meaningless. Some of the images showed cranial fragments arranged for assessment and display. Others, time- sequenced photographs excerpted from the thousands taken by the motion-picture camera, documented in slow motion the effect of a bullet penetrating and passing through a human head. The results, in practical terms, were identical. Each and every head containing a pseudobrain flew apart. Government scientists might care to count pieces and decide which rifle was more lethal, as if such measurements were revelatory. Soldiers would have better sense. Heads were heads, not ammunition cans, watermelons, cubes of gelatin, or blocks of pine. When attached to living men they contained brains. The shooting tests had established, unsurprisingly, that bullets fired from some military rifles caused human heads to fragment into more pieces than bullets fired from other rifles. But what exactly did this mean for either a rifleman or his victim? There is, after all, but one degree of death.
If General Wheeler wanted to know about the relative effectiveness of the rifles under review, this was not a measure. The lethality tests, in the end, offered little of obvious value. Every human bone struck by a rifle bullet had broken, every gelatin-filled cranium had shattered. The relative differences in damages were academic. And yet the test results still might have had a use in restoring a more realistic conversation about the AR-15. Dziemian and Olivier’s final report subtly but clearly revealed that the army’s terminal-ballistics experts were suspicious of the Project AGILE findings. “More shots were made with the AR-15 on legs than with the other bullets,” it noted, “because of the rather startling results from limb wounds in combat described in the A.R.P.A. report.”24 The Biophysics Division’s gunners shot legs at various orientations, with standard bullets and with bullets that had their tips trimmed off. No matter what they did, they were unable to reproduce the effects that the participants in Project AGILE claimed to have seen. At Aberdeen Proving Ground, the traumatic amputations simply did not occur. This result, coupled with the observation that all heads shattered when struck by any of the bullets, might have been a basis for the Pentagon to question the objectivity and methods of Lieutenant Colonel Hallock’s team. But instead of having such a practical value, the lethality tests underlined other things: the risks of secrecy and the deep dysfunctions in McNamara’s supposedly highly functional systems-analysis approach. There was no peer review of this kind of hushed work, and events that followed ensured that almost no one would ever find out about it.25
The deadline for an initial report was November 30,26 seven weeks after Secretary McNamara raised the issue of American rifle choices with Secretary Vance. An understandable curiosity rose through the Pentagon. The possibility that scientists might tell soldiers which new American rifle was more dangerous was intriguing. Lethality data for the AK-47, a Soviet weapon the United States was beginning to face, added spice. But rather than allow a wider set of minds to examine the study and glean what they might from it—in other words, rather than being systematic—the army restricted its circulation reflexively and fiercely. Sometimes supervisors have to be wary of what they ask for, especially in institutions that mix secrecy, cash, and guns. Away from the Biophysics Division, where shooting body parts had become part of the job, the realization that federal employees were performing gunfire tests on human heads was at once unsavory and politically risky. Someone, it seemed, felt that this was something the United States did not want to get caught doing. In spring 1963, the staff of Charles Hitch, the Pentagon’s comptroller, asked the army chief of staff’s office for two copies of the lethality report. Worry and embarrassment crept in. The cover-up began. In a memorandum stamped SECRET the army chief of staff ordered Lieutenant General R. W. Colglazier, Jr., the deputy chief of staff for logistics, not to share the report, even with Hitch’s office, “in view of the sensitivity and potential sensationalism with the use of human cadavers from India.”27 General Colglazier sought help from Vance. “Although this is not the first use of elements of the human cadaver for this purpose, I consider such use to be extremely sensitive,” he wrote, in another secret memorandum, adding that the army would like Vance to ask Hitch to withdraw the request in “recognition of the potential sensationalism which could arise from public disclosure of this information.”28 The results of the lethality study remained hidden for forty-six years. And at the most important time, during the early and mid-1960s, the Project AGILE report, with its suspicious observations and false conclusions, remained uncontested.29 The AR-15 continued its rise, boosted by a reputation for lethality and reliability that it did not deserve.
General Wheeler’s larger set of worldwide tests also failed to provide a fully useful or enduring evaluation. This was in part because of the undue rush, but there were darker subcurrents, too. The study collapsed under the weight of allegations of bias and a climate of mutual recriminations between various camps. Proponents of the AR- 15 accused the army and its evaluators of efforts at sabotage. The acidic climate prompted Vance to order an investigation by the army’s inspector general, which found that at least part of the testing, a tactical assessment at Fort Benning, had been rigged against the AR-15. The results tainted the entire effort and provided the AR-15’s backers with another argument they would use well: that their weapon was never treated fairly. Absent anything definitive to say, General Wheeler chose the safest path available. On January 14, 1963, based on the results of the worldwide tests, he delivered his recommendations. The army was most clear on one point—the pitfalls of the Soviet assault rifle. In his summary, General Wheeler dismissed the value of the AK-47, calling it “optimized for the submachine gun role” and declaring its “overall inferiority” to both the M-14 and M-16. Thus ending discussion about that, the general assessed the American arms. Being politically astute, he found merits in each. Both weapons, according to the army’s assessment, had superior accuracy as well as acceptable durability, ease of maintenance, and other desirable characteristics. But the AR-15 had a consequential fault. The evaluation unequivocally rated it unsatisfactory in the important category of reliability. This was because, General Wheeler wrote, stoppages during firing were “so troublesome that soldiers might well lose confidence in the weapon.” The general added that the army was optimistic that the deficiencies “can be readily corrected,” though he was silent on what these corrections might be.30
Given the ferocity building in the argument over which rifle was best, General Wheeler was careful not to offend any camp too greatly. He seemed eager to step aside and let the defense secretary choose the next steps. The defense secretary, after all, was the one pushing the question. General Wheeler offered three recommendations: continue with the M-14 program, terminate the M-14 program and adopt the AR-15 as the new standard arm, or defer a drastic step and mollify all camps via what he called Option C. In this option, the army would buy AR-15s for helicopter-borne, airborne, and Special Forces soldiers and reduce its M-14 purchases accordingly. It would decide later which weapon would be the American shoulder-fired arm of the future: the AR-15, the M-14, or the SPIW, if the SPIW ever became feasible. Predictably, McNamara leaned toward Option C.
Now came the matter of readying the AR-15 for service. In early March, the Army Materiel Command formally opened an AR-15 development program, with an office at Rock Island Arsenal in Illinois headed by Lieutenant Colonel Harold W. Yount. This was not to be a normal program, and Colonel Yount’s managerial control was titular. A special committee with oversight duties was formed. The committee included representatives from all the services but also put the program under the defense secretary’s control: Two of McNamara aides were given seats on it, along with veto power over all its decisions.31 This gave the AR-15 a high priority. It also left the program vulnerable to political interference on technical matters and introduced fresh tensions between the defense secretary’s office and the ordnance service. McNamara had already expressed his dissatisfaction with the army’s weapons experts, and the inspector general’s investigation of the handling of the worldwide tests only added to the feelings of distrust. The defense secretary wanted to put his stamp on the AR-15 program and place it under his protection. But the top-heavy assignment of political appointees to the committee risked alienating or even removing people with weapons expertise from participation in the AR-15’s development. McNamara’s whiz kids were smart. But they had almost no experience in either war or weaponry, and were not necessarily an able substitute for those whose careers had been a study of ordnance and guns. One of the government’s ballistic experts was appalled at their role. “Their qualifications,” he said, “consisted of, and apparently were limited to, advanced academic degrees, supreme confidence in their own intellectual superiority, virtually absolute authority as designated representatives of OSD [Office of the Secretary of Defense], and a degree of arrogance such as I have never seen before or since.”32
Firearms and their ammunition form a system. While that system can seem exceedingly simple compared to a fighter jet or a tank, it often is not simple at all. Automatic firearms and cartridges work together in complex ways, and changes to either a weapon or a cartridge can create pervasive disharmonies that can be difficult to pinpoint; ghosts can readily inhabit these machines, and they do. The services and the committee, working through both the government’s arsenal and Colt’s, proceeded to make changes to the AR-15 and the .223 round—more than one hundred in all. Many changes were minor and arcane. Others, including a change to the rifling in the AR-15’s barrel and the addition of a device that could push the bolt forward manually, were significant and consumed
