A recurring theme in the study of George McClellan’s record with the Army of the Potomac is his apparent propensity for overestimating the size and mettle of the opposing force. I really don’t have a good understanding of why this was or what his sources of information were. Except for a nodding acquaintance with Pinkerton (below), I don’t understand his intelligence apparatus at all.

A. Pinkerton

I’m now reading Edwin Fishel’s The Secret War for the Union [more at Amazon], and I think there might be some help there. I have to admit I skipped ahead when I first got the book, Sunday last, looking for tidbits on the Maryland Campaign. Fishel has some very fine specifics on events of early September 1862 that I’d read of in summary elsewhere. I appreciate well documented details, though it doesn’t make for light reading in this case. Now that I’ve got a good feeling from the author in my small area of expertise, I’ll go back and begin at the beginning of the book. I promise.

Side note: I went to school and played music with a Fishel in the early 70’s in Arlington (Va). Great trumpet player. I knew his father, slightly, as a Dixieland bandleader and jazz musician. Had no idea he was a spook. Now I know “the rest of the story”.

There has been a discussion recently on H-CivWar about Antietam. About whether it was a draw or victory, and some of the common “what-ifs” have made their appearances: if only McClellan had insert cliches here … the usual suspects.

In an outgrowth from this, Tom Clemens and others have pointed out that it was what McClellan believed at the time about his opponent that guided his actions, not what we know now to be the reality. I think Professor Joe Harsh – although more concerned with Lee – and other recent writers have made this point well. I’m inclined to that school of thinking myself.

McClellan sincerely believed he was outnumbered at Antietam. Whether this was self-inflicted or based on factual information, delusional or reasonable, I don’t have the facts to appreciate. More digging is required.

G.B. McClellan

In his post-battle report of October 15, 1862, by way of example, he said:

“… With the day [17 September] closed this memorable battle, in which, perhaps, nearly 200,000 men were for fourteen hours engaged in combat …” [note: McClellan counted just under 90,000 of his own, here crediting Lee with 110,000, more than twice as many as he probably had]

“…The object and results of this brief campaign may be summed up as follows … The army of the Union, inferior in numbers, wearied by long marches, deficient in various supplies, worn out by numerous battles, the last of which had not been successful, first covered by its movements the important cities of Washington and Baltimore, then boldly attacked the victorious enemy in their chosen strong position and drove them back, with all their superiority of numbers, into the State of Virginia…”

Even if this is crafted for public consumption, or in defense of criticism, it does give some feel for McClellan’s perspective. I need to better understand where this came from.


Extra credit

Among otherwise cogent arguments he’s offered in that H-CivWar exchange, Jim Epperson has posed a bizarre thesis: that 1860 Federal Census data gave McClellan the basis for knowing, in fact, that Lee’s army couldn’t have been as large as he claimed.

Yesterday he summarized a previous post:

“… My point was that a trained soldier who knew the census figures should have known it was not likely he faced an enemy army of 200,000.”

I have high regard for Mr Epperson’s previous scholarship – I know him best from his Causes of the Civil War site – and I know he’s been around, but this sounds absurd to me.

How should this have worked, exactly?

  • McClellan and/or staff, (or, previously, someone at Army HQ or the War Department in Washington) have census data on hand
  • They actively recognize that there are relationships between population data and military service figures
  • They know the rates at which populations joined the CS army (by county, state; minus those serving in local militia/home guards, essential industry, etc)
  • They know the portions of the total enlisted who were assigned to units in the ANV
  • McClellan knows which units of the ANV are across the creek from him at Antietam
  • McClellan knows what portion of the theoretical total of these enlistees are actually present for duty.

If someone went through all this, and Mr Epperson isn’t claiming that anyone actually did, the margin of error would be too large to make the number useful.

I think this is getting out of hand :)

5 Responses to “McClellan and intelligence”

  1. Harry says:

    Tell you true, Brian, I think methods of estimating enemy strength at this point in the war were more related to estimates of numbers of brigades and regiments than to estimates of actual feet on the ground. As on the Peninsula, IIRC McC’s info on the number of Rebel regiments with Lee was pretty good. The problem came in estimating the mumbers of troops in those regiments. And in order to reach the (traditionally accepted) low numbers with the ANV at Antietam, a ridiculously low “conversion factor” would have to be employed, one that I think no reasonable military man could accept as realistic. And let’s not even get into the differences in “effective” strengths as reported by the two armies.

    In the early 1990’s a GMU grad student did some pretty good work with Confederate numbers on the Peninsula, so good that the old saw that Lee was outnumbered there has been pretty soundly trashed. I sure would like to see someone do a similar study for the Maryland Campaign. Maybe we could put to rest the notion, as expressed by Francis Palfrey, that Lee fought the battle with only himself and a one-armed orderly.

  2. Brian Downey says:

    Hi Harry,

    Good points, thanks. I agee about counting units, though I note that GBM’s ‘final’ report counted many that apparently didn’t exist (ie., “Forty-six regiments not included in above – 18,400”).

    The Confederate numbers on the Maryland Campaign are elusive. I like what Prof Harsh has done in this area (Taken at the Flood, Sounding the Shallows) for the ANV. He figures just under 40,000 of all arms actually fighting for Lee at Sharpsburg. I don’t know precicely how this compares to “effectives”.

    There’s a fat opportunity for research here in figuring how exactly Lee got to that point, having had some 70,000 troops on the books on September 1st. After losses, that looks like one hell of a lot of straggling!

  3. Will says:


    The extent of straggling by the ANV during the Maryland campaign was recongized as massive. If I recall correctly, Lee reported to Davis that it had reduced his strength by at least 1/3. If one statrst with the 75,000 Lee is suppossed to have had on Sept 1 and reduces it by 1/3 for straggling then subtracts casualties from actions between Sept 1 and 17, you arrive at about 45,000, which is close to the number often stated for the ANV’s effective strength at Antietam.

    The numbers that GBM gave in his √Ä√∫final' report were based on a memorandum prepared in October 1862 by John Clark, who was a Virginian on the staff of Nathaniel Banks. Clark has conducted interviews of 250 prisoners captured during the “late battles in Virginia and on the Potomac”. The total arrived at was 97,000. Clark’s tabulation included 18,400 from 46 regiments not included in the known organization of the ANV. If you discount these unknowns, the number is pretty close to Lee’s supposed strength on Sept 1. There are a lot of unknowns in Clark’s informationbut McClellan embraced the data without qualification and used it to validate his claim of the size of Lee’s army at Antietam.


    Will Keene

  4. Brian Downey says:

    Thanks for the details, Will. I expect I’ll find more about Mr Clark as I read further. Plenty of room for more work here, yet.

  5. behind AotW » Blog Archive » Fishel: Secret War for the Union says:

    […] As mentioned in an earlier post, I’ve been reading Edwin C. Fishel’s The Secret War for the Union : The Untold Story of Military Intelligence in the Civil War (Houghton Mifflin, 1996). In particular, I was hoping to gain some insight into how General McClellan arrived at the strength figures he used for General Lee’s forces in Maryland in 1862. […]

Please Leave a Reply