On War

by Carl von Clausewitz

5. If there is one or if there are several fortresses near the frontier, then the great question arises, whether the defender should seek an action before or behind them. The latter recommends itself—

a, by the superiority of the enemy in numbers, which forces us to break his power before coming to a final struggle.

b, by these fortresses being near, so that the sacrifice of territory is not greater than we are compelled to make.

c, by the fitness of the fortresses for defence.

One principal use of fortresses is unquestionably, or should be, to break the enemy’s force in his advance and to weaken considerably that portion which we intend to bring to an engagement. If we so seldom see this use made of fortresses, that proceeds from the cases in which a decisive battle is sought for by one of the opposing parties being very rare. But that is the only kind of case which we treat of here. We therefore look upon it as a principle equally simple and important in all cases in which the defender has one or more fortresses near him, that he should keep them before him, and give the decisive battle behind them. We admit that a battle lost within the line of our fortresses will compel us to retreat further into the interior of the country than one lost on the other side, tactical results in both cases being the same, although the causes of the difference have their origin rather in the imagination than in real things; neither do we forget that a battle may be given beyond the fortresses in a well chosen position, whilst inside them the battle in most cases must be an offensive one, particularly if the enemy is laying siege to a fortress which is in danger of being lost; but what signify these nice shades of distinction, as compared to the advantage that, in the decisive battle, we meet the enemy weakened by a fourth or a third of his force, perhaps one half if there are many fortresses?

We think, therefore, that in all cases of an inevitable decision, whether sought for by the offensive or the defensive, and that the latter is not tolerably sure of a victory, or if the nature of the country does not offer some most decisive reason to give battle in a position further forward—in all these cases we say when a fortress is situated near at hand and capable of defence, the defender should by all means withdraw at once behind it, and let the decision take place on this side, consequently with its co-operation. If he takes up his position so close to the fortress that the assailant can neither form the siege of nor blockade the place without first driving him off, he places the assailant under the necessity of attacking him, the defender, in his position. To us, therefore, of all defensive measures in a critical situation, none appears so simple and efficacious as the choice of a good position near to and behind a strong fortress.

At the same time, the question would wear a different aspect if the fortress was situated far back; for then it would be necessary to abandon a considerable part of our theatre of war, a sacrifice which, as we know, should not be made unless in a case of great urgency. In such a case the measure would bear more resemblance to a retreat into the interior of the country.

Another condition is, the fitness of the place for defence. It is well known that there are fortified places, especially large ones, which are not fit to be brought into contact with an enemy’s army, because they could not resist the sudden assault of a powerful force. In this case, our position must at all events be so close behind that we could support the garrison.

Lastly, the retreat into the interior of the country is only a natural resource under the following circumstances:—

a, when owing to the physical and moral relation in which we stand as respects the enemy, the idea of a successful resistance on the frontier or near it cannot be entertained.