Abstract
The relationship between Jefferson Davis and Joseph E. Johnston is the counter-case to the partnership examined in the preceding study. Where Davis and Lee built and maintained a working relationship over three years, Davis and Johnston allowed an early grievance to ossify into a pattern of mutual distrust that degraded Confederate command at two of the war’s decisive points: the defense of Mississippi in 1863 and the campaign for Atlanta in 1864. This paper reconstructs the mechanics of that dysfunction. It argues that the relationship failed not because either man was simply incompetent or malicious, but because each behaved in ways that confirmed the other’s worst expectations, producing a feedback loop in which Johnston’s reticence invited Davis’s interference and Davis’s interference justified Johnston’s reticence. The paper traces this loop through three episodes, the Peninsula in 1862, the Vicksburg crisis in 1863, and the relief outside Atlanta in July 1864, and weighs the competing postwar narratives the two men constructed to explain a failure for which neither would accept responsibility.
Introduction
The seniority dispute of 1861, examined in the first paper of this volume, established a grievance at the center of the Davis–Johnston relationship but did not by itself determine that the relationship would fail. Grievances can be managed, set aside, or outgrown; the Davis–Lee partnership succeeded despite Lee’s own early failures and the absence of any natural intimacy between the two men. What distinguishes the Davis–Johnston dyad is not the presence of an initial wound but the absence of any subsequent effort, by either party, to prevent that wound from governing all that followed.[^1]
The argument of this paper is that the Davis–Johnston relationship is best understood as a self-reinforcing system rather than as a series of discrete quarrels. Each man entered their interactions with an expectation about the other, Davis that Johnston was a sullen and uncommunicative subordinate nursing a grievance, Johnston that Davis was a meddling and hostile superior bent on his humiliation, and each then behaved in ways that elicited from the other precisely the conduct the expectation predicted. Johnston, expecting interference, withheld his plans; Davis, receiving no plans, interfered; Johnston, interfered with, was confirmed in his suspicion and withheld further. The loop did not require fresh malice at each turn. It needed only the initial distrust and the two men’s failure to break the pattern, and it ran, with worsening consequences, from 1862 to 1864.[^2]
This framing matters because it resists the partisan accounts that have dominated the subject. The Lost Cause tradition, and Johnston’s own memoir, cast Davis as the villain who destroyed a capable general out of personal animus. Davis’s defenders cast Johnston as a chronic retreater whose secretiveness and want of aggression forced the president’s hand. The systems view advanced here declines both verdicts in their simple form. It holds that both men contributed to the failure, that each had genuine grounds for distrusting the other, and that the tragedy lay in the interaction rather than in either character taken alone.
The Peninsula, 1862: The Pattern Established
The first wartime test of the relationship came in the spring of 1862, when Johnston commanded the army defending Richmond against George McClellan’s advance up the Virginia Peninsula. Johnston’s conduct in this campaign displayed, in embryo, the trait that would define his relations with Davis for the rest of the war: a reluctance to disclose his intentions to his own government.[^3]
As McClellan pressed toward Richmond, Davis and his military adviser, Lee, found themselves repeatedly uncertain of Johnston’s plans. Johnston withdrew up the Peninsula without keeping Richmond adequately informed of his reasoning or his intended line of stand, and on at least one occasion the capital learned of a major movement after the fact rather than before.[^4] To Davis, this was both a practical embarrassment, the government could not coordinate the defense of its own capital without knowing the army’s plans, and a confirmation of the suspicion planted in 1861, that Johnston was a man who held himself apart and could not be trusted to cooperate.
The crucial point is that Johnston’s reticence was not, in his own mind, insubordination. It reflected a settled conviction that the details of operations were a soldier’s province, that premature disclosure risked leaks, and very probably an unwillingness to expose his plans to a president he already distrusted. From Johnston’s side, withholding was prudence; from Davis’s side, it was sullenness. Each man read the same conduct through the lens of the 1861 grievance, and each found in it what he expected to find. When Johnston was wounded at Seven Pines on May 31 and Lee took the army, the immediate problem resolved, but the pattern had been set and observed. Davis had now seen with his own eyes the behavior the seniority quarrel had led him to expect, and the lesson stayed with him.[^5]
The Vicksburg Crisis, 1863: The Pattern Compounded
The Western theater in 1863 supplied the conditions under which the Davis–Johnston dysfunction produced its first catastrophic result. In an attempt to coordinate the defense of the Mississippi Valley, Davis assigned Johnston to a department command spanning Tennessee and Mississippi, with nominal authority over the forces of both Braxton Bragg and John C. Pemberton.[^6] The arrangement was meant to impose unity on a sprawling theater; in practice it placed a general who distrusted his president, and was distrusted in turn, in charge of a command structure whose lines of authority were ambiguous and whose constituent armies faced threats in opposite directions.
When Ulysses S. Grant’s campaign against Vicksburg reached its climax in May 1863, the defects of the arrangement became fatal. Pemberton, commanding at Vicksburg, faced conflicting imperatives: Davis, a Mississippian who regarded Vicksburg as politically and strategically indispensable, wished the city held; Johnston, surveying the theater, judged that the army’s survival mattered more than the fortress and counseled Pemberton to avoid being trapped within it.[^7] Pemberton, caught between his department commander and his president, and temperamentally inclined toward Davis’s view, made decisions that led to his army’s investment inside Vicksburg, where it was besieged and ultimately surrendered on July 4, 1863.
The Davis–Johnston dysfunction shaped this disaster at several points. Johnston, characteristically, regarded his department command as nominal and was reluctant to exercise vigorous control over forces he had not raised and a situation he had not created; his instinct was to husband, to consolidate, and to avoid commitment, an instinct Davis read as passivity.[^8] Davis, for his part, never trusted Johnston’s grasp of the theater and continued to communicate his own wishes directly into the situation, deepening Pemberton’s confusion about whose orders governed. The command structure that was supposed to produce coordination instead produced paralysis, and the distrust between president and department commander was a principal reason the structure could not function. A department head who believed his superior hostile, and a superior who believed his department head feckless, could not generate the mutual confidence that unified command requires.[^9]
It is necessary to the honesty of this study to note that the Vicksburg failure cannot be laid wholly at the door of the Davis–Johnston relationship. Grant’s campaign was brilliantly conducted, the Confederate forces were outnumbered and divided by geography, and Pemberton’s own judgment was a factor. But the relationship was a genuine cause, not merely a backdrop. Had Davis trusted Johnston, he might have given him real authority and stood back; had Johnston trusted Davis, he might have exercised that authority decisively rather than holding himself at a wary distance. Neither condition obtained, and the absence of both was a Confederate liability at the moment Vicksburg most needed coherent command.[^10]
Atlanta, 1864: The Pattern Concluded
The final and most consequential episode came in the summer of 1864, when Johnston commanded the Army of Tennessee against William T. Sherman’s advance on Atlanta. By this point the relationship had no reserves of trust to draw upon, and the campaign exposed the dysfunction in its starkest form.[^11]
Johnston conducted a campaign of skillful defensive withdrawal, trading space for time and preserving his outnumbered army as Sherman maneuvered him back from the mountains of north Georgia toward Atlanta. To Johnston, this was sound generalship under unfavorable odds: he would not be drawn into a destructive battle on the enemy’s terms, and he would keep his army intact. To Davis, watching Atlanta, a vital manufacturing and rail center, draw ever nearer to Sherman’s grasp, the campaign looked like the same passivity he had perceived at Vicksburg and the same reticence he had perceived on the Peninsula, now threatening the loss of one of the Confederacy’s last great cities.[^12]
The fatal element was, once again, communication. Davis repeatedly sought from Johnston a clear statement of his intentions, in particular whether and where he proposed to stand and hold Atlanta. Johnston, true to the pattern established in 1862, declined to commit himself in terms that satisfied Richmond, offering little assurance that he would defend the city rather than continue withdrawing.[^13] To Johnston, withholding a definite commitment preserved his operational flexibility and reflected the honest uncertainty of a commander facing a stronger foe. To Davis, the silence was the final confirmation of a three-year suspicion: Johnston would retreat indefinitely, would not fight for Atlanta, and could not be induced even to say what he meant to do.
On July 17, 1864, Davis relieved Johnston of command and replaced him with John Bell Hood, who was known to favor the offensive.[^14] The decision and its catastrophic consequences are the subject of the seventh paper in this volume; what concerns the present study is what the relief reveals about the relationship. Davis removed Johnston not on the strength of a single failure but on the accumulated weight of a pattern, the Peninsula, Vicksburg, and now Georgia, in which Johnston’s reticence and apparent passivity had repeatedly, in Davis’s reading, endangered the cause. Johnston, for his part, regarded the relief as the culmination of the persecution begun with the seniority slight of 1861, the final act of a president who had hated him from the start.[^15] Each man, at the moment of rupture, understood the event entirely through the framework of distrust the two of them had spent three years constructing.
The Symmetry of Fault
The interpretive heart of this paper is its refusal to assign the failure to one party. The systems view requires that both contributions be stated plainly.
Johnston’s share is real. His reticence was genuine and chronic; he withheld his plans from his own government repeatedly and across years, in a manner no civil authority could be expected to tolerate indefinitely. His instinct for the defensive and for the preservation of his army, defensible as generalship, was not paired with any corresponding effort to keep his president informed or reassured, and his failure to communicate was, at least in part, a choice rather than a necessity. A general who will not tell his government what he intends forfeits the trust on which his latitude depends, and Johnston forfeited it more than once.[^16]
Davis’s share is equally real. His distrust of Johnston, rooted in the 1861 grievance and the Peninsula, hardened into a settled hostility that disposed him to read every ambiguity against the general. He interfered, communicated past his own department commanders, and ultimately could not extend to Johnston the patience and the benefit of the doubt he extended freely to Lee. The contrast with the preceding paper is the indictment: Davis could keep a general informed of his confidence, could refuse a resignation with warmth, could grant operational latitude to a commander he trusted. He did none of these things for Johnston, and his failure to do so was a failure of the same man-management that the Lee partnership shows he was capable of when he chose.[^17]
The tragedy, properly understood, is that each man’s faults activated the other’s. Johnston’s reticence would have mattered less under a president disposed to trust him; Davis’s distrust would have mattered less under a general disposed to keep him informed. The two faults together produced a dysfunction larger than the sum of its parts, and that dysfunction cost the Confederacy dearly at Vicksburg and Atlanta.[^18]
The Source Problem
The evidentiary discipline of this volume is tested severely by the Davis–Johnston dispute, because both principals left lengthy postwar accounts written for the express purpose of fixing blame on the other. Johnston’s Narrative of Military Operations (1874) is, in large part, a sustained indictment of Davis, presenting the general as the victim of a president’s unreasoning malice from the seniority dispute onward.[^19] Davis’s Rise and Fall of the Confederate Government (1881) answers in kind, presenting Johnston as a chronic retreater whose secretiveness and passivity left the president no responsible choice but to remove him. Each memoir is a brief for the prosecution, and each is least reliable precisely where it is most emphatic.
The method applied throughout these studies is to anchor the narrative in the contemporary correspondence, the orders, dispatches, and letters of 1862 through 1864 gathered in the Official Records, and to read the two memoirs as evidence of what each man wished posterity to believe rather than as trustworthy reports of what occurred. Read this way, the memoirs become useful in a manner their authors did not intend: as documentation of the very distrust this paper describes, each man’s account demonstrating, in its selective indignation, the framework through which he had processed the relationship all along.[^20] The later Lost Cause historiography, which tended to adopt Johnston’s version because it served the larger project of explaining defeat by blaming Davis rather than the cause, must be treated as historiography to be analyzed and not as testimony to be believed, a treatment the ninth paper in this volume undertakes in full.[^21]
Significance
The Davis–Johnston dyad earns its central place in this volume because it is the clearest demonstration of the book’s organizing thesis: that Confederate command effectiveness turned on the interaction between Davis’s temperament and the particular general, not on either in isolation. Set beside the Lee partnership, the Johnston relationship isolates the variable precisely. The same president succeeded with one general and failed with another, and the difference lay in the dynamics of the relationship, a dynamic of accumulated trust in the one case and accumulated distrust in the other, each maintained by the conduct of both parties.[^22]
The relationship also demonstrates the strategic cost of command dysfunction in concrete terms. The distrust between Davis and Johnston was not a private unhappiness with no operational consequence; it degraded the unity of command in the Mississippi Valley at the moment of the Vicksburg crisis and shaped the relief that handed the Army of Tennessee to a commander whose offensive instincts would shortly destroy it. The dyad shows, in other words, that the quality of a command relationship is a military variable in its own right, capable of losing campaigns as surely as a flawed plan or an inferior force.[^23]
Conclusion
The relationship between Jefferson Davis and Joseph E. Johnston failed not through a single act but through a pattern, a self-reinforcing cycle of reticence and interference in which each man’s conduct confirmed the other’s worst expectations and deepened the distrust that had begun with the seniority dispute of 1861. Traced through the Peninsula, Vicksburg, and Atlanta, the pattern shows both men contributing to the failure: Johnston through a chronic unwillingness to keep his government informed, Davis through a hostility that denied Johnston the trust he extended to Lee. The dysfunction was real, it was mutual, and it was costly, contributing to the loss of Vicksburg and shaping the fatal change of command before Atlanta. Read against the preceding study of the Davis–Lee partnership, the Johnston dyad completes the volume’s central contrast: it shows that Davis was capable of both the best and the worst of command relationships, that the difference lay in the interaction rather than in the president alone, and that a relationship left to corrode could cost a nation its campaigns as dearly as any defeat in the field.
Notes
[^1]: For the framing of the Davis–Johnston relationship as a failure of maintenance rather than an inevitable consequence of the 1861 grievance, see Woodworth (1990) and Symonds (1992); the seniority dispute itself is treated in Paper 1 of this volume.
[^2]: On the self-reinforcing character of the distrust, see Woodworth (1990), whose study of Western command treats the relationship as a recurring pattern rather than a series of isolated quarrels.
[^3]: On Johnston’s conduct during the Peninsula Campaign and his reluctance to disclose his plans, see Symonds (1992) and Freeman (1942–1944, Vol. 1).
[^4]: On Richmond’s uncertainty about Johnston’s intentions and movements during the withdrawal up the Peninsula, see Symonds (1992) and Woodworth (1995).
[^5]: On Seven Pines, Johnston’s wounding, and Lee’s assumption of command, see Freeman (1934–1935, Vol. 2); on the lasting impression Johnston’s Peninsula conduct made on Davis, see Woodworth (1990).
[^6]: On Johnston’s 1863 department command spanning Tennessee and Mississippi, see Woodworth (1990) and Symonds (1992).
[^7]: On the conflicting imperatives facing Pemberton and the divergence between Davis’s wish to hold Vicksburg and Johnston’s preference for preserving the army, see Woodworth (1990) and Ballard (2004).
[^8]: On Johnston’s reluctance to exercise vigorous control over his department command, see Woodworth (1990) and Symonds (1992).
[^9]: On the breakdown of unified command in the Mississippi Valley and the role of distrust within it, see Woodworth (1990).
[^10]: On the multiple causes of the Vicksburg disaster, including Grant’s generalship, Confederate numerical and geographic disadvantages, and Pemberton’s judgment, see Ballard (2004) and Woodworth (1990).
[^11]: On the Atlanta Campaign and the state of the Davis–Johnston relationship by 1864, see Woodworth (1990), Symonds (1992), and McMurry (2000).
[^12]: On Johnston’s defensive withdrawal toward Atlanta and Davis’s growing alarm, see Symonds (1992) and Woodworth (1990).
[^13]: On Davis’s efforts to obtain a clear statement of Johnston’s intentions regarding Atlanta and Johnston’s refusal to commit, see Woodworth (1990) and Symonds (1992).
[^14]: On the relief of Johnston on July 17, 1864, and his replacement by Hood, see Woodworth (1990) and McMurry (1982); the consequences of Hood’s elevation are the subject of Paper 7 of this volume.
[^15]: On Johnston’s interpretation of the relief as the culmination of Davis’s longstanding hostility, see Johnston (1874) and Symonds (1992).
[^16]: On the reality and consequences of Johnston’s chronic reticence, see Woodworth (1990) and Symonds (1992), the latter of whom, though sympathetic to Johnston, acknowledges the pattern.
[^17]: On the contrast between Davis’s handling of Johnston and his handling of Lee, see Woodworth (1990, 1995); the Lee partnership is examined in Paper 2 of this volume.
[^18]: On the interaction of the two men’s faults as the source of the dysfunction, see Woodworth (1990).
[^19]: On the apologetic and accusatory character of Johnston’s memoir, see Symonds (1992) and the historiographical treatment in Paper 9 of this volume.
[^20]: On reading both memoirs as evidence of the distrust they describe rather than as reliable reports, see Woodworth (1990) and the method established in Paper 1.
[^21]: On the Lost Cause tradition’s adoption of Johnston’s version, see Connelly (1973) and the discussion in Paper 9 of this volume.
[^22]: On the Davis–Johnston dyad as the clearest demonstration of the interaction thesis, see Woodworth (1990, 1995).
[^23]: On the operational consequences of the relationship’s dysfunction, see Woodworth (1990) and Ballard (2004).
References
Ballard, M. B. (2004). Vicksburg: The campaign that opened the Mississippi. University of North Carolina Press.
Connelly, T. L. (1973). The marble man: Robert E. Lee and his image in American society. Alfred A. Knopf.
Cooper, W. J. (2000). Jefferson Davis, American. Alfred A. Knopf.
Davis, J. (1881). The rise and fall of the Confederate government (Vols. 1–2). D. Appleton.
Freeman, D. S. (1934–1935). R. E. Lee: A biography (Vols. 1–4). Charles Scribner’s Sons.
Freeman, D. S. (1942–1944). Lee’s lieutenants: A study in command (Vols. 1–3). Charles Scribner’s Sons.
Govan, G. E., & Livingood, J. W. (1956). A different valor: The story of General Joseph E. Johnston, C.S.A. Bobbs-Merrill.
Johnston, J. E. (1874). Narrative of military operations, directed, during the late war between the states. D. Appleton.
McMurry, R. M. (1982). John Bell Hood and the war for Southern independence. University Press of Kentucky.
McMurry, R. M. (2000). Atlanta 1864: Last chance for the Confederacy. University of Nebraska Press.
Symonds, C. L. (1992). Joseph E. Johnston: A Civil War biography. W. W. Norton.
U.S. War Department. (1880–1901). The war of the rebellion: A compilation of the official records of the Union and Confederate armies (Series I–IV). Government Printing Office.
Woodworth, S. E. (1990). Jefferson Davis and his generals: The failure of Confederate command in the West. University Press of Kansas.
Woodworth, S. E. (1995). Davis and Lee at war. University Press of Kansas.
