Alliances, we had always felt, were not our sort of thing. They would involve us in obscure quarrels and sordid rivalries which were none of our concern. They seemed to be both undesirable and unnecessary in view of our special geographic and political circumstances.

"It is our true policy to steer clear of permanent alliance with any portion of the foreign world": it was George Washington's Farewell Address to us. The inaugural pledge of Thomas Jefferson was no less clear: "Peace, commerce, and honest friendship with all nations-entangling alliances with none."

It became more than a policy; it became an expression of a national point of view about ourselves and our place in the world, a view which contrasted the simple virtues of our Republic with the subtle and complex qualities (some said corruptions) of Europe. From 1789 until the Second World War, excepting only our relationship with Panama, the United States refused to enter into treaties of alliance with anyone. In the 25 years since the end of the war, however, in a dramatic reversal of national policy, we have allied ourselves with half the world.

Was this wise? There has been a growing national sense of unease about the extent of our commitments, and more than a suspicion that we were imprudent to disregard the counsel of those who founded the Republic. There has been a feeling that we may have taken on too much in the way of military obligations abroad, especially in Asia. There has also been a sense of bafflement and frustration in trying to ascertain exactly what these commitments are which have sent our young men into the jungles and bogs of Annam and Tonkin, Laos and Cochin China.

After entering into the multilateral treaty which created the United Nations (the United Nations Charter), the United States negotiated three groups of alliances. The first-the Rio Pact-was contracted with the states of Latin America in 1947. The second-the North Atlantic Treaty-was contracted with the states of Western Europe in 1949. The third was the series of treaties with respect to Asia and the Pacific (including the SEATO treaty) which were contracted in the decade 1950-1960.

Our alliance relationships are not, of course, as tidy as this outline into groupings makes them seem. In addition to the three series of treaties described above there are some special cases: for example, a treaty with Liberia, reflecting historic ties; a treaty with Panama, because of the canal; a treaty with Pakistan, a remainder of which would become the CENTO Pact.

The Department of State informed the Senate Foreign Relations Committee in 1967 that in sum all of our treaties created alliance relationships with 42 countries. The count can never be really exact, because there is an ambiguity about the current status of some of the national signatures: Cambodia, for example, signed a Protocol to the SEATO treaty but subsequently indicated a lack of interest in receiving whatever protection the treaty might afford. Whether one counts Cambodia as a treaty ally or not, therefore, is open to question.

These, then, are our treaties of alliance. Were we wrong to conclude them? Was it because of them that we found ourselves pulled into the quicksands of Southeast Asia? Will they draw us into other conflicts we may not now foresee? What can or should we do about them?


Treaties of alliance are overvalued. In the century or more before 1941 the United States and Great Britain had no treaty of alliance, but acted as allies none the less. Where identity of interests exists between the parties, a treaty usually is not necessary; where it does not exist, often a treaty is not kept.

The first question to ask about any alliance commitment-that is, a commitment to go to war in given future circumstances-is: Why make it? In public as in private life, it is best to reserve decision until the event is known. The only thing we know with certainty about the future is that we do not know what it will bring. Therefore any advance commitment to take future action-especially the most extreme type of national action, going to war-should be contracted only with the utmost reluctance and only where there is an overwhelming necessity for doing so. As Palmerston once told Talleyrand:

We have no objection to treaties for specific and definite and immediate objects, but we do not much fancy treaties which are formed in contemplation of indefinite and indistinctly foreseen cases. We like to be free to judge of each occasion as it arises, and with all its concomitant circumstances and not to be bound by engagements contracted in ignorance of the particular character of the events to which they are to apply.

This gives rise to the standard which should be applied in reviewing commitments to defend allies against future attack: such commitments should be terminated unless it can be clearly shown they are vital to our national interests.

In this discussion, I shall divide our alliance commitments into three categories, which I shall call (1) de jure commitments; (2) apparent commitments; and (3) de facto commitments.

1. By "de jure commitments" I mean the category of promises which are binding at law and for the nonperformance of which, should we consent to be sued before an appropriate international tribunal, we would be held in default.

Contrary to what appears a widespread misapprehension, there is no outstanding de jure commitment by the United States, whether by treaty or otherwise, to go to war in defense of any foreign nation or nations whatsoever.

It is often and erroneously said that the NATO treaty pledges us to automatic defense of an attacked ally (see, for example, the caption of The New York Times diagram of "U.S. Commitments Abroad" of Thursday, February 19, 1970). This is not the case. Testifying in 1949 on behalf of the administration before the Committee on Foreign Relations of the United States Senate in support of ratification of the North Atlantic Treaty, Dean Acheson, then Secretary of State, stated: "This naturally does not mean that the United States would automatically be at war if one or the other signatory nations were the victim of an armed attack." Appearing as a private witness in support of the treaty, John Foster Dulles similarly stated: "Certainly the treaty does not impose any automatic duty to declare war."

The language of the North Atlantic Treaty seems stronger than it legally is. It states (Article 5) that "The Parties agree that an armed attack against one or more of them in Europe or North America shall be considered an attack against them all . . . ," and the Parties pledge that in the event of such attack each signatory shall take "such action as it deems necessary, including the use of armed force, to restore and maintain the security of the North Atlantic area" (emphasis added). It is evident that a signatory of the treaty may deem it unnecessary to take action of any sort to meet such an attack. Legally speaking, no signatory of the treaty is therefore bound by Article 5 to take any action whatsoever. The Rio Pact with our Latin American allies uses similar language with similar legal effect; it makes explicit (Article 20) that ". . . no State shall be required to use armed force without its consent."

It may well be that the most significant legal impact of the Rio and North Atlantic treaties will be felt in constitutional law. It is a constitutional principle that the President of the United States is empowered to repel enemy attacks on us but requires a Congressional Declaration of War in order to take other or further measures of war. If it is now agreed by treaty that an attack on a Latin or NATO ally is deemed an attack on the United States, then it can be argued that the President is empowered without Congressional authorization to send us to war in such case; and the significant legal effect of the Rio and North Atlantic treaties would be to transfer from the Congress to the President the constitutional power to declare war, in so far as European and Latin American events are concerned.

The treaties with our Asian allies are somewhat more cautious in tone. The bilateral treaties (with Japan, the Philippines, Nationalist China and South Korea), the ANZUS pact[i] and the SEATO pact[ii] (much discussed because, by Protocol, South Vietnam falls within the purview of the provisions here described), recognize merely that an armed attack on a signatory in the relevant geographical area would be dangerous to the peace and security of the others. Each signatory declares that in the event of such attack it "would act to meet the common danger in accordance with its constitutional processes." The action each party might take is unspecified. Explicit reference to "constitutional processes" indicates an awareness that even should an incumbent President of the United States of America determine that going to war was the proper response to the common danger, the ultimate determination would be made by the Congress.

There is no pledge in these treaties that the United States will go to war. Neither is there such a pledge in the Declaration on the Neutrality of Laos (July 23, 1962), in which we pledge (Article 4) nothing stronger than consultation. This Declaration is essentially an undertaking to respect the neutral character of Laos and the particular policy of neutrality outlined in 1962 by the Laotians; should anyone endanger that neutrality, we agree to consult with other signatories about the situation.

Senators and others who seek the explanation of our foreign military involvements in treaty provisions are looking in the wrong place. There are no treaties which require us to go to war. We have no legal commitments to send troops to defend foreign countries.

To some extent, in thinking so, we fall victim to a semantic confusion. Administration officials will often say troops have been sent abroad "under" or "pursuant to" a given treaty without adding, as they should, that sending troops is not, however, required by the treaty in question.

2. By "apparent commitments" I mean those obligations we believe ourselves to have undertaken, or which others believe us to have undertaken, even though no de jure commitment has been undertaken. These may become political, as distinct from legal, commitments.

One finds something of the same sort in the law of contracts. Although our traditional American authority, Professor Williston, emphasized objective criteria in determining the validity and proper interpretation of contracts, the other leading view, which stemmed from Professor Corbin and found its most advanced exponent in Professor Sharp, stressed subjective criteria, arguing that determination of these issues should be consistent with the legitimate expectations of the parties.

It is useful to apply the perceptions of the Corbin-Sharp school to the area of treaty commitments, because what people in fact expect is politically more significant than what they have a legal right to expect. Apparent commitments are actually of inordinate importance in the real world of international affairs.

Apparent commitments are contracted by treaties like ours, which seem to commit us but legally do not. If we consider an attack on any European ally to be an attack upon us (as we say we do in the NATO treaty), surely anyone would assume we mean to go to war in the event of such an attack. Again, when we agree with Liberia that "In the event of aggression or threat of aggression against Liberia, the Government of the United States of America and the Government of Liberia will immediately determine what action may be appropriate for the defense of Liberia," it sounds very much as though we regard ourselves as responsible for the defense of Liberia, even though all we commit ourselves to do is make a determination: i.e. think about it. That is the trouble with our many and far-flung treaties: they promise nothing, but suggest everything.

Statements by members of our administration; executive agreements with foreign countries; communiqués after international conferences-if carelessly phrased, these too can seem to commit the United States to all sorts of actions throughout the world, even when legally they should have no such effect.

The tendency of all these apparent commitments is to push us into actions we are not legally bound to take. In terms of domestic politics, a President must be in a strong position indeed to refrain from armed intervention in defense of an ally whom the public considers us obligated to defend; he becomes a President who lost a war for us. Overseas, too, a President runs the risk that other allies will desert our cause if we seem to abandon a treaty obligation to meet force with force, even though legally there is no such treaty obligation.

Again, we may find (as Sir Edward Grey did with France in 1914) that an ally has made military dispositions on the basis of our supposed obligation to go to war; to leave such an ally to defend itself dooms it to defeat. No more than Sir Edward do we want to have foreign ambassadors "wait to learn if the word 'honour' should be erased from the English dictionary."

In sum: to the extent that we find ourselves drawn into wars, it is not because of legal obligations, but in spite of the limits we have placed on our legal obligations.

3. By "de facto commitments" I mean such material dispositions as make it practically unfeasible for us to refrain from action. These are usually, though not always, military dispositions.

A prime example is our maintenance of substantial American armed forces in Europe. An attack by Warsaw Pact countries against Western Europe would necessarily strike at American armed forces, and we would perforce be at war. This, rather than the language of the North Atlantic Treaty, binds us to defend the West-a fact that is very well understood by our European allies.

As with apparent commitments, it is difficult to assess the extent of our de facto commitments. The Senate Foreign Relations Committee was unable to gauge the depth of the military involvements described to it in 1967 by the State Department: ". . . U.S. military installations in some 30 foreign nations added an unknown number of nations to the list of those that the United States might feel obligated to defend."

We have no de jure commitments to send troops abroad. But sending troops abroad creates de facto commitments. How do we know the extent of such commitments? Only by knowing what will happen in the future. How can we know what will happen in the future? We cannot.


Precision is the aim of legal draftsmanship, in the formulation of a treaty as in the formulation of any other legal document. Elegant discriminations are not merely possible-Professor Morgenthau tells us that Sanskrit has 16 different words designating distinct types of alliance-but desirable, because each party must understand the treaty terms and obligations with exactitude if the legitimate expectations of the other parties are to be fulfilled. Clarity is especially needed in the key provisions: (1) the extent and duration of the commitment; (2) definition of the commitment; and (3) casus foederis. Judged by this standard, our treaties of alliance are thoroughly inadequate.

1. Our treaties are exact in describing the geographic areas to which their provisions relate. However (apart from the NATO treaty), they are largely silent about their duration. That is, they precisely delineate their limits in space but not in time.

This is not to say that our alliance commitments run forever. While each treaty must be interpreted individually, in the case of treaties of alliance it is "not difficult . . . to establish that the parties did not intend a perpetual duration, and therefore that the treaties contain an implied power to denounce upon reasonable notice."[iii]

It cannot be said that duration provisions are de rigueur. Some alliance treaties have them, others not. The postwar military pacts between communist countries are mostly 20-year treaties; the NATO treaty, too, provides a right of denunciation by any signatory after 20 years. In times past, alliances have been permanent or temporary, depending on the nature of the case. It would seem better practice, considering that world conditions change, to contract only for a set term of years; treaties can, after all, be renewed if the parties desire.

2. With respect to definition of our commitment, the treaties are vague, and fail to specify what action, if any, is required. Testifying in 1967 before the Committee on Foreign Relations of the U.S. Senate, the then Under Secretary of State, Mr. Nicholas Katzenbach, said:

. . . the underlying framework concerning collective security in the past 20 years has been a series of treaty obligations and legislative provisions- the United Nations Charter, coupled with treaties with 42 countries. . . .

Let me emphasize the constitutional quality of these commitments. By their nature, they set only the boundaries within which the United States will act. They cannot and do not spell out the precise action which the United States would take in a variety of contingencies. That is left for further decision by the President and the Congress (emphasis added).

If, however, the treaties fail to specify what action the signatories are committing themselves to take, what is the point of having treaties? The question is fundamental, for the ambiguity at the very center of these treaties accurately reflects an ambiguity in policy.

On the one hand, neither the Senate nor the President has been prepared to bind us in advance to go to war in future circumstances under conditions of which we can now know nothing. On the other hand, the then Chairman of the Senate Foreign Relations Committee, Senator George of Georgia, declared in 1955 when approving the SEATO treaty that it "is inspired by the conviction that a potential aggressor may be deterred from reckless conduct by a clear- cut declaration of our intentions."

There is, if not an absolute contradiction, at least a tension between the desire to reserve judgment and the desire to spell out in advance what our future actions will be. They can both be done, but it requires careful consideration of how and in what context statements about future actions are made.

Treaties are the wrong instrument to use in carrying out the particular policy of deterrence indicated by Senator George. They would be appropriate only if we were following a policy of deterrence-through-certainty: a policy of publicly, clearly and definitely indicating in advance where we are prepared to fight, in order to keep other nations from invading such areas in the mistaken belief that we will not defend them. Within its own terms such a policy makes good sense, for it prevents other countries from blundering into war with us. To work, such a policy requires that we succeed in persuading others that we will really do what we say: contracting and publishing clear and unequivocal treaty commitments is the way to do this, for the solemn and binding character of such commitments compels belief.

The problem is that the United States is not prepared to do what is necessary for such a policy. It is a political fact of life that our Senate will not commit itself in advance to declare war conditioned on future circumstances and happenings. Therefore we must resign ourself to the conclusion that, desirable as the policy of deterrence-through-certainty may be, it is not a policy which we are able to follow because we are not prepared to pay the necessary price for it.

It is also evident that we do not intend to pursue the contrary policy, the traditional European policy of deterrence-through-uncertainty, whose ambiguities are sometimes called a cause of World War I. This is the policy of keeping the enemy from acting by keeping him unsure where, when and how we will react. Clearly we do not pursue this policy, for we publish our intentions to the world.

Our actual policy is to state our intentions so that all will be aware of the risks inherent in attacking particular areas of the world which it is likely (though not certain) we will choose to defend. There is much to be said for this policy but not much to be said for the way we carry it out.

Our policy is to deter aggression by publicly clarifying our intentions, but we employ an inappropriate form of communication which obscures them. Instead of using treaties we should be using the forms normally employed in communicating governmental thinking, such as diplomatic notes, statements by the President, and Senate resolutions. These are inherently credible, for they represent current policy and they do not pretend to be what they are not; they are not masquerading as commitments.

Using treaties to communicate intentions, however, as we now do, defeats our purpose; it leads to misunderstanding. Treaties, like any other legal documents, are documents of definition and therefore of limitation. Of necessity they indicate not merely what we are obligated to do but also what we are not obligated to do. Use of the treaty format therefore emphasizes that we do not undertake any legal obligation to defend our allies; and this can lead to the belief that in fact we will not defend them, which may not be the case at all.

It can also lead to the contrary misunderstanding: the belief that we definitely undertake to defend them. Treaties have traditionally embodied binding commitments and are intended and generally understood to embody binding commitments. Therefore statements of mere intention, when placed in the form of a treaty, appear to imply legal commitments we are not making.

If we are not prepared to decide in advance when we will go to war, we ought not to have treaties which purport to be treaties of mutual defense or alliance. Such treaties begin by fooling others, who mistakenly think we have agreed to defend them; they end by fooling us, when apparent treaty commitments generate political pressures which impel us to defend foreign countries despite the lack of any legal treaty commitment to do so.

3. "Casus foederis" is a technical phrase used in international law to designate the circumstances which bring a treaty obligation into play. For example, if we agree to assist A should he be attacked by B, an attack by B upon A is the casus foederis giving rise to our obligation to come to A's assistance.

The casus foederis in treaties of alliance, though phrased with precision, sometimes must be phrased with delicacy. Though alliances from most ancient times have frequently specified the countries they are aimed against, it often seems more diplomatic to not name names. Thus Germany, Italy and Japan in 1940 made reference to the United States in a treaty provision as "a Power not at present involved in the European War or in the Sino- Japanese War." Similarly, the military pacts between communist countries are normally directed against Germany and Japan or their allies (which is supposed to mean us).

Our own treaties of alliance are vague with respect to the casus foederis. Only in the SEATO treaty do we make it clear whom we propose to defend foreign nations against ("only . . . communist aggression") ; for the rest our loose phraseology reads as though we were concerned to protect everyone against everyone else. Again, this is a vagueness which accurately reflects a vagueness in policy.

There are four mutually inconsistent American policies with respect to casus foederis. Quite apart from the fact that at least two of them (A and D) are politically and militarily absurd, it is interesting to note that the two relevant policies with respect to casus foederis (what we say and what we do) are reflected in none of our treaties; while the two policies which are reflected in our treaties run counter to both what we have done in the past and what we say we will do in the future.

Policy A is reflected (SEATO aside) in almost all of our treaties. It pledges us to do something against armed attacks from any source against any of our allies. Article 43 of the United Nations Charter may go further; read broadly, it may pledge us to send troops to maintain peace anywhere and everywhere in the world.

Policy B is reflected in the "Understanding of the United States" appended to the SEATO treaty. It pledges us to act against armed attack only where it constitutes "communist aggression" against our SEATO allies.

Policy C, which is not reflected in any of our treaties, is the policy which the present administration states that it intends to follow. It was outlined by President Nixon in his first annual foreign affairs message to Congress, submitted on February 18, 1970, entitled United States Foreign Policy for the 1970s: A New Strategy for Peace. In this message President Nixon asserts that the communist world is no longer monolithic and implies that we will respond to different communist countries in different ways. He also suggests that American resources must be conserved for use against the major and vital dangers threatening us. Therefore, with respect to Asia and the Pacific, he asserts that "We shall provide a shield if a nuclear power threatens the freedom of a nation allied with us, or of a nation whose survival we consider vital" (emphasis added). In all other cases, President Nixon limits our role to providing aid, not armies. Thus his policy is to send troops only in case of attack by the two communist superpowers, China or Russia, and not in case of attack by any other communist nation.

Policy D, which also is not reflected in any of our treaties, is the policy which the United States has actually followed in practice under every administration since World War II. Russia and China have invaded several countries during this period of time. Russia invaded Czechoslovakia and Hungary; China invaded Tibet and India. These were important areas: Czechoslovakia, for example, for its industry, and Tibet for its strategic position. We did not intervene in any of these cases. On the other hand we have gone to war against North Korea and North Vietnam.

This policy is exactly the opposite of the policy enunciated by President Nixon. In practice the communist invasions we have not opposed have been those of Russia and China; we have intervened only against their satellites, that is to say, against only those communist nations which constitute no material threat to our security and whose defeat would not substantially improve our position in the world.

Thus with respect to casus foederis it can be seen that our treaties as such are defective-they are misleading, and vague to the point of meaninglessness-but since we do not follow the policies they reflect, our predicaments in the world are not caused by them. The fault, if fault there be, lies elsewhere.


A review of our alliance commitments has shown that we have no legal obligations to defend our allies, but that we do have substantial political and military commitments to friendly nations throughout the world.

In Europe, we have made our most solid commitment: our troops are stationed there indefinitely. This seems to represent a commitment which, in conscious consensus, we deem necessary.

In Latin America our far-ranging political commitment probably is an accurate reflection of a special concern and position in matters which affect our own hemisphere. They are the other aspect of the Monroe Doctrine.

In Asia and the Pacific, however, our treaties fail to reflect accurately our current intentions. President Nixon intends to defend our allies in that area only against armed attack by a nuclear power; but the treaties give the impression that we will defend them against others too. At the very least, therefore, even if our government deems it inadvisable to revoke or revise our other treaties of alliance, these should be revoked, or else they should be revised so as to eliminate all reference to action by us or use of our armed forces, if we are not to be politically burdened with responsibilities we are unwilling to assume. A Resolution of the Senate could appropriately express a present intention-not a commitment-to meet force with force should a nuclear power attack one of our allies in Asia or the Pacific.

This would perform a useful function in curbing the inflated expectations aroused by our present Asian treaties.

In evaluating our remaining commitments around the world, the criteria set by the Nixon administration are as follows: "We will view new commitments in the light of a careful assessment of our own national interests and those of other countries, of the specific threats to those interests, and of our capacity to counter those threats at an acceptable risk and cost" (emphasis added). This does not go far enough. These sensible criteria, which President Nixon proposes to use only in contracting new commitments, also must be applied, as Professor Morgenthau has pointed out, to the review of existing commitments.

How can this be done? How do we know the width and depth and implications of the commitments arising from our present situation and policies? How can we ascertain their extent in order to limit them in accordance with the President's criteria?

Senator Percy has proposed the issuance of an annual list-a sort of balance sheet on the liability side-of national commitments. It is an interesting suggestion, but nonlegal commitments are a question of perception rather than accounting: they have to be sensed rather than enumerated. Senator Ervin, more realistically, has proposed that the Senate assert a constitutional right to determine when and whether troops may be sent abroad. This surely would focus attention on the more obvious commitments we contract, and require conscious decision before each such commitment is made. The problem of political commitment remains unsolved, however, and no legal measure will answer it.

Recognition that the problem is not a legal one-is not, therefore, a problem merely on the surface of things-troubles us with questions about the nature of man and political society which most of us prefer not to face, but which we have to face in order to formulate effective policies. In all likelihood those who blame our legal treaty commitments for our involvements in tangled world affairs hope their allegation is true. It would make things easy; treaties, unlike passions and problems, can simply be revoked.

But the forces which draw us into the tides of international policies are of the real world; they are not mere pedantries or legalisms. The causes of our involvement are more complex, the cures more uncertain, than words or phrases or clauses in the corpus juris. Our fall from grace-if we have fallen from grace-is not due to our treaties of alleged alliance. The problems we face are not legal. They are practical problems, and can be solved only through prudent, experienced judgment combined with a good bit of luck. The law and the treaties, in this respect, are genuinely of no importance. The alliances did not entangle us; either life itself did it, or we did it to ourselves.

[i] Security treaty between Australia, New Zealand and the United States (1951).

[ii] South East Asia Collective Defense Treaty (1954). Signatories were the United States, Australia, France, New Zealand, Pakistan, Philippines, Thailand and the United Kingdom. Cambodia, Laos and South Vietnam were included by Protocol for some purposes, but Cambodia and Laos subsequently disclaimed the benefits of the Protocol.

[iii] Arnold Duncan McNair, "The Law of Treaties." New York: Columbia University Press, 1938, p. 373.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now