6/07/2005

Numbers, losses, surgeons

The pop ACW historian has less interest in numbers than in any other part of the work. Where used, numbers tend to be wonderfully round, with loads of zeroes, usually adjusted from pay (muster) rolls with a dash of whimsy added to make them "better."

Good history depends not at all on numbers - not even military history needs them - history is about people, events, decisions, betrayal, hope, triumph, faith, disaster, and all those things that make life itself interesting.

The peculiar thing about Civil War history is that numbers are made absolutely central to every story - the same numbers that the pop writer has no time to analyze dominate the material. In these best-sellers, numbers test character, numbers mark chances, numbers chart progress, numbers tell "the tale of the tape." Usually the number is ennumerated, in all its roundness; sometimes it is just implied with vague terms like "overwhelming superiority." However vague or ill-gotten, numbers form the basis for sweeping judgements about the men of those times.

The numbers on which so much in pop history depends are wrong, even by the very low standards many Civil War authors set for themselves.

If, for instance, you were to seriously decide that the payroll musters were the rational basis for estimating the size of a command, you would at least owe it to yourself and your readers to present the very best payroll-based estimates available - not some raw numbers plucked from the Official Records (OR) but those refined by Thomas Livermore in Numbers and Losses a century ago. Livermore corrected errors found in the OR, calculated which units were present for battle, how casualties affected the moving trendline of relative strength on the field, how to deduct for details and absentees, and a few more things besides.

The problem with Livermore versus raw OR data is that his thought-out numbers screw up a predetermined storyline. If you reach into the raw OR figures, for instance, you can present McClellan outnumbering Johnston at Fair Oaks; Livermore won't allow that (see p 81). In the Seven Days, leading up to Malvern Hill, McClellan fights outnumbered, 83,345 total engaged vs. Lee's 86,748 (see pp 84 – 85). He fights outnumbered throughout the Seven Days. There's not much our military historians can do with that without refresher courses in dramaturgy.

So today's historian stays with some flavor of raw OR data, without repudiating Livermore in a monograph; without acknowledging him; it is as if Livermore never wrote a word.

And yet even Livermore is very wrong, because any payroll-based headcounting is always very wrong.

As I mentioned yesterday, the medical staff of the Army of the Potomac was not allowed to use our historians' beloved muster rolls to count noses. Regimental surgeons were instructed to take the daily regimental present-for-duty tallies and average them to come up with a monthly regimental present-for-duty figure. This would serve as the statistical basis from which to figure illness and recovery.

Those who have served since the ACW immediately see the "honesty" in this approach. How hard we, as commanders, have fought administrative battles to remove dead weight from our own rolls; how bitter we have been at having so many "slots" filled by no-shows while work went undone. And this in a "professional" (not militia) army! Muster rolls are simply lists of people collecting pay through an affiliation with some unit. Men present for duty is an entirely different matter.

We get a view of the surgeon's numbers, aggregated at army level, through the Medical and Surgical History of the War of the Rebellion, Appendix to Part I. What follows is from the 1875 edition, Second Issue.

J.J. Woodward, Assistant Surgeon, U.S. Army writes in the introduction, "This figure [the surgeons' strength figures] was invariably obtained by adding together the mean strengths given in the individual reports consolidated for the month." For those baffled by the difference between the Surgeon General's strength figures and the Adjutant General's muster rolls, he says: "... the returns of the Adjutant General represent the strengths on the day of their [submitted] date." That is, they represent a notional count for pay purposes of people assigned the unit on that one day (only) during the month. It's the paymaster's snapshot of one day.

Looking at this volume's Table VI for the Army of the Potomac, we revisit Livermore's well-thought-out battle tally for the Seven Days: McClellan fought outnumbered, 83,345 total engaged vs. Lee's 86,748, says Livermore.

But the surgeons tell us a different story. In the month of June, 1862, the mean effective strength of the AoP (pre-sickness, pre-death) was 78,733. McClellan cannot have had 83,345 men on the field. Furthermore, that 78,733 suffered 24,690 medical cases of treated illness during the month (this does not include combat wounds or deaths; it may include multiple illnesses in one person); additionally 705 non-combat deaths were suffered.

The difference between the surgeons' view of events and even the most careful reconstruction from pay records is striking. It is not clear how we might fairly deduct the numbered medical cases from McClellan's maximum number of boots on the ground (78,733), however if we do that, we are then ready to make further deductions, as Livermore did, for details, detachments, shirking, picket duty, the base at White House, etc.

In July, McClellan is reinforced, bringing the surgeons' mean to 106,069 men available for operations against Richmond. However, these men reported 42,811 cases of sickness, 371 of them dying.

Drew Wagenhoffer mentioned to me that AoP illness on the peninsula gets mere "lip service" from authors eager to move their stories forward. I agree. The root of that comment speaks to Civil War historians not wanting to do their research while still committing to numeric interpretations of virtue. The treatment of numbers and the dedication to story structure are the two most obvious indications we have that Civil War history as a field is broken.

For even the most casual ACW reader, neither the name Livermore nor The Medical and Surgical History of the War of the Rebellion should be a discovery on some fool's blog.