I finally got the chance to meet – briefly – Dr. Joseph Glatthaar at the Gettysburg College Civil War Institute conference this past summer. At the time I expressed some interest in his new book, Soldiering in the Army of Northern Virginia: A Statistical Portrait of the Troops Who Served under Robert E. Lee. I didn’t purchase it then – it’s a little pricey – but I eventually did get myself a copy, and Dr. Glatthaar was good enough to take some time to answer a few questions about the book.
BR: I’m sure most of my readers are familiar with your work, Dr. Glatthaar, but would you mind filling them in a bit on your background?
JG: My training is in both Civil War and American military history. I received an MA at Rice University under Civil War specialist Frank Vandiver and a PhD at the University of Wisconsin-Madison under American Military history specialist Edward “Mac” Coffman. Mac Coffman was a pioneer in the “New” Military history, which intrigued me since I read Bell Wiley’s books on Johnny Reb and Billy Yank years earlier. My dissertation was The March to the Sea and Beyond: Sherman’s Soldiers on the Savannah and Carolinas Campaigns. I wanted to study soldiers on the march to see what factors motivated them and encouraged their behavior. Among my arguments was that Sherman’s soldiers were veterans with strong ideological ties to restore the Union and destroy slavery and through years of military campaigning had established powerful bonds of camaraderie within their units. My second book was Forged in Battle: The Civil War Alliance of Black Soldiers and their White Officers . What intrigued me was the intersection or friction between races within the confines of military units. At the U.S. Army War College immediately after the First Gulf War, I taught a course entitled “Command Relationships in the Civil War.” The object was to use history to get senior officers to think about the types of subordinates they should be seeking to work under them in future assignments. That was the basis for my next book called Partners in Command: Relationships Between Leaders in the Civil War. I then wrote the volume on black soldiers for the National Park Service. After that, I could not resist writing The Civil War in the West, 1863-65, so that I would be part of a larger volume with friends Gary Gallagher, Bob Krick, and Steve Engle. Shortly thereafter, I published Forgotten Allies: The Oneida Indians in the American Revolution with James Kirby Martin. Jim and I have been great friends for decades, and the story of the Oneidas was so dramatic that we could not resist the topic. A consortium is currently working on a movie based on our book. Since 1989, I had been working on Lee’s army, and it finally came together in 2008 under the title General Lee’s Army: From Victory to Collapse. It was under contract so long with The Free Press that my editor told me they would have terminated the contract years before but the money was so small that it was not worth the paperwork! To my mind, it is my best piece of scholarship to date. With all the extra statistics, Gary Gallagher urged me to publish a book, which became Soldiering in the Army of Northern Virginia. I am currently working on my companion volume to Lee’s army, one on the Army of the Potomac, which is quite fun, since I love research and I am in the research phase.
BR: Your book General Lee’s Army made quite a splash when it was published in 2008, and it’s now considered by many – myself included, for whatever that’s worth – as the must read book on the Army of Northern Virginia. More than a modern version of D. S. Freeman’s Lee’s Lieutenants, it’s a study of the makeup of the army. Much more of a “social history”, for lack of a better term, it concerns the personality of the army, as opposed to a simple chronology of events. What did you learn about the army in the process of writing and researching General Lee’s Army?
JG: General Lee’s Army was an eye-opener for me. I suspected those soldiers bore great burdens, but I was shocked to learn how many severe hardships they endured, how many losses they sustained, and how much of the fighting they bore for the Confederacy. Rich, poor, and middle classmen came out to fight in the army. Seventy percent who ever served in Lee’s army was killed, died of disease, wounded at least once, captured at least once, or discharged for disability. That statistic does not include overlap, so many were wounded one time and later captured, or wounded in one battle and later killed in action. Again, we knew they had clothing shortages, but in the wintertime, soldiers were eager for combat because if they won the field they could get shoes, blankets, and overcoats from the Yankees. Some men appeared for inspection in late 1864 without pants because theirs were worn out! In the last half-year of the war men lived on 900-1,200 calories per day. Their intake of vitamins and minerals was so insufficient that they could not extract the nutrition from the food they consumed. Yet so many of Lee’s men continued to fight. I also love Robert E. Lee’s revolutionary way of thinking. Contrary to the arguments of J.F.C. Fuller and so many scholars since, Lee was an extremely creative problem solver with ideas that were far ahead of their time, a truly innovative thinker. In Partner’s in Command, I got a sense of how good Lee was, but in researching General Lee’s Army, I learned that he was truly exceptional. Quite frankly, those who think otherwise simply have not done their homework.
BR: The conclusions you drew from an impressive amount of data concerning the soldiery were not without controversy. Can you describe how that book was received in various circles? Were there any reactions that surprised you?
JG: The reaction to the book was a bit surprising. Some people dismissed it immediately because I point out the ties to slavery. Others know just enough statistics to voice an opinion but not enough to understand them properly. One person suggested I did not weight my sample, which of course I did. One academic criticized me for mentioning Lee’s temper but not exploring the psychological dimensions of it—in effect, the root cause. It is hopeless to explain that the book was not about Lee’s psychological makeup or speculations about his relationship with his father! Rather, that is a book about an army. One reader criticized me for not blaming O.O. Howard for Jackson’s successful flank attack. Rest assured I shall do so in my study of the Army of the Potomac, but blaming a Union general in a book on Lee’s army still makes no sense to me. Another scholar who wrote a regimental history complained that my sample was small—he, after all, had an entire regiment—or that none of his soldiers in the Tennessee Veteran’s questionnaire admitted that the war was about slavery. Of course, he only had to look at four or five reels of microfilm for his service records data. It took me months and months just to get the names of my sample and then I had to look in perhaps 800 or more reels of service records to gather my soldiers’ data. And then I had to find them in census records and gather all sorts of other data, and this information probably totaled ten or twenty pages in a 600-page book! I guess it never occurred to him that the reason no one admitted they fought for slavery was because by the second decade of the twentieth century most of the world had come to the irrevocable conclusion that slavery was immoral and that no veteran wanted to admit that they fought a war with 600,000+ dead for an immoral cause! No doubt, Confederate soldiers fought for all sorts of reason—defense of hearth and home, their rights (which, incidentally, included their right to own slaves), spirit of adventure, community and government pressure, and other factors–but to deny slavery as a consideration is absurd.
Fortunately, the people whose work and opinion I value, such as Gary Gallagher, Jim McPherson, Bud Robertson, Jack Davis, Bob Krick, Bill Cooper, and Emory Thomas, to name a few, were pleased, and that thrilled me.
BR: Your new book, Soldiering in the Army of Northern Virginia, is a different cup of tea. Not a traditional narrative, it’s a more in-depth look at the data behind the conclusions drawn in General Lee’s Army. I liken it, somewhat, to Joseph Harsh’s Sounding the Shallows which was a look behind the curtain of his Taken at the Flood. However, that book focused more on strictly military matters. Soldiering examines, among other things, the socio-economic backgrounds of the officers and men of the AoNV. First off, what is your object in publishing this as a separate work?
JG: As I mentioned earlier, General Lee’s Army utilized a mere fraction of the statistics I generated. For each chapter, I probably calculated close to 100 pages of tables, sometimes with three tables per page. All this data was unused, and Gary Gallagher had a hunch it would be very interesting to publish on its own. I viewed it as a chance to slice the army in various ways to catch a glimpse of these soldiers, their experiences, and their lives. What I wanted to do is show readers some of the possibilities that were available from a research perspective. Decades ago, historians moved toward statistics and then dismissed it. Some of it was quality work, and other elements were not well done. Lately, I felt like scholars had been cherry-picking evidence to support one argument or another, when the preponderance of evidence indicated otherwise. Having read so many Civil War letters and diaries over the decades, I have a good sense of their contents. What I hoped to do is to generate hard data that will help guide scholarly research—in effect, work in conjunction with more traditional, qualitative sources.
For that reason, I am trying to get funding to launch a massive research project on Confederate soldiers, with a purely random sample of 4,000 men. In Soldiering in the Army of Northern Virginia, I gathered all my own data. In that instance, with a sample of 4,000, I would need help from graduate research assistants.
BR: Now, to be clear, yours is not a compilation of data that attempts to support sweeping conclusions about subjective things like soldiers’ opinions or feelings, a somewhat dubious practice we’ve been seeing a lot of over the past few years. You focus on specific, quantifiable, objective points.
JG: I cannot calculate attitudes or other such things in Soldiering. I merely explore what is measurable in the data. The data that I have collected is from Compiled Service Records, Census Records, and some other sources. I am only generating statistics that one can quantify from that data, such as year of entry, rank, how the individual left the service (killed in action, discharged for wounds, discharged for disability, deserted, prisoner at Appomattox, etc.), date they left the service, desertion, desertion dates, length of desertion, illness, wealth, slave ownership, and other things. Attitudes and motivations are more a product of traditional sources.
I used a stratified cluster sample—a complex way of generating a random sample—on the recommendation of my friend, Kent Tedin, an expert in sampling. Since we had no list of the soldiers in Lee’s army, and I wanted to compare branches, we chose this method. It included 150 artillerists, 150 cavalrymen, and 300 infantrymen. At 150, the data would be reliable for artillery and cavalry, but because the infantry was so large we needed a larger number.
I compiled a list of every unit that ever served in Lee’s army or its predecessor, the Army of the Potomac, by branch. I then generated 50 random numbers each for the artillery and cavalry and 75 for the infantry. I then generated three random numbers per artillery and cavalry unit and four for each infantry unit selected. At that point, I simply went through and counted, so if the 652nd soldier was in my sample, I waded through the service records until I reached the 652nd CSR! Fortunately, Tom Broadfoot published his roster for some of the states, so I was able to use that, which, needless to say, was much quicker to identify soldiers for inclusion in the sample. I then gathered data from their service records, located them in census records, checked county histories, obituaries, family histories, pension files, and other sources for personal data, and loaded that into an ACCESS document. Based on a tabulation of strength throughout the war, 81.8% of Lee’s army were infantrymen, 11.3% were cavalrymen, and 6.9% were artillerists. These factors were used for weighting purposes. The statistical weighting took into account the larger size of the infantry sample. I then converted the database to STATA and wrote code to crunch the numbers in STATA. The results were nearly as accurate as if the sample was purely random. Several statistics experts have complimented me on the skill and sophistication of the sampling. I cannot take credit.
There are some chapters where the sample size was not large enough for firm conclusions. For example, only 3.4% were foreign-born. I calculated foreign-born and northern born against southern born, but the results for those two categories do not allow for confidence limits. Of course, they do for southern born. In those instances I am explicitly clear about confidence limits.
BR: Pretty much all of the socio-economic data in the book is fascinating to consider, and the graphics help the numbers pop. The lightning rod as always is slavery – in this case the percentage of men and officers in the army who were stakeholders in the institution. Of course it’s impossible to quantify the number of people whose lives and livelihoods were dependent on chattel slavery, but you seem to have at least put to rest the age-old argument that “only blankety-blank percent of Confederate soldiers owned slaves” by providing hard numbers regarding those who belonged to slave-holding households. And that number is a lot higher than what normally gets tossed around. While doing the research into this particular aspect, was there anything that surprised you or confirmed previously held notions?
JG: Certainly I was surprised that 37.2% of all soldiers either owned slaves or their parents with whom they lived owned slaves. I was also surprised that 44.4% of all soldiers came from slaveholding households. Other things, though, surprised me. For example, I compared soldiers in Lee’s army with males of comparable ages in the states from which Lee drew his troops. The results were that Lee’s soldiers had a considerably higher median wealth and had more people in the wealthy class and fewer people in the poor class. I was surprised about the comparatively low percentage of middle-class folks in Lee’s army. Upon thinking it over, though, many skilled workers whose talents were needed by the army and the people at home would have been in the middle class. The very heavy casualties also surprised me. All of us would assume infantrymen would bear the brunt of combat, but when 83.1% of those who joined the infantry in 1862 were KIA, WIA, died of disease, discharged for disability, or POW at least once, and 74.4% of those who joined the army in 1861 were, it is quite startling.
BR: “Soldiering” has a fairly narrow target audience, but how has it been received so far?
JG: I have no idea how it has been received. As you know, it is not an easy read; it is not the kind of book that you pick up and read straight through. Despite all my efforts to write it clearly, the numbers are dense. Plus, there are not all that many people who have backgrounds in both the Civil War and statistics. As a result, it has not gotten much “play.” I think that people who are willing to take their time and go through it with care will find it rewarding. There is some fascinating information in there. Ultimately, it will, I hope, have a real impact on how we perceive Civil War soldiers. Recently, I was attending a talk by Jim McPherson and in the Q and A he began discussing issues on desertion and wealth, which he derived in part from Soldiering. So, I have hope that it will have an impact.
BR: I’m very interested in how writers go about their business. Can you describe your research and writing process?
JG: Much of research for a book like this is fairly tedious, but I actually don’t mind it. In fact, I find all research wonderful. The idea of discovery still charms me. I love to go through archives and read letters and diaries.
When I embark on a project, I try to read all the primary materials I can find. In a big project, no one can ever find everything, but I still try to be extremely thorough. I go from archives to archives and see as much as I can find. Needless to say, I am dependent on great archivists like Dick Sommers, John Coski, Lee Shepard, and so many, many more. I also love researching at the National Archives (NARA). Over the years I have been blessed with great archivists: Sara Dunlap Jackson, Mike Musick, Mike Meyers, Trevor Plante, Mike Pilgrim, Connie Potter, and a host of others. They have helped me find new ways to exam old and new questions. Because the research material is so vast, I still use 3×5 notecards. Taking the equivalent of 10,000 or 15,000 notecards on a computer would be an organizational nightmare. Documents that offer information that won’t fit on a 3×5 get a numbered file and a notecard that summarizes briefly the information and the relating file number.
Once I have finished gathering my evidence I go through all the notecards and keep reading and sorting. Then, I compile topical lists from the notecards and begin to try to outline the book, chapter by chapter. In General Lee’s Army, the book had a chronological thrust so I had to plot matters carefully. I did not want to quote an 1864 source when I was writing about 1862. Moreover, I had to introduce themes early and leave them because later in the book they would become important.
For me, organization is the key. When I have a good, clear outline and carefully sorted notecards, I am ready to write, and although I am no Jack Davis—Jack gets 16 pages of finished prose per day!—I am able to write reasonably quickly.
BR: What’s next for you? Do you have a particular project in mind?
JG: Way back in the in the mid-1980s, I began research for a book on the Army of the Potomac. I have now turned my attention back to it. Right now I am fairly far along on my sample. I’ve looked at the Compiled Service Records and am two-thirds through with pension files at NARA. I’ve barely scratched the surface of manuscript collections—perhaps 60 or so. And, of course, there are fabulous amounts of records at NARA, and then published materials. In short, I’ve got a ways to go.
As I mentioned earlier, I am also trying to get funding to develop a statistical database for 4,000 randomly selected Confederate soldiers. I hope to get data from Compiled Service Records and Census Records first and then collect data on them for their entire life course, as well as the life course of their widows and children. Not only would a project like this generate fabulous data (the database would be posted on the website at the Odum Institute at the University of North Carolina at Chapel Hill) for all to use, but it would allow us to explore topics like postwar adjustment and impact of soldiers and their families, widows, disabled veterans, orphans, and all sorts of other economic, health, and social issues. Once it is open to the public, we could add photos, letters, diaries, and other materials, so that the database could be used in schools as well as by researchers.
It sounds like there’s more groundbreaking work ahead for Dr. Glatthaar. The Confederate soldier study is intriguing, but I for one am excited that he’s also turned his attention to the Army of the Potomac. I think you should be, too.