Like most historical generalizations, these three phases of the empire cannot be taken as precise demarcations. There is some blurring of the distinctions and some overlapping of the time segments. Edward Luttwak presents the three phases as he devises them:
With brutal simplicity, it might be said that with the first system the Romans of the republic conquered much to serve the interests of a few, those living in the city — and in fact still fewer, those best placed to control policy.
Even within this first phase, there was development, both of the basic principles of this phase, and of what would become features of the next two phases. “During the first century,” Luttwak writes, “Roman ideas evolved toward a much broader and altogether more benevolent conception of empire.”
Although the first phase was characterized by expansion, it was in 9 A.D., during this first phase, that a German chieftain, known as Arminius or Herrmann, successfully attacked, defeated, and largely destroyed “three legions and auxiliary troops” commanded by Varus. Thus features of the latter two phases — maintenance and defense — are seen already in the first.
The clever act of bestowing fully Roman citizenship on many people living in the colonies or provinces marks the second phase. Enjoying their citizenship, and the rights and privileges it brought, these people developed an attachment to the empire, and were willing to support it and serve in its defense.
Under the aegis of the second system, men born in lands far from Rome could call themselves Romans and have their claim fully allowed; and the frontiers were efficiently developed to defend the growing prosperity of all, and not merely of the privileged. The result was the empire of the second century, which served the security interests of millions rather than of thousands.
During the third phase, in which Edward Luttwak sees the Roman military as largely defensive during the gradual contraction of the empire, echoes of the first phase are seen in, e.g., attempted Roman expansion eastward into Persia and Persian-controlled territories.
Under the third system, organized in the wake of the great crisis of the third century, the provision of security became an increasingly heavy charge on society — and a charge very unevenly distributed, which could enrich the wealthy while certainly ruining the poor. The machinery of empire now became increasingly self-serving, with its tax collectors, administrators, and soldiers of much greater use to one another than to the society at large. Even then the empire retained the loyalties of many, for the alternative was chaos. When this ceased to be so, when organized barbarian states capable of providing a measure of law and order began to emerge in lands that had once been Roman, then the last system of imperial security lost its last source of support, men’s fear of the unknown.
In addition to dividing the military history of the empire into these three phases, Luttwak also conceptualizes the fall of the empire as having one of its many roots in the growing awareness on the part of the Roman populace of the fact that the Germanic tribes could be, and often were, reasonable. The end of the empire would not be a descent into total chaos and destruction, but could be a someone orderly transition of power.
Indeed, when the Germanic chieftain Odoacer ended the empire in 476 A.D., he was careful to observe a number of customs and phrases as he presented himself, and a number of Roman institutions were left functioning in place, like the Senate.