Barbarians and Romans, A.D. 418-584

Ulpian’s model form for census declarations makes no reference to buildings or orchards (Dig. 50. 15. 4), which are likewise omitted from the surviving Asian census inscriptions;¹ a prefectoral ordinance of 512 (Justinian Novel. 168) specifies, “in censum seu descriptionem praedia solum referantur, non autem domus vel aliae res.” Although another model for a plena describtio includes the entry “quis aedificiis ac possessionibus ornatus,” this is with a view, not to a census declaration, but to incorporation with the imperial res privata (CTh 9. 42. 7 [369]; cf. CTh 10. 9. 2 [395]).

I can find nothing to suggest that the state set a value upon country houses. Imperial indulgence toward house property in cities is nicely illustrated in CTh 13. 6. 7. 1 (375), which insists that, in the contribution of houses to the navicularian munus, no account is to be taken of improvements and observes that houses have value for the adornment (cultus) of cities rather than for their revenues (fructus).² The levy of one year’s rent imposed in 405 on all city buildings—from warehouses and stores to dwellings—was altogether exceptional (CTh 11. 20. 3). See also CTh 15. 1. 41 (401). The lesson suggested by the privileges listed in Chapter II n. 12 is that the quartering of soldiers and officials was the unique “tax” regularly imposed upon city houses. This resource was exploited in earnest during the fifth century in the East (note particularly NTheod II 25 [444], in which quartering gives rise through exemption to a money revenue). Justinian Novel. 43 and 59 (536-537) attest that taxes and services were levied upon shops in Constantinople from at least the time of Anastasius, but there is no apparent connection between these dues and the general fourth-century scheme of taxation.

The distinction made by LBurg between woods and wastes (silvae, campi) and “arable” (terrae) seems consistent with the general direction of assessment in the later empire. Although prata, pascua, and silvae (caeduae) were subject to census declaration according to Ulpian (Dig. 50. 15. 4),³ they do not appear in so many words in the Asian census inscriptions. In the surviving late Roman scale for conversion into assessment units (iuga), a special procedure, eschewing tax units, is laid down for assessing mountain lands sown with wheat and barley and for imposing a low payment in money (to the res privata?) on mountain pastures.⁴

Almost all the evidence in the law codes is concerned with so-called “desert” land, rather than with the descriptive categories of LBurg. The two need not correspond, for the fiscal concept of “desertion” has no precise agrarian meaning (above, Chapter IV n. 18). Nevertheless, LBurg 72 uses deserta as the antonym for cultivated land. As shown by the taxpayer’s right to indulgence for land that unavoidably went out of cultivation (Dig. 50. 15. 4. 1; CTh 13. 11. 15-17 [417]), the state recognized a distinction between productive and nonproductive land; simultaneously, it insisted that new owners must take bad land with good (for example, CJ 11. 62. 3, 59. 5, CTh 5. 18. 8, 14. 30, 33, 34, and many others) and that conlationes and munera were borne by gleba inutilis together with terrulae fertiles (CTh 11. 24. 62 [415]). These rules illustrate the coexistence of divergent viewpoints: on the one hand, that taxes were owed by the total unit of ownership, at rates varying with the productive value of the land (rather roughly graded); on the other hand, that taxes were owed exclusively by the lands effectively producing the obvious commodities—grain, oil, and wine. Although the latter is characteristically a taxpayer’s viewpoint, the state’s practice of converting professiones into uniform assessment units (iuga, capita, millenae, and so forth) ought logically to have encouraged everyone concerned to identify the units at which an estate was rated with its productive arable, to the exclusion of its less profitable parts, such as woods and wastes.⁵ A similar effect would have resulted from the tendency for tax liability to devolve from persons to land:⁶ when a person is the unit of contribution, he is understood to assist the state on the basis of his total fortune, but when "land” is expected to “pay,” its contributory potential is narrowly limited by fertility. It seems correct, therefore, to conclude that, by the fifth century, woods and wastes would not normally appear in the registers of professiones or be thought to be referred to by the standard units of assessment.

The state, however, never surrendered its interest in these lands. Fourth-century veterans had been promised lands ex vacantis (CTh 7. 20. 8 [364]), and they were invited to appropriate “loca absentium squalida et situ dissimulationis horrentia, de solida fructuum indemnitate securi,” with the owners being denied any right to collect rent from them (CTh 7. 20. 11 [368]). It would have been consistent with such laws as these to complement an award of terrae to barbarian troops with regulations specifying the rights of the beneficiaries to lands classified otherwise than as taxpaying arable.

¹ These inscriptions are generally ascribed to the fourth century; they are analyzed by Déléage, Capitation, pp. 164-194; Cérati, Caract. annon., pp. 385-412; and Goffart, Caput, pp. 113-121.

² This munus is not precisely assimilable to a tax. A naviculary was tax exempt on the grounds that his entire property was obligated to the munus—an essential public service; when he alienated part of his property, the obligation devolved pro rata upon the acquirer even if he was not a naviculary. The law of 375 in question here limited the liability incurred by those who bought city dwellings from navicularies.

³ The qualification that the silvae to be declared are caeduae, i.e., under exploitation, strongly suggests that all three items were assumed to contribute actively to the declarant’s wealth; they were not “waste.”

⁴ Syro-Roman Lawbook, leges saeculares 121, FIRA 2, p. 796; for the assessment of arable according to this schedule, Goffart, Caput, pp. 33-34. Gregory of Tours Virtutes s. Iuliani 17 affords a startling parallel to the tax on mountain pastures described in the Lawbook.

⁵ Cf. my remarks in Caput, pp. 95 (Theodoret’s misunderstanding of iuga), 108-110. When the tax declaration of a large and complex farm, expressed in the ordinary, and very diverse, measurement units of land survey was sifted through the schedules that converted real measures into abstract units of fiscal assessment, the result was a homogeneous figure of, shall we say, eight and three-quarters iuga. (Note that a very large proprietor like Sidonius appears to have been rated at a small total for all his declared assets: something between fifteen and thirty fiscal units, Caput, p. 131 n. 26.) Even though, in origin, our abstract 8.75 was the conversion figure for every exploitable resource that the declared farm contained, anyone converting backward from tax units to real assets might easily reach the simplified conclusion that the iuga—a small number—referred only to the most obviously productive sector of the total farm.

⁶ I argue in Caput that this devolution took place; for a summary, see pp. 94-98.