The unsinkable ship

What actually went wrong in the North Atlantic in 1912, and what the wreck is still arguing about

A The ship that sank in the North Atlantic in April 1912 has proved unusually resistant to being remembered accurately. Almost everything the public knows about it is either an exaggeration of something true or an invention that has survived because it improves the story. The vessel was not, for instance, advertised as unsinkable in the confident manner usually described; the claim appeared in trade publications in a qualified form and was elevated to the status of a boast largely after the event, when it supplied the disaster with the moral shape a catastrophe seems to require.

B Nor was the vessel racing to break a record, as the popular account maintains; she was not capable of the speed her rivals could manage and her owners knew it. The most consequential of the genuine failures was the provision of lifeboats, and the striking feature of that failure is that it broke no rules. The regulations then in force calculated the required number of boats from the tonnage of the ship rather than the number of people aboard, and had been drafted when vessels were very much smaller. A ship of this size was obliged to carry sixteen boats. It carried twenty, exceeding the legal minimum, and the twenty had capacity for barely half of the roughly two thousand two hundred people on board. Nor was the shortfall an oversight nobody had noticed: the deck space had been designed to take more boats, and the question of adding them had been raised and set aside, partly because a full complement was thought to clutter the promenade and partly because boats were understood as a means of ferrying passengers to a rescuing ship rather than of holding everyone at once. The owners were not being negligent by the standard of the day. The standard was the negligence.

C A second failure was procedural rather than material. Several ice warnings reached the ship by wireless during the day, and not all of them reached the bridge. The wireless operators were employees of a telegraph company rather than the shipping line, and their commercial priority was the backlog of passenger messages waiting to be sent. One warning from a nearby vessel, sent shortly before the collision, was cut off by an operator busy with that traffic. When the distress calls began an hour later, the same nearby vessel had shut down its wireless for the night, and its operator was asleep.

D The evacuation itself was hampered by the same assumption. Crews had not drilled the procedure, passengers were reluctant to exchange a warm and apparently solid ship for an open boat in the dark, and several boats were lowered well short of capacity. The physical design has attracted the most speculation, some of it well founded. The hull was divided into compartments that could be sealed off, but the bulkheads dividing them did not extend to the upper decks. Water entering the forward compartments therefore pulled the bow down until it spilled over the top of each bulkhead into the next compartment, in the manner of an ice-cube tray tilted at one end. The arrangement was adequate for a collision that breached two compartments and inadequate for one that opened six, which no designer had thought a plausible accident.

E Whether the materials were at fault is argued to this day, and the argument is a good illustration of how difficult it is to assign blame to a component recovered from three kilometres of water eighty years after it failed. Steel and rivets recovered from the wreck have been tested and found to become brittle at low temperatures, and some researchers concluded that the hull failed more readily than it should have. Others reply that the samples are indistinguishable from what any shipyard of the period would have used, that the metallurgical knowledge required to do better did not yet exist, and that a hull opened along ninety metres by a collision at speed would have flooded regardless of the quality of its rivets. The dispute is unresolved and probably unresolvable.

F What is not in dispute is the effect on regulation, which was rapid and comprehensive, and which stands as the strongest argument against treating the event purely as a tragedy. An international convention on safety of life at sea was concluded within two years and has been revised ever since. Lifeboat capacity was tied to the number of people carried. Continuous wireless watch was made compulsory. A patrol was established to track ice in the shipping lanes, and it has operated every season since. Measured by the crude arithmetic of lives, the disaster has almost certainly saved far more people than it killed.

G The wreck itself was not located until 1985, after decades of failed searches, and its condition has since become an argument of its own. The hull is being consumed by bacteria that feed on iron, and the structure is collapsing at a rate that makes detailed survey urgent. Salvage, meanwhile, divides opinion sharply: one view holds that objects recovered from a grave belong in a museum where the dead can be remembered, another that a site containing human remains should be left undisturbed regardless of what can be learned. Both positions have been argued by people who have descended to the wreck, which suggests the disagreement is not about information.