Amphibious operations are among the most complex of military endeavors. Due to that complexity, communications between key leaders is paramount to any success. There are simply too many moving parts when placing troops on a hostile shore.
That was true on September 8, 1863 as Rear-Admiral John Dahlgren and Major-General Quincy Gillmore looked to recapture Fort Sumter. But instead of easy, direct communication, the two commanders experienced great difficulty with even the simplest messages. Around 11:45 a.m., Dahlgren first sent a message to Gillmore about plans of a boat assault on Fort Sumter. Emphasis on “sent.” That message was not received until 1 p.m. At that time, Gillmore sent an aide to inform Dahlgren that the Army intended to assault Fort Sumter that night. So at least the two commanders were communicating.
But the communication lacked any details until very late in the day. At around 7 p.m. Gillmore started that ball rolling:
Your dispatch by signal that you intend to assault Sumter to-night reached me about an hour after I had sent my letter by one of my staff, informing you that I intended to do the same thing. In an operation of this kind there should be but one commander to insure success and prevent mistakes. Will your party join the two regiments that I have designated and let the whole be under command of the senior officer, or will the two parties confer and act in concert? The former method is, I think, much to be preferred.
This citation is from the Naval Official Records. A dispatch, found in the Army records, with a time given of 6:55 p.m, is worded similarly, but notes the force is “two small regiments.” Minutes later Dahlgren responded (and again there are different versions and time stamps in the records):
I have assembled 500 men, and I cannot consent to let the commander be other than a naval officer. Will you be kind enough to tell me what time you move, and what will be the watchword to prevent collision?
That message was sent at either 7:10 p.m. or 8:10 p.m. The discrepancy here seems to be the means of transmittal. The message went by way of signals (lamps or flags) and courier. In the Army’s record, Dahlgren up the ante with regard to the command of the operation:
The rank of my commanding officer will be lieutenant-colonel, but if it will be of service in avoiding any question of rank, I will send one of the rank of colonel who has done the duty of commodore, the equal of brigadier-general.
That, of course, picked at Gillmore’s rank. The general responded in kind:
You decline to act in concert with me or allow the senior officer to command the assault on Sumter, but insist that a naval officer must command the party. Why this should be so in assaulting a fortification, I can not see. I am so fearful that some accident will take place between our parties that I would recall my own if it were not too late.
Gillmore added the “watchword,” to use in case the two forces came into contact, was already with Dahlgren’s staff. Then he remarked, “We must trust to chance and hope for the best. No matter who gets the fort, if we place our flag over it.” In the naval record, an annotation points out that only the first sentence arrived by signals. The remainder of the message came by courier. Such explains the next message from Dahlgren, at 9:30 p.m.:
I have found it impossible to communicate with you. Mr. Adams must have left your ship and an incompetent signal officer taken his place….
And that message left off incomplete. The next message went out from Dahlgren thirty minutes later demanding Gillmore provide the hour his assault planned to move. Compounding a disagreement between commanders was a terrible problem relaying messages. At 10:25 p.m., Dahlgren, who must have sensed wires were crossed (perhaps literally), sent anther message indicating he had dispatched a staff officer to confer. But still at 10:45 p.m., Dahlgren remained impatient, “I am waiting an answer.”
Given the lack of response, at 11:30 p.m. Dahlgren prepared an order for Captain T.H. Stevens, who would command the assault force. In the order, he let the friction between the Army and Navy be known:
I have been unable to yet agree on a plan with the general. You will therefore take a position near Cumming’s Point, so as to assist the assault of the troops if any is made on Sumter. Let a boat connect at once with the troops on your left. The watchword is “Detroit.” Toward morning let not your boats be seen, but draw them off in proper season.
But that message was not sent. Word came back from shore that Gillmore would accommodate the Navy’s preferences in command – separate and independent commands that is. But the hour was getting late. The USS Daffodil was already towing the Navy’s assault boats up the main channel.
On the Army side, verbal orders went to Colonel Francis Osborn, of the 24th Massachusetts and commander of the expedition, to retire if there were any indication the Navy had started its assault. Gillmore was very cautious, lest the two forces stumble into each other in the fort and cause unnecessary casualties. Better the Navy took the fort than the two parties kill each other off in front of the Confederates. And as things worked out, Osborn’s obedience to orders, which sprang from Gillmore’s concern about fratricide, left the Navy to assault Fort Sumter alone.
The “friction” in this case might be written off as Army-Navy rivalry. But there was more to it than that. Poor communication channels exacerbated misunderstanding. And the root of the misunderstanding was two services which operated with different sets of procedures and priorities. The entire matter might have been resolved with a mid-afternoon staff meeting (with a couple dozen PowerPoint slides of course). But seriously, that is what a good staff does – coordinate the particulars. That was not done on September 8, 1863.
And for those who tally these things, the issue of who should command a landing force remained a point of contention between the services (and even between the Navy and the Marines) well into 1943.
(Citations from ORN, Series I, Volume 14, pages 608-10; OR, Series I, Volume 28, Part II, Serial 47, pages 87-89.)