, 1963; Parker, 1985; Spykman, 1944; Gray, 1977). Taylor (1989), in his general work Political Geography, discusses long cycles and geopolitical theory in tandem, but not much by comparison or with regard to whether, relatively, they complement or contradict each other. Among the numerous definitions of what geopolitical theory is or has been about, Osterud (1988: 191–192) offers the following: In the abstract, geopolitics traditionally indicates the links and causal relationships between political power and geographical space; in concrete terms it is often seen as a body of thought assaying specific strategic prescriptions based on the relative importance of landpower and seapower in world history. or This geopolitical tradition had some consistent concerns, like the geopolitical correlates of power in world politics, the identification of international core areas, and the relationships between naval and terrestrial capabilities. or The endemic antagonism between the British-American seapower and Russian landpower; the inherent dangers of the German “drang nach Osten”; the strategic 962 R.E. Harkavy / Political Geography 18 (1999) 941–972 importance of different geographical areas; the re-shuffle of geostrategic relationships by technological innovations in warfare and transport% or, the debate between the Blue Water school of strategists and the advocates of vast continental areas as the strategic key to world power. What then is the relationship of long cycle theory to geopolitical theory and to what extent can a review of basing networks over the past half millenium or so inform us about this nexus? First off, of course, it is clear that the long cycle theorists (but also Colin Gray in his eminent work on seapower) come down on the side of Mahan in seeing seapower as having been the key, consistently, to global military and commercial hegemony. Stated otherwise, and except for brief transition periods in which major wars have been decisive in establishing long-term hegemonies, long cycle theorists are less inclined to see long term, bipolar dualisms between landpowers and seapowers (indeed, Thompson and Modelski, and also Padfield, stress that dominant European landpowers are normally, crucially tied down by continental military requirements and, hence, can not easily conjur up the resources required for seapower dominance). But as noted elsewhere here, the long cycle periods of successive maritime hegemons have often seen real contests both for maritime supremacy and for basing networks between predominantly naval and predominantly land powers. The Mongol Empire, centrally a land-based military power, built a navy, invaded Japan and the East Indies, presumably involving basing access in what now is China, Vietnam and Indonesia. France, a landpower in the 18th and 19th centuries, challenged British naval dominance and constructed a fairly formidable basing network in the Caribbean, North America, and in the Indian Ocean and India. Germany’s challenge to British naval dominance involved (before 1914), bases in Africa and the Central Pacific, and indeed, even after Versailles, in the Canary and Balearic Islands in the late 1930s. The Soviet Union, of course, succeeded in building an elaborate global basing structure hinged on Cuba, Guinea, Angola, Somalia, India, South Yemen, and Vietnam, among other places. Napoleon and Hitler embarked on their conquests after their nations had lost major wars (1759 for France, 1918 for Germany) and had largely been stripped of overseas basing assets. Long cycle phases and global basing networks George Modelski, William Thompson (and Richard Rosecrance emendating the former) have devised schemes delineating the phases of long cycles. Though a simple version of this has the serial hegemons deemed dominant for approximately a century apiece (Portugal from 1494–1580; the Netherlands from 1580–1688; Great Britain from 1688–1792 and again from 1792–1914; and the U.S. from 1914 onward), the periods of dominance have contained several phases in a consistent manner. Modelski claims actually that each of these periods has four phases: 1) global war; 2) the emergence of world power; 3) delegitimation of power; and 4) deconcentration of R.E. Harkavy / Political Geography 18 (1999) 941–972 963 power. These are summarized for the period since 1494 in Table 1 drawn from Rosecrance’s review of long cycle theories (note Venice is absent here). This scheme then raises some interesting questions about the relationship of the long cycle phases to the ebbs and flows of basing access. For instance, it raises the question of whether the periods of “global war phase” have been those where basing structures have been developed or where one hegemon has taken over basing networks previously belonging to the earlier hegemon. Or, have some maritime hegemons lost some of their basing assets during periods of where one hegemon has taken over basing networks previously belonging to the earlier hegemon? Or, have some maritime hegemons lost some of their basing assets during periods of delegitimation and deconcentration? Generally, has the loss of bases led to delegitimation, deconcentration and then the loss of hegemonic wars, or has the loss of bases resulted mostly at the close of hegemonic conflicts? Have there been consistent tendencies throughout the period since 1494, or have different cycles seen different relationships between basing access and the phases of long cycle hegemony? On a general level, has the acquisition of global basing networks tended to be rather sudden and compact in time (during the global war phases), or has that been more stretched out over the whole of the cycle? Further (and beyond the scope of this paper), how have the sizes of navies, measured by capital ships, varied according to basing structures as well as the phases within the cycles? There is no easy or simple way to summarize a response to these questions, nor any obvious consistent and overall patterns. Indeed, there have been strong variations across the various long-cycle hegemonic periods and the transitions between them. Table 1 World Events Since 1494, According to Modelskia Global War Phase World Power Phase Phases of Delegitimization and Deconcentration 1. 1494–1516: France is the challenger during Italian and Indian Ocean Wars 2. 1580–1609:Spain is the challenger during Spanish–Dutch Wars 3. 1688–1713: France is the challenger during wars of Louis XIV 4. 1792–1815: France is the challenger again during Revolutionary and Napoleonic Wars 5. 1914–1945: Germany challenges during World Wars 1516–1540: Portugal is leader 1540–1580 a Source: Rosecrance (1987). 1609–1640: Netherlands is leader 1640–1688 1714–1740: Britain is leader 1740–1792 1815–1850: Britain is leader again 1850–1914 1945–1973: United States is leader 1973–? 964 R.E. Harkavy / Political Geography 18 (1999) 941–972 The Portuguese long cycle For instance, to the extent Portugal replaced Venice as a hegemon in the early 16th century, there was virtually no direct military confrontation between them; rather, political and commercial rivalry, accompanied by diplomatic maneuvers and covert action. As noted, while Venice was part of an essentially quintipolar struggle for the Mediterranean around the early 1500s (Venice, Genoa, Spain, France, Ottoman Empire), Portugal was building its maritime empire that eventually stretched from Recife to Nagasaki (west to east). But it did not defeat any previous hegemon; rather, it established an empire gradually via conquest and diplomacy in what had been somewhat of a strategic vacuum, or rather, in a large swathe of global territory controlled by myriad empires and principalities in Africa, Asia, Brazil, etc. Hence, what Modelski cites as the “global war phase” of Portuguese hegemony, from 1494 to 1516, was really the peak of Portugal’s conquests in a series of clashes, not with another major power, but with smaller, weaker and unrelated political entities all around the globe. Some of these conquests actually preceded this “global war phase”. Ceuta was taken in 1415, which acted as a jumping-off place for operations further south in Africa. By around 1470 Portugal had occupied various fortified trading posts and slave pens in West Africa; had seized the Canaries and taken over the Guinea trade. Regarding the latter, they erected a castle at Sao Jorge da Mina on the Gold Coast. But in the early 1500s, there was the major burst of Portuguese expansionism, with bases and entrepôts established in Benin, Diu in India, Axim in West Africa (1503), Sofala (1505), Mocambique (1507), Goa (1510), Malacca (1511), and Ormuz (1515). Goa was the site of a major dockyard and repair facility for Portugal, the only one in the Indian Ocean area. There were also failures to capture Socotra Island in 1510 and Aden (1513) from the Muslims, as Portugal tried to close the spice route through the Red Sea (note both of these would become Soviet bases in the 1970s). Portugal also failed to extend its power along China’s coast, although it did establish a foothold later at Macao in 1577 (Boxer, 1965). This ended up a very extended empire that peaked between 1516 and 1540, and which at its peak involved a chain of about 40 forts and coastal settlements between Sofala in east Africa and Nagasaki, and many more in West Africa (Ceuta, Tangier, Mazagao, Cape Verde, etc.) (Boxer, 1965: 53). Some bases were established during what Modelski defines as Portugal’s phase of delegitimization, i.e., Luanda in 1575, Macao in 1577. Much of it was acquired and maintained via local alliances and related balance of power politics. Portugal acquired Mombasa and Malindi on the Swahili coast, but Portugal later allied with the latter against the former. Its position in India depended on playing off the rulers of Calicut and Cochin; likewise, the sultans of Ternate and Tidore in the Moluccas region, three different rulers in Ceylon, Sunni Turkey vs. Sh’ia Persia etc. And meanwhile, the Pope-enforced treaty of Tordesillas in 1494 effectively separated Portugal’s seaborne empire from the corresponding and co-existent Spanish seaborne empire stretching from South America and the Caribbean across the Atlantic to a basing network inside the Gibraltar Straits in the Mediterranean. R.E. Harkavy / Political Geography 18 (1999) 941–972 965 The Dutch long cycle Later, Portugal’s seaborne empire would be attacked and substantially cut back in size by the Dutch in an extensive and lengthy war that some have characterized as the first global hegemonic war. This one was often head to head, with the Netherlands fighting both Portugal and Spain, now joined by dynastic relations, but still functioning as somewhat separate powers. The fall of the Portuguese seaborne empire actually occurred over 80 years, though Modelski points to a “global war phase” from 1580 to 1609 as leading to a period of Dutch ascendancy up to 1640. In Brazil, the Dutch captured Bahia in 1624–1625 and Pernambuco in 1630. Indeed, many of the Dutch conquests of Portuguese imperial possessions took place during Modelski’s characterization of a “world power phase”, after the Dutch had defeated Spain. For instance, between 1638 and 1658 the Dutch conquered Portuguese settlements in Ceylon and its Asian strongholds in Cochin and along the Malabar Coast. They blockaded the Straits of Malacca from 1634 to 1640, with Malacca falling in 1641. In Africa, they took Sao Jorge da Mina in 1638, the coast of Angola and Benguela in 1641, but Dutch attacks failed also against Macao and in the Lesser Sunda group (Timor, Solor, Flores) and Mocambique Island (Boxer, 1969). But the Portuguese empire during its phase of decline was also attacked by Omani seapower (expedition against Mombasa in 1660–1661; Diu sacked in 1688, Bombay raided in 1661) and by the Marathas in West India. The overall picture here is of a very protracted maritime conflict that ultimately resulted in the supercession of one empire by another, yet which was not entirely conclusive with the Portuguese retaining some remnants of empire, indeed, all the way up to the 20th century in the cases of Macao, Goa, Timor and Mocambique, albeit later without the trappings of empire. The bases and settlements, and related trading benefits, were a major focus or object of the conflict. And, the changes in control of bases were, cumulatively a major reason for the outcome of the struggle as the Dutch used captured bases as jumping off points for new conquests in the East Indies, India, Ceylon and Brazil, much as had the Portuguese in their early conquests over a variety of “local powers”. The British long cycle(s) Next, Britain moved toward the assumption of the role of maritime hegemon. It was involved in several naval wars with the Dutch in the later 1600s. Following that, with the assumption of the British monarchy by the House of Orange, there was a reasonably peaceful transition from the Dutch to the British maritime empires, with Modelski citing the period 1688–1713 as one of “global war phase” with Britain and the Netherlands fighting France, followed by a British “world power phase” from 1714 to 1740 following the Treaty of Utrecht. But actually, the British accrual of a global network of bases was a gradual process amidst a multi-sided, complex struggle for colonial empire involving Britain, France, the Netherlands, and declining but still powerful (in terms of empire) Spain (Portugal 966 R.E. Harkavy / Political Geography 18 (1999) 941–972 had by this time become a marginal actor, albeit often allied with Britain). And each of the powers retained extensive basing assets even as Britain, during the 18th century, assumed the role of the leading maritime power. As it happens, the early phase of expansion of British naval access saw some use of base acquisition by diplomacy rather than by conquest. Britain under Blake’s leadership used Lisbon as a base to blockade the Spanish coast, and received as a dowry for the 1662 marriage of the British King to the sister of the Portuguese monarch, fortresses at Tangier and Bombay. Negotiations with Norway allowed for use by British ships of the port of Bergen. Still in the 1600s, the British established positions in New York and Cape Verde and Gorée Island in Africa. But in the period 1688 to 1713, Britain moved toward maritime supremacy vis a vis primarily French power, with the Netherlands having by now become aligned with Britain. Britain also nibbled away against the remnants of Spanish maritime power, as France and Spain came most often to be aligned against Britain. Around the turn of the 18th century, Britain concentrated on establishing dominance in the western Mediterranean, particularly involving control over Gibraltar and Port Mahon on Minorca. By the Treaty of Utrecht in 1713, Britain retained Gibraltar and Minorca; also retained Acadia, Newfoundland, and Nova Scotia (seized in 1710), and the coveted slaving Asiento in Africa. Britain emerged from the wars as the dominant naval power; according to Padfield, “her retention of key naval bases, her acquisition of the slaving Asiento and her penetration of both the Portuguese (1703) and Spanish trading areas were the rewards of her great naval power” (Padfield, 1982: 191) (referring, respectively, to the Indian Ocean and Caribbean areas). British power surged again in the 1740s and 1750s (a period characterized by Modelski as one of deconcentration) amidst a protracted, back and forth struggle for naval supremacy with France and Spain. There was a mixed and shared pattern of control and basing access in the key areas in contention, i.e., the Indian Ocean and South and Southeast Asia, the Caribbean and North America, as of 1750. Even then faded Dutch and Portuguese empires retained important factories, colonies and bases. Britain failed to take Cartagena, and temporarily lost Minorca and Gibraltar. One war ended inconclusively in 1747, causing Britain to hand back Louisbourg (Cape Breton) to France while the latter handed back Madras to Britain. Britain achieved effective control of much of India at this point. But subsequently, the tides turned, and by the time of the Treaty of Paris in 1763 (following major British victories at Quebec, Lagos, and Quiberon Bay), Britain’s winning position was made clear. She got the whole of Canada; in the West Indies, Grenada, St. Vincent, Dominica and Tobago, Senegal and Minorca, also Florida, and in exchange returned Belle Isle, Martinique, Guadeloupe, St. Lucia, and Gorée and part of the St. Lawrence area. Also, Britain had to give back to Spain its conquests of Cuba (it had stormed Havana in 1762) and Manila (Padfield, 1982: 248). As Britain became aligned with the Netherlands after 1688, it did not capture and take over the hitherto Dutch basing network as did the Dutch previously from Portugal in the Indian Ocean and Brazil. Some exceptions: Britain’s takeover of New York before 1688, and its later takeover of former Dutch positions in India and Ceylon, particularly the major naval bases at Colombo and Trincomalee in the latter. R.E. Harkavy / Political Geography 18 (1999) 941–972 967 At the end of the American War of Independence, Britain suffered some setbacks, which no doubt account for Modelski’s attribution of this period as one of power deconcentration. Florida and Minorca went back to Spain, Ceylon and Senegal to the Dutch, St. Lucia and Tobago to France. Then came the Napoleonic wars, during which Britain established control of the seas and was able to expand its overseas empire and corresponding basing network. At the end of it, and as part of the settlement, Britain acquired bases that were the “keys which locked up the globe”: Heligoland, Malta, the Ionian Islands, Gambia, Sierra Leone, Ascension, Capetown (perhaps the most strategic position in the world in the age of seapower), Mauritius, the Seychelles, Ceylon, Malacca, St. Lucia, Tobago, Guiana. It by then possessed major bases in every part of the world save perhaps in the central Pacific. There were later additions as the 19th century progressed: Singapore (1819), the Falklands (1833), Aden (1839), Hong Kong (1841), Lagos, Fiji, Cyprus, Alexandria, Mombasa, Zanzibar, Wei-hai-wei (Kennedy, 1976: 154– 155). By the turn of the 20th century, Britain had an unrivalled global network of bases and coaling stations, all of it on the basis of raw conquest and/or protectorates. The main bases were Esquimalt (British Columbia), Halifax, Bermuda, Jamaica (Kingston), St. Lucia, Capetown, Mombasa, Malta, Gibraltar, Freetown, Alexandria, Aden, Bombay, Hong Kong, Singapore, Sydney, and Auckland (Kennedy, 1976: 207). As it had during the Napoleonic wars a century before, Britain used its bases at the outset of World War I for the favorable prosecution of the war, while also translating its (in a coalition) victory into still more extended basing access. At the war’s outset, Britain was able to cut communications cables to Germany’s outlying few colonial possessions in Togo, Cameroon, Southwest Africa and the island groups of the Central Pacific (Kennedy, 1971). Britain’s basing structure was crucial to the prosecution of a blockade against Germany and its allies. Its basing structure in the Mediterranean–Middle East–South Asia region enabled it to conduct military operations in the Mediterranean (including the failed invasion at Gallipoli) and against Turkey in the Middle East. At the war’s close and on into the interwar period, Britain was availed of important new basing access, particularly in Iraq, Jordan, and Palestine (air bases in Iraq and at Jordan’s Aqaba, naval access at Haifa, etc.). Germany lost all of its overseas possessions at the war’s close, while France, the Netherlands, Spain and Portugal all retained colonial possessions and bases, albeit much less than Britain’s. But Britain had, according to Modelski, entered into a phase of delegitimization and deconcentration in the period 1850–1914. This was not then reflected in a diminished global basing network. Rather, it pertained to the growth of German industrial and military power on the European continent (as well as the growing German naval challenge embodied in Tirpitz’s “risk strategy”); to a lesser degree, perhaps, Russia’s rising power, albeit the maritime weakness so openly revealed in the defeat by Japan in 1905. After World War I (reflected by, among other things, the ratios established by the Washington Naval Treaty), Britain abandoned its long-term policy of fielding a navy 968 R.E. Harkavy / Political Geography 18 (1999) 941–972 equal to the next two possible rivals. In effect, it conceded naval parity with the U.S. In the process, during the late 1920s (a matter as much related to budgets as sheer strategic considerations), Britain abandoned its naval bases in Jamaica and Trinidad (always thought to provide a basing point in proximity to the Panama Canal) and at Esquimalt in British Columbia. There was a transition from British maritime dominance to a joint Anglo-American condominium, howsoever merely implicit amidst Washington’s return to a form of isolationism. As World War II began, the U.S.–U.K. Lend-Lease Agreement traded some aging destroyers for a 99 year American lease on a string of naval bases along the U.S. Atlantic littoral, from Labrador to Newfoundland, Bermuda, the Bahamas, Antigua, Trinidad, and British Guyana. It was the first big step towards a U.S. global basing network, and in Thompson’s terms of “system lineage”, reflected a shift in maritime hegemony, albeit between allies, as had earlier been the case in the hegemonic transition from the Netherlands to Britain in the 17th and 18th centuries. Noteworthy is the fact that colonial control remained primarily the basis for basing access up to and beyond World War II. The American long cycle After World War II (Modelski sees the U.S. entering its “world power phase” after 1945 and after a “global war phase” running from 1914–1945), the U.S. quickly acquired the world’s most elaborate basing network, corresponding to its role as preeminent naval power. Much of the basing structure was acquired during World War II, though it is worth noting that the U.S. was made to withdraw from some basing points after 1945 by recalcitrant wartime allies such as Brazil and Australia, respectively involving naval and air bases in northeast Brazil and some southwest Pacific island bases such as Rabaul. And to only a small degree relative to Britain in 1815 and 1919, the victorious U.S. benefitted only minimally from “imperial pickoff”, constrained by new “norms” of international organization regarding the illegitimacy of wartime conquests. The exceptions were to be U.S. control via a U.N. trusteeship over important island bases in the former Japanese island mandate groups in the Central Pacific, i.e., the Marshalls, Marianas, Carolines, and Belau. The subsequent history of the U.S. elaboration of a giant global basing network ran on two levels, as previously noted. First, the U.S. was able to utilize, at least for a decade or two, the ever-dwindling basing networks of its allied European colonial powers, i.e., Britain, France, Spain, Portugal, and the Netherlands. That was a wasting set of assets, particularly as Britain’s empire gradually eroded as had the Portuguese and Dutch empires in earlier centuries. But the U.S. compensated via arms transfers, economic assistance and formal alliances, including a host of European and Asian members of the NATO (North Atlantic Treaty Organization), CENTO (Central Treaty Organization), SEATO (Southeast Asia Treaty Organization), and ANZUS (Australia–New Zealand–United States) alliances, plus important bilateral relations with nations as disparate as Japan, South Korea, Taiwan, Ethiopia, Liberia, Morocco, Bahrein, Oman, etc., not without various ebbs and flows regarding gains and losses of access. R.E. Harkavy / Political Geography 18 (1999) 941–972 969 Beginning in the late 1950s, but accelerating and continuing on to the end of the Cold War, the USSR acquired a formidable and near-global network of bases: naval and air bases, submarine bases, technical facilities in various areas of C3I. Those bases were in all cases hosted by ideological client states that came to be allied with the USSR: Syria, Egypt (up to around 1975), South Yemen, Somalia (up to 1978), Ethiopia (after the Horn War), Cuba, Vietnam, Mongolia, Angola, Mozambique, Guinea, Algeria, etc., primarily, with some additional states providing for temporary access (port visits, air overflights) — India, Cambodia, Peru. Correspondingly, the Soviet navy grew into a formidable “blue water” force, matching the U.S. Navy in terms of deployments (measured in ship-days) in the major bodies of water, i.e., Indian Ocean, the Med, Caribbean, etc. As had become the case for the U.S., arms transfers greased most of the basing arrangements. A few major points emerge, retrospectively, concerning the Cold War “base race” as they relate to long cycle history and previous cycles of maritime supremacy. First, virtually none of either the U.S. or Soviet basing assets were taken by force or conquest; rather, access resulted from diplomacy and as a quid pro quo for security assistance, albeit on both sides with an ideological “cement”. That appears to have no historical precedent. Second, by comparison with the previous Portugal–Spain bipolar basing structures, the U.S. and Soviet bases were not sharply demarcated geographically, i.e., there was no analogy to the Treaty of Tordesillas. Rather, the basing structures overlapped in various regions, were cheek by jowl (the Soviets in Cuba and the U.S. in Puerto Rico; the Soviets in South Yemen and Somalia, the U.S. in Ethiopia). There was, however, little if any direct combat or military confrontation involving the two superpowers. The U.S. used its bases in conflicts against Soviet clients (Vietnam, Korea, various efforts at coercive diplomacy); the Soviets used theirs in relation to conflicts — sometimes by proxy — in Ethiopia and Angola. Both sides utilized strings of air transit bases to conduct long-range arms resupply operations on behalf of clients and friends. It was indeed, from this perspective, a “cold” war. Myriad pundits routinely characterize the current period as one of overwhelming U.S. global hegemony — economically, militarily, and culturally — following a period, after 1973, in which “declinism” was in vogue. Meanwhile, the Cold War over, the U.S. has seen an extensive stand-down of its hitherto far more extensive global basing network, what with the considerable withdrawal of forces from Europe, and the shutting down of previously critical facilities in Spain, Greece, Turkey, Thailand, and the Philippines, etc. Whereas earlier, in the cases of Portugal, the Netherlands and Britain, victories in global war cycles had left the victors with enhanced and unrivalled basing networks, the U.S. “victory” in the Cold War resulted in its losing much of its basing structure, although as the USSR lost all of its overseas bases, the U.S. ended up in a more hegemonic position, relatively speaking. But with alliances frayed and the increasingly prevalent “norms” about sovereignty and control of others’ territories, it was not surprising that in the 1998 crisis over Iraq, that the U.S. had evidenced great difficulty regarding access in places like Saudi Arabia, Egypt, and Turkey. Whether that constitutes the beginnings of “delegitimiz- 970 R.E. Harkavy / Political Geography 18 (1999) 941–972 ation” or “power deconcentration” in Modelski’s terms may be argued, pro and con, in historical context, with the jury still out. Conclusion Recent years have seen a number of research foci in the general domain of International Relations Theory (IR) related to the concepts of hegemony and international systems structure. More specifically, with respect both to conceptualization and empirical measurement, this has involved such questions as the measurement of the existence and degrees of global hegemony, the thesis of hegemonic stability (stability is posited as more likely in periods of hegemony), measurements of comparative power at the big power level, the attribution of constructs such as unipolarity, bipolarity and multipolarity to the structure of the international system, and measurements of “polarization” (the amount of conflict between the poles). In one way or another, all of these problems are related to or addressed by the foregoing discussion of long cycle theory, and its (ideological?) rival, world systems theory. Confusion and disagreement are rampant. Heretofore, hegemony has tended to be measured by various combinations of GNP, military expenditure and other standard measures of national power, albeit the near absence of historical data in these categories prior to the 20th century. As noted, Modelski and Thompson utilized capital ship construction and orders of battle as a primary measure of naval power and, hence, global hegemony through several cycles. And, as noted, rival long cycle theory and world systems theorists have disagreed somewhat on the identification of hegemonic periods, particularly in the instance of 16th century Portugal and Spain, wherein these powers constituted somewhat of a bipolar system albeit with a low degree of “polarization”. This paper has bruited the approach to these questions from a newer angle, that of global basing systems and “global reach”; i.e., the geography of power projection through several long cycles running from Venice (or earlier, the Mongol Empire) up to the current American hegemony. It is a subject that has received too little attention, theoretically speaking, and where there is a total dearth of organized historical data. An analysis of the basing systems of the successive hegemons and their rivals should shed more light on the phases of long cycles proffered by Modelski and Thompson; i.e., how basing systems have been reflective of phases of “rise and fall.” Such an analysis might, for instance, shed some light on the rival claims of earlier Portuguese and Spanish hegemonies; i.e., which had a more developed global reach and with what implications. An analysis of how bases have been acquired, whether via outright conquest or diplomacy, may also inform us of major shifts in the practice of international politics, including the evolvement of the current international system. For geographers, further research on global basing systems might be particularly valuable in illuminating the aforementioned problem of “system leader lineage”, whereby certain regions or places in the world — islands, choke points, littorals — have maintained a consistent strategic centrality through centuries of hegemonic R.E. Harkavy / Political Geography 18 (1999) 941–972 971 cycles, even as earlier rivalries based primarily on surface naval fleets have now given way to a spatially more multidimensional situation, involving the relationships between outer space, the underseas and the surfaces of the oceans and large continents. In a way, this brings us back to traditional geopolitical theory, Saul Cohen’s concept of “the world that matters”. Note: Lack of space precluded inclusion of what would have been a lengthy tabular presentation of all the hegemons’ bases with dates, types, etc. The author is in the process of constructing a data base that will be available in subsequent publications. Acknowledgements I would like to thank Thomas Kozlik, former Penn State undergraduate, for research assistance in connection with this work. References Benson, D.S., 1995. Six emperors: Mongolian aggression in the thirteenth century. D. Benson, Chicago. Boxer, C.R., 1965. The Portuguese seaborne empire, 1415–1825. Knopf, New York. Boxer, C.R., 1969. The Dutch seaborne empire 1600–1800. Knopf, New York. Cipolla, C., 1965. Guns and sails in the early phase of european expansion. Collins, London. Cohen, S., 1963. Geography and politics in a world divided. Random House, New York. Dismukes, B., McConnell, J., 1979. Soviet naval diplomacy. Pergamon, New York. Epstein, J., 1984. Horizontal escalation. Sour notes of a recurrent theme. Brookings Institution, Washington report. Goldstein, J., 1985. Kondratieff waves as war cycles. International Studies Quarterly 29, 411–444. Graham, G., 1965. The politics of naval supremacy: studies in British maritime ascendancy. Cambridge University Press, Cambridge. Gray, C., 1977. The geopolitics of the nuclear era. Crane Russak, New York. Gray, C., 1992. The leverage of sea power. Free Press, New York. Harkavy, R.E., 1975. The arms trade and international systems. Ballinger, Cambridge, MA. Harkavy, R.E., 1982. Great power competition for overseas bases. Pergamon, New York. Harkavy, R.E., 1989. Bases abroad. Oxford University Press, Oxford, UK. Hopf, T., 1991. Polarity, the offense-defense balance, and war. American Political Science Review 85 (2), 475–493. Kemp, G., Harkavy, R.E., 1997. Strategic geography and the changing Middle East. Brookings, Washington. Kennedy, P.M., 1971. Imperial cable communications and strategy, 1870–1914. The English Historical Review 86 (141), 728–752. Kennedy, P.M., 1976. The rise and fall of British naval mastery. Charles Scribner’s Sons, New York. Lane, F., 1973. Venice: A maritime republic. Johns Hopkins University Press, Baltimore. Mackinder, H.J., The geographical pivot of history, Geographical Journal 23 (4) (April 1904), pp. 421–444. Mahan, A.T., 1980. The influence of seapower upon history, 1600–1783. Little, Brown, Boston. Millett, A., Murray, W., 1988. Military effectiveness, 3 vols. Unwin Hyman, Boston. Modelski, G., 1978. The long cycle of global politics and the nation-state. Comparative Studies in Society and History 20, 214–235. Modelski, G., 1987. Long cycles in world politics. Macmillan, London. 972 R.E. Harkavy / Political Geography 18 (1999) 941–972 Modelski, G., Thompson, W.R., 1988. Seapower in global politics, 1494–1993. University of Washington Press, Seattle. Osterud, O., 1988. The uses and abuses of geopolitics. Journal of Peace Research 25 (2), 191–199. Padfield, P., 1979. Tide of empires, vols. I and II. Routledge and Kegan Paul, London. Padfield, P., 1982. Tide of empires, vols. I and II. Routledge and Kegan Paul, London. Parker, G., 1985. Western geopolitical thought in the twentieth century. St. Martin’s Press, New York. Parry, J.H., 1967. The Spanish seaborne empire. Knopf, New York. Porter, A.N. (Ed.), 1991. Atlas of British overseas expansion. Routledge, London. Roberts, S.S., 1977. The decline of the overseas station fleets: the United States Asiatic Fleet and the Shanghai Crisis, 1932. Arlington, VA: Center for Naval Analyses, Professional Paper No. 208, November. Rosecrance, R., 1987. Long cycle theory and international relations. International Organization 41 (2), 283–301. Rosinski, H., 1957. In: Simpson, B.M. (Ed.), The development of naval thought. Naval War College Press, Newport, RI. Scammell, G.V., 1989. The first imperial age: European overseas expansion c. 1400–1715. Unwin Hyman, London. Spykman, N., 1942. America’s strategy in world politics. Harcourt Brace, New York. Spykman, N., 1944. The geography of the peace. Harcourt Brace, New York. Swanson, B., 1982. Eighth voyage of the dragon: A history of China’s quest for seapower. Naval Institute Press, Annapolis, MD. Taylor, P.J., 1989. Political Geography. 2nd ed. Longman, Essex, UK. Thompson, W., 1997. Passing the torch in a manner of speaking: The system leader lineage. Paper presented at meeting of International Studies Association, Toronto, Canada. United States Congress, annual throughout the 1980s. Congressional Presentation for Security Assistance Programs: Fiscal Year 19 . Washington, DC: U.S. Government Printing Office. Viotti, P., Kauppi, M., 1993. International relations theory: Realism, pluralism, globalism. Macmillan, New York. Wallerstein, I., 1979. The capitalist world economy. Cambridge University Press, Cambridge. Watson, B.W., 1982. Red Navy at sea: Naval operations on the high seas 1956–1980. Westview, Boulder, CO.