The size of the invading army

Jumièges says that William built a fleet of 3000 ships. It is usually discounted as being unrealistically high because Normandy did not have enough lumberjacks, carpenters and shipwrights to build that many sea-going vessels in the time between William’s commitment to an invasion and their departure from Dives. Indeed, being clinker design, it is difficult to believe that Normandy had enough high-quality 200-year-old oak trees or grown timbers for more than a hundred or so new ships.

Lawson makes a case that many of William’s mercenaries might have brought their own ships and that they might all have been several times larger than normally believed. This is part of his argument that the armies might have been much larger than usually assumed. It is possible but we think unlikely. If the English army had 20000 or more men, as Lawson suggests, we think Harold would have encircled the battlefield 30 deep, with no possibility of the Normans breaking through in a day.

Wace says that there were 696 ships in the Norman battle fleet. Brevis Relatio says 782. Both exclude cargo carriers. The ‘Ship List’ supplement to the Battle Manuscript says that William had at least 1000 ships at his disposal for the invasion. Gillmor has verified that 700 was roughly the upper limit of troop and horse carriers that could have left St Valery on one tide at that time of year. Perhaps the discrepancy between Wace, Brevis Relatio and the Ship List is those lost on the crossing or, as Wace says, some of the non-war horses were probably brought on skiffs. We will assume that the number of troop carriers and horse carriers was close to Wace’s figure, say 700. Wace says that 400 of them were horse carriers, leaving 300 troop carriers.

Trying to assess the size of the Norman army is more subjective. Reputable estimates vary widely: between 3000 and 12000 for the infantry; between 1000 and 3000 for the cavalry. Peter Poyntz Wright estimates that Normandy could field 1200 knights on their own. Wace infers that Normans made up more than half the invading army. Even if all the Norman knights were mobilised, William’s cavalry could not have been much more than 2000. The horse carriers depicted on the Tapestry have 10, 8, 4 or 3 horses, although it might only be figurative. Gillmor has calculated that most of the Norman fleet was re-purposed from existing cargo vessels. Typical cargo ships of the day would carry no more than four horses. An average of five, equating to 2000 cavalry, does not seem an unreasonable upper limit.

Rupert Furneaux estimates the number of Norman troops from the size of the ships shown on the tapestry.  He comes up with 7500, including 2000 knights. Poyntz Wright compared the Norman fleet to other contemporary battle fleets to arrive at 3000 infantry and 800 archers. Domesday and other records suggest that William gifted land to 7000 of his supporters. This figure includes more than a thousand non-combatants. It presumably excluded some fatalities, but we think that number was low, probably no more than 100. This suggests the number of invading combatants was near 6000.

Carmen comments on the number of invaders singing as they leave St Valery: “Quippe decem decies[ , ] decies et milia quinque diversis feriunt vocibus astra poli". Thierry, Giles, Barlow, Morton & Muntz and others use a transcription with a comma after the first ‘decies’, making the translation: “for, indeed, a hundred and fifty thousand conflicting voices struck the firmament”. Kathleen Tyson re-transcribed the original manuscript without a comma, making the translation: “Surely ten times ten times ten and five thousand men in varied voices strike the pole star"; in other words, 6000 men were singing as they sailed north.

We like the look of Tyson’s 6000 men. She confirmed to us that there is no comma in the original manuscript. Her figure is specific and it ties in with the more rational of the other estimates. “Ten times ten times ten plus five thousand” might seem a very odd way to say 6000 but Carmen is a Latin poem. Latin poems are not constructed to rhyme but to flow as ‘iambic pentameters’. The unusual wording was presumably for poetic purposes.

We will assume in the rest of this document that the Norman army had roughly 2000 knights with coursers or destriers, plus 3000 infantry and 1000 archers.

It would be helpful to calculate the amount of shore space needed to land the Norman fleet. Neumann used Froud’s hydrodynamics on the Channel crossing speed to estimate that the troop carrying longships had an average beam of 2.77m. However, he did not allow for how long the Normans moored near St Valery or near the English coast. The average beam was probably greater.

If there were 6000 troops on 300 longships, as we estimate above, each vessel would have carried 20 troops plus attendants and sailors, as most often depicted on the Tapestry. That would put them in the category of 20-oar Snekka style longships, like the Helge Ask at Roskilde (Figure 11). It has a 2.5m beam. Most of the Tapestry horse carriers brought three or four horses. That puts them in the category of Byrding or Karvi style fishing and cargo boats, with a typical beam of 2.5m and 5m, respectively. There were some larger ships and more horse carriers than troops carriers. We will assume the average beam for the battle fleet was 4m.

Wace says that the Norman fleet lands together. It does make military sense to land simultaneously, thereby stretching the defence to give the best chance of establishing a bridgehead. This was as true for D-Day as it was for William.

The only way to affect a simultaneous mass landing in an estuary or inlet was to anchor line-astern midstream, then to simultaneously sail, row or pole ashore. It makes no odds here, but Tapestry Panel 39 (Figure 7) is traditionally interpreted to be showing Norman boats being poled ashore. We are unconvinced. The ship is already aground. It seems to us that the man in the stern is holding the boat steady with a pole while the horses are unloaded. The ship’s oar ports are closed, so it was not rowed ashore, but its mast was up when it ran aground, so it might have sailed ashore. The other boats that are already out of the water have open oar ports. They might have rowed, sailed or poled ashore.

Regardless how they came ashore, the first wave of ships would be separated on shore by at least the difference between their length and width. Perhaps the gaps were filled by cargo vessels. They would need to be separated by at least 3m, in order to make space to unload horses and cargo over the side and/or to deploy oars if they needed to leave in a hurry.

In summary, we think the Norman army had roughly 2000 cavalry and 4000 infantry, and that they arrived on 700 longships plus several hundred cargo skiffs and barges. We estimate that the longships had an average 4m beam. So, 700 longships with an average 4m beam separated by a minimum of 3m means that the Norman battle fleet would have needed at least 4.9km of landing space. Cargo vessels might have needed another 2km.

Here is Figure 4 again. Note that Pevensey Lagoon (now the Pevensey Levels), Combe Haven and the Brede estuary were open to the sea in those days and that the estuaries were deeper and much wider than they are now. Note also that shingle bars retained Pevensey Lagoon and the Romney Marshes. The only estuaries or inlets surrounding the Hastings Peninsula that were anywhere near big enough to hold the Norman fleet are the Brede, the Ash Bourne, Combe Haven and Hooe Haven. We are convinced that the landing must have been in one of them. The last two would be tight, each having less than 4km of contiguous navigable strand, Hooe Haven less than 3km, but perhaps the fleet was a tad smaller than we think. We will give them the benefit of the doubt.