The Colonate in Justinian'sReign

A. J. B. SIRKS

Every studyof the colonate of theLater Roman Empire is confronted with the legal consistof fragments sourcesinTheodosius' and Justinian's of imperial Codes. Since these Code with addressee, date and issuing constitutions, usuallyenteredintothe emperor,it is verytempting as fragments toconsiderthem per se, as unrelated historicalfacts which are oftenbut feebleimagesof theoriginal. They have been put, it is true,intotitles which are said to deal eachwith a particularsubject,but does this make the individualfragments relateto each other? a sentenceina languagethe Codes would be comparable to reading Such a viewof these grammar ofwhich one does not know: there may be individual words which one might understandbut coherence is lacking. And thereare certainly collectionsof legal rules which do indeedlackcoherence, which are nothingbut a disorderly mess. Yet even fora law collectionas earlyasHammurabi's Code ithas been argued thatit is to some extent That is certainly thecasewith the Roman law structured and thatcaution is necessary.1 codes.The orderof thepraetorianedict may have been a historical growthonly; itnever for commentaries, theless served for centuriesas a structure and theGregorian and HermogenianCodes, aswell as thetwoabove-mentioned Codes, inpart followeditsstruc was made toorganize the ture. Butwith each code also an effort material systematically. were used in legal Moreover, twootherstructures writings:thatof Sabinus' treatment of thecivil law, and the institutional ofwhich Gaius' gained greatpopularity. The system, on theedict and Sabinus' treatise, as well as the institutional commentaries works, are Roman lawwas an intellectual evidenceof thefactthat And as in thecase of the system. sentenceina foreign languagetheremedy presentsitself (thegrammar),so here it isbetter to take thecodification and toconsideritfromthelegalpointof view, thatis,as seriously Justinian himselfsaid for theDigest, 'sit una concordia una consequentia adversario nemineconstituto'(CJ I.I7.i.8), ratherthandiscardingthecodification as such.2
I JUSTINIAN'S CODE

were builtaround a structure. For Justinian's Those twocodifications3 Code, thestructure ofTheodosius' Code4 first servedas an example,but it was completely overhauledforthe second editionof A.D. 534. The compilersput the textstheyconsideredrelevantat the place in thestructure they considered appropriate. This had important consequences.

to resolve long-standing questions. Decisiones' In the Digest there are indeed still differences of 'Quinquaginta opinion, but these served to deepen legal discussion and doctrine; the Code was primarily meant for practice. As to the in c. Cordi 3, see A. J. B. Sirks, 'From the Theodosian regards the similarities and contrarities mentioned in Atti dellAccademia 6 (1983 [1986]), 299-300, Romanistica Costantiniana and idem, 'The Justinian Code', in the ms. Vat. reg. Lat. 886', ZSS-rA 113 (1996), 257-9, 267. Summaria Antiqua Codicis Theodosiani 3 use I the word codification in the wide sense of a compilation of rules, brought into a more or less systematic structure, and not in the far more restricted modern sense of a code. 4 On the Theodosian the Law (2000); F. Millar, A Greek Roman Empire: Code, see J.Matthews, Laying Down II (408-450) Power and Belief under Theodosius Code. A Study (2007). (2006); A. J. B. Sirks, The Theodosian JRS 98 (zoo8), pp. I20-143. ? World Copyright Reserved. Exclusive Licence to Publish: The Society for the Promotion of Roman

1 See R. Westbrook, 'The nature and origin of the XII Tables', ZSS-rA 105 (1988), 74ff., who strongly argues that various sanctions inHammurabi's Code must be read as theoretical discursions. But a lot of European medieval town statutes are indeed mere compilations of enforceable social rules without much or any coherence. 2 This was certainly not meant as a mere aspiration. Between A.D. 529 and 534 Justinian took the so-called

Studies 2008

THE

COLONATE

IN

JUSTINIAN

S REIGN

121

Firstof all, thefactthatat a certain moment in timeall data considered particularto a subjectare collected is alreadya statement aboutwhat people at that momentconsidered relevant to thatsubject. These datamay vary inage, and they may certainly also be judged upon their value at their date of issue, but they have an independent and perhapsdifferent value forthe moment they were collectedand selected. Within thatlatter contextthey may and shouldbe considered.Itdependson thedesignof thecodification what valuemay be attached to thiscontext.Justinian was veryexplicitabout his Code: all constitutions collected in theGregorian,Hermogenian and Theodosian Codes, and all constitutions issuedafter A.D. 438 (thePost-Theodosian Novels), shouldbe collected inone code, but as they were found. not simply They had tobe examined; what was superfluous shouldbe removed, as shouldwhateverwas similaror contradictory or had become obsolete, and to the likewise fortheintroductions Novels. Texts could be combinedand shortened. The constitutions shouldbe put underan appropriatetitle and there shouldbe no doubt about It followsfromthisthatthetexts their generalvalidity.5 of Justinian's Code presentin the first place rulesof generalvalidityand that in any case within a titleruleson the same have been gatheredtogether. subject(indicated by thetitle'srubric) This isnot to say that containsall theruleson thatsubject, sucha title because itcontainsthem only in so faras there had been constitutions issuedabout it. to an actual case. It could Many textsrefer happen that when a case was put before theemperorto decide, he, or laterjurists, dis cernedin the decision theapplicationof a newgeneralruleor theadjustment of an existing one. In thatcase a constitution to this (an edictor edictal letter) was issued, or the purport in this text(e.g., a rescript) was interpreted way. In view of thenatureofRoman legisla we may expect that were dealtwith ina general tion, arose and problems way when they a general remedy, required thatperhaps somepotentialproblems were includedtoo,but not thattheadministration triedina perfectionist way to think up all theoretical problems inadvance and dealwith themin a general way in theconstitution. Such approachesdate from thedays of the But it ispossible,and in the Natural Law scholars. privatelaw indeed veryoften thecase, thatother rulesare contained in the writingsof jurists,in thiscase collectedinJustinian's Institutes also gained theforce of Digest. Furthermore, Justinian's law.Thus to findthelaw on a specific subjectas valid in theyearsA.D. 530-534, one has to checkall thesethree himselfindicates(CJ I.I7.1.II, z.II).6 works, as Justinian It is still possible that thereare some pointswhich were governedby customary, unwritten or unrecordedlaw,but normally we may in this all thelaw way expect to find intention was indeedaccomplished. and that Of course, this must be accompaniedby a structure of the law (thedogmaticsof law) as prevalentat knowledgeof the intellectual words and concepts does one get a pictureof the law mightbemisunderstood. Only then at that What theoriginalreason forissuing a constitu particular pointor period in time. tion was no longer mattered in thecodification do not present unrelated process: thetexts of previouslyissuedconstitutions but have become (if they were not already) fragments That picturecan and shouldof course be put into thecontextof the pieces of a system. and culture of the moment.7 society
5 c. Haec 2; c. Cordi 3. 6 For an example of this method see Sirks, op. cit. (n. 4), Nr. 34. 7 The texts in codifications can also be treated as historical sources and used as theywere issued in their own time, order in the codifications. Within that context it is sensible to try to join conveniently collected in chronological fragments of the same original constitution. Thus it is quite usual to see a treatise on the colonate start with the has to realize, for that reason later aware of the context One of the codification ones

as itwas valid at that moment that time

and only then

no obsolete

rules. That was Justinian's

after all, Justinian's entire codification had to serve legal education -

or else

earliest known text on this, CTh order. 5.17.1, to be followed by others in chronological however, that they were originally only issued or interpreted as general rules for legal use and on selected and collected in the codifications or other collections. Consequently one has to be within which these texts are transmitted and 'deduct' the potential layers and changes Code the debate is still on-going, whether process(es). With Theodosius' or not; see Sirks, op. cit. (n. 4), Nr. 44 ff. for a survey. the texts preserved

include obsolete

122

A.

J. B.

SIRKS

I shall followthis approach in theensuingresearchintotheagricolaecensitivel coloni, on theseand connectingthese whereverappro partingfromthe textsin thecentraltitle priatewith textsfrom otherparts of theCode, and by this I hope to presenta concise I shall restrict underJustinian. myselfto texts survey of thecolonate as a legal institution word 'colonate' for what in the sources is relevantfor thepresentargument,8 using the with theadscripticii) and thecondicioof called thecondicioadscripticia(theadscripticiate meaning of word colonus in itsoriginal thecoloni liberi (the 'free'colonate). I use the Itwill depend on thecontext was the or farmer. whether this farmer cultor,cultivator or a farmlabourer;also itdepends on thecontext owner of his plot of land, a tenant whetherhewas subjectedto a particular(public law) condicioor not. summarize the discussion, without intending to Regarding thecolonate, I briefly present In thecourse of the enterhere extensively into thedebate on thecolonate as institution. we see constitutions issued with thepurpose of tying workers fourth century agricultural of land and poll tax. Some authors (coloni) to the land inorder to facilitatethe raising of the taxationsystem under Diocletian. assume thishappened in thecourseof a reform a category of coloni The statusof these workers graduallydeclined and underJustinian must have been considerable, (theadscripticii)is even compared to slaves.Although it evidenceof free about theiractual number.There is certainly there is no information labouralso in thistime. The discussionon thecoloni has been dominatedby twoviews. as thesuccessorto One sees thecolonate, within thecontextof agricultural exploitation, and as the Late Republic (whencolonuswas used fora tenant) precursor thetenancy of the Marxist's theory of a transi of thetiedserf of the Middle Ages. In thisview itproves the tionfrom slave societyto feudalsociety.In theotherview thecolonate isbut one illustra Roman Empireby itssocial petrification and bureaucracy. Some tionof thedeclineof the of thegreatdomains as semi authorsalso discern in thecontextof thistheemergence public institutions.9
II THE PLACE OF THE TITLES ON THE COLONI WITHIN BOOK II

Code (CJ), The titles which in on thesecoloni are placed in the iith book of Justinian's on theorganization main certainservices of thethree of public interest (CJ ii.z-I8), titles of towns(CJ II.29-47), on theorganization citiesof the Empire (CJ ii.i9-z8), and titles CJ II.48-53 are about agricolaecensitivel coloni, thecapitatio,coloni censitiand coloni
I have dealt with the colonate in, particularly, A. J. B. Sirks, 'Reconsidering the Roman colonate', ZSS-rA no colonatus: the origo of the Late Roman Empire', JRS (1993), 331-69, but it appears from C. Grey, 'Contextualizing 97 (2007), 155-75, a stimulating contribution, that it is necessary to deal with the question in a methodologically research will have to be done on, for example, different way; where necessary I shall re-examine my position. More There were also coloni on the imperial estates, but from the rules on the autopragia and the estate management. them (in CJ 11.63-4 anQl 68-9) itwould appear that their situation was basically identical. 9 in Cambridge Ancient History A summary of the status quaestionis with B. Ward-Perkins (CAH), Vol. XIV in ibid., 465, refers to them as formerly free tenants who did not own the land they (2000), 343-4. P. Heather worked, but also points out the rise of the landlord class. For more detail on the colonate, also referring to the existence of free labour, see C. Whittaker and P. Garnsey inCambridge Ancient History (CAH), Vol. XIII in A. H. M. colonate', (1998), 287-94. Some of the many works on this question: A. H. M. Jones, 'The Roman in Storia di 'II lavoro nelle campagne', Jones, The Roman Economy, ed. P. A. Brunt (1974), 293-307; A. Marcone, in Storia d?lia societ? italiana, Vol. 4, 'II colonato Roma, Vol. 3.1 (1993), 823-43; G. Giliberti, tardo-antico', e destrutturazione nella tarda antichit? (1998), 24-30; A. Demandt, Die Sp?tantike Restaurazione (2007), 398-401. Against the concept as such and other aspects sound arguments have been raised since 1981 by Carri?: see J.-M. del Basso Impero": la resistenza del mito', in E. Lo Cascio Carri?, '"Colonato (ed.), Terre, proprietari e contadini continued 8

general deals with public

law. Placed

after titles on groups of persons obliged

to perform

delVimpero romano (1997), 75-150 (resuming his thesis with responses to criticism); by Sirks, op. cit. (n. 8), 358?61; and other authors (see CAH as cited above); further scepticism with P. Sarris, Economy and Society in the Age of Justinian (2006), 150-4, and n. 85 regarding the idea of the great domains as semi-public institutions. See for further in Lo Cascio, op. cit. literature Sirks, op. cit. (n. 8), 331 n. 1 and the Bibliography

THE

COLONATE

IN

JUSTINIAN'S

REIGN

I23

of particular provinces. After thatfollowCJ II.54-57 on rusticani and their villages, CJ ii.58-9 on correction of thetaxationrateand distribution of desertedlands (peraequatio, impositio agrorum desertorum), CJ ii.6o-i on border-and pasture-lands, CJ ii.6z-5 on patrimonial and other lands,Cf ii.66 on landsof theresprivata,Cf ii.67-8 on landsof In theselastfourtitles theresdominica,and Cf ii.69 on landsof thetreasury. thereisalso mentionof coloni on theselands.'0 Book ii finishes with some titles on theleaseof public on particularsubjects(CJ II.75-8). and imperiallands (CJ II.70-4) and some titles This surveyalreadymakes clear that,notwithstanding that there may have been a connection with taxes,fortheJustinianic compilersthe main feature of thecolonatewas servicesto be rendered, within thecontextof agricultural exploitation.If thecompilers main characteristic, a connection had considered with taxesas the would presumably they have placed thosetitlesinBook iO,where inCf IO.I-30 thetaxesare dealtwith. the rubricimpliesthat thereare normal agricolae Regarding theagricolae censiti as well and that we are dealingherewith a particular called (farmers) groupof farmers, - thesequenceof thetitlesis systematic. It starts who have been censused with a coloni, on theabolitionof the general title(CJ II.48), thenCJ II.49 with a singleconstitution capitatiohumana fortheurbanplebs in the East, Cf II.50 with thesubjectof litigation by coloni censitiagainst their masters (thisis theconsequenceof therulesas collected inCf on thecoloni inPalestine, II.48), and finally threetitles eachwith a singleconstitution, Thrace, and Illyricum(whichconcern the agricolae censitiof certainprovinces).The on coloni is a sign thatthistax placement of CJ II.49 in the middle of theconstitutions was relevantto thegroup of agricolae censiti,a connection confirmed byCJ II.48.I0 in which the ratesof thecapitatio humana are rendered. Though the rubricof Cf II.48 besides speaks of agricolae censitivel coloni,we meet in its textsvarious designations colonus: coloni originales (c. 4), rustici(c. 5), adscripticii(coloni) (c. 6, 2I, 22, 23, 24), condicionis(c. ZZ, z3, 24), censibus originarii(c. 7, ii, i6), tributarii (c. iz), adscripticiae Greek ?vaur'Ypwpo; adscripti(c. i8 andCJ ii.5o.z), colonariaecondicionis(c. z3), and the who were considered'free' and different Therewere also coloni censiti fromthe (c. ig).11 coloni adscripticii(c. I9). We shall use, in accordancewith the title'srubric,the term colonus for bothkindsof coloni censiti. are a number thesetitles, there ofothertextsinthe Code orNovels in which Apart from arementioned.'2In somecases thisconcerns coloni in colonior adscripticii (?vcutypawpoi) inCf 4.44.65 and otherplaces and ifand in tenants. thesenseof independent They figure In some or liberi were tenants, thosetexts also apply to these.13 as faras coloniadscripticii
10 I leave these coloni out here: it is likely that their position did not substantially differ from the two (other) kinds of coloni. 11 But this text (see below, n. 40) is a restitution from the Basilica and the word may be a later hellenism, or the antecessor used in his Greek summary or translation of the current Greek equivalent, in both of which cases the original text most likely had adscripticius. 12 in CJ 1.2.14, 1.3.36, 1.4.24, 1.12.6.9, 2.7.22, 2.7.24, 3.26.7, Apart from CJ 11.48-69: colonus and colonarius 3.26.8, 3.26.11, 3.38.11, 4.10.3, 4.10.11, 4.21.19, 4.26.13, 4.65.5, 4.65.9, 4.65.27, 4.65.35, 5.34.13, 5.62.8, 6.4.2, 7.30.1, 7.32.5, 7.32.12, 7.38.1, 7.38.2, 8.5.1, 8.15.8, 8.51.3, 9.24.1.5, 9.2.7-4? 9-2-7-5? 9-49-7-1? 10.7.1.3, 11.8.7, 11.26.1, 11.75.1, 2.4.43, 3.38.11, 7.24.1.1, 8.51.1, 12.10.2,12.19.12,12.21.8,12.33.3; adscripticius in CJ 1.3.20,1.3.36,1.4.24,1.12.6.9, inNov. 12-3.35, 162.2 pr, 3, 4, 5, 162.2.1; ?vaTCOypaipo? inNov. 22.17 Pr? 54 8.51.3, 12.19.12, 12.54.3; Kotaovo? pr-i, 123.4, 12-3.17.1, 12.3.35, 12-8.14.3, 156.1, 162.2-3. 13 4.21.19 pr.; 4.65.5, 9, 27, 35 pr (here the conductores have taken it upon themselves to find farmers CJ 4.10.3,11; to till the rented lands as subtenants, which may therefore concern free tenants, but perhaps also coloni adscripticii or liberi); CJ 5.62.8 (where coloni is explained by the interpolation 'id est conductores'; otherwise confusion might be created with CJ 5.34.13 where the coloni patrimoniales enjoy immunity from the guardianship); CJ 7.30.1 (where colono vel may be interpolated); CJ 9.24.1.5 (assistance with counterfeiting); CJ 11.58.3 (the owner resists a fiscal re-evaluation of his lands by retracting his procurator or by dismissing his colonus: itwould not help to dismiss an so itmust concern a tenant; further, if it concerned tied farmers, the text would rather have read adscripticius, since meadows colonos); CJ 11.61.3.1 (which concerns tenants (conductores) of provincial and res privata meadows, are not cultivated); coloni or (which concerns a reassignment of lands, deserted by previous CJ 11.62.5 emphyteuticarii).

124

A.

J. B.

SIRKS

cases theadscripticiate which shows how texts has been interpolated, were adapted for use.14 we meet tapotKot, contemporary Furthermore meaning accordingtoZepos coloni, we are dealinghere with inquilini tiedto land;but it ispossible that or casarii (who could also be tiedto an estate).15
III THE TEXTS ON THE COLONATE

Within C II.48 fourtexts date fromtheyearsA.D. 529-534, all issuedby Justinian: Cj II.48.20-4. Consequently we may assume thatthey are coherent and accuratelyrepresent thesituation of A.D. 534 incontemporary idiom.They deal with thefollowing questions: when can theowner of an estate (dominus terrae)claim somebodyas his colonus or And connectedto this: when is somebodyconsidereda colonus or adscript adscripticius? icius? What can an estateowner claim from his colonus or adscripticius? Does thestatus of colonusor adscripticius And there are collateralproblemsimagin pass on tooffspring? able. Ifcoloni are attachedto an estate, what iftheestate isdivided?Ifsomebody has two one to theother? coloni from estates, may he transfer What ifa colonus runsaway and enterstheadscripticiate with anotherestateowner,orworks as an independent or tenant Or runsto a town, or seeks immunity in theimperial services or the day-labourer? Church? What about thepayment of thetax incase of flight? What ifa colonusor colona (wehave to think of the of a colonus) marries thecolona or colonusof another? Can coloni daughter estateowners? litigate against their I take thesefourtexts as the pointof departure, butwill refer toother texts of the Code and Justinian's Novels where appropriate in order to consolidate theexamination,and texts on coloni.The other textsinCJ II.48 relating then dealwith theremaining tocoloni were all issuedbefore Justinian. Theoreticallyit ispossible thattheir original meaning and differed fromtheir setting meaning and settingin Justinian'stimes. Because Justinian obsoleteand allowed themtochangeand interpolate orderedhis compilersto leaveout the we may, as faras their thetexts, application inA.D. 534 iSconcerned,leaveanypotentially different original meaning aside and interpret themin accordancewith what we know from Cf II.48.20-4 and otherJustinianic texts.
IV THE CONDICIO ADSCRIPTICIA AND THE ORIGO

An estate owner could claim somebodywas his colonus adscripticius,scil. with the intention thatthis would render him services(seebelow,Section ix), if person he provided proofof thelatter's status, This could be, forexample namelyby at least twodocuments. a conductioor conductionaleinstrumentum and a copyof thepublici (a labourcontract)16

e.g., CJ 7.30.1. 120. See P. J. Zepos, 'Servi e paroeci nel diritto bizantino e postbizantino', RAL 35 CJ 1.2.24, Nov. 7 pr., Nov. (1980), 421, 424; Zepos (427) assumes the rcapoiKOl absorbed later on the coloni and adscripticii. But of such a third category of farmers nothing is known in any other way, and also inquilini had an obligation to remain on the land. as 'adscripticius avant la CJ 9.49.7.1 has casarii next to coloni. I therefore doubt Sards' interpretation of Apphous 15

14

lettre' (Sarris, op. cit. (n. 9), 151-2). 16 Conductio may refer to a lease, inwhich case the landlord is the locator and the tenant is the conductor, or to a contract inwhich somebody (the conductor) hires another person (the locator). Although the firstmeaning ismore ? in view of the duties of the adscripticii, namely to cultivate fields? it common, cf., e.g., CJ 4.65.9, in this case is not the right one. If they were tenants their duties would have been circumscribed differently, e.g. to do what tenants have to do; and CJ n.48.19 speaks of uio"0a>xoi, hirelings (see n. 40). Until now we possess only one tenancy contract by an adscripticius (a.d. 505)), but he probably was already adscripticius by birth (P.Oxy. LXVII.4615 through his origo.

THE

COLONATE

IN

JUSTINIAN

S REIGN

I25

censusadscriptio,theenrolmentinto thepublic tax register of the land,17 evidently the or his father or further enrolment of himself ascendant,or a documenttowhich a later, voluntarily-made acknowledgement of his statusby thecolonuswas added (as in thecase of an adscripticius by birth); justa contract,forexample, was not enough."8 The census will have been theestateowner's census,scil. thatof theestate towhich theadscripticius was said to be attached,cf.CfJJ.7*7.19 How preciselytheadscripticius was registered we do not know; perhaps itwas done aftertheexampleof theslaves on an estate,under a since thecoloni arementionedafterthesein theexpositionof thecensus special titulus, fromtheestate iscomparedto theflight declaration(Dig. 50.I4.4.5 and 8) and their flight of slaves (CJ II.48.23 pr.; see also below,Sectionx).20 These requirements shouldprevent free persons from beinggroundlessly drawn into theadscripticiate (CJ II.48.22 pr.-z). If was reached, it could not be annulled (CJ during such a claim procedure a settlement to a contractimpliesthatbyJustinian's timethestatusof adscript 2.4.43). The reference iciuswas not derived fromone's fatherin everycase, but that it could stillhave been ex novo. It also impliesthattheadscripticiate was not the logicalconsequenceof entered a mere labouror othercontract. The originof all cases of theadscripticiate21 (and thus oftenof the 'free' have been a separate colonate, see below,Sectionv) must consequently whose conditionscould not be changed lateron (CJ II.48.z3.3). There isone agreement to suchan agreement with reference papyrus (itat leastsuggestsit; see n. I7). could be claimedaswell. Ifthey Childrenof coloni scil.adscripticii raisedtheexception or forty of limitation of thirty years against the claim, itwould be of no avail (CJ of prescription II.48.zZ.3-5). This general limitation applied both inprivateand public law (e.g.,CJ 7.39.4I), with theexceptioninpublic law regarding thesummons of people As forthechildren for birthstatus.22 publicdutieson basis of their of an adscripticius, the basis of thesummons was not a contract, but their origo (on thebasis of their birth;see below). Theymighthave been away fromtheestate fora long timeand could have pur suedoccupationsother thanfarming, as CJ II.48.22.3 says.Their obligations would have been thoseconnectedto thecondicioadscripticia(or,as it is said in thepapyri,the-n5X'9 seeSection Ix). We see indeedin thepapyriadscripticii with variousoccupa &vacutypawpi; when theyare named ?vaioypcupoiycopyo, it tions; while merelydenotes theirstatus, mean theyare farmers.23 Likewisewe see that theyown 7rop670;does not necessarily assets which they pledge.24

17 Of the land: since the references are always to being tied to a piece of land, itmust have been the registration the land, which had to be done in the town in whose territory the land lay. It therefore cannot have been the (6th century), Pieous asks to be accepted again, with registration of an individual. In P.Oxy. XXVII.2479 as the census had under children. Such an enrolment would not have had the same personal law consequences

of tax his the

Republic. 18 III.8 (4th century) such a document. The reading of the text is, Carri?, op. cit. (n. 9), 946 sees in P.Ross.Georg. III.8', JJP however, very difficult, see P. van Minnen, 'Patronage in fourth-century Egypt. A note on P.Ross.Georg. 27 (1997), 67?73, and conclusions are actually not possible. 19 This text deals with fledmetallarii, state miners, whose condicio was similar to that of the adscripticii and who were harboured by private persons who enrolled them on their census. 20 See Sirks, op. cit. (n. 8), 348-51, 366. 21 we assume it (colonatus perpetuus) on able-bodied beggars of Constantinople; CJ 11.26.1 imposes the colonate since that condicio was also imposed in other cases, but this should be further concerned the 'free' colonate duties. The other e.g., in CJ 7.39.5 for curi?is and their children when summoned for municipal exceptions are unimportant here. 23 ismoved from the It is therefore correct that in the latest editions of the P.Oxy. the designation ?va7toypa(po? general list of words; it should be under statuses or liturgies. 24 ? in legal texts of their peculium and the prohibition on selling without consent for Apart from the mention ? a striking example of this is P.Oxy. LXX.4794, where Ieremias, former headman and which see Section xn examined. 22 Rejected,

?va7toypa(po? ye pyo?, gives surety and pledges all his belongings present and for the future, in particular and in general, by way of security and by right of mortgage. Another text inwhich adscripticii pledge for the contract of tax collection: P.Oxy. LXII.4350 (a.D. 576). For this see below, Section xn.

iz6

A.

J. B.

SIRKS

This birthstatus,theorigo,derivedfromthetimethatthe Mediterraneanworld was a of independent and autonomouscities, patchwork which could summontheir citizensfor public tasks. According to rulesof international private law (iusgentium),itdetermined one's public and private law system.Itwas one's home town (patria)and consequently retainedin theRoman Empire, not so much forcitizenshipas such (almosteverybody A.D. ziz also a Roman citizen), but forthepublic tasksnow calledmunera and beingafter honores. The origopassed on in thesameway as ithad done previously:in legitimate mar riages fromfather onto his children(CJ 8.47.7), otherwisefrom mother onto her child ren.25 Likewise freedmen took their manumitter's origo (cf. CJ 7.I4.I).26 A wife kepther own origo or else theapplicationof theSC Claudianum (see below) would have been more complicated. The situationis,however, The origowas established on thebasis of citizen one's descent. of a towncould be summonedforall publicobli Theoretically every with theorigo,but inpractice this gationsconnected was not thecase. For example, with who had a decurionas father decurionsonly those or sometimes were eligible, grandfather amountofwealth inorder toperform and inaddition they had todisposeof a certain their duties.Only ifnot enough new candidateswere found in this way were hominesnovi chosen (whoseexistimatio, had to be good also). This eligibility reputation, might further as regards, entailrestrictions forexample,otherfunctions. It is thisensemble of origo, lia dutiesand restrictions which definedthecondiciocurialis.28 It bility, eligibility, might seem as if thecondiciowas 'inherited' but that is not thecase. Birthwas thecriterionfor the origo and theorigowas one of thecriteriaforthe imposition of thecondicio (in thiscase criterion since itdefinedthe town whichmight by thehome town). Itwas an important to thesystem, claim and as such itwas fundamental but itdid not have to be theonly criterion. as such,but also to thestatus in itstotality. Condicio may point to the liability It is the samewith thecondicio adscripticia. Here also thedescentdefinedtheorigo, which again is theessentialcriterion forimposition of thecondicio.Furtherrequirements were apparently not set,but duties and restrictions were present.29 Yet was theorigo a town?In view of itbeing thedominus terrae theadscripticius, and inview of summoning aswe saw above, thattheadscripticiate was tobe provedby theadscriptiocensus thefact, publici, thecolonus' origomust have been theestate inquestion,namely the terra of the dominus terrae. This means thatin thesecases the townas origo had been substituted by
unnecessary.27

states that the colonus was not tied to the estate owner, but to the taxation, and that the origo was a fiscal category. But he does not specify what such a tie to the taxation meant. 26 This explains the application of the colonate in Illyricum (CJ 53.1.3) to freedmen of coloni. 27 See below for the SC. Further CJ 10.32.36, where the request to be transferred to the maternal origo is rejected, which implies that the mother of the applicant kept it; CJ 10.32.11. In Sirks, op. cit. (n. 8), 367 itwas mistakenly assumed that a wife took the origo of her husband: she took his domicile. 28 See on this in general Sirks, op. cit. (n. 25, 1993). 29 in CJ 11.48.22.3 were actual objections the objections mentioned that for lack of experience in Perhaps agricultural labour the condicio should not be imposed on the children, but this argument did not play a role.

the case that the owner of the slave agrees not to enslave the woman, but that in that case, due to the she will bear slaves. Hadrian corrected this: in such a case she would bear free children. senatusconsultum, 'Du r?le Unfortunately, this has not been taken into account by Grey, op. cit. (n. 8), 156 and 170-1. Ch. Saumagne, de Y origo et du "census" dans la formation du colonat', Byzantion 12 (1937), 506 sees the land as the dominus and the colonus as its servus. Legally this is nonsense, as is Saumagne's assertion that the land is the subject of a right to the colonus. D. Vera, 'Schiavit? rurale e colonato nell'Italia 317 imp?riale', Scienze dell Antichit? 6-7 (1992?1993),

25 M. Kaser, Das r?mische Privatrecht, I. Teil (1975), 279. After A.D. 212 Rome became communis patria, which implies that itwas the patria next to one's own patria. But a change in domicile could imply additional eligibility for mu?era of one's residence, cf. Dig. 50.1.20. See on this further A. J. B. Sirks, 'Did the Late Roman government try to tie people to their status or profession?', Tyche 8 (1993), 165?6, where the significance of the origo for public duties is set out, next to other criteria important in this respect, and idem, op. cit. (n. 8), 347, where the link ismade as well. Also A. J. B. Sirks, 'Der Zweck des Senatus Consultum Claudianum von 52 n.Chr.', ZSS Rom. Abt. 122 effected a deviation from the international private law (2005), 138, where it is set out how this senatusconsultum rule on status as regarded unions between free women and slaves. The inequity Gai. 1.84 refers to in this context is

THE

COLONATE

IN

JUSTINIAN

S REIGN

I27

an estate (terra, possessio).Thus theestateownercould summon a colonus toperform his duties in thesameway as a towncould summonitscitizens and curials toperform public duties (e.g., and othercoloni CJ IO.3z.z, 5);what duties theorigoof thecoloniadscripticii implied we shalldiscussbelow inSection ix. Moreover, theuse of thecensusas origo made itpossible forthe personof theestateowner tochange,e.g. by sale, with thecolonusbeing obliged to thenew owner. If theadscripticiate was entered itsformalization byagreement, by the adscriptioreleaseditfrom theconstraints of thelawofobligations whichwould not recognize a change ina contractual Thus itcould be said thatthecoloniwere relationship. alienable together with theestate (CJ II.48.2 pr., zi.i) and in thissense theyresembled who were attachedas instrumentum slaves to a plot of land. A townas origomust be distinguished froma townas centrepoint in the taxationsys tem. On thebasis of thecensusdeclarationstheexpensesof thestate were proportionally apportionedtoprovinces, again to towns, and, in theend, to theindividuallandholdings, and subsequently leviedaccordingly.30 If somebody owned land in more thanone town's hewould have to declare each estate in the town territory, where it lay (Dig. 50.I5.4.2), and pay thetax likewisein more thanone place; his origohad nothingtodo with this.31 It is therefore ina censusdeclarationof an estate wrong to assume thattheenrolment made a colonus originariusresponsible forthe taxes imposedon this land: theownerwas res ponsibleand did not becomeoriginarius by this.32 Therewas nevertheless a difference: an estatedid not equal a town, notwithstanding the use of origo forboth.Normallywe would expect thatin thecase of amarriagebetweena man, subjected to the adscripticiate, and a woman, not subjected to thisor another condicio but citizenof a town, thechildren would follow theorigo of the father.(Such unsubjected personsare usuallycalled ingenui, of obligations;likewisein the law i.e., free were not subjectedtoobligationsaswere liberti of persons ingenui their manu regarding missors.) But we know that theSC Claudianum was applied to theseunionswhen the husbandwas an adscripticius and the woman was not.Originally thesenatusconsultum made thechildren bornout of a union betweena slave and a free woman slave instead of their being freeborn. theabove-mentioned By thisit reversed iusgentiumruleon thecivil status. Here theapplication made children of such a marriage follow thestatusof their father instead of thatof their motherand they became adscripticii resp.adscripticiae(the latter have been attested).3This means thattheoriginesof father andmotherwere not

30 See on this J.-M. Carri?, 'Diocl?tien et la fiscalit?', Antiquit? Tardive 2 (1994), 33-64. 31 See Giliberti, op. cit. (n. 8), 31?2; Giliberti is incomplete when he states in his otherwise instructive survey that e Diocletian's reforms 'richiedava per tutti i contribuenti l'individuazione di una residenza fiscale obbligatoria definitiva (origo). Ogni individuo sottoposto a imposizione era registrato (adscriptus) nell'unit? amministrativa cui lo legava la sua origo: citt?, villaggio, latifondo'. CJ 11.48.4.1 (= CTh n.i.14.1) proves precisely for the adscripticii that one could have one's origo inA, but have to pay one's land tax in B. Grey has likewise not seen this point (see n. 32). Further CTh n.3.5: if somebody acquires part of a property, he has at once to register his name in the is not restricted to one's own origo. If the taxes on the paginae censuales, seil, where the land lay. The acquisition sometimes widely spread possessions of a magnate were all levied in his origo, itwould have meant an unnecessary of the system of distributing the taxable sum over the Empire and implied a complicated and complication unattested system of administration. 32 Grey, op. cit. (n. 8), 171-2 states this, then mentions slaves who are originarii (but slaves never paid taxes or then mentions registered coloni and finally connects these with the mu?era on the land as performed muneral), taxpayers. Yet his argument is inconsistent. The owner of an estate did not become originarius of the estate by his census declaration. A colonus, as any other person, was, with or without an estate, liable for the poll tax. But he

was

not, unless he was the landowner himself, liable for, e.g., the munus viae sternendae as his estate owner could be if the land was next to a road, or for the munus angariae. The obligations he was liable for were, on the other hand, not mu?era of the estate as such: no owner of an estate was obliged to cultivate the land (but he would still have to pay the land tax). The passage, on which Grey builds his view of origo, does not well support his conclusions. 33 Adscripticiae ?varcoypaipo? in P.Wash.Univ. yec?pyo? (a.D. 593), 4801 (a.D. 617). They 1.25 (a.D. 530), P.Oxy. LXX.4797 (sic), which would imply this was a status and not an occupation. call themselves

iz8

A.

J. B.

SIRKS

was not equal to thatof a consideredequal, i.e., theorigo of an adscripticius, theestate, town because the application of the senatusconsultum without it the ius implied that gentium would have applied: thechildren would not have been subjectedto thestatusof theirfather. That would mean a loss fortheestateowner, and that must therefore have been thereason thattheSC Claudianumwas applied analogously.34 Itwas forthissame reason thatJustinian, when he abolished theSC Claudianum inA.D. 53I-534 (CJ7.24.I; Nov. 22.I7), at once introduced the rule thattheestateowner of theadscripticius could a 'free' woman. Ifhe did nothing, forcibly takehis adscripticius away from he had himself to blame for the loss ofmanpower (CJ II.48.24).35 In A.D. 539 Justinian changed this Now children arrangement. bornout of amarriagebetweenan adscripticius and a woman freefromthecolonatewould be 'free'coloni (see below, Sectionv), so that they would have to stayon theestateof theirfather. Otherwise they would have been freeto leave. Yet, ifsuch a son,now a 'free'colonus,acquired a piece of land largeenough forhim to cultivateand not have timetowork elsewhere, he could go, live there and have his origo there(Nov. i6z.2).36For an adscripticiasuch a measurewas, of course,unnecessary. But of different thenatureof theorigo posed no problemwith marriages between coloni37 estates:they were so to speakon thesame leveland here the paternalorigo and in this way thiscondiciowas decisivefortheoffspring (CJ II.48.I3 pr.).38 This application of theorigomeant that,as stated above, thedesignationcolonus no longer meant thatthis was a farmer andwe do indeed (adscripticius) necessarily person - steward, find a variety of occupations in the scribe, guard in the legal texts;functions imperial service;in the Church.39 His duties,however, could involve work on an estate,but it isnot impossiblethatan adscripticius with funds could have contracted somebodyelse toperform this dutyforhim (cf. CJ I.3.I6). The applicationof thesenatusconsultum Claudianummeant no change in thepersonal law statusof theadscripticii, but remarkably measure takenafteritsabolition enough the inA.D. 53I/534did. It implied a potential marriage restriction, whichwas remediedinA.D. Novel (Nov. i6z.z). That Novel restored 539 byJustinian's thepersonal law statusof the adscripticii.

Africa the same rule as had been introduced in Illyricum by Justinian, namely that children of an adscripticius and an ingenua would be free, but be subjected to the 'free' colonate on the estate of their father, hence be coloni liberi 13 of A.D. 582). Apparently the possibility of thwarting a union (Nov.Just. 6, confirmed later by Tiberius: Nov.Tib. itwas considered an inappropriate curtailing of the personal status. may not have been effective or, more probable, The abolition did not profit children born previously (Nov.Just. 54.1 of A.D. 537). 36 See M. Bianchini, 'Sul regime delle unione fra libere e adscripticii nella legislazione giustinianea', Studi in onore di Cesare Sanfilippo, Vol. V (1984), 61-107. Bianchini does not enter deeply into the question how the adscripticiate was precisely imposed and does not make the connection with the origo. Further W. Schmitz, 'Appendix 1 der Wende der Politik Justinians gegen?ber adscripticii und coloni?', Historia justinianischen Novellen-Eine 35 (1986), 381-6. 37 And inquilini, by whom probably dwellers on estates are meant, who had to perform services as well: CJ 11.48.13 states that substantially there is no difference between coloni and inquilini. 38 There was no coniugium inaequale 22.1.8) or coniugium non aequale (Nov.Theod. (CTh 14.7.1). 39 (steward, scribe: contracting the collection of a tax); P.Oxy. XVI.1979 P.Oxy. LXII.4350 (guard); P.Oxy. VI.996 (steward). For the legal texts: see below, Section x.

In previous publications I was still uncertain about the role of the origo in this context and the function of the of the SC Claudianum, and only recently have I come to this interpretation. I have to withdraw any I may have made. J. L. Murga, en el 'Una extra?a aplicaci?n del senadoconsulto Claudiano previous explanation de Teodosio', Studi in onore di Cesare Sanfilippo, Vol. I (1982), 417-42, merely accepts the application C?digo without explanation. It is furthermore an argument against the theoretical possibility that the estate owner exercised his rights as representative of the town as origo and as such in a public or semi-public function. 35 The rule of CJ 11.48.24.1 was introduced everywhere in the Empire; also inAfrica as soon as itwas reconquered in a.d. 534, but its introduction there caused a shortage of labourers. Therefore Justinus introduced in A.D. 570 for application

34

THE

COLONATE

IN

JUSTINIAN

S REIGN

I29

CHANGE

OF

STATUS

AND

THE

FREE

COLONATE

Although thereis no texton thesubject,release fromtheadscripticiate must have been possible by a unilateralact of theestate owner, because ifan adscripticius wanted to become a priestormonk, his estateownerhad to approveof this (CJ I.3.36.pr.).Hence But ifthisdid not happen,his thelatter's consent was required and apparently sufficient. statemight be mitigated, be it only in the long term. CJ II.48.z3.I says that if an adscripticius has performed his dutiesforthirty years,hewill be 'free'(the meaningof this we will discussbelow), but cannotmigrate fromtheestate.CJ II.48.23.1 attributes this statuschange to a constitution ofAnastasius (A.D.49I-518), reproducedinCf II.48.ig.40 colonate to thoseadscripticii who have servedthirty years,a CJ II.48.I9 grantsthe 'free' of prescription of thirty period which equals the limitation Zepos and later years.41 this textdifferently: Mirkovic and Carrie interpret theyassume that free,unindebted coloni who had spent thirty years on an estatewere made liberiadscripticii, viz. on account of CJ II.48.I9.42Mirkovic uses the word adscripticius forboth kinds of coloni coloni.The otherkindof coloni (for and she means thatsuchcoloni became 'free' whom we reserved came intoexistence the nameof adscripticii) either because they fellintorental arrears as originarii, were born on theestateand had inherited or because they, debt and In this view all farmers became coloni censiti of one or another kindand the dependence.43 But objectionscan be raisedagainst colonatewould have been a universal phenomenon. it. In Justinian's law thereisno questionof rental arrearsas a cause of thecolonate.And notmigrate after,forexample, twenty-five why did farmers years to escape thisconse and tenancy were not twocommuta quence? It isclear that underJustinian adscripticiate tiveconcepts.To describeadscripticiias 'registered tenants' is consequently wrong for on an estate,but ofmen Justinian's reign.44 CJ II.48.23.i does not speak of freetenants 'heldin thecondiciocoloniaria forthirty annos coloniaria years' ('homines qui per triginta detentisuntcondicione'). SinceAnastasius and Justinian wanted to give thema better fromtheadscripticiate to the 'free' condiciocoloniaria, itcould onlyconcerna transition colonate and so it isunderstoodin the text('sancimusliberoscolonorumesse quidem in condicionepraegravari'). perpetuum secundum praefatamlegemliberoset nulla deteriore The legalquestiondealtwith in thetextiswhethersuch a grantalso applied to children would not have raised the who had not (yet)completed thirty years of service: they

Whittaker and Garnsey, op. cit. (n. 8), 292. 43 Mirkovic, op. cit. (n. 42), 70. There is no reason for them to have inherited debts: they could have renounced the inheritance, with debts and all. 44 As in, e.g., Grey, op. cit. (n. 8), passim, particularly 173, where he cites me in n. 104 for this characterization of adscripticii as 'indebted, landless tenants'. Yet I expressed in Sirks, op. cit. (n. 8), 334-5, at n. 9, grave doubts about adscripticii as being in general tenants and pleaded for an interpretation as hirelings, which he apparently did not see. That Grey does not find debt mentioned in the sources is only possible ifhe interprets it as private law debts, is always connected with the taxes and taxes are a debt the but that is precisely the point. The adscripticiate adscripticii

40 A.] T?>v ye py v oi u?v evarc?yacpoi eiaiv Kai x? xo?xoov 7i?KOUA.ia CJ ii.48.19: [A?xoKpaxcop Avaaxaaio? uicjG?oxo? y?vovxai ?XeuGepoi |i?vovxe? jiex? x?v xo?? ?ecnroxai? ?vf|K6i, oi ?? xp?vco xf|? xpiaKovxaexia? Kai xfjv yf|v ye pye? Kai x? xea.o? rcap?xeiv. xo?xo ?? Kai x<p Tipayu?xcov auxcov Kai ouxoi ?? ?vayKa?ovxai ?eaTioxn Kai xo?? ye pyo?? A.i)aiX?A.??. 'Of the farmers some are ?varcoypacpoi (= adscripticii) and the peculia of these belong to the masters, others become by thirty years free labourers, although remaining with their assets; and these are forced to till the land and provide the tax. This profits the master as well as the farmers.' (= Dig. 55.1.18). 41 10 it is put under the term of thirty years and it is said that they have In the later Byzantine work Ai Vonai performed their obligations and are no (longer) evarcoypacpoi. 42 The Later Roman Colonate (1997), 65, 69?70. The same error in Zepos, op. cit. (n. 15), 422; M. Mirkovic,

had to pay and which their estate owners guaranteed or took over. As to being landless, the fact that the adscripticii were not prohibited from owning land does not necessarily imply that they regularly owned land. CJ 11.48.4.1 suggests that itwas an exception. Grey further, in thewake of this, states that there were not two different condiciones, referring in n. 105 to a critique by Carri?, op. cit. (n. 9), 113, which apparently is convincing to him; but see my comments in n. 45.

130

A.

J. B.

SIRKS

was to because a fugitive benefice make flight lessattractive, Perhaps thereasonforthis whereas now he liveunder thethreat of being recalledas adscripticius, would permanently (potestas)of could be sure thatafterthirty yearshewould quit thecontroland authority were away for of thisalso. Ifhis children would be free theestateowner,and his children in itself, froma limitation but iftheirfather became thirty years, they would not benefit would followhis new statusas well (CJ II.48.z3.I). We must be 'free'in this way they were enrolledor summonedat notmanywill have profited fromthisunless they realistic: was a long timeinantiquity. an earlyage.Thirtyyears Apart fromthissourceofwhat we shall call, in accordancewith the texts,the 'free' of 'free'coloni over thoseprovinces colonate,which must have caused a distribution were some provinces inwhich the 'free'colonate where theadscripticiate existed, there on adscripticii: Palestine, Thrace, and Illyricum (CJ ii.5i, 52 had been imposed generally in their context as the legal These threeconstitutions functioned Justinianic and 53).46 had fled, fromtheir harbourers and definedthecon these basis forrecalling coloni, ifthey Those who admit fugitives colonate foreach region. are, as seen inCJ ditionsof the 'free' is the fact that in Illyricum the rules for II.48, fined(CJ II.52.I.2, 53.1.z). Interesting (CJ II.53.J.3).47Since adscripticii children of 'free'coloni also applied to theirfreedmen own plots of land, they could own land,and since they were liable forthe taxeson their could become 'free' coloni, there could have been 'free' coloni who owned land. when inA.D. 535 Justinian forbade money-lendersto take the land from Consequently itcould concern independent but also thesecoloni, now farmers, theirfarmer-debtors,48 own land.49 'free', or adscripticii with their A third source of the 'free' colonate was afterA.D. 539 themarriages between whichwe discussedabove, inSection iv. and ingenuae, adscripticii for theclaimingof a 'free'colonus.This was Nothing is said about therequirements possible (CJ II.48.23.5). Presumablyithappened in thesameway aswith adscripticii.

CJ I I 48.23 .I .45

conditions. This point isnot seenby theaforementioned question ifithad notmeant better nor byCarrie,who uses CJ I1.48.19 to interpret theterm condicio coloniaria in authors,

I cannot enter here into a complete discussion of Carri?, op. cit. (n. 9), but his approach suffers greatly from not taking the entire legal dossier into account in a proper way. Thus he does not analyse CJ 11.48.22: not on p. 113, nor on p. 99 (on p. 119 he neglects the contractual part of the adscripticiate and only mentions the adscriptio); as I did. It becomes understandable then that he neither does he consider Dig. 50.15.4 for the census declaration reading of CTh reproaches me for speaking of the census as a 'lista' and accuses me of taking over a misunderstood too that he denies any contractual origin 11.1.26 by Goffart (p. 113), when I did not do this at all. Understandable of the adscripticiate and the difference with the 'free' colonate, adhering to the mentioned interpretation of CJ a text which neither he nor Mirkovic I consider absolutely 11.48.19, which incompatible with CJ 11.48.23.1, analyses. He merely mentions this text on p. 122: but instead of accepting the meaning of (condicio) coloniaria as adscripticia on the basis that the Latin text is in any case reliable, he attributes another meaning to coloniaria, on the basis of the less reliable Greek text of CJ 11.48.19 (which is a Greek summary of the Latin constitution), which saves his interpretation. Apart from that, this is a circular argument (the wrong interpretation of one text is proved rather questionable by applying it to another text); this seems to me a methodogically approach. The legal content

45

It would be different, probably, if they had possessed would have been part of their peculium. land with slaves 'free' coloni itwas different. They were not subjected to the peculium. Freedmen would attached to it. With take the origo of their manumissor. 48 Which they must have bought at auctions through straw men, or bought themselves at a low prices, since the lex commissoria prohibited creditors from becoming owners of the pledges in case of defaulting. 49 The Novels 32-34 (a.d. 535) forbid this and order the return of the seized lands in the provinces of Thrace, Illyricum and Mysia secunda.

of the text is not discussed at all. This is precisely the approach I think it necessary to avoid. I shall deal elsewhere more extensively with Carrie's article. 46 See n. 51. 47 freedmen had the origo of theirmanumissors (see above, Section iv) and on the basis of this rule theirs Normally But could adscripticii have slaves and set them free? Not if these would have been the estate of their manumissors.

THE

COLONATE

IN

JUSTINIAN

S REIGN

I3I

VI

SALE OF THE

ESTATE

AND

REALLOCATION

OF COLONI

Regarding the tie to theestate, it was forbidden, ifan estate was transferred, to separate thecoloni attachedto itand tiethemtoanother estate,even ifthebuyeragreed to this(Cj II.48.2). Likewise coloni could not be alienated separately fromtheir estate.An evident trick tocircumvent while trans this, namelyby selling only a verysmallpart of theestate was suppressed with itall coloni on theestate, thatincase of a by theprescription ferring over theparts (Cj divisionof theestate, thecoloni should be distributed proportionally consisted II.48.7 pr.-z). The trick presumably of registering at the momentof divisionall whichwas thensold and transferred thecoloni in thecensusof thesmallpart, with them estatenow devoid attached to it. If itnevertheless happened, theownerof theremaining of coloni could reclaimthem, while thenew possessor could not raise limitation of pre whethertheownerof twoestates, scription againsthim (CJ II.48.7.3). The questionarises somecoloni from one estate to the each ofwhich had coloni attachedto it,could transfer other if thishad a shortageof labour.This is indeedalso dealtwith in the texts.It is hands, theowner of theestate from allowed, but if the two estatespass intodifferent of the transferred which coloni were taken could claim the offspring coloni (CJ with theconceptbehindCJ II.48.2.I, which says that II.48.13.I). This tallies whoever on a parcelof land shouldkeep themthere, thinks thatcoloni are useful and it is inaccord ancewith theprinciple of theorigo.
VII TAXES

in thecensushad as consequence, or as condition,thattheestateowners The registration inperson or through had to acknowledge,eitherby themselves theirstewards(procura whichwill oftenhave been thecase withmagnates), the liability for the tax their tores; once havingbeen censiti, had topay (CJ II.48.4 pr.).According toDig. colonioriginales, taxeseven if would be liablefortheir he had notdone so.50 This 50.I4.4.8, theestateowner tax was thecapitatiohumana, as followsfrom(a) thetariff forthisinCJ II.48.I0, (b) the of theurbanplebs in the East and someprov ofCJ II.49.I with theexemption positioning on coloni, and (c)CJ II.52-3.51The papyri AsiaMinor in the middle of thetitles incesin n. 85). But ifsuch a also indicatethis(seebelow, Section IXat n. 65 and SectionXII after
50

without

counteracted by CJ 11.48.8 pr. 51 in A.D. 386 would release the coloni tributariae sortis of their The abolition of the capitatio humana in Thrace bonds. This must refer to the adscripticii, and itdemonstrates that the designation tributarii must also refer to these coloni. It confirms that itwas the guarantee of their estate owners for the capitatio humana which made them subjected to him and, parallel to the pledge of a creditor, which ends when the debt is paid or cancelled, their bond must have ended as well (I think the parallel is fullywarranted in view of the language of the constitution: it assumed

mean, vide CJ 11.48.4 pr., that he would be liable anyway and that he could not recover the tax from 'Dalla locazione al colonato', Annali Dipartimento di phrase must have been interpolated. P. Rosafio, antico 13 (1991), 264 thinks it concerns tenants and the tributum soli, Studi del Mondo classice del Mediterr?neo since it would be absurd to pay the capitatio for an unregistered colonus. Yet that is precisely the point: not registered means that the estate owner profited from them, but that he would avoid the tax. It is a situation also Itwould them. The

11.53.1, with reference to the tributarius nexus), and in Palestine, where the text underlines the faculty of the estate owner to recall them (CJ 11.51.1). Although the text suggests that this colonate was introduced for all coloni, the phrasing suggests that it concerned those adscripticii who were released: reference ismade to estate they once took upon them to till and 'nullus omnino colonorum suo iure velut vagus ac liber exsultet': it cannot be that every farmer rejoiced, but only those who were released. These three constitutions functioned in their Justinianic version as the legal basis for recalling these coloni. Those who admit fugitives are, as we saw before in CJ 11.48, fined (CJ 11.52.1.2, 53.1.2). Since adscripticii were liable for the taxes on their own plots of land, they could own land.

reserve that the abolition of the tax (a debt to the state) generally ends all cases of the adscripticiate, regardless of the individual agreements). In order to prevent a massive migration, these coloni (it does not say that are made into coloni liberi. They are retained originario iure, which shows that it concerned all coloni of Thrace) this designation comprises the two kinds of coloni (CJ 11.52.1). The same was done in Illyricum in A.D. 371 (CJ

132

A.

J. B.

SIRKS

inhis own name in thecensusand his ithad to be registered colonus owned landhimself, of thetaxon thatland (the tributum, estateownerwas not liable forthepayment due by theowner of land): thecolonus had to acknowledge thisdutyhimself(Cj II.48.4.I).52 Nov.Just. 128.14ofA.D. 545 repeatsthis, ifhis estateownerhas adding thatit isdifferent forthat obligedhimself(as surety) plot of landof his colonus.CJ II.48.4 pr.makes itclear was based on an accept thattheestateowner's liability forthepoll taxof his adscripticius ance of thison his part (recepta),bywhich he would become theonly liabledebtor as regarded thecollectors(but theadscripticius remainedthe personuponwhom thisliability was based). Conversely,seeCJ I.3.i6, theadscripticius and itsamount would have to ack nowledge again thecapitatio when releasedfromtheadscripticiate. The consequenceof was (a) thattheadscripticius did not have toworryabout statetaxcollectors, this but (b) now owed him thisamount and the that if theestateowner had paid, theadscripticius estate owner could recoverthis.53 This is confirmed by CJ II.48.zo.3a and by P.Oxy. XXVII.z478 and LXIII.4398, where the vcvanoypcpot promise to renderthe tax to their estateowner. as guarantorfor the taxesof his registered The fact that theestateowner functioned of fugitives coloni explains thedispositionofCj II.48.8 pr., thatanyharbourer who uses were his own coloniwithout paying thema loan, must pay the taxesdue themas ifthey was that whoever profits fromthemalso must carrythe by them:therationaleevidently estateownerhad nevertheless burden.The question,of course, is whether thefirst topay longertodo so. This makes sense.The poll taxwas leviedon persons: ifthey were gone, was no longer a subjectto levythetax from and bydefault there of theprimary obligation thatof thesurety would fallaway.54 It also followsfrom in thecensus, to theregistration CJ II.48.4 pr. thatcensitusrefers thusthata colonus censitusis thesame as a colonus originalis who has been enrolled. It would be thesameprocedureaswith public obligationssuch as the munera,where liable duties formally. As such the persons are summonedand thenhave to acknowledgetheir inJustinian's toboth categories word originalis(ororiginarius) of coloni may refer, times, are obligedon thegroundsof their which often will have been thecase) (in thatthey origo, aremeant because theothercategory but here theadscripticii of coloni had always topay thetaxes themselves. with regardto thecapitatioand iugatio Immunities, especiallygiven (whichcould also profit coloni),were unacceptable ifthey had beenmade without proof most likely, (i.e., through bribery; CJ II.48.9). Their tie to theestateensuredthecoloni at leastof one advantage, were not liablefortheunpaid fiscal namelythatthey debtsofother Nov.Just. I08.I4 of A.D. 545).Normally those with the people (CJ II.48.I5, repeatedin in the fiscalsystem. same origowould be subjected to thissecondaryliability, inherent They, however,had as their origo only theestate (CJ II.50.I). Their estateownerwas liable forthis estateand fortheotheradscripticii(thisarguesagainst theideaof theestate owner as public taxcollector;see below, Sectionxiii).
We the tax in case of flight. have to infer that as soon as the coloni had fled, he had no

52 Grey, op. cit. (n. 8), 172 n. 94 reads this text as if the coloni (for him: tenants) became responsible for the estate's taxes by their registration as originales. It is different: the registration as originales makes their estate owner liable ? as owners, not as for the taxes of the coloni. But if the coloni possessed land themselves, they were responsible tenants or land labourers ? for the taxes on that land regardless of their being originales. He cites me on p. 172 n. tax declaration on a particular origo made tenants 98 for his assertion that 'registration though a landowner's visible in the municipal or imperial tax rolls, so that they could be held responsible for the mu?era of that land' but

see n. 32. It does not seem to me that tenants would become liable in this way (Sirks, op. cit. (n. 25, 1993), 165). The text he cites in n. 98, CJ 11.48.11, does not say anything on this point. 53 See P.Oxy. XXVII.2479 (6th century) with Pieous, who laments that he could not pay his estate owner and that the owner's procuratores came and seized all his goods. 54 See CJ 11.48.8 pr.-i as interpreted below, Section x, for both statements.

THE

COLONATE

IN

JUSTINIAN

S REIGN

I33 OTHER TENANTS

VIII

EXPLOITATION

OF

ESTATE AND

LAND, FARM

COLONI-TENANTS, LABOURERS

If litigation were initiated about the identity of theestateowner, thecoloni of theestate to escape their duty to pay their dues, and so shouldnot be able to use this opportunity Justinian regulatedthis(CJ II.48.20 pr.-5). He did this not somuch out of concernover of thetaxes (publicaefunctiones). what was due to theestateowner as over thepayment iftherevenues(reditus) Regarding thefirst, were paid ingold, thecoloni had toprovide would claim theseand deposit thesumsat a surety fortherevenues or else thegovernor were paid inkind, thegovernor would sell themand suitabledepository.If therevenues sum forafterthetrial. likewise deposit theresulting two situations. As to the taxes,Justinian again distinguished Either thecoloni used to over a lumpsum (totasumma) to or it was custom thattheyturned pay thesethemselves, theestateowner, who thentook thesumdue fortaxesout of thisand paid it to the tax collector, while he kept the restas income (sui reditus).In the first case, even if the tax shouldhave been paid through theestateowner, theexisting practiceshouldcontinuetill theend of the law suit. In theothercase, however,ifthecoloni had alreadyappointeda for also had topay theowner surety what they owed to theestateowner, thesame surety was time) thesum (scil. when it due as tax,and thelatter had to turnthis over to thetax collector.If thecoloni did not give surety, theentirerevenues were setaside, or sold and set aside, and thegovernor would takeout the sum due as taxes and give theowner a receipt forthis(CJ II.48.23.3-4). What were thesetaxes? They are called publicae or tributariae functiones.55 Itwill in any case have concernedthecapitatiobumana, thepoll tax,of thecoloni, since it is this With adscripticiithis was regularly taxwhich is dealtwith inother texts. paid by their men estateownersbut it remainedtheirtax (see above, SectionVII). The first situation concern'free' coloni (of whom itis said thatthey tioned abovewould then pay taxesthem own person sincethey were registered for independently selves,i.e.,aswe seehere,in their mentionof theland tax.Since thelitigation was about the We do not see any special this). estateownerwas indeedtherightful estateowner, who questionofwhetherthe purported must also have been a question as towhether the land tax was liable for the land tax, it must have dealtwith thisas well. In thesecondabove would be paid. Thereforethetext mentionedsituationit was usual thatthecoloni, i.e.,adscripticii, turned over a lumpsum out of therevenues worked on, in of the land they money or inkind.Although thiscould aswe shall see (seebelow,SectionXII). Itcertainly concernshare-cropping, it isnot likely The form of exploitationthrough coloni (ofbothkinds)Justin was not a regulartenancy. ian dealtwith here must apparently have been one in which theestateownerput land in thechargeof people forthepurposeof cultivation and expected inprinciplea lumpsum inanycase of thetaxhe had tobe turned over tohim, in thecase of adscripticii comprising topay forthem.Itpresumably also includedtheamountof the land taxdue fortheplot, fortheland tax. Ifthis but this of coursedid notmake thefarmers responsible exploitation have likedtodeduct their mode concerned'free' coloni, these peoplewould certainly poll tax fromthe lumpsum. In thecase of adscripticiitheestateownerwould pay their poll as recompense tax. We do not knowwhetherhe also leftthecoloni a part of therevenues accounts inhis book-keeping. but it ispossible. Itmay also be thathe creditedtheir and agricultural workersnot tiedto an estate, Nothing is said inCJ II.48 about tenants inCJ 4.44.65 not tobe expected. but thatisnotwhat thetitleisabout and thus They figure

55

A definitive answer

to this question

awaits a new treatment of taxes and taxation

in Late Antiquity.

134

A.

J. B.

SIRKS

were tenants(weknow of one case,P.Oxy. Ifcoloniadscripticii or liberi and other texts.56 to them, becausewe do not see inCJ II.48.20 would also apply those texts LXVII.46I5), of tenancy thecontract regarding coloni a particularregime and 'free' thatforadscripticii forthe This constitution onlyprovideda specialarrangement (locatioconductio)existed. of therentin a specialcase. payment
IX OBLIGATIONS IN THE ADSCRIPTICIATE

He had to performtaskson theestate What were theobligationsof an adscripticius? work of a colonus, he had to (ruraliaobsequia, CJ I.3.i6), till the land,or performthe and he, and his remain on theestate (CJ II.48.I5), he disposedof assetscalled peculium57 stood under thepotestasof theestateowner (CJ II.48.-2.I); hewas prohibited children, againsthis estateowner (CJ II.50.2.4, see below fortheexceptionto this fromlitigating and a subordinate socialposition as betweenfreedman rule).Such a prohibitionindicates which his became a priestinhis own hamlet (vicus) (for Ifan adscripticius manumissor.58 see Section v), his estateowner estateowner had to releasehim fromtheadscripticiate, by thathis could requirethathe acknowledgedthe liabilityfor thecapitatio, releasing estateowner fromthis obligation,and thathewould have his estatedutiesperformelby of his own choice (CJ I.3.i6). A 'free'colonus had as his onlyobligation to a substitute remain on theestateand tilltheland (CJ II.48.I5 and 23.1),whichmust have impliedthat or as farmlabourer and thathe had topay his taxes, he had tocontinueto tillitas tenant as stated,thatan adscripticius did not have i.e. thecapitatiobumana.From thisitfollows, his estateownerdid so (seeabove,Section which again impliesthat topay thistaxdirectly, had theuse of his labour, thisperson vii). If thepersonwho harboured theadscripticius was liable forthesetaxes.59 theduties is further Some evidenceregarding providedby thepapyri. In a numberof Fikhman has ycwpyoi,i.e. coloni adscripticii. papyrimention ismade of evcur6ypctpoi A.D. 44I (or, inany case, These papyricover theperiod from thoroughly analysed them.60 codification and legisla the A.D. 6i6,61thusroughly period coveredbyJustinian's 469) till He observes that with lawsof Justinian's reign. makes comparisons tion.Fikhman indeed or but alsowith otherprofessions with farmers is found not only v?vanr6pU(pos (y0cop76q),
56 Some texts: CJ 4.10.3, 11; 4.21.19 pr.; 4.65.5, 9, 27, 35 pr (here the conductores have taken it upon them to find farmers to till the rented lands, which may concern free tenants, but perhaps also or just coloni adscripticii or liberi); otherwise confusion might be created CJ 5.62.8 (where coloni is explained by the interpolation 'id est conductores'; with CJ 5.34.13 where the coloni patrimoniales CJ 7.30.1 (where 'colono enjoy immunity from the guardianship); vel' may be interpolated); CJ 9.24.1.5 (the owner resists a fiscal (assistance with counterfeiting); CJ 11.58.3 re-evaluation of his lands by retracting his procurator or by dismissing his colonus: itwould not help to dismiss an must concern a tenant; further, if it concerned tied farmers, the text would more likely have read adscripticius, so it are since meadows colonos); CJ 11.61.3.1 (it concerns tenants (conductores) of provincial and rei privatae meadows, not cultivated); CJ 11.62.5 (it concerns a reassignment of lands, deserted by previous coloni or emphyteuticarii: because coloni adscripticii and liberi may have owned lands, but not from the crown, itmust refer to free tenants

(conductores)). 57 for the adscripticii, 'agros col?re, aliquid colonarii operis celebrare, agriculturam peragere' CJ 11.48.23.3: 'free' coloni do not (CJ 11.48.19); 'terrain col?re' for the 'free' coloni; adscripticii have peculium (CJ 11.48.23.5), both groups have a duty to remain on the land and consequently can be recalled (CJ 11.48.22 and 23). 58 Who was restricted in his capacity to sue his enfranchiser: CJ 7.6.3. See further on this below, Section xn. 59 'functiones sive terrenae sive animales'. CJ 11.48.23.5: ? 60 I. F. Fikhman, 'Coloni adscripticii yecopyo? in den Papyri', in I. F. Fikhman, Wirtschaft und 8va7i?ypa(poi im sp?tantiken ?gypten. Kleine Schriften (2006), 190?250. This article was originally published in 1984 Gesellschaft in this volume. The article is not only important because of Fikhman's inRussian, but is now accessible inGerman observations, but also for the Russian articles he refers to and discusses, which enable the reader who is not capable to form an opinion of Russian of reading Russian scholarship. Further Sarris, op. cit. (n. 8), 60-6; B. Palme, Form und Funktion der byzantinischen Gestellungsb?rgschaften (forthcoming). 61 Fikhman, op. cit. (n. 60), 194-5.

THE

COLONATE

IN

JUSTINIAN

S REIGN

I35

This tallies occupations likeguard and gardener.62 with our observationthata colonus might be tied through his origo and not be a farmer. A numberof papyri are about guarantees given fora fugitive vacouypa(poq, usually imprisoned; othersare receiptsfor The surety must guaranteethatthegvanoypa(po; received parts forirrigation machines.63 will answer for all thatpertains to his person or ratherhis status as &vwoypacuo;:
gig aLavTa Ta optivTU TO a 2t pO[ojontov acrOKplVOgF1VOV lTOttip TOV ?avoypapou 64 thatt e isremiasonfou as in that formulated is, that he will remain continu P.Oxy. extensively L.I35, TUXlV,, TO a6TO KT1fLa pts1 aclUTOV jTrjV gEO[[E]]icvcTaOalt ously on the estate (ll6ajdi KaTakCJJWaL

andwife,cattleand utensils,that K-cpovTunov) with his children hewill be answerable gig; and thathewill not go to forall that pertainsto hispersonor his statusas 9vcu6ypc~po;, on himself(it must have been anotherplace; oftenalso, thathe will pay the tax ((popop) fleeif he could terminate thecontract within thecapitatio;not a rent: why shoulda tenant thesurety a year?);65 will deliverhim;or else thesurety thatin thecase of a second flight must have been the or will do his duties.66 What pertains to his status will pay a fine toguard the services which his estateownercould requireinconnection with agriculture: land (&ypo(ptXaR), to be fruit groweron the estate orchard (cotcuapitr7, K1pLoUpo().67 were required, Perhaps seasonal jobs such as harvesting butwe do not know; thereare the irrigation otherpossibilities.In thepapyriconcerning machines it isusually said that In the same way a certain irrigation work is 'in the charge' of the adscripticius.68 a plot of land in their could have been charged with farming adscripticii byputting charge, someor all revenues(with while in return would have to render which thetax forthe they land could be paid also); they may also have been paid by credits in the estate book into 'thechargeof' an adscripticius who then had to exploit it Such transfers keeping.69 dues of coloni.There isa group in this would agree withCJ II.48.20,which dealswith the to theestateowner (seeabove, Sectionvii). Could itbe text over all revenues which turns in thecase of theirrigation to charge works thattheestateownerused theadscripticiate

62

Fikhman,

op.

cit. (n. 60),

194: P.Lond.

III.778

(aypoqr?taxC, guard),

P.Oxy.

XXVII.2478

(7t?)fiapixr|?, fruit

in Late Antiquity. Gold, Labour and Aristocratic Dominance (2001), 97 considers that in P.Oxy. Agrarian Change this refers to the rent. That does not make sense. Why should a tenant flee, when he could simply XXVII.2478 terminate the lease? See CJ 4.65.34, but the constitution says that this could be excluded in the contract, and that see J. Beaucamp, indeed happened, (ed.), Egypt and the 'Byzantine Egypt and imperial law', in R. S. Bagnall Byzantine World 300?700 (2007), 283. As in other texts, the orchard must have been put in his charge. Banaji is right ? see CJ 11.48.20 ? in PSI 1.62 as taxes and not as rent, but itmay well be that Petrus was in interpreting ?r|nocjia on the plot of land. In P.Oxy. tax so as to to to in of the volume levied able work hard be case, any produce, expected (a.d. 566, an account of an estate), there is explicit mention of the poll tax being received (xo? KOiv(o?) LV.3804 x v y?C?p(Y(X)v) urc?p ai)VT??,(eia?)-K8<p(aA/n?)) but this does not conflict with the above. The adscripticius may have been obliged to promise in a general way to perform all payments he had to make, whereas the steward had to make a precise account of what he collected. 66 himself, even Fikhman, op. cit. (n. 60), 202-6. In P.Oxy. XVI.1979 (a.d. 614) the surety, an ?vaTtoypaipo? guarantees that he himself will perform all that is required from the still imprisoned ?va7toypacpo?. Apparently the duties were in this case not so burdensome that he could not take somebody else's upon him. 67 Fikhman, op. cit. (n. 60), 194. See also Sarris, op. cit. (n. 9), 63-5 for the variety of labour and the flexibility of of labour. the Apion management 68 4698, 4799, 4801 4782, 4784, 4788, 4793, 4797 (adscripticia), LXX.4781, P.Oxy. LXVIII.4697, LXIX.4755, (adscripticia). 69 and Rural Society in Third-Century See now for this system, D. Rathbone, Economic Rationalism Archive and the Appianus Estate (1991); for the system in the 6th century, see P.Oxy. the Heroninos 4797, 4800; cf. also CJ 11.48.8.2 and Sirks, op. cit. (n. 8), 362-4. AD Egypt: LXX.4788,

inXVII Congr. intern, di 'Figures du "colonat" dans les papyrus d'Egypte: lexique, contextes', (1984), 945, observes this remarkable restriction. As I shall explain, this might not be so remarkable. Papirologia 64 Fikhman, op. cit. (n. 60), 195. 65 In P.Oxy. LXIII.4398: ?lern Kai xo? (popiKO? (?ou (p?pot) (see Section xn), in P.Oxy. I.133 (a loan) ix?x? Kai xo? (a deed of surety) x?v (popiKO? fijLicov(p?poi) (translated by the editor as 'tax payable by us'), in P.Oxy. XXVII.2478 ?7t?p a?xo? (p?pov (translated by the editor as 'the tax due on it', seil, the land). But this tax cannot concern the plot of land in his charge; itmust refer to the poll tax and be translated as 'with payment of my tax'. J. Banaji,

grower). 63 J.-M. Carri?,

136

A.

J. B.

SIRKS

with the Since capable individuals managementof thisimportant agricultural machinery? two would do heavy could managers arewomen, it isnot likelythatthey work, but they have been good managers.The advantage for theestateownerwould be thatthesesub ordinatescould neverquit theirjob andwould thusbemore underhis control,as would of theearlyPrincipate. be their assets, like the freedmen Cj ii.68.z orders the recallof to imperial coloniwho are fit manage (ratiocinia was gerere)or to farm. Apparentlythere concernto use thetalents of coloni. It was thehabitof the Fikhman Apiones to requiresureties with all kindsof contracts.70 inobservingthatthesurety is right reflects thedutiesof theadscripticius, but thequestion is:are thesethedutiesof an adscripticius who entereda contract and thecensus,or of one we have are for who was of that status on thegrounds of his origo?Many sureties We see thatin thesecases theenumerated who are inprisonbecause they fled. adscripticii duties reflect the legaldescription of thestatus: todo what pertainsto their which status, otheragricultural included work on thelandor toperform services(CJ II.48.2z.3: 'neque Thus it seemsbetterto assume agrumcoluit neque aliquid colonariioperis celebravit'). reflects the thatthesurety previousor simultaneous (confessio, compar acknowledgement law ofmunera) by theadscripticius ablewith theagnoscere inadministrative of his public theobligation tokeep his goods on theestatefitted obligations. Furthermore, theprohibi tion against alienatingfromhis peculiumwithout theconsentof theestate owner (Cj momentof entering At the theadscripticiate, it II.50.2.3; seeSectionXII).71 may have been The agreement betweenestateowner and prospective different. must have adscripticius For thefarmer consistedof severalelements. these would have been toprovide labour,to with all and everything remain on theestate (whichresembles theparamoneagreement or clause),72 topay thepoll tax to theestateowner- perhaps these were simply comprised in thephrase: thedutiesof an adscripticius. For theestateowner these would have been to Theremay have been special clauses,or it pay thepoll taxof thefarmer (and his family). may have been leftto custom (cf. CJ II.48.5, II.50.I). Furthermore, of extraordinary performances CJ II.48.I does not allow therequirement ifthatcould also be done at a more opportune when thecoloni are harvesting moment, and CJ II.48.5 remindstheestate owners that they may not demandmoney payments fromthefarmers butmust acceptpaymentsinkind,unless thisis thecustomof theestate (thiscould happen if a new owner took over). CJ II.50.1 prescribestheprocedureby which colonimay contestincreased claims,as does CJ II.50.z.4 foradscripticii.
X FLIGHT

on theestate, which layon every The obligationto remain colonus who had acknowledged his status,impliedthatanyunauthorized absenceamountedto a flight. Itcomes as no sur was a permanent concernfortheestateownersand emperors. prise, therefore, thatflight would seem,consideringthe legal textsand papyri, to Coloni, particularly adscripticii, have been prone to flight, presumably because of theharshconditions of thefarming life. from This is to be distinguished absence as such, which, as longas a person liable forthe colonate by origo had not been formally summonedby his estateowner to perform his was possible. If a colonus had fled, his estateowner could coerce him to return, duties, with theassistanceof theprovincialadministration More (CJ II.48.6, on adscripticii). was forbidden toharbour fugitive coloni (CJ II.48.z3.4-5). It out that over, it might turn

70 Sarris, op. cit. (n. 9), 60-1. 71 '... ne quid de peculio sui cuiquam colonorum CJ 11.50.2.3: ignorante domino praedii vender? aut alio modo alienare licet.' 72 in E. Lo Cascio See A. J. B. Sirks, 'Continuit? nel colonato?', (ed.), Terre, proprietari e contadini dell'impero romano. Dall'affitto agrario al colonato tardoantico (1997), 183-4; Sarris, op. cit. (n. 9), 65.

THE

COLONATE

IN

JUSTINIAN

S REIGN

I37

thecolonus (to be) had availed himself of a position which provided immunity from obli gations,based on theorigo (likecurialduties).The imperial services, the Church and the monasteriesofferedsuch positions,and in ordernot to harm theestateowners, itwas toacceptcoloniwithout their estateowner'sconsent(CJ II.48.18, I.3.36 pr., z) forbidden or at all (CJ I2.43.I, I2.54.3).When itnevertheless happened, the immunity was ineffec on accountof their tive(CJ II.48.II, mentioning originarii, i.e.,personssubjected origo). This indicates again thepublic law sideof thecolonate. But aswe saw previously, at thebasis of all theseregulations lay theconcernand need forsufficient labouron theland.Therewere alwaysestateowners inneed of hands.Thus itappears thatcoloniwho had fled wereworkingon thelandof otherpeople. Itcould be dealtwith coloni and used themto their thatestateownersknew they profit. Theywould over itsproductsto them, or by having accomplishthisby having themtilllandand turn them otherlabourand notpayingthem thatthey perform any wages. The textimplies used status.In thatcase they with their their knowledgetoblackmail them had topay fugitive the taxes whichwere otherwiselost (CJ II.48.8 pr.).However, itcould also be thatthe estate owners did not know of the flight and that the coloni deceivingly offeredtheir services asmen, free of thecolonatebond (ingenuiis thetermin thesourcesforthosefree or as tenants In that fromsuchpublic obligations),as farmlabourers or share-croppers. case the taxes should be claimed directlyfromthecoloni themselves since,as the texts And so, as theconstitution will state,itconcernsa contractus privatus. says,all fugitives what they be recalledtogether with their emolumentatributaria had i.e., ('fiscal benefits', II.48.8.I). In all cases any debts should be settledbefore the return(C) II.48.8.z).73 If theserulesconcern 'free' coloni,CJ II.48.8.I would state theobvious, since these people were alreadydirectlyresponsible fortheir poll tax; but theobvious is indeedsometimes to a contractusprivatus implies that the text refersto the stated.Yet the reference It followsthatincase of flight theobligationof theestateowner topay for his adscripticii. taxes).Thus ifadscripticiifledand were used by adscripticii stopped (as regardsfuture a loan,neither who did not pay them estateowner fromtheoriginalestateowner another nor fromtheadscripticii(orcoloni) themselves Hence the could thepoll taxbe reclaimed. on the mala fide estateowner. In thecase of the 'private forthis imposition of theliability was no reason contract'there themselves. why itshouldnot be claimed fromthefugitives It is obvious that these measures did notmake theharbouring and use of coloniwho while itburdenedthedeprivedestateowners with thesearchand had fledlessattractive,74 a problem,as thenumber was evidently of sureties reclamation. Flight proves (seeSection not only theclaimed Ix). A fine of iz pounds for the fiscand theobligation to render colonusbut also anotherfarmer and hisprice (probablytheadditionalvalue of landby the towork on it),should such as of a slave as instrumentum, appendageof an adscripticius, have deterred potentialharbourers(CJ II.48.I.z2). A colonamight fleeto a townor anotherplace andmarry an ingenuus(i.e., someone was found, shecould be recalled with her not subjectedto thecondicioadscripticia).Ifshe could be claimedaswell was due to inequality offspring (CJ II.48.I6). That heroffspring as regardstheorigoof townand estate (seeabove, atCJ II.48.24). Otherwise thechildren father and not have been subjectedto thecondicio would have followedtheorigoof their an ingenua was alreadydealtwith above. The case of an adscripticius coloniaria. marrying Ifan estateownerpossessed ingood faith coloniwho were claimedby somebodyelse it was the (as in thecase before),and thecoloni fledtoavoid theoutcomeof thelitigation,
to pay as tax the capitatio humana and which was a benefit to the state; Cl

73 By this ismeant the system we know from theHeroninos Archive, that on an estate account books were kept in which entries were made for things bought by the labourers from the estate and for salaries earned. See on this Sirks, op. cit. (n. 8), 362-4. The settlement would have made clear the amount of the peculium of the coloni, see below. 74 The harbouring estate owner would upon discovery pay the taxes which he had not paid, but he would have saved them otherwise; and in the second case he only had to pay their wages.

138

A.

J. B.

SIRKS

estateowner tohim, who could claim them back and not the otherperson;havingreturned theactual procedurecould start (CJ 11.48.I4, in accordancewith theuse of possessory interdicts in theprivate law).
XI DUTIES IN THE 'FREE COLONATE

on theestateor land they Those subjectedto the 'free' condiciohad to remain had as origo were originalesor originarii as well) and had to tillthis(CJ II.48.23.I, (in thissense they 5I.I, 52.I.I, 53..I), according to theconditionsoriginallyset for theirancestors (CJ estateowner to performtheservice II.48.23.3). Since they were obliged towardstheir of were in thisrespect subjectedtohis power (CJ ii.5z.i.i; unless this was agriculture, they only valid inThrace). They could be recalled in case of absence (CJ II.48.z3.4-5, 5I.I, 52.I, 53.I). Itwas forbiddento reduce theirstatus, i.e., tomake themadscripticii(CJ possessions (Cf II.48.23.I). They had to pay the tax, and theycould dispose of their held the land is unknown (apart fromthe II.48.i9, 52 pr.). Underwhat condition they special case ofNov.just. i6z.z), but CJ II.48.20 pr.-z would suggesta charge. It is not estateowner could release them. knownwhether their
XII THE DIFFERENCE BETWEEN THE TWO CONDICIONES: THE PECULIUM

inCJ II.48.z3.3. . Whereas adscripticius is used both as noun and as adjective to colonus, there is no such word to distinguish the other coloni otherwise than the reference that they are 'free', and we have to interpret colonus in its context to see to which category it applies. Likewise we have to do this for originarii and originales, terms which express the liability for a condicio by way of the origo. The question iswhat precisely the difference was between these two condiciones, apart from the way of paying the poll tax, since in both cases there is, basically, an obligation to be on a particular piece of land and render agricultural services. Perhaps with the adscripticii the estate owner could ask other services as well, whereas we see that for 'free' coloni the obligations must have been circumscribed precisely and were not to be changed (CJ II.48.23.2). adscripticiate, the colonus held his peculium for the estate owner. In the 'free' colonate, the colonus was 'free' and with him his assets (scil. what otherwise would have been a To^v icpuyjiaTov owuv). The meaning of Ta tokTwv peculium; ?6O?fpot gPVoVT-C t?&1b

Cf 11.48.23.I proves thatthere were twokindsof colonate inA.D. 534, as does theplural

evidentlyimportantfor theByzantines. In the CJ II.48.I9 defines the differences

the estate owner's adscripticii could not alienate anything in the peculium without consent. These assets could be claimed after an adscripticius' death by the estate owner as of this peculium ismade in the context of the faculty of coloni to peculium.76 Mention

JciKOUXWL statethatthe toi5 6crnrTrnv tV1K?inCf II.48.I9 isnot clear.Two other texts

Grey denies the existence of two condiciones coloniariae, but he does not analyse CJ 11.48.23.pr. and 1, he only mentions CJ 11.48.19: Grey, op. cit. (n. 8), 173 and 173 n. 104. See, however, my refutation of this in n. 43. But it concerns op. cit. (n. 42), 69, whom Grey criticizes, is correct in assuming regarding CJ 11.48.4, Mirkovic, adscripticii. It is evident from these texts that Justinian considered two different statuses: one the adscripticiate, a deterior condicio (and 'worse' implies that there is another, 'better' status), and another, the 'free' colonate, less or not deterior. 76 CJ. 11.50.2.3: 'Cum enim saepissime decretum sit, ne quid de peculio suo cuiquam colonorum ignorante domino contra eius personam aequo poterit consistere praedii aut vender? aut alio modo alienare liceret, quemadmodum tantum, non etiam transferendi iure, quem nee propria quidem leges sui iuris habere voluerunt et adquirendi

75

potestate permissa, domino et adquirere et habere voluerunt?'; CJ 1.3.20.1 deals with the inheritances of clerics, left to the monastery or church. If somebody has a claim, as the patron has a claim to a part of his freedman's estate, it In this context the estate of a censibus adscriptus ismentioned: '... bona seu peculia, quae aut has to be honoured. cui quis eorum fuerat adscriptus ...' patrono legibus debentur aut domino possessionis,

THE

COLONATE

IN

JUSTINIAN

S REIGN

139

betweenthetwo litigate againsttheir estateowners: CJ ii.50.2 pr. describesthedifference way, but itcomesdown to this. The coloni censibus categories of coloni in a veryelliptic in thecensus'), i.e. the 'free'coloni, are not sub dumtaxatadscripti ('merelyregistered are jected to theirestate owners,whereas theother category,thatof theadscripticii, which obliged (obnoxii)because of theyearlytaxesand theobligationsof their condicio, kindof slavery. For thatreason their faculty to sue makes them subjectedas ifina certain masters their estateowners is very limited:thensuch a thingis unbearablewhere their have thepower to alienate themtogether with theestate they are on (CJ ii.5o.z.i). The word obnoxius canmean 'subjectedto another'spower', 'liable'or 'obliged'. In thelatter nature sense it isoften used in thecontext ofpublicobligations(munera),"as here,but the thatthefirst of these meaning creptin.Furthermore, obligationscould have had theeffect he has thetextrhetorically puts thequestionof how sucha colonus may alienateanything inhis peculium without hismaster's knowledge, when the law did notwant him tohave an independent todo so and allows himonly to acquire and possess forhismaster faculty recallsa slavewho had a peculiumbutwithout the libera This strongly II.50.2.3). (CJ over it (which withoutpriorpermission administratio would have allowed him to alienate but it isbeyonddoubt thatone of his owner, seeCJ 4.z6.io). The languageis rhetorical, ormore constitutions theadscripticius' as owneror confirmed had eitherrestricted faculty a private law construction with a restriction regarding alienation,and thatsubsequently with thebond to the landhad led to the interpretation and con thisrestriction together as being in thepotestasof theestateowner. firmation of thepositionof an adscripticius One has to realize that inprivate law an estateownerwould always have been able to privateassets, afterhaving recover debts of his adscripticii by seizingand sellingtheir If these assets included sucha peculium,itassuredhim obtaineda judgement againstthem. were there. But thequestion is, what was thepeculium, thattheassets whywas thissepar Did itextend to all possessionsof an atemeasure introduced andwhat was itspurport? This questioncannotbe solvedherebut severalarguments would seem? adscripticius as it fit better with thelaw. can be raised make theinstitution whichmight were given chargeover irrigation works (n. 68). It is We have seen thatadscripticii which they had to cultivate. with plots of land The estate possible thatthesamehappened ownerwould have retainedownership.In thissituationtheadscripticius would act as over a lump actor (representative) fortheestateowner:he tookcare of hisplot and turned sum. If he borrowed money, thecreditorcould not recoverthisfromtheestateowner had givenan order (iussum),inwhich case he could proceed against the unless the latter estateowner with theactio quod iussu.78 ispledging his possessions. in a case in which an adscripticius The question is reflected Victor giveshis possessions in InP.Oxy. LXIII.4398 (A.D.553) the &vaco6ypapoq 7yopy70q cdovgtoi6zapX(O(6vov))to his estate ownerDioscorus for a loan of surety(Ktv6UVv(q) inhis charge (Vi'?t?).He promisesto return theloan wheat forsowingon theestatefields ofhis tax (pac& with the together payment KaicTOO(POlKOt) gou Popou) inkindof thesame himselfismeant;why shouldVictor be With pOpoqthe tax on theadscripticius quality. sense (i.e., liable forsomebodyelse's land? Ifhis possessionswere peculium in thestrict Neither is it likely that he ofDioscorus, inchargeofVictor) he could not do this. property ofpeculiumin the wider senseas discussed could pledge a potentialclaim to theremainder above. It is thesame inP.Oxy. LXVI.4s3s (A.D.6oo), where the ?xcucoypcpo; yeopny6
77 See Sirks, op. cit. de la S1HDA (1993), 78 This would mean applied to slaves and

inQuestions de responsabilit?, XLV?me Session (n. 25,1993), 164; A. J. B. Sirks, 'Obnoxietas', 32.5-32. that he could recover the amount as comprised in the order. Normally we see the quod iussu sons in potestate, but ratification (ratishabitio) was possible anyway. By this the represented become liable for the dealings of his procurator. This would work also for a colonus who dominus would ? which restricted the liability of the dominus to the range functioned as actor. The variation of action quod iussu ? was sense See 'Ratihabitio' if done. for the ratishabitio and iussum, A. Kacprzak, this of the iussum would make nel diritto romano classico (2002), 91-113.

I40

A.

J. B.

SIRKS

Elias acknowledges a debt to his estate owner, forwhich he gives a security on his
property, mortgaged pay this and the tax of the estate orchard he has to work for the taxes, but that he will pay (in any case) this amount

Anothercase: ina deed of surety Zacharias guaranteesthattheadscripticius tofito).79 will

on ('Ocv 67Tp ciauTou (popov; to the estate owner, since he

to this purpose

(KtV6<6V>p TO)V?ioti 6.cpn6va

vV]ZTOKM1tEvOV Ciq

P.Oxy. XXVII.z478 of A.D. 595/596). We shouldnot read thisas ifhewere liablehimself like Victor cannothave been liable forthetaxesof another person. The cases conform with thepracticeattestedinCJ II.48.20 pr. thatcoloni turnover most or all revenues, out ofwhich the landowner pays all taxes.Here it seems thatthey could retainpart of the revenues. Was thispart the reward for their work and conse
quently

could also recoup from it the taxes he paid

estateowner? Did they way, over a floating work capital, sometimes but dispose, in this with a loan, towork as smallentrepreneurs tressed on their plot? If so, theestateowner
for the colonus if the latter had not turned

their property? Or was

itworking

capital provided

to them in this way

by the

This is certainly thecase forimmoveables and herewe must assume thattheadscripticii were not underany lien;as followsfrom CJ II.48.4.I with itsregistration of adscripticii in thecensusof the territory, own plot of land lay. where their And indeed, nowheredo the textssay thatall theadscripticii had was peculium: they only say thattheadscripticii - which leaves open thepossibilitythat they cannot freely dispose of their peculium
owned other assets, not part of the peculium

owner, or else a pledge would make

to suchfinancial CJ 4.z6.I3.4 andCJ II.48.8.2 refer enoughover.80 practices with a period ical settlement of accounts. Next to that there are clearlyassetswhich do not fallunder thecontrolof theestate
little sense, and Elias is quite assertive in his assurance.

other than his wife or children. Such a person could be the son of another who had been to him in the course of an emancipation, mancipated or a debtor which had been

organizedand settled accordingto theusual private law ruleson loan.The yearlysettling of accounts would havemade them ownerof their saldo. The situationrecalls remarkably theclassicalRoman law figure of a person inman cipio.Mancipium was thepower (potestas)a pater familiasexercisedover a free person

contract. If they exploited

With the 'free' did not have thesameeffects. coloni, theenrolment Since they had topay their theneed fora guaranteefromtheestateownerwas evidently poll tax themselves, Their obligation was solelybased on a public obligationand not on a private superfluous.
the land with capital from the estate owner, itmust have been

and that thiswas

settled by law.

adjudicated tohim through addictio by the magistrate.81 Theywould remainfree persons and capable ofmarriage.The pater familias could acquire through them property but not what they possession (Gai. z.90), thus hadwould be peculium. Under Justinian thesetwo sourcesofmancipium had disappeared and with themthe mancipium as personal law very possible thatit was modelled afterthe mancipium,sincethis must stillhave existedat
status,82 but if the adscripticiate had been created in the firsthalf of the third century, it is

formulations are almost in perfect accordance with what Justinian in CJ 8.16.9 wrote as necessary to a hypothec: 'fide et periculo rerum ad me pertinentium'; cf. P.Oxy. LXX.4794 (a.D. 580), also a hypothec of all belongings. Itmay look strange that in a case of peculium there could be a debt of the adscripticius to his estate owner, but the peculium functioned as a separate entity and was usually leftwith the dependent person after the dependency ended, after a final settlement to establish its net value. 80 Not unlikely, since a harvest had to supply food for his family, seed for the next year, and pay the taxes. The first two expenditures would have priority. In view of the loans for seed, attested harvests were not always sufficient. 81 For the first, see Kaser, op. cit. (n. 25), 302. For the latter, see M. Kaser and K. Hackl, Das r?mische Both establish (1996), 387-8. Zivilprozessrecht 82 because it had fallen into disuse by his time, emancipation Justinian abolished the mancipatio being performed now before the magistrate; and condemnation to pay a debt no longer led to addictio in the power of the creditor.

79

THE

COLONATE

IN

JUSTINIAN

S REIGN

I41

that time.83 The mancipium could be imposedby a public officialand consequently by public law, so to assume thattheact of inscription intothe (public)census register drew isdefensible since thecolonus into theadscripticiate to sucha statute. CJ II.50.2.3 refers Yet there aremore aspects to theadscripticiate. It isevidentthatit was connected with of thepoll tax.The estateowner tookupon him theduty topay thisand the thepayment was bound because of this must have been a bond tohis estate adscripticius yearlytax. It The colonate served(also) toguaranteethelatter owner.84 thecounter-performance of the theestateowneras lessor colonus. If thecolonushad been a tenant, would have had a tacit on theproperty hypothec of the tenant(CJ 8.I4.7), whichmeant thateven ifthe tenant Therewould have been no need forthe sold his goods, his claims could stillbe satisfied. privilegesthe adscripticiate gave the estate owner. This (again) is a strongargument to theadscripticiate. as being fundamental were against the idea of tenancy Adscripticii or farmers who had a plot of landof theestateowner then eitherland labourers originally incharge(inprecario?).It is likelyinviewof their situationthatthey destitute presumably of thison theestateowner forloansand would have been dependentfortheexploitation in times as they of adversity; did forthepoll tax. In thiscase theprivatelaw did leniency of security. not offertheestateowner a form Mancipium would have been a solution. Regarding those who could only offertheirlabour, it securedtheauthority of theestate owner to summonthemand to forcethemtoworkwithout theneed to go to court first. Regarding thosewho had some assets, it securedhis claim by thesebeing treatedas would be built in by denying the adscripticiithe libera peculium.A further security in it over thepeculium, whichmeant thatthey administratio could not alienateanything inhaving Protection of theestateowner's interest without theestateowner'spermission.85 must have led to thelegalisation of thetaxpaid byhim as guarantor of thisand recovery would have been an effect of theenrolmentin thecensus as consequentlythepeculium An exceptionexistedfortherealestate they had adscripticius. possessed in as faras they in thecensusof the to pay the tax forthis (the tributum) themselves and had to register estateowner apparently where the land lay (CJ II.48.4.I). Since their did not territory did not apply to this carryany liability here, it is likelythat thepeculiumconstruction land. Here thetaxcollectorcould seize theland incase of default. was liableto theadscripticiate owner Ifsomebody byorigo,hewould have unrestricted shipof his assets untilhe was summonedand had acknowledgedthisobligation, in the must thus same have become subjected way aswas donewithmunera.His moveable assets to thepeculium. of theadscripticiate. The otherpossibility This is one way to explain thepeculiarities hispossessions infiduciary owner would be thattheadscripticius would have transferred He would keep possessionor receive ship tohis estateowner, inorder toprovidesecurity. was not uncommonwith his propertyback in precario. The practice of transferring of a powerful sociallyvulnerable people, sinceas property person their(few)assetscould assets was not easily be seized by the authorities.This practice of transferring

It is said that we do not have a constitution introducing the colonate or adscripticiate. This is true and itwould If itwas done after A.D. 311, there would have been a general constitution indicate an introduction under Diocletian. Code and after that in Justinian's Code, and probably also in issued, which would have figured in the Theodosian the Breviary of Alaric. It does not. On the other hand, we possess from Diocletian's reign up till A.D. 295 only, or almost only, rescripts, and such an introduction is not likely to have been done by way of this expedient. Yet it is for the period A.D. 285-305 is possible too. still possible, and a constitution of Diocletian 84 the fisc, everybody would have been bound and itwould not have merited a separate mention. Regarding 85 was to a slave: he could then alienate things out of itwithout the It possible to grant the libera administratio in the Later prior permission of his owner, who was also the owner of the assets in the slave's peculium. Although Roman Empire filii familias got a libera administratio, Justinian went back to the old system (M. K?ser, Das r?mische Privatrecht II (1971), 215).

83

142

A.

J. B.

SIRKS

suchgoods peculium would fitthesituation of social condemned.86 To call untechnically have to assume thatit would have been recognized later dependence. We would, however, The first which isunlikely. on inpractice, hypothesis gives a better explanation,although in thecensushave taken there remain questions,suchas: howwould theregistration place? in litigation,imposed upon the Did it indeed automaticallyentail the restrictions
adscripticius?87

XIII

SOME

CLOSING

OBSERVATIONS

on Justinianic our investigation textsand We have concentrated legal and documentary view on thecoloni censiti. seen thatit ispossible to get a coherent in First, there were two condiciones regarding coloni censiti,thepersons registered somebody'scensus as coloni: the condicio coloniaria or adscripticia,and the condicio was based on an origo estab whichwe called the 'free' colonate.The condicioadscripticia or unwed a father on an estate,itself mother who had already lished based on descentfrom intothecensusof theestateowner fol been subjectedto this condicio,or on an enrolment to thiseffect. The othercondiciowas also based on an origo estab lowingan agreement a father on an estate,and thisagain on thedescentfrom or unwedmother who had lished of it. alreadybeen subjectedto thiscondicio,or on theimposition The consequencesof each condiciodiffered considerablyinprivate law. In thedomain of personal law, thecondicio adscripticialed inA.D. 53I/534-539 fortheadscripticiusto restriction on marriagewith a woman not subjected to thiscondicio. In thedomain of propertylaw, it implieda grave restriction regarding his assets (his moveables and prob most of theadscripticii their ably for onlycapital). Itwas called peculiumand considered as ifpeculium without liberaadministratio, which impliedtheprohibition againstalien todispose without the of theestateowner, including thefaculty atinganything permission was of his peculium. If an adscripticius possessed real estate, thisproperty by testament not affectedby it.Regarding procedural law, in this condicio the possibilityfor the colonus to sue theestate ownerwas reduced to theposition of a freedman against his was almostnil. Itwas theserestrictions manumissor,i.e. it whichmade himconsideredas ifinpotestateand a subjected was not alreadya feature of his condicio.The person, ifthis were called,with othercondiciodid not have these consequencesand thosesubjectedto it justice,'free'. In public law, theconsequencesof both condicioneswere originallythe same: both kindsof coloni could be summoned or, incase of absence,be recalled,together with their and peculium. within theexploitation Both could be ordered toperform services of family theestate that was their was thesame: they could sue their origo.Regardingprocedures,it estateownersfor their unjustclaims regarding condicio.But theadscripticii had three pos sibilitiesto leave their condicio.They could exit their condicio if theestateowner con fulfilled their sentedto this.If they years,their condiciochanged into obligationsforthirty thatof the 'free'colonate, and thisapplied to their childrenas well. This will not often have been thecase. FromA.D. 539onwards, ifa male adscripticius entered amarriage with awoman not subjectedto thecondicioadscripticia,thechildren of sucha marriage would own estate ifthey be 'free' coloni and could even transfer their origo to their acquiredone largeenough forthemto be fully occupiedwith it. In both condiciones,thecoloniwere taxed forthepoll tax. If they possessed land, they had tomake a professio in thecensusof the territory where it layand pay the tax on it

86 See CJ 2.13-14 on this. Cf. further CTh 13.7.1, which describes this practice with ships, whose owners wanted to escape the compulsory transportation; and CTh 12.1.6 fin. for a decuri?n. 87 I therefore offer this hypothesis with reservation. In view of the present state of information a good analysis of the taxation system in the Later Roman Empire is indispensable for further research.

THE

COLONATE

IN

JUSTINIAN

S REIGN

143

In thecase of the themselves. poll tax (capitatio humana), theestateownersof adscripticii on responsibility paid thistax to thetaxcollectors, because they had taken forit, while the in theirturn were topay it to their estateowner. Itwould go too farto see in adscripticii a role for theestateowner as public tax collector,as Gascou does.88It thisconstruction was rather a case of taking over as primary and thus co-debtor the beingobliged to render amountof the tax as revenue. The otherkind of coloniwas in thisrespect'free'as well: suchpeoplewere primarily forthistax and had topay it themselves. responsible Justinian underlined thecontractual originof theadscripticiate and although we do not was thecase inhis time(manycases of theadscripticiate know how oftenthis inhis reign will have been based on origo bydescent,aswill have beenmost or almostall cases of the 'free' colonate), it must have happenedor else his emphasis would not havemade sense. to theadscripticiate And we know of a case of re-admittance (seen. I7). This means that was not a universal theadscripticiate which afflicted all farmlabourers. phenomenon And as a statusbased on origomeant thata colonus on theother side, the implementation was not necessarily a farmlabourer, censitus but could pursueotheroccupations. In view of therestricted scope of thisarticleit isnot possible to draw conclusionsfor theentire debate on thecolonatehere.But something can be said nevertheless. The many references to coloni censitibeing away fromtheestate and occupyingpositionswhich shieldedthemfrom theemphasisof Justinian on thecontractual being summoned, origin of theadscripticiate, theoccupationsof thoseadscripticii who were of thisstatusappar ently onlyon accountof their origo, indicatethatfor Justinian's reignthecharacterization of societyas a sociallypetrified does not hold well for this level. Neither is it system possible to state that thepersonalpositionof thecolonus censitushad deterioratedinto serfdom.89 This corroboratesthecaution as expressedby severalauthors.Perhaps society and thecolonatewere indeedso, or to an extentso in theperiod up untilJustinian; but one has to surmise momentbetween then a considerable Diocletian change inbothat some and Justinian. Oxford n.sirks(law.ox. ac.uk boudewij

et M?moires Les grands domaines, la cit? et l'?tat en Egypte byzantine, Travaux 9 (1985), 22-3. J. Gascou, Following Gascou: M. Kaplan, Les hommes et la terre ? Byzance du Vie au Xle si?cle (1992), 162. Challenged by, e.g., Sarris, op. cit. (n. 9), 155-9. F?r a summary of the increased reservations, see B. Palme, 'The imperial presence: government and army', in Bagnall, op. cit. (n. 65), 263-4. See also n. 31 and Section vu. 89 see n. 9, without any reservations. As Demandt,

88