INTRODUCTION
Interpreters and translators, and their professional activities, are mentioned only infrequently in historical sources from the Roman Empire.Footnote 1 They are also difficult to trace in the epigraphic and documentary records. Semantic ambiguity in the Latin and Greek terms for ‘interpreter’ often makes discerning the precise role of any individual ‘interpreter’ rather difficult. There is a small corpus of Roman soldiers from provinces along the Danube commemorated as ‘interpreters’ (Latin interpres) in inscriptions, and various other officials named as ‘interpreters’ (Greek hermêneus) in the papyri of Roman Egypt. In addition, there are two occurrences of the verb ‘interpret’ in documentary texts from the north-western marches of the Empire which merit further consideration. These references to ‘interpreting’ — one from northern Britain in the Vindolanda Tablets (Tab. Vindol. II, 213) and the other from the Netherlands on a stilus tabletFootnote 2 — offer, at first glance, a chance both to enlarge the small corpus of attested Roman official interpreters, and to shed light on their professional activities and position within administrative hierarchies. The reference in the Vindolanda text is not, however, to an individual interpres, but is an instruction to an unnamed individual that he should ‘interpret’ with regard to a commercial transaction. Likewise, the person who ‘interprets’ in the document from the Netherlands appears to be acting as an intermediary of some sort. It is, therefore, difficult to associate these persons directly with those Roman army officers who served as interpretes in the Danube provinces. However, these two references fit neatly with the evidence we possess for both the para-interpreting activities of language-mediators — who might run commercial ventures during or after their military service — and for those individuals who are designated as ‘interpreters’ without fulfilling any obvious linguistic role at all.
The discussion begins with the Vindolanda tablet in question and its parallel, the document formerly known as the ‘Frisian Ox Sale’, followed by a brief review of the evidence for ‘interpreters’ in the Roman Empire as a whole; an assessment of the Danubian evidence for military interpreters (who might also be engaged in commercial activity); and finally a summary of the evidence, mostly from Egypt, of ‘interpreters’ whose primary role was participation as agents in economic transactions. Although those ‘interpreters’ whose role was primarily economic may well also have drawn on linguistic skills, it is evident that both the Latin and Greek words commonly rendered into English as ‘interpreter’ were semantically rather flexible, and covered a range of functions — and permit a range of translations.
TAB. VINDOL. II, 213: ‘UT INTERPRETERIS’
The relevant tablet dates to Period 2 of Vindolanda (c. a.d. 92–97), and is therefore roughly contemporary with the earliest of our Danubian interpretes, Q. Atilius Primus, whose epitaph dates to the second half of the first century a.d. (on which see below).Footnote 3 There are, however, some important differences, which mean that we cannot consider the Vindolanda reference as part of the Danubian group, or as a product of the same administrative and social context. Rather than interpres as a military or professional title, we find here interpretari as a verb. The context is clearly commercial, and the text concerned is documentary, a letter relating to the day-to-day business of the camp, not a stone epitaph which reflects the professional self-definition of an individual, or the ascriptive choices of his commemorators.
The letter takes the form of a diptych, complete at the head and at both margins, but damaged at the foot. It belongs to the correspondence of Cassius Saecularis,Footnote 4 an officer who was stationed at Vindolanda at least in Period 2, but quite possibly for longer. We do not know his rank, but the editors suggest that he was an optio, decurio or centurion, with his concern with supplies indicating that optio may be most likely.Footnote 5 Curtius Super is not known from elsewhere in the tablets. The letter relates to a transaction involving barley, which Curtius Super suggests some persons should obtain from Saecularis.
The crucial phrase ‘ut interpreteris’ is translated by the editors as ‘so that you may explain’. As they note, there are three broad senses — in approximate order of probability — in which the verb may be read here: in line with the sense of interpres as ‘agent’ or ‘go-between’; in a general sense as ‘explain’; or with the specific linguistic meaning of ‘interpret’ with non-Latin-speakers. We may, in the view of the present author, be more confident in the reading and treat the sense here as ‘so that you may act as intermediary’. As the material considered in the following sections will demonstrate, there are present good parallels for interpres/interpretari and hermêneus/hermêneuein being used to relate to activities which are solely or partly commercial in nature. The interpres Q. Atilius Primus in Pannonia Superior was also, at some point in his working life, a negotiator. The clearest parallels are those Egyptian documents where hermêneus has a commercial sense.
The present text obviously relates to commercial activities, although their precise nature remains unclear: are we dealing with transactions on the open market, or with systems of official requisitioning? On commercium, the editors ‘presume that it is here used in apposition, “barley as commercial goods”, in distinction to something for personal use’. (There is another possible occurrence of commercium in the same sense in Tab. Vindol. III, 581, l. 41, but this is very doubtful and should more probably be read as the beginning of a personal name.) Although we do not know much more about the individuals mentioned in this text, others from among the published material from Vindolanda testify to the role of civilian traders and military officials in provisioning the army.Footnote 6 Tab. Vindol. II, 180 (http://vindolanda.csad.ox.ac.uk/TVII-180), for example, is a lengthy private account apparently kept by a civilian supplier rather than a bureau within the military — although it was discovered in a room within the barracks. ‘The evidence for the involvement of civilians in army supply, and especially in the supply of wheat, the basic commodity, is of considerable importance and contributes to filling an important gap in our knowledge. The unknown author of this account must have been a crucial link between the producers and the army personnel who authorised the distribution within the unit.’Footnote 7 Many other accounts relate to supplying the fort with food and other necessities, often through intermediaries whose status as military or civilian is not clear. Tab. Vindol. II, 192, for example, records foodstuffs and textiles received from an agent named Gavo, who appears in a couple of other accounts.Footnote 8
Given Vindolanda's position in Britain, and moreover on the northern frontier, it is tempting to posit that the reference here is to interpreting with native Britons, locally or in the territories to the north. Whatever role military interpreters of indigenous languages may have played in Britain, however, this text is not evidence for them. As in some of the other texts which will be discussed below, a person engaged in ‘interpreting’ in a commercial sense may also have been capable of operating in more than one language, but this is not the sense of the present text.
A PARALLEL (OR PRECURSOR) FROM FRISIA: ‘QUADRATUS INTERPRETAVIT’
A new reading of an old text now supplies an interesting parallel for Tab. Vindol. II, 213, in a Latin loan-note on a fragmentary stilus tablet excavated in the early twentieth century at Tolsum (Frisia) in the Netherlands.Footnote 9 Digital imaging techniques have been used to yield a fuller and more accurate text than was hitherto possible. Like the Vindolanda document, the context is both military and economic, but the Frisian text is rather earlier, with a consular date of a.d. 29, supported by radiocarbon dating.Footnote 10 The tablet was previously known as the ‘Frisian Ox Sale’, but the new edition reveals that it was a loan-note in which oxen did not figure: the ‘bovem’ of earlier editions is now to be read as a simple ‘quem’.
This text is sixty to seventy years earlier than Tab. Vindol. II, 213, and from another part of the Empire, but there are certain parallels to be drawn. The transaction is a somewhat more private one than in the Vindolanda document, but the persons involved were apparently all members of or attached to legio V Alaudae, at a time when it was based at Vetera on the Lower Rhine. Once again, we know little of the one who ‘interprets’, other than his name, Quadratus. His role in the transaction may be that of literal ‘interpreter’ between languages. Like the Vindolanda document, this loan-note comes from a region on the north-west marches of the Roman Empire at that date, and the slave Carus(?) — or even his creditor, whose name is not preserved — may not have spoken Latin. But, again, the most economical (in both senses of the word) explanation is to make Quadratus an agent or go-between, whose role in the transaction was to facilitate and mediate between the parties. Quadratus may well also have spoken more than one language and put this expertise to use in his role as mediator — the ambiguity of interpretavit may, indeed, be conscious — but the description of his activities here does not absolutely require this to be the case.
INTERPRETERS AND THE ROMAN AUTHORITIES
Searching for interpreters of foreign languages in the Latin and Greek literary, historical, epigraphic and documentary records as a whole is a frustrating business. Two circumstances conspire to make language mediators ‘invisible’: the semantic ambiguity in the Greek and Latin terms (hermêneus, interpres) used to refer to interpreters and translators; and a general neglect or reluctance on the part of our sources to record that mediation between languages, through active agents, had taken place at all.Footnote 11 Despite the linguistic expertise needed to control and administer a multilingual empire,Footnote 12 and manage relations with peoples beyond imperial frontiers, we find only a handful of explicit references to individuals who served as interpreters within Roman military and civilian administrative hierarchies: this is true of both the literary-historical and the epigraphic-documentary records. These references have a slight but notable tendency to ‘clump’.Footnote 13 As already mentioned, there are inscriptions recording a number of army interpreters in the provinces along the Danube, mostly of the late second and early third centuries a.d. Cicero provides accounts of the doings of various interpreters within late Republican provincial bureaucracies: loyal and trusted interpreters whom he himself employed in Cilicia, such as M. Marcilius who he considered his ‘amicus atque interpres’;Footnote 14 and the nefarious agents of C. Verres in Sicily, who were interpreters also in the broader sense of agents and go-betweens.Footnote 15 Other, more isolated inscriptions record the presence of interpreters in imperial service in Egypt,Footnote 16 SyriaFootnote 17 and Germania Inferior.Footnote 18 Those sources in which we might expect to find more frequent reference to interpreters or translators who mediated between local people and the Roman authorities in fact furnish very few examples. The archives of Babatha (c. a.d. 94–132) and Salome Komaise (c. a.d. 125–31) from Arabia contain signposted translations, such as the translated subscriptions to census declarations of both the parties and the presiding Roman official, but no indication of the authors of such translations.Footnote 19 In Egypt, we find occasional reference to interpreters employed in law cases before Roman courts.Footnote 20 Only one of these interpreters is identified by name, and this is most likely because he was also one of the presiding officials at the hearing in question.Footnote 21 We also sometimes encounter translations of Latin legal documents in which legal experts claim authorship and responsibility for the accuracy of the translations,Footnote 22 but such references are really very few.
ROMAN MILITARY INTERPRETERS IN THE DANUBIAN PROVINCES
There follows a relatively brief summary of the Latin inscriptions from the provinces of Pannonia Superior, Pannonia Inferior and Moesia Superior which mention interpretes, without venturing into questions of the ethnic background of these men or the peoples whose languages they spoke (Dacians, possibly also Sarmatians and Germans). They all, at any rate, bear ‘colourless names typical of soldiers’Footnote 23 including some imperial gentilicia. These men are without exception Roman soldiers, and most of them hold ranks above that of simple miles. Another inscription, mentioned above, from Germania Inferior may also commemorate a Roman military interpreter, but this epitaph is fragmentary, of uncertain date, and apparently no longer extant.Footnote 24 The recorded text lacks any indication of military position, but given that Roman soldiers were the most prolific generators of Latin inscriptions in the provinces, it seems likely that the man commemorated was a soldier.
The military ranks held by these interpretes were as follows (in bold text):
-
1. Q. Atilius Primus (Kolnik Reference Kolnik1978; AE 1988, 938; Mosser Reference Mosser2003, no. 13, 171; Pannonia Superior, near Carnuntum; second half of the first century a.d.).
-
Q(uintus) Atilius
-
Sp(uri) f(ilius) Vot(uria) Pri-
-
mus inter(p)rex
-
leg(ionis) XV idem (centurio)
-
negotiator an(norum)
-
LXXX h(ic) s(itus) e(st)
-
Q(uintus) Atilius Cocta-
-
tus Atilia Q(uinti) l(iberta) Fau-
-
sta Privatus et
-
Martialis hered(es)
-
l(iberti?) p(osuerunt)
-
-
2. M. Ulpius Celerinus (AE 1951, 103; Barkóczi and Mócsy Reference Barkóczi and Mócsy1972, vol. 2, 590; Pannonia Superior, Brigetio; late second–early third century a.d.).
-
M(arco) Ulp(io) q(uondam) Romano mil(iti)
-
praet(oriano) et primoscr-
-
inio praef-
-
f(ectorum) qui vi-
-
xit an(nos) XXXV
-
M(arcus) Ulp(ius) Cele-
-
rinus sal(ariarius)
-
leg(ionis) I Ad(iutricis)
-
P(iae) F(idelis) interprex
-
Dacorum vivus sibi
-
et filio suo s(upra) s(cripto) carissimo
-
f(aciendum) c(uravit)
-
-
3. M. Aurelius Flavus (CIL III, 10505; Pannonia Inferior, Aquincum; late second–early third century a.d.).
-
D(is) M(anibus)
-
M. Aur(elio) Flavo m[il(iti) leg(ionis) II Ad(iutricis) du]pl(icario)
-
[e]t interpetri [sic.] Ge[rmanoru]m
-
off(icii) co(n)s(ularis) et M. Aur[(elio) … an(norum)] XIII
-
filio communi Aurelia Q-
-
u[i]aeta (?) marito et filio dulcissim-
-
is scribi in memoriam iussit
-
-
4. ? Gaius (CIL III, 143495; Pannonia Inferior, Aquincum; late second–early third century a.d.).
-
…]PAIL[…]
-
[…] et Iulio Dig[…]
-
[…] patri paga[(no) …]
-
[…] Ulpi(a)e Homenat[…]
-
[…]n(a)e so(ro)ribus […]
-
[…] Gaius mil(es) leg(ionis) [II Ad(iutricis)?]
-
[…] interprex S(armatarum?) e[x…]
-
[…o]ffici(i) co(n)s(ularis) sib[i et]
-
[…] parentibus H[or]-
-
mus filius eius
-
-
5. C. Valerius Valens and M. Ulpius Martialis (CIL III, 14507; Mirković Reference Mirković1986, 53; AE 2004, 1223; Moesia Superior, Viminacium; after a.d. 195).
-
l. 11: C(aius) Val(erius) Valens int(erpres) R(atiaria)Footnote 25
-
l. 40: M(arcus) Ulp(ius) Martial(is) int(erpres) Sc(upis)
-
With the exception of No. 5, a list of discharged veterans of legio VII Claudia, all are funerary inscriptions.
Q. Atilius Primus (No. 1) was in addition a negotiator. Whether this was concurrently with or subsequent to his military career is not specified, which leaves the two equally interesting possibilities that he was a negotiator for the army, involved in provisioning in some way,Footnote 26 or that he was a retired soldier-turned-trader on the frontier. Since he lived to the age of eighty, Atilius had ample time to have established himself as a trader after his retirement, whether or not he had also been involved in provisioning the army during his active service, or in trading as a sideline. Atilius was buried beyond the Danube, in an area where the archaeological evidence reveals considerable adoption of or engagement with Roman architecture and material culture by the local population at this date.Footnote 27 As well as providing an environment in which Atilius' commercial venture might flourish, in this ‘contact zone’ language skills will have been key in enabling him to conduct business. His professional life as an interprex may, therefore, have stood him in good stead.
M. Ulpius Celerinus (No. 2) appears in another inscription from Brigetio,Footnote 28 a dedication to Jupiter where he is described simply as salariarius legionis I Adiutricis, without mention of the fact that he was an interpreter. The salariarius was an officer who, from the time of Septimius Severus, was responsible for stores and provisioning.Footnote 29 His son, M. Ulpius, was a person of some distinction: a member of the Praetorian Guard (effectively disbanded by Septimius Severus in a.d. 193 and re-formed with heavy recruitment from the Danubian provincesFootnote 30 ) and primiscrinius, a chief financial officer in the Prefectoral administration.
M. Aurelius Flavus (No. 3) was a duplicarius, on double pay. Flavus and ?Gaius (No. 4) served on the the staff of the provincial governor.Footnote 31 The officium consularis, in addition to its regular personnel, could include other officials with special functions, such as haruspices, or those general dogsbodies who described themselves simply as milites of the officium.Footnote 32 Six inscriptions from Pannonia Inferior, including the two inscriptions here, record officials off. cos. The others are a miles of legio I Adiutrix a responsis officii consularis;Footnote 33 two adiutores officii consularis;Footnote 34 and a miles of legio II Adiutrix who was a strator (‘groom’) officii consularis.Footnote 35 The language skills of our interpretes were evidently as useful to the governor's office as they were to the army.
These inscriptions, aside from their shared military context, display a fair degree of correlation between designation as an interpres, bureaucratic roles and trading or provisioning. These commercial activities may be part of their military or administrative duties, or personal ventures. We should suspect that the dividing line between these spheres was not always clear: M. Ulpius Celerinus (No. 2), the salariarius, for example, would have been ideally placed to conduct his own operations on the side, or to set himself up as a businessman after his discharge from the army. The location of the inscriptions of all of these linguistic-commercial (or commercial-linguistic) mediators on the Danube frontier is significant. This area was a contact zone where the Roman Empire and its inhabitants interacted with groups beyond the frontier — and vice versa — and where the Roman army and Roman officialdom might seek to monitor and control such interaction.
Negotiatores, interpretes and the Roman army and bureaucracy could be bound up in any number of ways. For example, Roman civilian negotiatores and their Greek colleagues on Delos in the second century b.c. mostly dealt with the language contact issues inherent in trading enterprises through bilingualism, a fact which is reflected in the complex intersections of languages in their epigraphic production.Footnote 36 In the first century b.c., in Sicily, at least one of Verres' interpreters is also described by Cicero as a negotiator (see above), and language skills and expertise as a middleman went naturally together. Epigraphically-attested negotiatores in Pannonia Superior of the first century a.d. were often freedmen of Aquileian trading families, sent to manage regional operations along the Amber Route.Footnote 37 However, the professional title negotiator also commonly appears applied to men who had dealings with the army, often supplying common military requirements such as weapons and clothing, or belonged to it themselves.Footnote 38
COMMERCIAL INTERPRETERS–MIDDLEMEN
The most frequent attestations of ‘interpreters’ who play no obvious language-mediation role, although they were possibly, or indeed probably, also multilingual, come from Greek papyri from Egypt. ‘Interpreters’ (hermêneis) appear in various connections in texts throughout the Hellenistic and Roman periods, and it is not possible to see much evolution in the semantic range of the term: at all periods it may have either linguistic or commercial implications, or both. In the Zenon Archive, of the third century b.c., most if not all of the hermêneis mentioned are demonstrably brokers or commercial agents, involved in carrying out or facilitating transactions.Footnote 39 A ‘measure of the hermêneus’ is used to regulate dry goods in a small number of texts from the Arsinoite nome, dating between a.d. 14 and 151.Footnote 40 In texts of the Roman period, a good proportion of the few references to hermêneis implicate them in commercial activity. A tax-inspector's report of the early second century a.d. mentions a ‘tax on brokerage (telos hermêneias)’.Footnote 41 In Theadelphia in a.d. 155 ‘clothing-merchants and brokers (himatiopôlai kai hermêneis)’ are found involved in the manufacture of garments for the state, paid through the public bank.Footnote 42
Such hermêneis are patently go-betweens, middlemen, but this does not compromise the reading, in other texts, of hermêneus as ‘interpreter of foreign languages’. The semantic range of the Greek term, just like that of the Latin interpres, was rather different from that of the English ‘interpreter’, although the English term encompasses its own range of meanings, concerned with the exposition of difficult or arcane information. An ‘interpreter of Trogodytes’ appears in a text from Thebes of 134 b.c.Footnote 43 The ‘interpreter and secretary (hermêneus kai grammateus)’, Papiris, who dedicates to Isis on behalf of the imperial family at Berenike on the Red Sea coast is a literate bureaucrat,Footnote 44 although it would be interesting to know a little more about his precise position within the Roman administration. In Hellenistic and Roman legal documents, translation is indicated by variants of the verb hermêneuein, and very similar formulae are used to introduce such translations, of whatever period.Footnote 45 The unnamed interpreters mentioned in Roman trial protocols are described as hermêneis, and all are quite clearly assisting individuals who cannot speak the language of the proceedings (for references, see above).
Since the overwhelming majority of papyri are in Greek or Egyptian, with a much smaller number in Latin, there is a lack of primary source material to examine applications of the Latin interpres or interpretari in documentary texts. But even from the literary and epigraphic material already cited, it should be clear that the semantic range, and contextual application, of the Latin term is similar to that of the Greek. Interpres may imply an interpreter of foreign languages, an interpreter of dreams or oracles,Footnote 46 or an agent, often in an expressly commercial sense (e.g. Verres' interpretes, see above).
It is, therefore, unnecessary for the ‘interpreting’ transaction in Tab. Vindol. II, 213 to bear any implications of language mediation, although such a commercial middleman, involved in the supply of barley to the Roman fort, will doubtless have either spoken Latin and the local language himself, or worked with other bilingual agents who did. This person is unlikely to be a formal army interpres, like those commemorated along the Danube frontier, or at least is not likely to have shared the bureaucratic side of their professional role. The unnamed recipient of the letter, who is to ‘interpret’ and get hold of grain, is hardly a former interpreter in the officium consularis down on his luck. He is not, it should be recalled, named as an interpres at all. Where he does have something in common with these Danubian interpreters, however, is in his place in the army supply chain, like the negotiator Q. Atilius Primus and the salariarius M. Ulpius Celerinus. Even within the Roman military, and even within a single military career, ‘interpreting’ could take place on a number of levels.
