Battle of the Somme

Background

By the end of 1915 the Western Front was a continuous line of trenches from the Channel to Switzerland. At Chantilly in December the Allies agreed to attack together on all main fronts in 1916, and Joffre and Haig chose the Somme, where the British and French sectors met, for a combined offensive in which the French were at first to provide the larger share.

The German attack at Verdun on 21 February 1916 changed that. As French divisions were fed into the mincer on the Meuse, the French contribution on the Somme shrank, and the British, who had been expected to play a supporting role, became the main attackers north of the river. The offensive also became, in part, a means of relieving Verdun, and its date was brought forward.

The ground suited the defender. The Germans had held the chalk uplands north of the Somme since 1914 and had dug deep dugouts, fortified villages and a second position behind the first, with a third under construction. The British army that would attack was largely new: Kitchener's volunteer divisions, many of them 'Pals' battalions raised from single towns, workplaces or sports clubs, with little experience of a set-piece attack.

South of the river, and on the British right north of it, the French Sixth Army under Fayolle, part of Foch's Northern Army Group, faced a weaker German line held by the 2nd Army under Fritz von Below. Haig hoped for a breakthrough towards Bapaume; Rawlinson, commanding the Fourth Army, preferred a methodical 'bite and hold'. The plan that went forward sat uneasily between the two.

The armies

The British Expeditionary Force under Haig attacked mainly with Rawlinson's Fourth Army, with a diversion by the Third Army at Gommecourt and, from July, Gough's Reserve Army (later Fifth Army) on the northern flank. Its divisions were a mix of regulars, Territorials and New Army volunteers, joined over the summer by Australians, New Zealanders, Canadians, South Africans, Newfoundlanders and Indian cavalry. Its artillery fired over 1.5 million shells in the preliminary bombardment, but much of it was light shrapnel, ineffective against wire and dugouts, and up to 30 per cent of the shells may have failed to explode.

The French Sixth Army (Fayolle), joined from September by the Tenth Army (Micheler) south of the river, was a veteran force with a far denser concentration of heavy guns per yard of front and infantry that had learnt to advance in small groups by bounds.

The German 2nd Army, split on 19 July into a 1st Army north of the river under Below and a 2nd Army south of it under Gallwitz, defended from deep positions with machine guns sited for interlocking fire. Falkenhayn's insistence that lost ground be retaken by counter-attack, carried out by Below's chief of staff Fritz von Loßberg, cost the Germans heavily. New weapons on both sides included the British Mark I tank and, from September, German Albatros fighters.

The battle

CDB90 records the whole battle as engagement 304 (FIRST SOMME) and four of its days as separate engagements; the narrative below names them as it reaches them.

1 July: the first day. After a bombardment begun on 24 June, the Hawthorn Ridge mine went up at 07:20 and the rest at 07:28; at 07:30 the infantry left their trenches on a front of about 15 miles. CDB90 engagement 305 (SOMME-FOURTH ARMY ATTACK) covers the Fourth Army's day. From north to south: the Third Army's diversion at Gommecourt failed; VIII Corps was cut down in no man's land before Serre and Beaumont-Hamel; the 36th (Ulster) Division broke into the Schwaben Redoubt north of Thiepval but could not hold it; and astride the Albert–Bapaume road the 8th and 34th Divisions were shattered before Ovillers and La Boisselle. CDB90 engagement 306 (SOMME-OVILLERS) is the 8th Division's attack: 5,121 casualties against 280 in the defending Infantry Regiment 180. In the south XV and XIII Corps took Mametz and Montauban. Beside them, the French Sixth Army overran the German front on both banks, took Assevillers and Herbécourt south of the river and some 2,000 prisoners, for about 1,590 casualties. British losses were 57,470, 19,240 of them killed.

July: grinding forward. Through early July the British cleared the woods and villages of the first position at high cost, while the French pressed on south of the river towards Péronne. CDB90 engagement 307 (SOMME-BAZENTIN RIDGE) is the dawn attack of 14 July: four divisions formed up in no man's land by night and went in at about 03:25 behind a short hurricane bombardment, taking the German second position at Bazentin-le-Grand, Bazentin-le-Petit and the south of Longueval. Cavalry of the Deccan Horse and 7th Dragoon Guards rode towards High Wood that evening, but the wood was not held. Then came weeks of attrition: the South Africans in Delville Wood from 15 July, the Australians at Pozières from 23 July, Guillemont (3 September) and Ginchy (9 September). On 12 September the French made their largest attack of the battle north of the river.

15 September: Flers–Courcelette. CDB90 engagement 308 (SOMME-FLERS-COURCELETTE) is the opening day of the third phase. At 06:20 the Fourth Army and the Canadians of the Reserve Army attacked with tanks for the first time: 49 were earmarked, 32 reached the start line and about 18 went into action. Flers, Martinpuich, Courcelette and at last High Wood fell; the advance was about 2,500 yards, 3,500 in the centre.

Autumn. Morval (25 September) and Thiepval Ridge (26 September) followed, and Combles fell to a joint Anglo-French advance. In October rain turned the chalk to mud; the attacks towards Le Transloy and on the Ancre Heights made little headway, while the French fought for Sailly-Saillisel. The last effort, on the Ancre from 13 November, took Beaumont-Hamel, a first-day objective. Haig closed the offensive on 18 November.

Aftermath

The Allies had advanced about six to seven miles on a front of some 16 miles, short of Bapaume and Péronne. British Empire casualties were about 420,000 and French about 200,000; German losses are the most disputed figure of the battle, with serious estimates ranging from about 240,000 to 680,000 and most modern accounts settling between 450,000 and 600,000.

The battle cost Falkenhayn his post at the end of August; Hindenburg and Ludendorff replaced him, ordered the Siegfriedstellung (the Hindenburg Line) built behind the Somme from September, and withdrew to it in March 1917. German veterans remembered the Somme as the first true Materialschlacht, a battle of material, and some historians argue it was a German defensive victory (Barton, via s1).

In Britain it became a byword for futility. The image of brave soldiers sacrificed by incompetent generals took hold between the wars and again in the 1960s. Historians since Terraine, including Sheffield and Philpott, argue that Britain had no strategic alternative in 1916, that the battle helped save Verdun and wore down the German army, and that the BEF learnt there the methods it used to win in 1918; critics such as Prior and Wilson find Haig's conduct of it largely wasteful. The Thiepval Memorial carries more than 72,000 names of British and South African soldiers missing on the Somme. France's considerable share of the battle is often left out of the British story.

Wargaming it

The Somme is a campaign, not a battle. The natural game is a single sector on a single day at battalion or company scale (a stand as a platoon or company): Ovillers or La Boisselle on 1 July, Montauban for a British success, the French south bank for a contrast, or a tank-supported attack on Flers. For the operational picture, a division-level map game of the July–November attrition, with fatigue, rotation and replacement modelled, shows why both sides bled.

Key terrain is low but decisive: the spurs and fortified villages from Thiepval to Fricourt, the woods (Mametz, Trônes, Delville, High Wood) that became traps, and the Ancre valley splitting the British front. Command friction should dominate: fixed artillery programmes, barrages that lift on a timetable whether or not the infantry keep up, and messages that take hours to reach anyone who can act on them.

The decisions are worth arguing over. Should the British have attacked in rushes rather than at a walk, or gone for Rawlinson's limited objectives? Could the French success on 1 July and the Bazentin break-in on 14 July have been exploited with reserves close at hand? Should the tanks have been held back until there were hundreds? On the German side, a player can test Falkenhayn's rule of no ground yielded against an elastic defence. A scoring system based on casualties inflicted and divisions worn out, rather than ground taken, captures the attritional logic better than one built on breakthrough.

Maps and pictures

Pictures from the time

Photographs by the armies' photographers and the press, and drawings made at the time: free-to-use pictures from the battle's category on Wikimedia Commons. Each opens its file page there, with the full caption and source.