The Concept of Military Alliance

Stefan Bergsmann
The Concept of Military Alliance
1. Introduction
1.1. Alliances are a central phenomenon in world politics
Alliances are a central and constant phenomenon in international politics throughout
history. Whether we look at ancient periods, at the Middle Ages or at the centuries of
Bismarck or Napoleon, we find states forming alliances. As George Liska has put it, "It is
impossible to speak of international relations without referring to alliances; the two often
merge in all but name."1
1.2. Yet there exists no accepted definition of the concept of alliance
Reflecting this important role of alliances in world politics, the literature in international
relations has produced quite an impressive list of interesting studies, articles and analyses in
this area of research.2 However, it seems striking that despite this scholarly assiduity not
much thought has been given to the question "What is a military alliance?"3 Now one might
argue that there exists such a broad consensus about the concept that no further analysis is
needed. But exactly the opposite is the case. As Edwin Fedder stated in his conceptual
analysis in 1968, "the concept of alliance in the literature of international relations is
ambiguous and amorphous."4 Five years later, Holsti/Hopmann/Sullivan came to the same
conclusion by observing "the lack of an accepted definition of alliance."5 Although quite
many outstanding studies have been published in the area of alliance research since then, we
are still lacking a concise, theoretically useful and practical definition of the concept of
military alliance.
1.3. This article wants to contribute to fill this gap
This article is an exercise in conceptual analysis and wants to contribute to fill this gap.
First, it will present some considerations about concepts in general. Following this, the
1 George Liska (1968): Nations in Alliance. The Limits of Interdependence, paperback edition, Baltimore, p.3.
A similar statement is made by Holsti/Hopmann/Sullivan (1973), who consider alliances to be „a universal
component of relations between political units, irrespective of time or place“, Ole R. Holsti/Terrence P.
Hopmann/John D. Sullivan (1973): Unity and Disintegration in International Alliances, Lanham/New
York/London, p.2.
2 A compilation of the most important alliance literature is found in Stefan Bergsmann: Warum entstehen
Bündnisse? Konzepte und Theorien der Allianzbildung in Europa, Sinzheim: ProUniversitate Verlag, and in
Holsti/Hopmann/Sullivan (1973).
3 A careful query in the literature brought about 35 different definitions, only one conceptual analysis done by
Edwin Fedder in 1968 and two lexical articles by Stephen Walt and Arnold Wolfers. Edwin H. Fedder
(1968): "The Concept of Alliance", In: International Studies Quarterly V.12 N.1, pp.65-86; Arnold Wolfers
(1968): "Alliances", In: International Encyclopedia of the Social Sciences, ed. By David L. Sills, V.1,
pp.268-271; Stephen M. Walt (1993): "Alliance", In: The Oxford Companion to Politics of the World, ed. By
Joel Krieger et al., New York/Oxford, p.20.
4 Liska (1968), p.70.
5 Holsti/Hopmann/Sullivan (1973), p.3.
21
common meaning of the term "alliance" and a few prominent definitions of the concept that
are found in the alliance literature are discussed. On the basis of this analysis, a new definition
of the concept of military alliance will be put forward, defining an alliance as an explicit
agreement among states in the realm of national security in which the partners promise mutual
assistance in the form of a substantial contribution of resources in the case of a certain
contingency the arising of which is uncertain.
Thus, the definition leaves aside other forms of cooperation in the military and nonmilitary
field and allows the researcher to focus on the core element of military alliances: the
assistance clause. While this might seem too narrow a focus for researchers who deal with
alliances in a broader sense, it is argued that concepts have to be that narrowly defined in
order to allow further theorizing and comparisons and to avoid confusion in the theoretical
discussions. Thus, researchers that focus on other aspects of cooperation or alliances in a
broader sense are encouraged to define their objects of research in a similar way with the
vision to get to a more general but still concise concept by generalizing from the new
definitions at a later point in time.
2. Theoretical Considerations about Concepts in General
2.1. Concepts can be seen as parts of the triangle concept-term-referent
Concepts are the bricks out of which theories are built. However, their definition is quite
often difficult and contested. In the classical view, concepts are ideas of something that exists
in the world. Therefore, they are related to a term which denotes the concept, on the one hand,
and referents which correspond to it in reality, on the other hand.6
Figure 1: The Triangle Concept-Term-Referent
concept
alliance
term referent
alliance, pact, NATO, WEU
security agreement, Dual Alliance
etc. etc.
2.2. Confusion arises from the different conception of this triangle among authors
Problems mainly arise because this triangle varies a little bit from author to author. First,
the concepts are defined differently among authors. Second, some terms are sometimes used
synonymously or for different concepts. Thus, we find in the literature, for example, the terms
alliance, coalition, pact and bloc sometimes used interchangeably, while other scholars
6 For a different view of concepts see for example the very inspiring study of George Lakoff: Women, Fire,
and Dangerous Things. What Categories Reveal about the Mind, Chicago/London 1987. However, for the
task undertaken in this paper the classical approach to concepts was found to be more useful than other
approaches such as those developed by Lakoff and others.
22
distinguish among them along various criteria. Besides these terms, the related concepts of
entente, alignment, neutrality and non-aggression pact are also used widely but not uniformly.
Finally, even the referents which are seen as the central elements or exemplars of the defined
category differ from author to author: some see, for example, NATO as the prototype of an
alliance, others the alliances against Napoleon or those of Bismarck. So, while this analysis
concentrates on the definition of the concept of alliance, we always have to keep also the
other two related elements of the triangle above in mind.
2.3. A concept has to cover the referents but also to be theoretically useful
It is obvious that it is necessary to define a concept in such a way that it covers the
phenomena one wants to deal with. In other words, it has – to some extent – to correspond to
reality. Less common sense, however, is the necessity of the definition of a concept to be at
the same time theoretically useful. This means that it has to be as clear as possible and to
contain only a few essential elements. This is important because only then clear conclusions
can be drawn during later exercises in theorizing. Imagine, for example, a concept that is
defined by a great number of qualities with different consequences following from each of
them. Even if such a definition may cover the reality better than another, more sparse one, it is
of only little theoretical use. To illustrate the point one just has to look, for example, at the
"operational typology of international military alliances" developed by Bruce Russet in 1971.7
Russett did a good job in identifying all kinds of characteristics of different types of alliances,
e.g. duration, nature of commitment, etc. Such a multitude of elements seems necessary to
differentiate between different kinds of alliance arrangements in a typology. A definition of
the concept that shall be promising for theoretical purposes, however, must concentrate on
only a few characteristic elements on the basis of which theoretical conclusions can be
developed. Therefore, Russett’s typology, while being a useful tool for describing actual
alliances, is not very promising for theoretical purposes. A definition must try to capture the
most essential characteristics of a phenomenon that are important for further theorizing
efforts. Only then conclusions can be drawn from the defined qualities that can be subjected
to empirical testing. If such a theoretical finding – in case it would be confirmed – would then
be valid for a specific referent in the real world, depends on how close this referent
corresponds to the definition of the concept and, in particular, to the few isolated defining
characteristics.
2.4. The theoretical usefulness ranks higher than the empirical accuracy
Thus, everybody trying to define a concept faces a dilemma: on the one hand, one has to
try to cover the corresponding referents which common sense has designated by using the
term that denotes our concept as good as possible; on the other hand, one has to leave aside
the multitude of attributes and to concentrate on only a few essential characteristics in order to
have a useful basis for further theorizing.8 Being in this dilemma, I would argue for the
greater importance of the second point: the concentration on only a few essential qualities in
order to have a concept that is theoretically useful ranks higher than its claim to cover all the
referents that exist in the real world. The reason for this is quite simple: firstly, concentration
on certain qualities makes it easier to identify causal relations: conclusions can be drawn from
crucial qualities of the concept; secondly, by referring to these qualities instead of the real
7 Bruce Russett (1971): "An Empirical Typology of International Military Alliances", In: Midwest Journal of
Political Science V.15, pp.262-289.
8 This problem can be observed with almost any concept. A very intensive conceptual debate was going on,
e.g., over the definition of the concept of small states.
23
world category it might be possible to explain some apparently deviant cases: cases, that lack
certain effects or outcomes just because they lack the crucial quality although belonging to the
real world category according to common sense; thirdly, it becomes easier to transfer possible
findings from one realm to another: other cases that belong to a very different category but are
followed by the same effects might thus be explained by looking if the crucial quality is
present.
Figure 2: A Theoretically Useful Concept
"Deviant Cases"
Referents in
reality
other realms
and
phenomena
Concept
3. The Practical Usage of the Concept of Alliance
Apart from the theoretical usefulness and the accuracy two other important criteria have to be met:
1) the definition has to stick as closely as possible to the common meaning of a term in ordinary
language so as to minimize the danger of confusion, and
2) the concept shall relate to existing definitions of the most celebrated writers in the area of
research.9
This section will try to meet these two criteria: first, the common meaning of the term
"alliance" and its roots will be identified; following this, some of the important definitions of
the concept of alliance put forward by leading scholars will be discussed.
3.1. The common meaning of the term "alliance": a connection between actors
As for the first, the ordinary meaning of the term, a short look into Webster's Encyclopedic
Dictionary may suffice. There, an alliance is defined as10
1a) the state of being allied
b) a bond or connection between families, parties, or individuals
2a) an association (as by treaty) of two or more nations to further their common interests
b) a treaty of alliance
The essential point here seems to be the element of a connection between two or more actors.
9 Compare the rules of conceptual analysis as developed by Malthus, Oppenheim and Machlup. David A.
Baldwin (1980): "Interdependence and Power: a conceptual analysis", In: International Organization V.34
N.4, pp.471-506.
10 Webster's Encyclopedic Dicitionary, revised edition, New York 1995, p.27.
24
This meaning also corresponds to the historical development of the term "alliance". The
corresponding German term "Allianz" developed in the 17th century out of the French word
"alliance", which meant "connection", "pact" or "association between states". This French
term, in turn, goes back to the old French verb aleier (= to connect, to combine, to join),
which is itself rooted in the Latin verb alligare. Alligare, finally, is a compositum of Latin
"ligare" (= to bind) and could be translated as "to bind" and "to combine".11
As shown, the common usage of alliance defines it as some kind of connection between
actors to achieve some common goals. However, it is quite evident that this definition is much
too broad for scientific purposes: according to this almost everything – e.g., such different
associations as the International Postal Union, the G-7 or the Non-Proliferation-Treaty –
would count as alliances, although all of them have very different consequences. Thus, the
common meaning has to be restricted.
3.2. Its usage by famous political scientists: a specific kind of relation between states to
achieve certain goals
In order to do this, we shall proceed by looking at the usage of the concept by some of the
leading writers in the area of alliance research.
A short look at the different definitions found in the alliance literature shows that most
authors use the concept in a sense that is very similar to the common usage identified above. All
authors see alliances as a specific kind of relation between states to achieve certain goals.
Concerning the concrete type and intensity of the relation and concerning the kind of goals an
alliance serves to achieve, however, the authors differ widely. As for the intensity, the spectrum
reaches from some kind of loose cooperation12, to concerted action13 and a closer association14,
and, finally, to very specific agreements in the form of a formal treaty15. The same is found
concerning the goals of such an alliance: some authors just speak of some general policy as the
common goal16, others restrict this policy to the realm of national security17 or – in some cases –
name some very specific goal, such as the fighting of a war together18.
11 Duden. Etymologie. Herkunftswörterbuch der deutschen Sprache. Ed. By Günther Drosdowski, Paul Grebe
et al., Mannheim/Wien/Zürich 1963, pp.19-20.
12 E.g. Jack S. Levy/ Michael M. Barnett (1991): "Domestic Sources of Alliances and Alignments: The Case of
Egypt, 1962-1973", In: International Organization V.45 No.3, p.370; Stephen M. Walt (1987): The Origin
of Alliances, Ithaca/New York, p.12; Stephen M. Walt (1993): "Alliance", In: Joel Krieger et al. (Eds.): The
Oxford Companion to Politics of the World, p.20.
13 E.g. Fedder (1968), p.68.
14 E.g. Wichard Woyke (1983): "Militärbündnisse", In: Handlexikon zur Politikwissenschaft, Ed. by Wolfgang
M. Mickel, München, p.292; Jellinek and Despagnet, both cited in Rafael Erich (1907): Über Allianzen und
Allianzverhältnisse nach heutigem Völkerrecht, Helsingfors, p.23 and p.24; Glenn H. Snyder (1990):
"Alliance Theory: A Neorealist First Cut", In: Journal of International Affairs V.44 N.1, p.104.
15 E.g. Ken Booth (1987): "Alliances", In: John Baylis/Ken Booth/John Garnett/Phil Williams (Eds.): Contemporary
Strategy, Vol.1, 2nd edition, London/Sidney, p.258; Erich (1907), p.15; Wilhelm G. Grewe
(1970): Spiel der Kräfte in der Weltpolitik. Theorie und Praxis der Internationalen Beziehungen,
Düsseldorf/Wien, p.105; Holsti/Hopmann/Sullivan (1973), p.5; Dan Reiter (1994): "Learning, Realism, and
Alliances. The Weight of the Shadow of the Past", In: World Politics V.46 N.4, p.495; Russett (1971),
pp.262-263; David J. Singer/Melvin Small (1966): "Formal Alliances, 1815-1939: a Quantitative Description",
In: Journal of Peace Research V.3 N.1, p.4; Gerald L. Sorokin (1994): "Arms, Alliances, and Security
Tradeoffs in Enduring Rivalries", In: International Studies Quarterly V.38 No.3, p.423; George A. Lopez/
Michael S. Stohl (1989): International Relations. Contemporary Theory and Practice, Washington, p.367.
16 E.g. Funk-Brentano/Sorel, Bonfils and Pradier-Fodére, all cited in Erich (1907), p.23; Stohl/Lopez (1989), p.367.
17 E.g. Levy/Barnett (1991), p.370; Booth (1987), p.258; Fedder (1968), p.68; Holsti/Hopmann/Sullivan
(1973), p.5, Walt (1987), p.12; Woyke (1983), p.292.
18 Barry Posen (1984): The Sources of Military Doctrine, Ithaca/New York, p.62.
25
3.3. The definition of George Liska: a formal association between two or more states against
the threat of a third
One of the most important early works on alliances, George Liska’s study on Nations in
Alliance, does not offer the reader an explicit definition of the concept at all. However, from
the text it becomes apparent that Liska sees an alliance basically as a formal association
between two or more states against the threat of a third, more powerful state. The association
itself is what Liska calls an "alignment" and corresponds to the predictions of the balance of
power theory. Alliances for him – although he uses the terms almost interchangeably – merely
formalize these alignments.19 Conflicts are thus for him the primary determinant of
alignments and alliances, with threat and power potential used as synonyms as usual in the
balance of power literature: "Alliances are against, and only derivatively for, someone or
something".20
The problem here is that Liska already includes a possible reason for the formation of
alliances in his (implicit) definition: threat. If an alliance is stipulated as an association against
a threat from outside then balancing against threats is what one is going to find by analyzing
the alliance policy of states. Because the definition already includes one of the hypotheses to
be tested, it is not useful for theoretical purposes.
3.4. The implicit definition of Melvin Small and David Singer: a data set
A second definition that was very important in the alliance literature was put forward by
Melvin Small and David Singer. Actually, they did not really formulate a definition of the
concept but collected a data set on formal alliances which was widely used by other scholars.
By using their data set these researchers also accepted the criteria according to which the two
authors had collected and selected their data.21
Small/Singer differentiate three types of alliance in their collection of data: 1) defense
pacts, 2) neutrality and non-aggression pacts, and 3) ententes.22 In addition, it was required
that at least two of the signatories of an alliance treaty hat to be independent nation-states with
a population of more than half a million people and their sovereignty recognized by the two
leading nations in the period analyzed: Britain and France. Second, they included only
alliances that were in the form of a written, formal agreement. And finally they also excluded
all alliances that were formed during a war or within three months before war broke out and
also a rather large class of treaties "because they did not reflect, in any appreciable fashion,
the coalitions and divisions in the system."23
This data set is problematic in several respects. First, it mixes up treaties of a very different
nature: defense pacts provide for mutual military assistance in the case of an attack against
one of the partners; neutrality pacts bind the signatories not to intervene in a conflict between
19 Liska (1968), p.3.
20 Liska (1968), p.12.
21 E.g. George T. Duncan/Randolph M. Siverson (1982): "Flexibility in Alliance Partner Choice in Multipolar
Systems", In: International Studies Quarterly V.26 No.4, pp.511-538, Randolph M. Siverson/Juliann
Emmons (1991): "Birds of a Feather. Democratic Political Systems and Alliance Choices in the Twentieth
Century", In: Journal of Conflict Resolution V.35 N.2, pp.285-306, David Lalman/David Newman (1991):
"Alliance Formation and National Security", In: International Interactions V.16 N.4, pp.239-253.
22 Small/Singer (1966), p.5.
23 Small/Singer (1966), p.5. Excluded were a) collective security agreements, b) charters and constitutions of
international organisations such as the ILO, c) treaties of guarantee to which all relevant powers had given
their consent like the Locarno Pact of 1925, d) agreements limited to general rules of behavior such as the
Briand-Kellogg Pact, and e) unilateral guarantees such as the US commitment to the protection of the
Isthmus of Panama of 1903.
26
others; in non-aggression pacts, which are mostly formed between hostile nations, the partners
promise not to attack each other; an entente, finally, provides just for consultations in the case
of one of the partners being attacked by a third state. So all of these treaties provide for a
certain behavior in the case of a conflict. However, the kind of behavior provided for is so
different in these four cases that – for general theoretical purposes – it does not seem to be
legitimate to put them all into one data set,24 even more so, as each behavior is followed by
very different consequences. But similar to other arrangements, alliances are formed because
of their expected consequences. If, because of the definition of the concept, the consequences
vary so widely, however, an analysis of the reasons for their formation, e.g., is no longer
possible.
A second problem with the data set is the exclusion of all war time alliances. Although for
the purpose of the authors this exclusion was important, for other research efforts it might
cause serious problems. Because alliances are evidently connected to national security and
conflict situations, part of the relevant data is thus excluded from the beginning.25
In sum, the usage of the data set of Small/Singer has to be decided case by case. For the
problem dealt with here – i.e. the definition of the concept of alliance –, however, it is not
very useful.
3.5. The definition of Ole Holsti, Terrence Hopmann, and John Sullivan: a formal treaty
between nation-states concerned with national security issues
Another important definition in the literature was developed by Ole Holsti, Terrence
Hopmann and John Sullivan. For them three elements are essential for an association to
qualify as an alliance:
1) a formal treaty – open or secret
2) it must be directly concerned with national security issues
3) the partners must be nation-states
On the basis of these three elements the authors define the concept as follows: "an alliance
is a formal agreement between two or more nations to collaborate on national security
issues."26 Compared to the two definitions discussed before, the one here is clearly a step
forward: it is explicit and confines itself to a few objective criteria.
Nevertheless, here also some criticism is in order. First of all, this definition is simply too
broad: it covers for example the SALT treaties as well as Austria’s neutrality or the
Partnership for Peace Accord. Second, it requires a formal treaty, which is actually not really
necessary. The point is rather that both partners know of their commitment, i.e., that it is
made explicit among them. This usually is done by signing a treaty, however, one could also
imagine other forms of explicit statements. Third, – but this may be an individual preference
based on my German-speaking origin – it is not necessary that states are nation-states to form
an alliance. Alliances have been formed long before the rise of the nation as the stateconstituting
element in the era of Napoleon; they will be formed long after the end of
Nationalism.
24 For the purpose of the authors, which was to analyse the correlation between alliances and the outbreak of
war, the data set may nevertheless be useful.
25 A list of the ommitted war time alliances can be found in Holsti/Hopmann/Sullivan (1973) in appendix A.
26 Holsti/Hopmann/Sullivan (1973), p.4.
27
3.6. Stephen Walt's definition of alliances: arrangements for security cooperation among
states
The last important definition in the literature that shall be discussed here is the one
developed by Stephen Walt, who is without doubt one of the main researchers in the area of
alliances. He uses alignment and alliance as synonyms and defines the latter in his Origins of
Alliances as "a formal or informal arrangement for security cooperation between two or more
sovereign states".27
Again we face the same problem as with Holsti/Hopmann/Sullivan: the definition is so
broad that almost every security arrangement will qualify as an alliance thereby making
theorizing impossible. In order to be theoretically useful the concept has to be defined in such
a narrow way that it covers only phenomena that are sufficiently similar to allow conclusions
that are valid for all of them. The origins of the open skies agreement in the framework of the
conference on security and cooperation in Europe are probably not the same as those of the
treaty of Dunkirk concluded between France and Great Britain in 1947. Yet both would
qualify as an alliance according to Walt’s definition.
In a later article Walt modified this definition a little bit and wrote "an alliance is a
cooperative security relationship between two or more states, usually taking the form of a
written military commitment."28 However, the concrete content of this commitment is still not
clear, nor when it should come into force.
3.7. Summary of the critique: existing definitions are too broad and vague
The analysis of some of the most important definitions of the concept of alliance in the
alliance literature showed three important things:
1) there is no single definition that is accepted by all or most of the authors,
2) not much energy has been spent up to now to develop a theoretically useful and practical
definition of the concept of alliance,
3) the existing definitions are only of limited use because most of them are too vague and too
broad.
In the remainder of this article the attempt should therefore be made to develop a new
definition of the concept on the basis of the discussion above.
4. A More Concise Definition of the Concept
As said above, a concept is an idea of something formed by mentally combining its
attributes. The common meaning of the term alliance did this in a fashion that seemed too
broad for analytical purposes. The various definitions of the scholars mentioned did narrow
this somehow, but still not enough. None of them seems to be covering all the necessary
qualities that make an alliance. Clearly, we need a new definition that – on the basis of these
findings – tries to correct these shortcomings.
27 Walt (1987), p.12. In footnote 1 at p.1 he uses the word "relationship" instead of "arrangement".
28 Walt (1993), p.20.
28
But what are the essential elements of an alliance? After carefully analyzing some real
alliances and comparing them with other forms of cooperation and association29 I would
suggest the following eight elements:
1. Alliances are arrangements between states: important here is, however, not if a state is
formally recognized or accepted in the international community; it is only necessary that
there exists an independent authority which has the power to rule over a certain
population and territory. This is important because only then the possibility exists to
mobilize and dispose of power capabilities – a necessary precondition in an alliance.
2. Alliances are explicit agreements: it does not matter if the agreement is made explicit by
a hand-shake between statesmen or by a formal treaty. The important thing is only that
the participating parties themselves know with certainty that an agreement exists. Thus,
they can calculate on this basis and form their expectations accordingly. This criterion
distinguishes alliances from alignments which are only informal groupings of states
based upon interests that give rise to mere implicit expectations.
3. Alliances deal with a certain behavior for a certain contingency in the future. Although
most alliances also comprise some activities that take place for the whole time the treaty
is in force, such as coordination of doctrines or joint exercises, the main part of an
alliance is focused on a specific behavior that shall be followed in the event of a certain
situation, the so-called casus foederis. This element distinguishes alliances from mere
security cooperations or from non-aggression pacts which promise a certain behavior for
the full period of duration of the agreement.
4. In connection with the last element, it is essential that the event for which the specified
behavior is promised is uncertain: the partners do not know, when this occasion will
occur nor if it will occur at all. This separates alliances from actual coalitions, which are
formed in anticipation of a decision that will take place for certain at a more or less
known point of time – such as an election or a war, for example. This element of
uncertainty is very important because the pros and cons of the promise for a specific
behavior to be expected differ decisively compared to a situation of certainty: when a
state joins a war coalition, entanglement into the conflict is certain, when a state joins an
alliance, however, entanglement is only a possibility that does not necessarily have to
occur.
5. An alliance is a promise. Therefore, it has to be distinguished from the actual behavior
shown by the state once the casus foederis has occurred. From this element together with
element 3 described above follows the inherent insecurity of alliances and, therefore, –
from the perspective of the allying partners – the problem of credibility or the risk of
abandonment, which both cover just different views of the same problem.
6. The promise comprises an assistance in the event specified in the treaty (usually an attack
on one of the partners). This assistance comes up to the use of one’s own resources for
the defense of the other. How this is done in particular and exactly which kind of
resources are covered by the alliance is not that important, the point is, however, that
each of the partners can calculate with a substantial external contribution to its own
resources in the case of an actual occurrence of the casus foederis. This element
distinguishes alliances from neutrality pacts and from ententes: whereas neutrality pacts
promise only not to augment the adversary’s resources, the promise of an entente
comprises only the vague commitment of consultations in case of a crisis.
29 The definition was developed during my studies on alliances for my master’s thesis. See Stefan Bergsmann:
Warum entstehen Bündnisse? Sinzheim 1996.
29
7. The promise is a mutual one. This means that each of the partners has to calculate not
only the advantages of external assistance in the case of a serious threat but also the
disadvantages of the risk of getting entangled in conflicts of the partner and, thus, of
suffering high costs should this risk become reality. Unilateral guarantees are in many
respects very similar to alliances, however, they differ in this point because in a guarantee
relation one partner worries only about the risk of abandonment whereas the other is only
concerned about the risk of entanglement.
8. Last but not least, the agreement falls into the realm of national security. This element –
although maybe seeming obvious – is also very essential because only in this realm the
risk is so high as to cover the question of the further existence of a state as a sovereign
entity. This gives alliances a quality of seriousness that clearly distinguishes them from
agreements in other areas of foreign policy, such as finance or commerce. By covering
the question of national security the risks that are inherently entailed in an alliance
become not just matters of cost but matters of life and death.
Putting together these eight constituting elements an alliance shall be defined as an explicit
agreement among states in the realm of national security in which the partners promise mutual
assistance in the form of a substantial contribution of resources in the case of a certain
contingency the arising of which is uncertain.
5. Summary
This article tried to develop a theoretically useful definition of the concept of military
alliance. Starting from older conceptual efforts by Edwin Fedder the paper analyzed the
common sense meaning of the term as well as the definition and use of the concept by leading
researchers in the field of alliance theory. However, it is argued here, all definitions developed
so far are not clear, concise and narrow enough to be a useful basis for further theorizing.
Therefore, it was attempted to identify the eight constituting elements of the concept and to
put them together into a new definition in hopes that it may be of greater theoretical value
than its predecessors. Thus, it is proposed here to define an alliance as an explicit agreement
among states in the realm of national security in which the partners promise mutual assistance
in the form of a substantial contribution of resources in the case of a certain contingency the
arising of which is uncertain.
Although this definition will for sure not be the last one to be developed it tries to put
forward the idea of defining concepts more concisely by focusing on a small number of key
features hoping that such definitions will contribute to avoid confusion in theoretical
discussions and to get even more inspiring results out of theorizing and comparisons.
6. References30
Booth Ken (1987): "Alliances", In: John Baylis/Ken Booth/John Garnett/Phil Williams (Eds.):
Contemporary Strategy, Vol.1, 2nd edition, London/Sidney, p.258.
Boutros-Ghali Boutros (1963): Contribution a une Théorie générale des Alliances, Paris.
30 A full list of all the literature used during the development of this paper is given in Bergsmann Stefan:
Warum entstehen Bündnisse? Sinzheim.
30
Christensen Thomas J. / Snyder Jack (1990): „Chain Gangs and Passed Bucks: Predicting
Alliance Patterns in Multipolarity“, In: International Organization V.44 No.2 (1990),
pp.137-168.
Duncan George T. / Siverson Randolph M. (1982): "Flexibility in Alliance Partner Choice in
Multipolar Systems", In: International Studies Quarterly V.26 No.4, pp.511-538.
Erich Rafael (1907): Über Allianzen und Allianzverhältnisse nach heutigem Völkerrecht,
Helsingfors, p.23 and p.24.
Fedder Edwin H. (1968): "The Concept of Alliance", In: International Studies Quarterly V.12,
pp.65-86.
Gärtner Heinz (1997): Modelle Europäischer Sicherheit. Wie entscheidet Österreich?, Wien.
Holsti Ole/ Hopmann Terrence P. / Sullivan John D. (1973): Unity and Disintegration in
International Alliances, Lanham/New York/London
Keohane Robert O. / Martin Lisa L. (1995): „Toward an Institutional Theory of Alliances“,
Paper prepared for the Meeting of the International Studies Association in Chicago,
21-25 February 1995.
Lalman David / Newman David (1991): "Alliance Formation and National Security", In:
International Interactions V.16 No.4, pp.239-253.
Levy Jack S. / Barnett Michael M. (1991): "Domestic Sources of Alliances and Alignments:
The Case of Egypt, 1962-1973", In: International Organization V.45 No.3, p.370.
Liska George (1968): Nations in Alliance. The Limits of Interdependence, Baltimore
Lopez George A. / Stohl Michael S. (1989): International Relations. Contemporary Theory
and Practice, Washington, p.367.
Morgenthau Hans J. (1959): „Alliances in Theory and Practice“, In: Wolfers (1959): Alliance
Policy in the Cold War, Baltimore, pp.184-212.
Reiter Dan (1994): "Learning, Realism, and Alliances. The Weight of the Shadow of the
Past", In: World Politics V.46 No.4, p.495.
Rothstein Robert L. (1968): Alliances and Small Powers, New York.
Ruge Friedrich (1972): Bündnisse in Vergangenheit und Gegenwart, 2. Aufl., Frankfurt.
Russett Bruce (1971): "An Empirical Typology of International Military Alliances", In:
Midwest Journal of Political Science V.15, pp.262-289.
Singer David J. / Small Melvin (1966): "Formal Alliances, 1815-1939: a Quantitative
Description", In: Journal of Peace Research V.3 No.1, p.4. Siverson R. M. / Emmons
J. (1991): "Birds of a Feather. Democratic Political Systems and Alliance Choices in
the Twentieth Century", In: Journal of Conflict Resolution V.35 No.2, pp.285-306.
Snyder Glenn H. (1990): "Alliance Theory: A Neorealist First Cut", In: Journal of
International Affairs V.44 No.1, p.104.
Sorokin Gerald L. (1994): "Arms, Alliances, and Security Tradeoffs in Enduring Rivalries",
In: International Studies Quarterly V.38 No.3, p.423.
Verosta Stephan (1971): Theorie und Realität von Bündnissen. Heinrich Lammasch, Karl
Renner und der Zweibund (1897-1914), Wien.
Walt Stephen M. (1993): "Alliance", In: The Oxford Companion to Politics of the World, ed.
By Joel Krieger et al., New York/Oxford, p.20.
31
Walt Stephen M. (1987): The Origin of Alliances, Ithaca/New York, p.12.
Waltz Kenneth N. (1979): Theory of International Politics, New York
Wolfers Arnold (1968): "Alliances", In: International Encyclopedia of the Social Sciences, ed.
By David L. Sills, V.1, pp.268-271.
Woyke Wichard (1983): "Militärbündnisse", In: Handlexikon zur Politikwissenschaft, Ed. by
Wolfgang M. Mickel, München, p.292.