On War - title image

Carl von Clausewitz

[Table of Contents]

NOTE: This version of Carl von Clausewitz's On War is the long-obsolete J.J. Graham translation of Clausewitz's Vom Kriege (1832) published in London in 1873. The 1976/84 Howard/Paret version is the  standard translation today; for the most accurate text one should always consult the 1943 Jolles translation. Consider the more modern versions and other relevant books shown below.

cover image Vom Kriege, by Carl von Clausewitz, ed. Werner Hahlweg.
From Gebundene Ausgabe - Dümmler, Bonn. Erscheinungsdatum: 1991, 19. Auflage, Nachdruck.fl.
ISBN: 342782019X.

This is the 19th German edition published by Dümmlers, Clausewitz's original publisher. It was edited by the esteemed German scholar Werner Hahlweg and is considered the standard and most accurate edition.

Jolles translation, book coverBuy the best translation—recommended for serious readers. The Book of War (The Modern Library, February 2000). ISBN: 0375754776. Clausewitz's On War and Sun Tzu's Art of War in one volume. The translation of Clausewitz's On War is the 1943 version done by German literary scholar O.J. Matthijs Jolles at the University of Chicago during World War II—not today's standard translation, but certainly the most accurate.

On War, Princeton ed.Buy the standard English translation of Clausewitz's On War, by Michael Howard and Peter Paret  (Princeton University Press, 1976/84). ISBN: 0691018545 (paperback). Kindle edition. This quite readable translation appeared at the close of the Vietnam War and—principally for marketing and copyright reasons—has become the modern standard.

Book coverVanya Eftimova Bellinger, Marie von Clausewitz: The Woman Behind the Making of On War (Oxford University Press, 2015), ISBN: 0190225432. A rich biography of Countess Marie von Clausewitz that also sheds enormous light on the life, ideas, influences upon, and character of the great military thinker himself.


Relation of the Three Arms


WE shall only speak of the three principal arms: Infantry, Cavalry, and Artillery.

We must be excused for making the following analysis which belongs more to tactics, but is necessary to give distinctness to our ideas.

The combat is of two kinds, which are essentially different: the destructive principle of fire, and the hand to hand or personal combat. This latter, again, is either attack or defence. (As we here speak of elements, attack and defence are to be understood in a perfectly absolute sense.) Artillery, obviously, acts only with the destructive principle of fire. Cavalry only with personal combat. Infantry with both.

In close combat the essence of defence consists in standing firm, as if rooted to the ground; the essence of the attack is movement. Cavalry is entirely deficient in the first quality; on the other hand, it possesses the latter in an especial manner. It is therefore only suited for attack. Infantry has especially the property of standing firm, but is not altogether without mobility.

From this division of the elementary forces of war into different arms, we have as a result, the superiority and general utility of Infantry as compared with the other two arms, from its being the only arm which unites in itself all the three elementary forces. A further deduction to be drawn is, that the combination of the three arms leads to a more perfect use of the forces, by affording the means of strengthening at pleasure either the one or the other of the principles which are united in an unalterable manner in Infantry.

The destructive principle of fire is in the wars of the present time plainly beyond measure the most effective; nevertheless, the close combat, man to man, is just as plainly to be regarded as the real basis of combat. For that reason, therefore, an army of artillery only would be an absurdity in war, but an army of cavalry is conceivable, only it would possess very little intensity of force An army of infantry alone is not only conceivable but also much the strongest of the three. The three arms, therefore, stand in this order in reference to independent value—Infantry, Cavalry, Artillery.

But this order does not hold good if applied to the relative importance of each arm when they are all three acting in conjunction. As the destructive principle is much more effective than the principle of motion, therefore the complete want of cavalry would weaken an army less than the total want of artillery.

An army consisting of infantry and artillery alone, would certainly find itself in a disagreeable position if opposed to an army composed of all three arms; but if what it lacked in cavalry was compensated for by a proportionate increase of infantry, it would still, by a somewhat different mode of acting, be able to do very well with its tactical economy. Its outpost service would cause some embarrassment; it would never be able to pursue a beaten enemy with great vivacity, and it must make a retreat with greater hardships and efforts; but these inconveniences would still never be sufficient in themselves to drive it completely out of the field.—On the other hand, such an army opposed to one composed of infantry and cavalry only would be able to play a very good part, while it is hardly conceivable that the latter could keep the field at all against an army made up of all three arms.

Of course these reflections on the relative importance of each single arm result only from a consideration of the generality of events in war, where one case compensates another; and therefore it is not our intention to apply the truth thus ascertained to each individual case of a particular combat. A battalion on outpost service or on a retreat may, perhaps, choose to have with it a squadron in preference to a couple of guns. A body of cavalry with horse artillery, sent in rapid pursuit of, or to cut off, a flying enemy wants no infantry, etc., etc.

If we summarise the results of these considerations they amount to this.

1. That infantry is the most independent of the three arms.

2. Artillery is quite wanting in independence.

3. Infantry is the most important in the combination of the three arms.

4. Cavalry can the most easily be dispensed with.

5. A combination of the three arms gives the greatest strength.

Now, if the combination of the three gives the greatest strength, it is natural to inquire what is the best absolute proportion of each, but that is a question which it is almost impossible to answer.

If we could form a comparative estimate of the cost of organising in the first instance, and then provisioning and maintaining each of the three arms, and then again of the relative amount of service rendered by each in war, we should obtain a definite result which would give the best proportion in the abstract. But this is little more than a play of the imagination. The very first term in the comparison is difficult to determine, that is to say, one of the factors, the cost in money, is not difficult to find; but another, the value of men's lives, is a computation which no one would readily try to solve by figures.

Also the circumstance that each of the three arms chiefly depends on a different element of strength in the state—Infantry on the number of the male population, cavalry on the number of horses, artillery on available financial means—introduces into the calculation some heterogeneous conditions, the overruling influence of which may be plainly observed in the great outlines of the history of different people at various periods.

As, however, for other reasons we cannot altogether dispense with some standard of comparison, therefore, in place of the whole of the first term of the comparison we must take only that one of its factors which can be ascertained, namely, the cost in money. Now on this point it is sufficient for our purpose to assume that, in general, a squadron of 150 horsemen, a battalion of infantry 800 strong, a battery of artillery consisting of 8 six-pounders, cost nearly the same, both as respects the expense of formation and of maintenance.

With regard to the other member of the comparison, that is, how much service the one arm is capable of rendering as compared with the others, it is much less easy to find any distinct quantity. The thing might perhaps be possible if it depended merely on the destroying principle; but each arm is destined to its own particular use, therefore has its own particular sphere of action, which, again, is not so distinctly defined that it might not be greater or less through modifications only in the mode of conducting the war, without causing any decided disadvantage.

We are often told of what experience teaches on this subject, and it is supposed that military history affords the information necessary for a settlement of the question, but every one must look upon all that as nothing more than a way of talking, which, as it is not derived from anything of a primary and necessary nature, does not deserve attention in an analytical examination.

Now although a fixed ratio as representing the best proportion between the three arms is conceivable, but is an x which it is impossible to find, a mere imaginary quantity, still it is possible to appreciate the effects of having a great superiority or a great inferiority in one particular arm as compared with the same arm in the enemy's army.

Artillery increases the destructive principle of fire; it is the most redoubtable of arms, and its want, therefore, diminishes very considerably the intensive force of an army. On the other hand, it is the least moveable, consequently, makes an army more unwieldy; further, it always requires a force for its support, because it is incapable of close combat; if it is too numerous, so that the troops appointed for its protection are not able to resist the attacks of the enemy at every point, it is often lost, and from that follows a fresh disadvantage, because of the three arms it is the only one which in its principal parts, that is guns and carriages, the enemy can soon use against us.

Cavalry increases the principle of mobility in an army. If too few in number the brisk flame of the elements of war is thereby weakened, because everything must be done slower (on foot), everything must be organised with more care; the rich harvest of victory, instead of being cut with a scythe, can only be reaped with a sickle.

An excess of cavalry can certainly never be looked upon as a direct diminution of the combatant force, as an organic disproportion, but it may certainly be so indirectly, on account of the difficulty of feeding that arm, and also if we reflect that instead of a surplus of 10,000 horsemen not required we might have 50,000 infantry.

These peculiarities arising from the preponderance of one arm are the more important to the art of war in its limited sense, as that art teaches the use of whatever forces are forthcoming; and when forces are placed under the command of a general, the proportion of the three arms is also commonly already settled without his having had much voice in the matter.

If we would form an idea of the character of warfare modified by the preponderance of one or other of the three arms it is to be done in the following manner:—

An excess of artillery leads to a more defensive and passive character in our measures; our interest will be to seek security in strong positions, great natural obstacles of ground, even in mountain positions, in order that the natural impediments we find in the ground may undertake the defence and protection of our numerous artillery, and that the enemy's forces may come themselves and seek their own destruction. The whole war will be carried on in a serious formal minuet step.

On the other hand, a want of artillery will make us prefer the offensive, the active, the mobile principle; marching, fatigue, exertion, become our special weapons, thus the war will become more diversified, more lively, rougher; small change is substituted for great events.

With a very numerous cavalry we seek wide plains, and take to great movements. At a greater distance from the enemy we enjoy more rest and greater conveniences without conferring the same advantages on our adversary. We may venture on bolder measures to outflank him, and on more daring movements generally, as we have command over space. In as far as diversions and invasions are true auxiliary means of war we shall be able to make use of them with greater facility.

A decided want of cavalry diminishes the force of mobility in an army without increasing its destructive power as an excess of artillery does. Prudence and method become then the leading characteristics of the war. Always to remain near the enemy in order to keep him constantly in view—no rapid, still less hurried movements, everywhere a slow pushing on of well concentrated masses—a preference for the defensive and for broken country, and, when the offensive must be resorted to, the shortest road direct to the centre of force in the enemy's army—these are the natural tendencies or principles in such cases.

These different forms which warfare takes according as one or other of the three arms preponderates, seldom have an influence so complete and decided as alone, or chiefly to determine the direction of a whole undertaking. Whether we shall act strategically on the offensive or defensive, the choice of a theatre of war, the determination to fight a great battle, or adopt some other means of destruction, are points which must be determined by other and more essential considerations, at least, if this is not the case, it is much to be feared that we have mistaken minor details for the chief consideration. But although this is so, although the great questions must be decided before on other grounds, there still always remains a certain margin for the influence of the preponderating arm, for in the offensive we can always be prudent and methodical, in the defensive bold and enterprising, etc., etc., through all the different stages and gradations of the military life.

On the other hand, the nature of a war may have a notable influence on the proportions of the three arms.

First, a national war, kept up by militia and a general levy (Landsturm), must naturally bring into the field a very numerous infantry; for in such wars there is a greater want of the means of equipment than of men, and as the equipment consequently is confined to what is indisputably necessary, we may easily imagine, that for every battery of eight pieces, not only one, but two or three battalions might be raised.

Second, if a weak state opposed to a powerful one cannot take refuge in a general call of the male population to regular military service, or in a militia system resembling it, then the increase of its artillery is certainly the shortest way of bringing up its weak army nearer to an equality with that of the enemy, for it saves men, and intensifies the essential principle of military force, that is, the destructive principle. Any way, such a state will mostly be confined to a limited theatre, and therefore this arm will be better suited to it. Frederick the Great adopted this means in the later period of the Seven Years' War.

Third, cavalry is the arm for movement and great decisions; its increase beyond the ordinary proportions is therefore important if the war extends over a great space, if expeditions are to be made in various directions, and great and decisive blows are intended. Buonaparte is an example of this.

That the offensive and defensive do not properly in themselves exercise an influence on the proportion of cavalry will only appear plainly when we come to speak of these two methods of acting in war; in the meantime, we shall only remark that both assailant and defender as a rule traverse the same spaces in war, and may have also, at least in many cases, the same decisive intentions. We remind our readers of the campaign of 1812.

It is commonly believed that, in the middle ages, cavalry was much more numerous in proportion to infantry, and that the difference has been gradually on the decrease ever since. Yet this is a mistake, at least partly. The proportion of cavalry was, according to numbers, on the average perhaps, not much greater; of this we may convince ourselves by tracing, through the history of the middle ages, the detailed statements of the armed forces then employed. Let us only think of the masses of men on foot who composed the armies of the Crusaders, or the masses who followed the Emperors of Germany on their Roman expeditions. It was in reality the importance of the cavalry which was so much greater in those days; it was the stronger arm, composed of the flower of the people, so much so that, although always very much weaker actually in numbers, it was still always looked upon as the chief thing, infantry was little valued, hardly spoken of; hence has arisen the belief that its numbers were few. No doubt it happened oftener than it does now, that in incursions of small importance in France, Germany, and Italy, a small army was composed entirely of cavalry; as it was the chief arm, there is nothing inconsistent in that; but these cases decide nothing if we take a general view, as they are greatly outnumbered by cases of greater armies of the period constituted differently. It was only when the obligations to military service imposed by the feudal laws had ceased, and wars were carried on by soldiers enlisted, hired, and paid—when, therefore, wars depended on money and enlistment, that is, at the time of the Thirty Years' War, and the wars of Louis XIV.—that this employment of great masses of almost useless infantry was checked, and perhaps in those days they might have fallen into the exclusive use of cavalry, if infantry had not just then risen in importance through the improvements in fire-arms, by which means it maintained its numerical superiority in proportion to cavalry; at this period, if infantry was weak, the proportion was as one to one, if numerous as three to one.

Since then cavalry has always decreased in importance according as improvements in the use of fire-arms have advanced. This is intelligible enough in itself, but the improvement we speak of does not relate solely to the weapon itself and the skill in handling it; we advert also to greater ability in using troops armed with this weapon. At the battle of Mollwitz the Prussian army had brought the fire of their infantry to such a state of perfection, that there has been no improvement since then in that sense. On the other hand, the use of infantry in broken ground and as skirmishers has been introduced more recently, and is to be looked upon as a very great advance in the art of destruction.

Our opinion is, therefore, that the relation of cavalry has not much changed as far as regards numbers, but as regards its importance, there has been a great alteration. This seems to be a contradiction, but is not so in reality. The infantry of the middle ages, although forming the greater proportion of an army, did not attain to that proportion by its value as compared to cavalry, but because all that could not be appointed to the very costly cavalry were handed over to the infantry; this infantry was, therefore, merely a last resource; and if the number of cavalry had depended merely on the value set on that arm, it could never have been too great. Thus we can understand how cavalry, in spite of its constantly decreasing importance, may still, perhaps, have importance enough to keep its numerical relation at that point which it has hitherto so constantly maintained.

It is a remarkable fact that, at least since the wars of the Austrian succession, the proportion of cavalry to infantry has changed very little, the variation being constantly between a fourth, a fifth or a sixth; this seems to indicate that those proportions meet the natural requirements of an army, and that these numbers give the solution which it is impossible to find in a direct manner. We doubt, however, if this is the case, and we find the principal instances of the employment of a numerous cavalry sufficiently accounted for by other causes.

Austria and Russia are states which have kept up a numerous cavalry, because they retain in their political condition the fragments of a Tartar organisation. Buonaparte for his purposes could never be strong enough in cavalry; when he had made use of the conscription as far as possible, he had no ways of strengthening his armies, but by increasing the auxiliary arms, as they cost him more in money than in men. Besides this, it stands to reason that in military enterprises of such enormous extent as his, cavalry must have a greater value than in ordinary cases.

Frederick the Great it is well known reckoned carefully every recruit that could be saved to his country; it was his great business to keep up the strength of his army, as far as possible at the expense of other countries. His reasons for this are easy to conceive, if we remember that his small dominions did not then include Prussia and the Westphalian provinces. Cavalry was kept complete by recruitment more easily than infantry, irrespective of fewer men being required; in addition to which, his system of war was completely founded on the mobility of his army, and thus it was, that while his infantry diminished in number, his cavalry was always increasing itself till the end of the Seven Years' War. Still at the end of that war it was hardly more than a fourth of the number of infantry that he had in the field.

At the period referred to there is no want of instances, also of armies entering the field unusually weak in cavalry, and yet carrying off the victory. The most remarkable is the battle of Gross-gorschen. If we only count the French divisions which took part in the battle, Buonaparte was 100,000 strong, of which 5,000 were cavalry, 90,000 infantry; the Allies had 70,000, of which 25,000 were cavalry and 40,000 infantry. Thus, in place of the 20,000 cavalry on the side of the Allies in excess of the total of the French cavalry, Buonaparte had only 50,000 additional infantry when he ought to have had 100,000. As he gained the battle with that superiority in infantry, we may ask whether it was at all likely that he would have lost it if the proportions had been 140,000 to 40,000.

Certainly the great advantage of our superiority in cavalry was shown immediately after the battle, for Buonaparte gained hardly any trophies by his victory. The gain of a battle is therefore not everything,—but is it not always the chief thing?

If we put together these considerations, we can hardly believe that the numerical proportion between cavalry and infantry which has existed for the last eighty years is the natural one, founded solely on their absolute value; we are much rather inclined to think, that after many fluctuations, the relative proportions of these arms will change further in the same direction as hitherto, and that the fixed number of cavalry at last will be considerably less.

With respect to artillery, the number of guns has naturally increased since its first invention, and according as it has been made lighter and otherwise improved; still since the time of Frederick the Great, it has also kept very much to the same proportion of two or three guns per 1,000 men, we mean at the commencement of a campaign; for during its course artillery does not melt away as fast as infantry, therefore at the end of a campaign the proportion is generally notably greater, perhaps three, four, or five guns per 1,000 men. Whether this is the natural proportion, or that the increase of artillery may be carried still further, without prejudice to the whole conduct of war, must be left for experience to decide.

The principal results we obtain from the whole of these considerations, are—

1. That infantry is the chief arm, to which the other two are subordinate.

2. That by the exercise of great skill and energy in command, the want of the two subordinate arms may in some measure be compensated for, provided that we are much stronger in infantry; and the better the infantry the easier this may be done.

3. That it is more difficult to dispense with artillery than with cavalry, because it is the chief principle of destruction, and its mode of fighting is more amalgamated with that of infantry.

4. That artillery being the strongest arm, as regards destructive action, and cavalry the weakest in that respect, the question must in general arise, how much artillery can we have without inconvenience, and what is the least proportion of cavalry we require?


Return to the top of this page

Go to next chapter

Go to Table of Contents

Visit The Clausewitz Homepage logo Visit the
Clausewitz Bookstores!
US UK France Germany