I have finished my MA Thesis since I last responded to a post on this topic. Unfortunately, since responding to that post a year ago, my direction has changed. My focus is now on the question of whether or not a diagnosis of PTSD can be identified in the ancient world in general. As much as I had wanted to do this for soldiers specifically, the nature of the evidence makes it an extremely difficult argument to establish with any security. The reason for this is that 1) nearly all the sources we have which do hint at PTSD in soldiers are anecdotal and come laden with the author's biases and agenda, 2) among these anecdotes, there are actually very few which imply a link between trauma and the experience of combat, and none which directly state it, 3) material sources like inscriptions or papyri make little to no mention of trauma due to battle, and 4) in the field of Psychology itself the debate is still open whether the stimuli that can cause trauma are universal or are determined by individual experience. The closest we can get is the fact that in some papyri from around the Severan Dynasty there appears to be an acknowledgement of military discharge due to mental illness, but again, the sources do not tie this mental illness to experience in battle, explicitly or even implicitly. So if one wants to discuss battle-related trauma in the Roman army it is hard to do without speaking only in generalities. Nevertheless, there has been some good work done on this. Korneel Van Lommel, a Belgian historian, has published some good articles on mental impairment in Roman soldiers using the sources I just mentioned.
Here is where you can find his work.
As far as my own research is concerned, I am really interested in posing a solid argument for Post-Traumatic Stress in the Roman world which, despite the problems of finding PTSD in soldiers, is not impossible to do when it comes to the general populace. I still make use of the cross-cultural model for PTSD which focuses on intrusive memory symptoms as the common point for all trauma victims, regardless of culture. In the medical texts of Celsus and Aretaeus, the fact that some of their patients claimed to see the ghosts of the dead is mentioned in no uncertain terms. Outside of medicine, Lucretius mention this same phenomenon and further notes that it happens both when the sufferer is asleep and when he is awake. On the less reliable side of things, this exact phenomenon is a big topos for historians like Tacitus, Plutarch, and Josephus. The fact that the same thing appears in medical texts and that Lucretius feels the need to address it as a natural phenomenon lends some weight to the anecdotal evidence. More importantly, our Roman medical descriptions fit perfectly with our cross-cultural PTSD model's description of intrusive memory symptoms. What's more, there are modern comparisons one can make. Several African tribespeople diagnosed with PTSD choose to describe their symptoms in terms of ghost haunting or witchcraft. The symptomatology is all there, in both ancient and modern sources. One of the main difficulties is the differing prisms through which ancient and modern physicians organize and describe mental illness. We like to have neat recipes for psychological disorders. A and B plus C or D equals a patient with Borderline Personality Disorder. Ancient sources prefer to have all the ingredients laid out before them. A or B or C or D can happen when a patient suffers from
insania. These broad categories in ancient medicine disallow us from finding a clear-cut corollary for PTSD in ancient Rome, but they do allow us to say with a comfortable degree of certainty that PTSD existed, was noticed, was written about, and made its way into the public consciousness via myth and anecdote.
As to what could cause PTSD in ancient Rome, that is more difficult. Here we must rely mostly on stories in order to get a feel for what was and was not a trauma-inducing experience. Going off of the conclusions above, I've found very little which indicates that violence could induce trauma (that is, intrusive memory symptoms in the guise of seeing ghosts). Instead, this kind of thing crops up around individuals who have experienced a loss or violation of public honor or status. Going against the dictates of society or failing to live up to its expectations appear to have been a greater cause for trauma than risk of life and limb. Josephus gives my favorite example of this when he mentions a Syrian governor named Catullus, who harbors a fleeing Jewish rebel leader in the wake of the Great Revolt of 68-69. The governor uses intel from the rebel (which he knows to be false) as an excuse to put hundreds of Jewish families in his province to death and to seize their wealth. He then takes his booty and presents it to Vespasian, thinking that Vespasian will be pleased with his actions. Vespasian rejects the offer, calls the governor out on murdering innocent people, and proceeds to burn the governor's turncoat informant to death. Catullus himself gets treated with velvet gloves, even being given a room in the imperial palace. This is the point where the governor's mind breaks down. He begins to see the ghosts of the people he has killed and wakes up at night screaming. Ostensibly, we might link this development to guilt over ordering so many innocents to their deaths, but I think it is important to take note that Josephus has the governor lose his sanity only after Vespasian disapproves of his actions and relents from meting out the deserved punishment to Catullus himself. This series of events ties in very strongly to the differences in thinking between a "guilt culture" and a "shame culture." Time and time again, it is clear from our sources that public loss of face was much more devastating than personal dissatisfaction or guilt.
Therefore, if societal acceptance or rejection was the primary arbiter of mental trauma, it follows that killing on the battlefield would necessarily be less traumatizing in a society which approves of such behavior than in a society which does not. Based on what I've read, it appears that for a Roman soldier, running away from a battle (and being seen running away by your peers) would be immensely more traumatizing that the act of gutting a barbarian on the front lines.
But this approach has it's own issues: namely the ever-present problem that most of our sources speak exclusively from the aristocratic point of view. So it is difficult to say whether these conclusions could apply to Romans of every stratum. Anyhow, I hope that this answers your question somewhat. I wish I could give you more on the psychological impact of killing, especially since your question is exactly what set me on this path in the first place! Another source you may want to check out is
Mental Disorders in the Classical World, William V. Harris, ed. It does not have any solid material on PTSD, but is the most recent and thorough research on mental illness in antiquity that I have found. I'd love to fill that gap someday, but have been much too daunted by the prospect of publication. That's what New Years resolutions are for I suppose! W.V. Harris himself is also a fantastic writer on emotions in antiquity. He has a great book on anger in antiquity called
Restraining Rage.