Provincia and Imperium
The study of Republican imperial structures is currently in a state of flux, as much recent research has cast doubt on or even overthrown central points of the received wisdom. As yet no synthesis of the progress made in the last two decades has appeared, and therefore a lengthy digression on the nature of mid-Republican provinciae and imperium is unavoidable here. The reader should not, however, mistake the discussion that follows for an attempt at the sort of comprehensive synthesis that is needed; I have tried to restrict myself to what is needed to set the context in which the changes of the 140s in Macedonia should be placed.
In discussions of Republican imperialism, we hear much of "annexation," "creating" or "organizing" provinces, indeed even of "converting" places into provinces. These phrases imply that areas were reduced to units of a recognized Roman empire by a specific act of organization. This is of course precisely what was envisioned by juristically minded scholars of the last century, who satisfied the overwhelming urge to reduce the chaos of historical reality to order by finding a neat, legal definition of provinciae: a legal enactment of the conquering Roman commander was postulated that served as the charter of the newly founded province; the pseudotechnical term lex provinciae was invented (there is no ancient authority, it seems) to denote it; and the concept passed into the standard handbooks, where it remains to this day, serving as the theoretical underpinning of the phrases mentioned at the beginning of this paragraph.[34] But the lex
[34] See Marquardt, RStV , 1:500-502; Mommsen, RStR , 3A:727-28, 746-47; Arnold, RSPA , 26-29; Abbott and Johnson, Municipal Administration , 48-52; Stevenson, RPA , 68, 165; Meyer, RSS , 233; de Martino, Storia , 2:283-86; Bleicken, Verfassung , 212-13; Jolowicz, Historical Introduction , 69-70. Lintott, Imperium Romanum , 22-42 (cf. Lintott, GR n.s. 28 [1981] 58-61), has at last supplied a convenient antidote with which the following account agrees on many, though not all, points. See also Crawford, Storia di Roma , 2.1:91-121.
provinciae is a seriously misleading construct. The absence of ancient authority for the term itself is gravely troubling: at the very least, the ancients' lack of interest in such hypothetical provincial "charters" is likely to be significant. Of the only two ordinances claimed as leges provinciae about which we have substantial information—the lex Pompeia for Bithynia-Pontus to which Pliny refers in his letters to Trajan and the lex Rupilia for Sicily of 132 B.C. —neither can be shown to provide the legal foundation of the province as a whole. To judge from the evidence available to us, the first was concerned with the local constitutions of the cities that Pompey had established in the former kingdoms of Bithynia and Pontus, the latter with judicial arrangements for suits between citizens of different communities.[35] There is, then, not one "lex provinciae " that can be proved to perform the wide range of functions attributed to the institution. Rather, important recent work has shown that the construct of a lex provinciae conflates into one measure what was in reality the cumulative result over generations of enactments by a long series of magistrates and decrees of the Senate.[36] Moreover, not all provinces should be assumed to have had, at any point in this incremental process, a "comprehensive group of ordinances arranging [their] administration."[37] Further, the old notion that a lex provinciae imposed a distinct "subject status" upon those communities that were not "free" was exploded even earlier.[38] Nor is Roman
[35] Lex Pompeia data Bithynis Ponticisque: Pliny Ep . 10.79, 112, 114; cf. 115. Lex Rupilia: decreto, quod is de decem legatorum sententia statuit, quam illi legem Rupiliam vocant , Cic. Verr . 2.2.32 (cf. 34, 37, 38, 40, 42: for cases between citizens of different cities, the praetor appointed judges by lot; sortition within thirty days of institution of suit, 37; arrangements for suits between a community and a person, 59, probably; between citizens cases are to be judged under their own laws, 90). Other Rupilian leges besides judicial ones: see Cic. Verr . 2.2. 40; 125 (constitution of Heraclea). The alleged lex provinciae of Sicily came approximately a century after the island was "annexed" on the usual view!
[36] See Lintott, GR n.s. 28 (1981) 60-61; Hoyos, Antichthon 7 (1973) 49.
[37] Lintott, GR n.s. 28 (1981) 59. Crawford, Storia di Roma , 2.1:112-16, retains the concept of the lex provinciae but sees it as a novelty of the last decades of the Republic.
[38] By Kienast (ZSS 85 [1968] esp. 332-34), building on the work of Heuss (Völkerrechtliche Grundlage , 76-77), against Mommsen (RStR , 3A:726-20, 730-32). Dahlheim implausibly thinks of a status of perpetuated deditio as the norm in the provinces (Struktur und Entwicklung , 70, 109; cf. Gewalt und Herrschaft , 71-72). See Kienast, Augustus (Darmstadt 1982) 366 n. 1, for a response. The imperial jurist Gaius's doctrine that in eo (sc. provinciali ) solo dominium populi Romani est vel Caesaris (Inst . 2.7) is of no relevance here: see most recently Bleicken, Chiron 4 (1974) 359-414·
administration something that was simply put in place: recent studies of the early history of the Roman occupation of Sicily and Spain have demonstrated convincingly how gradually the administrative and judicial functions of Roman commanders in these provinciae developed from their originally essentially military duties.[39]
It seems dear, then, that provincial structures eventually emerged out of a long process of adaptation and experience and were not imposed at one blow by victorious Roman commanders. A conquering Roman general might indeed fix tributary obligations and make certain structural alterations in the pacified territory, but these acts in themselves did not amount to creating a province, as is clearly shown by the example of Paulus's settlement of Macedonia in 168/167, when the former kingdom underwent a drastic constitutional change and half of the traditional royal tribute was diverted to Rome. On the other hand, "for several provinces all that seems to have happened is that at some point the Senate recognized that the territory must henceforth be decreed as the province of a magistrate ... every year."[40] Talk of "annexation," "creating," and "organizing" provinces on any single occasion corresponds to nothing in our evidence of the second century B.C. ,[41] which helps to explain why we have no reliable explicit testimony to the "creation" of the provinces of Sicily, Sardinia, the Spains, Transalpine Gaul Cilicia, or Cyrene (to take a few prominent examples), and why the attempts to discern precisely when these provinces were "created" seem so many unconvincing exercises of scholarly ingenuity.[42] The basic working assumption was ill founded. It may be noted as
[39] Sicily: Dahlheim, Gewalt und Herrschaft , 12-73. Spain: Richardson, Hispaniae , 172-80 and passim. For the central role played by commanders in the field in the development of these provinciae , see, in addition, Eckstein, Senate and General , 103-232.
[40] Rich, LCM 10 (1985) 94.
[41] See Rich, LCM 10 (1985) 94, and Richardson, Hispaniae , 178-79. By the Principate the development of a strong notion of provincia as a unit of empire, and uniformity in the forms of administration, had led to both a recognizable provincial pattern and a distinct notion of annexation: hence the phrase redigere in provinciae formam , common in writers of the Principate. Already Caesar had used the phrase in provinciam redigere (BGall . 1.45.2, 7.77.16).
[42] For Sicily and Sardinia the assumption has traditionally been that the institution of additional praetorships in 227 implied the concurrent formal organization of provinces, but it is unclear why it implies more than recognition of the continuing need for an increased magisterial pool to cover expanded military commitments (see Harris, War and Imperialism , 136). For the same reason 197 has traditionally been accepted as the date of the organization of the Spains, with minority claims for 206/205 (see especially Harris, CAH 8 (1989) 121: "the first Roman annexation as that term is usually understood"; cf. War and Imperialism , 136 with n. 4) or even 133 (Bernhardt, Historia 24 [1975] 420; Sumner, CP 72 [1977] 129). Richardson, Hispaniae , 178-79, elegantly cuts the Gordian knot, as does also, for Cisalpine Gaul, Ausbüttel, Hermes 116 (1988) 117-22 (esp. 121). The dispute on the date of the annexation of Narbonese Gaul (for which see especially Badian, Mél. Piganiol , 901-3, and more recently Ebel, Transalpine Gaul , 75-95), and of Cilicia and Cyrene should similarly be laid to rest.
well that the Romans themselves stubbornly refused to make the distinction that we find so crucial between provincia as "permanent/formal/territorial province" and as "sphere of command."[43] Nor need the use of provincia even in a clearly geographical sense imply official demarcation or a more formal structure.[44]
The lack of evidence for a comprehensive structural reorganization in Macedonia in the 140s no longer looks therefore like a mere accident of our evidence. But more: we have seen that there was no specific act that answers to the idea of "annexation"; no benchmark that would allow one to say, "Now, and not before, this place belongs to the Roman Empire." The mere assignment of Macedonia to Roman commanders as a provincia ,
[43] A point effectively raised by Richardson, JRS 69 (1979) 159. Here Lintott is confounded by his belief in the formal demarcation of provinces: "A governor of Macedonia had opportunities for action outside Macedonia and indeed outside territory under direct Roman rule, which he could argue were part of his provincia in the sense of job, even if they were not part of his territorial provincia " (GR n.s. 28 [1981] 56). Lintott thus tries to evade the absurdity that arises from supposing that the proconsuls of Macedonia, Cilicia, and Syria in their normal course of activities continually left and reentered their provinces by postulating "a penumbra of responsibility outside the territory [magistrates] ruled directly" (p. 58). A far simpler hypothesis, given Lintott's own insightful arguments about the inconvenience of provincial boundaries, is that provinciae were strictly and formally demarcated only where it was necessary (1) to prevent commanders from stepping on each other's toes (e.g., division of the Spains in 197—in any case, not rigidly adhered to: Sumner, CP 72 [1977] 126-29; cf., however, Develin, Kilo 62 [1980] 355-68)—compare in this connection the lex Porcia of the end of the second century (Lintott, Imperium Romanum , 26-27)—and (2) for the functioning of maiestas laws at the regular points of entry and exit (cf. Cic. Fam . 3.5-6; cf. Lintott, GR n.s. 28 [1981] 54-55). Brunt rightly wonders whether provincial frontiers were regularly defined ("Laus Imperii," 324 n. 59). Lintott's trouble seems to stem from the notion that fines provinciae will have strictly distinguished "territory administered" (p. 55) from that not subject to "direct Roman rule" (p. 56); but that assumption is not defended and leads precisely to the bizarre results that Lintott describes.
even if in regular succession, was of course only a matter of Roman constitutional practice and did not affect the legal status of the area, any more than had previous assignments of Macedonia or even Graecia provincia during Rome's earlier wars in the East. The dear distinctions of the past between territory subject to Rome and the rest blur considerably. These preliminary conclusions raise a larger question of great importance for our understanding of the origins of Macedonia provincia . How did the Romans themselves define their empire in the second century?
Polybius is the sole contemporary source on second-century Rome that is available to us in any significant quantity; Livy cannot be considered a trustworthy guide to a conceptual world a century and a half before his time. Two central Polybian ideas about the nature of Rome's domination of the Mediterranean world stand out in sharp relief. The Romans had in 167 (with the defeat of Macedon and the abolition of the Macedonian kingship) completed the establishment of an arche over virtually the entire
Polybius is not, of course, Roman. Does it follow, as some have suggested, that he cannot be regarded as a useful guide to contemporary Roman perceptions of their supremacy?[47] A priori reflections are ultimately indecisive. It is not difficult to impugn some aspects of Polybius's view. It
[47] So Richardson, PBSR 47 (1979) 1-11; Gruen, HWCR , 278-79; cf. 343-51.
was perhaps all too easy for a young Achaean of the political class who came of age between the wars with Antiochus and with Perseus to attach inordinate significance to the matter of Roman orders, and in particular to what extent, and with what alacrity, they had to be obeyed. This was indeed the burning question of Greek leaders of the 180s and 170s[48] —and nothing may have seemed a more dramatic answer than the Roman deportation to Italy in 167 of all those who had seemed to be waiting on events rather than actively serving the interests of Rome.[49] Little wonder that from Polybius's perspective the matter was now dosed. But once the settlement and Rome's reprisals had passed, it was once again debatable whether its orders could be disobeyed or circumvented, and a study of Roman actions and Hellenic responses to them after 167 shows that at the least Polybius's view of the necessity to obey Rome is too starkly drawn.[50] But although this was an individual's fallible judgment of the factual situation, that does not diminish the significance of obedience to Roman commands as a basic and broadly understood criterion of empire. As for possible distorting effects of Polybius's Hellenic perspective, we must note that he was after all no ordinary Greek, unaware of Roman ways. During his detention in Rome, Polybius associated closely with prominent Romans for nearly two decades before returning to Greece, especially with P. Cornelius Scipio Aemilianus, the son of the conqueror of Macedon; and one of his main objectives, often explicit, is to explain the Romans to his Greek readership.[51] We should therefore need good reason to assume that Polybius so fundamentally misrepresents or misunderstands the Romans' own views about the nature of their arche , the very object of his lengthy and painstaking investigation. It is not enough simply to note that Polybius was capable of misunderstanding. We must be shown that such misunderstanding occurred.
[48] See especially Polyb. 24.8-13, in which the differing views of the Achaean statesmen Callicrates, Aristaenus, and Philopoemen are put on display and given great prominence. On Polybius's early education in Roman power, cf. now Eckstein, GRBS 26 (1985) esp. 277-82.
[49] Some 1,000 Achaeans alone, according to Paus. 7.10.11.
[50] Cf. especially Gruen, HWCR , 96-131, 192-98, 335-43, 517-20, 578-92, 660-71. See now Morgan, Historia 39 (1990) 37-76, for a vigorous attack on Polybius's "schematism"; but "perceptions" are hardly so easily distinguished from "realities," or so unimportant, as Morgan thinks.
[51] Cf. esp. Walbank, Polybius , 3-6. On Polybius's Roman associates, cf. now Dubuisson, Latin de Polybe , 260-62. Dubuisson is inclined to stress Polybius's Roman audience as well (pp. 266-67), but without diminishing the importance of the Greek.
One attempt to do so stresses the absence of Roman legal concepts such as imperium and provincia from Polybius's conception of the arche .[52] This line of argument would be valid only if we knew in advance that these concepts played a central role in Roman views of their empire in the middle of the second century, but that is precisely what is under investigation. Appeal to the ideas of Cicero, Strabo, Augustus, and others a century or more later will avail us not at all. Even Livy, who indeed gives much attention to the assignment of provinciae and grants of imperium , nowhere defines the imperium populi Romani as the sum of the provinciae . On the contrary, he can have a speaker say of L. Scipio, without a hint of paradox, that by his campaigns in Asia of 190/189 he had "extended the imperium populi Romani to the limits of the world" (38.60.5), although Livy knew well that not an inch of territory was "converted" into a province. Nor does he balk at having Cn. Manlius Vulso say in 187 that Greece and Asia were then under Rome's ius and dicio , and that the Taurus Mountains had become the finis imperii Romani (38.48.3-4). It will not do to assume that these statements (as well as similar ones that appear in Cicero) are mere oratorical fancies: if by the late Republic the imperium were sharply and strictly defined by reference to the provinces, such statements would simply be odd paradoxes rather than rhetorical extravagances.
Indeed, on the contrary, these statements illustrate well that even at the end of the Republic the imperium populi Romani was not inevitably linked with the provinciae .[53] After all, in the Augustan period allied kingdoms, whose legal status as socii et amici was no different from much of the Hellenistic world in the later second century, were regarded as part of the imperium .[54] In the Res gestae itself (26.1), Augustus unmistakably implies that there were peoples outside the provinciae who p [arerent imperio nos ]tro (cp. 30.2). If as late as the Augustan principate, when the military frontiers had roughly stabilized, the imperium was not strictly coextensive with the provinces, we should expect Roman conceptions of the imperium to be even more flexible nearly two centuries before, at a
[52] So Richardson, PBSR 47 (1979) 1-11.
[53] Cf. Lintott, GR n.s. 28 (1981) 64-66; Brunt, "Laus Imperii," 168-70; Harris, War and Imperialism , 105-6.
time when the permanent occupation of foreign territory overseas was far less extensive and the limits of Roman power still very much in flux.
An understanding of the imperium populi Romani begins with the phrase itself. It means fundamentally the "sway" or "supremacy" exercised by Rome over others; it does not mean "the Roman empire" in the way we use the term.[55] A Roman did not speak of a political community as "in the Roman empire" but as "under Roman sway" (sub imp trio populi Romani ) as a result of a Roman commander's action in "putting it under Roman sway" (subicere/subiungere imperio p. R .). What you do when you are "under Roman sway" (sub imperio p. R .) is "obey" it,[56] just as the citizen obeys the imperium of the consul, the son that of his father, the slave that of his master.[57] Imperium as "sway" is precisely the power to impose individual imperia or imperata , commands or orders.[58] Power and command are thus fundamentally linked in the Roman conception of imperium , a word that expressed above all the concrete relationship of power itself rather than an abstract concept of "empire" as formally and geographically defined. This brings us back to Polybius's stress on the significance of Roman commands and indifference toward concrete exploitation or possession of territory. His idea that the necessity to obey Romans defines and characterizes their arche fits too closely with the root meaning of imperium to be mere coincidence.[59] So natural was the link between obedience to any command and submission to the imperium that Livy can easily render Polybius's phrase
[55] Lintott, GR n.s. 28 (1981) 53. See now also Richardson, JRS 81 (1991) 1-9.
[56] E.g., orbem terrarum parere huic imperio coegit (sc. rei militaris virtus ) (Cic. Mur . 22). The verb parere (obey) is the natural complement in Roman contexts to imperium in this sense; cf. the passage from Augustus's Res gestae 26.1, quoted above.
[57] Cf. Plaut. Men . 1030: Nemp' iubes? Iubeo hercle, si quid imperi est in te mihi; Mil . 611: facilest imperium in bonis; Ter. Ad . 65-67 (in the context of father-son relations): errat longe mea quidem sententia, qui imperium credat gravius esse aut stabilius vi quod fit quam illud quod amicitia adiungitur .
[58] See OLD s.v. imperium .
[59] Noted in passing by Lintott, GR n.s. 28 (1981) 54.
[60] Polyb. 21.19.10 = Livy 37.53.4.
Hellenistic Greece as well as in Rome that the necessity to obey the victor's every command is the natural lot of those defeated in war.[61]
Although no conclusion can be absolutely decisive in view of the paucity of contemporary documentary evidence, those texts of Roman origin we do possess tend to confirm the essential identity of the Polybian and contemporary Roman conceptions of empire. In the Aetolian peace treaty with Rome, ratified in 189, the Aetolians committed themselves to preserve (?)
[63] Cic. Mur . 75: necesse enim fuisse ibi esse terrarum imperium, ubi ille esset .
[64] Orig . 5.3b Chassignet: Atque ego quidem arbitror Rhodienses noluisse nos ita depugnare, uti depugnatum est, neque regem Persen vinci. Sed non Rhodienses modo id noluere, sed multos populos atque multas nationes idem noluisse arbitror atque haud scio an partim eorum fuerint, qui non nostrae contumeliae causa id noluerint evenire; sed enim id metuere, si nemo esset homo, quem vereremur, quidquid luberet, faceremus , ne sub solo imperio nostro in servitute nostra essent. Libertatis suae causa in ea sententia fuisse arbitror (emphasis added).
Roman power in the region would leave the Rhodians sub solo imperio nostro , "under the sway of us alone,"[66] although there is clearly no suggestion that Rhodes might be made into a province. For him, as for Polybius, the imperium has no essential connection with provinces but consists in a practical relationship of power and domination that might extend well beyond the confines of Italy and even those areas in which commanders and Roman troops were present.[67] In light of these passages there is little temptation and less reason to suppose that Polybius has arbitrarily imposed an alien, Hellenic perspective when he reports Scipio Africanus's declaration in the 180s that the Romans enjoyed "control" (
Nor, I would add, was this exclusively an early phenomenon. A recent study of usage of the phrase has shown that before the first century B.C. an unambiguously "territorial" connotation of imperium populi Romani cannot be traced.[71] Notably, only in the late Republic do fines or termini of the imperium receive particular emphasis. Sulla was the first since the
[66] A phrase that suggests that one imperium need not even be exclusive of all others: a state can be dominated by more than one power (sub solo imperio nostro ). It is when a single imperium remains, with no remaining alternative to act as a balance, that servitus results.
[67] Gruen, however, seems to conclude from the usage of imperium in these passages that the Romans' "acknowledgment of empire" did not come until as late as the years between the Gracchi and Sulla (HWCR , 281; cf. 274).
[69] On which see especially Harris, War and Imperialism , 105-6; Lintott, GR n.s. 28 (1981) 54, 60-61.
[71] M. Awerbuch, Archiv für Begriffsgeschichte 25 (1981) esp. 171-84. Her adherence to a traditional conception of the nature of provinces (pp. 166-70) does not affect the validity of her observations. See now also Richardson, JRS 81 (1991) 5-7, who places the "full development" of a territorial connotation in the early Empire.
regal period to extend the boundaries of the pomerium , but the claim seems to have rested on territorial expansion in Italy rather than abroad.[72] Pompey made the proud claim, echoed by Cicero, that he had made the boundaries of the empire coextensive with those of the earth.[73] While it is true that in the prayer for the Secular Games of 17 B.C. the gods were called upon augere imperium , we do not know whether this phrase appeared in earlier versions, and in any case it is misleading to interpret this to mean "extend" rather than simply "increase": Cicero was quite happy in the Fifth Philippic (48) to include Flamininus among those who "increased the imperium of the Roman people" (populi Romani imperium auxerint ), although it was fifty years after Flamininus that Macedonia came to be assigned regularly as a provincia .[74] Even in the first century, then, the imperium was not regarded so much as a geographical expanse in which the Roman people exerted a dominant power but as that power itself; all the more so did this apply, we should suppose, in the second century.
These arguments do not commit one to the notion that all Romans had in mind a dear formulation of the nature of the imperium populi Romani , much less that they would all have agreed with Polybius's ideas of precisely when and how far Rome's supremacy was established. On the contrary, perhaps the most important result of the arguments above is precisely that
[72] Sen. Brev. vit . 13.8; Gell. NA 13.14.1-4; Tac. Ann . 12.23. Mommsen, RG , 2:355-56, who is often followed on this point (cf. Badian, Roman Imperialism , 34), guessed that Sulla's justification lay in a strictly formal extension of Italia . But the key point was surely the integration of virtually the entire peninsula into the ager Romanus as a result of the Social War: see now Sordi, CISA 13 (1987) 200-211. Note that only Tacitus speaks of extension of the imperium , while Seneca and Gellius both write of ager acquisitus or captus . Frei, MH 32 (1975) 75-76, justifiably speculates that the rule connecting extension of the pomerium with conquest may be an invention of the Claudian era. In a dialogue written in the late 50s but set in 129 B.C. , Cicero refers to monuments of generals on which was inscribed the boast of having extended the fines imperii (Rep . 3.24). But no examples of such inscriptions have yet been found. While it would be rash to discount a contemporary witness altogether (so Gruen, HWCR , 284-85), it is hardly safe to take the reference as secure testimony to second-century practice (so, apparently, Harris, War and Imperialism , 125), since Cicero could well have committed an unwitting (and for his purposes insignificant) anachronism on such a matter.
[73] Diod. 40.4; Cic. Cat . 3.26, Sest . 67, Balb . 13.
[74] Cf. Gruen, HWCR , 283, and Sherwin-White, JRS 70 (1980) 177, against Harris, War and Imperialism , 120-22. The haruspices on various occasions in the early second century predicted an extension of fines (see most recently Gruen, pp. 283-84); but imperium is not mentioned.
there was no neat, formal definition of empire for the Romans any more than there is for us.[75] The author of the Acilian extortion law of 123 or 122 has no pat phrase at hand to denote Roman-controlled territory and is forced to resort to a long circumlocution, in which provinciae and imperium nowhere appear, perhaps because these were insufficiently specific terms for inclusion in the law.[76]Imperium populi Romani was, of course, a phrase both concrete and not easily definable. The view of empire that I have argued was common to Polybius and to Romans in the second century focused on the actual capacity of the "metropole" to enforce its will upon the "periphery,"[77] rather than on legal structures (e.g., "provinces"). It was thus by its very nature nuanced and open to interpretation, unlike one that was formally or legally defined. It referred to norms of behavior and concrete means of control—the facts of power and the psychology of dependence—and thus was closely bound up with the real situation at any one time; but the "real" situation could change at any time, depending on precise circumstances. A Roman military presence could, of course, help to secure the imperium in any particular area; but as we have seen from Polybius and other texts, especially Cato, the imperium populi Romani prevailed no less where there were no Roman troops, as long as the dialogue of Roman command and indigenous obedience persisted. Senators in the middle of the second century B.C. will not have thought in terms of "annexations" or "organizing provinces."[78] The notion of the imperium as a spatial extent, geographically bounded, was little developed at this time. What counted was rather the maintenance and augmentation of Rome's power and supremacy—the imperium populi Romani in its original sense.
[75] Rightly stressed by Gruen, HWCR , 278-85 (cf. the comments of Devine, AJP 108 [1987] 785, and Rich, LCM 10 [1985] 95). Richardson's stress (JRS 81 [1991] 6) on the "co-existence of a pair of meanings" of the very concept imperium populi Romani is also apt. I differ from Gruen mainly in denying that the lack of a clear or official definition of a thing implies the absence of the thing. On definitions of empire, see above, p. 3.
[76] FIRA 7, line 1: [quoi socium no ]minisve Latini exterarumve nationum, quoive in arbitratu dicione potestate amicitiav [e populi Romani ]. Cf. Sherwin-White, JRS 70 (1980) 179; Gruen, HWCR , 281.
[77] Cf. Doyle's rather stricter, but similar definition, quoted in the Introduction, n. 5. Yoshimura's discussion of libertas and auctoritas (AJAH 9 [1984 (1988)] 1-22) similarly stresses the essential role of command and the evolution of "Machtverhältnisse" rather than legal concepts in the development of the Roman "empire."
[78] Harris, War and Imperialism , 105-6. See further on the Roman perception of empire, Crawford, Storia di Roma , 2.1:91-99.