meuiuuiiui iuissiuii wno were captured, sola, ana unven aciu&s uie nips irorn Moravia to Venice suggests that there is more to this than meets the eye. Nor is slavery absent from other accounts. Blaise himselfwas sold as a slave, and the old monk who hatched that idea must have known that beyond the imperial borders a market existed in Bulgaria where he might get away with selling a naive young man of good family and attached to the patriarchal clergy itself. Indeed, the notion of slavery was widespread enough to assume flexible forms: Blaise sold one of his disciples as a slave inside his monastery at Rome. Gregory Decapolite too sold himself as a kind of debt servant to another monk, to finance part of his travel to Rome. Real but surmountable dangers, vibrant piracy which supposes a lively shipping world, changing and complex routes, intercourse with the Muslim world, and an undercurrent of slavery as well as rich data on our travelers' broader life experiences: this is no meager harvest. The convergence with western patterns is striking and seems to argue that the shared features of these individuals' travel experiences are not coincidental. But again we must ask: are these travelers exceptional only by the eloquence of their stories? Once again the vividness of the anecdote gains depth and strength from the sterner discipline of examining the entire group of eastern travelers. 210 8 Easterners heading west: group portrait The distinctive faces of the half dozen eastern travelers we have just met stand up well against the overall group of 340 "eastern" travelers. "Easterners" are slightly more numerous than the "westerners." In their overwhelming majority, they were "Byzantine." Most (at least 247: 73 percent) were Greeks by political allegiance; a much smaller number (34:10 percent) were of Greek culture, although they resided in the Arab world. A handful (18: 5 percent) were Arabs, mostly attested in Latin or Greek sources; the others are of uncertain political and geographic background. The term "eastern" is no more than a convenient shorthand for this polyglot group, since it includes some Italo-Byzantines and a couple of Africans. We have included primarily "easterners" who crossed borders. Generally speaking, internal patterns of movement between the capital and western provinces of the Byzantine empire have not been taken into exhaustive account.1 Once the broad patterns of long-distance movement across boundaries have been clearly established, others may wish to analyze their correlations with the internal communications which wove together the Byzantine provinces and capital and, indeed, with similar communications inside the Islamic world. Byzantine Italy's shifting political geography complicates the issue. Places like Rome or Venice are obviously crucial to our theme in the ninth century, although they were still very much a part of the Byzantine empire at the outset of the eighth. So we have gladly broken our rule for them and the northern Adriatic coast, and taken into account activities of Byzantine officials involving them prior to their removal from the empire, if only to clarify what changed, and when. In this early stage of research, tracking for example the military governors (srratejjoi) of Sicily whose main activities were confined to that island would add more labor than insight. Administrative and other movements between the capital and provinces of the Byzantine empire can contribute to understanding the pattern of Mediterranean communications, but so vast a topic would require analysis on its own. For a preliminary outline of the problem, McCormick 1998b, 31-45. TAHLE 8.1 Types of eastern travelers: overview Type Number Percentage Envoys Pilgrims Officials Slaves Immigrants Missionaries Merchants Exiles/refugees Business Marriage Uncertain Total 148 92 19 19 17 9 8 5 3 3 17 340 44 27 6 6 5 3 2 1 1 1 5 100 Rome itself poses a special problem. Large numbers of Greek-speakers are documented there. But how are we to know where they came from? Its unique circumstances could well mean that many, even most, of them were native Romans, and this would change their significance to our overall inquiry. This chapter focuses on the two largest groups of eastern travelers I have turned up: those whose travel was state-sponsored, and those who moved mostly for religious reasons, chiefly as pilgrims. It includes also the data on other Byzantine officials who traveled to the west; Byzantine missionaries in particular served the state as well as God, as St. Constantine's dying words implied. Nonetheless, other eastern travelers help to answer questions about broader geographic horizons, wealth, age, and the like, and their contribution has been duly noted. But questions specific to the least well-documented types of travelers - essential categories like merchants, sailors, and slaves - are addressed in the next chapter. i. Basic facts As in western Europe, ambassadors and pilgrims dominate known "eastern" travelers. Envoys alone account for almost half the total, pilgrims for nearly a third. Beyond envoys, individuals who traveled under some sort of official sponsorship include officials, missionaries, and the three imperial princesses designated as diplomatic brides, making a total of 179 individuals. Together with the ninety-two pilgrims, these people account for most (271: 80 percent) of the eastern travelers. Table 8.1 summarizes the basic types and numbers. TABLE 8.2 Comparison of named and anonymous eastern envoys* and pilgrims ■--- Envoys State-related travelers Pilgrims Combined No. % No % No % No. % Mamed Anonymous Total 82 66 148 55 45 100 20 11 31 65 35 100 59 34 93 63 37 100 161 59 111 41 272 100 Note: ■'Including officials, missionaries and three imperial princesses sent to the west, A majority of all eastern travelers are known by name (195: 57 percent). The proportion of named to anonymous individuals stays roughly the same across the main categories of wayfarers (Table 8.2). Naturally, named travelers lend themselves to the most detailed analysis. Nevertheless anonymous travelers again furnish precious information. The data on easterners' geographic background are a little richer - and a little more complicated - than for westerners. It is possible, in a certain number of cases, to distinguish origin (main residence prior to travel), provenance (point from which we can discern someone traveling to the west), and even birthplace. For some we know enough about their other travels to describe their broader geographic horizons. And the whole issue is complicated by the large Greek community of Rome and the considerable contingent of voyagers from the House of Islam. 2. Geographic characteristics Taking the entire set of eastern travelers, including other, less well-attested kinds of travelers as well as the envoys and pilgrims, makes clear the relative contribution of various regions. The eastern travelers set out, overwhelmingly (275: 81 percent), from the Byzantine empire, including even a few residents of the Caliphate; a little under a sixth set out from the Arab world, including also a few subjects of the emperor; a small number came from Bulgaria.2 The figures will differ slightly, depending on whedier one classifies travelers by provenance, main residence prior to travel, or original residence. Looking at them by provenance - the place from which they set out on their travel - the other totals are; Arab world, forty-nine (14 percent); Bulgaria, twelve (4 percent); unknown, four. 212 213 ambiguity. Many of those whom I have initially classified as coming from Byzantium were monks resident at Rome. We will return to them in a moment. Travelers from the Arab world equally call out for closer attention. But let us begin with the largest single group, Byzantines, in the narrow sense of the residents of the Byzantine empire who set out from Byzantine territory, including, for the moment, Byzantine residents of Rome, but not including Byzantine residents of the Arab world. Is it possible to discern which regions of the empire furnished the travelers? Some kind of geographic connections can be detected for the overwhelming majority of these 275 individuals.3 Many concern the travelers' main place of residence prior to their journey (95 of 234 with some kind of known geographic connections -41 percent} and even, in a surprising tenth of cases, their birthplace.4 It is edifying to compare birthplaces with residence, as laid out in Table 8.3. Both the contrasts and the continuities between these two different series of data are revealing. First of all, Byzantine Italy appears frequently as a residence for the eastern travelers: proximity breeds travel, and the nearest region supplied a good many travelers. But it is noteworthy that Italian birthplace (and subsequent migration to another region of the Byzantine empire) correlates even more strongly with travel to thewest, Either Italian immigrants constituted a large share of the Byzantine elite in these generations or they were disproportionately disposed to travel back to the west and the borders of their native region.5 For Asia Minor, proportions of residents and natives are much closer. The European provinces show a smaller proportion of residences than birthplaces, which hints at emigration. The capital, on the other hand, displays far and away the biggest gap between births and residence: travelers were more than three times as likely to 85 percent; forty-one (15 percent) show practically no geographic data beyond where they appear in the west. As examples of the former, palatine charges are taken to imply residence at Constantinople, episcopal sees, residence in that cathedral town, although occasionally this assumption maybe unwarranted. On the other hand, the Byzantine emperors' ambassadors have not been counted automatically as residing in Constantinople, simply on the basis of their ambassadorial appointment. Although this is likely in many cases, there is at least one case of a churchman from Sicily being selected as an envoy and sent from Constantinople: Constantinus 4. Place of birth is known or can be deduced for twenty-eight travelers, including two whose residence prior to travel to the west is unknown. Only three of the fifty-seven envoys were probably of Italian birth so far as we can tell, and one of them did not represent the government: Epiphanius 2, Ioseph 1 and Theodorus 2. Tempting though it is, by itself this is insufficient to conclude that the imperial government made a practice of selecting envoys with specialized regional background. TABLE 8.3 Byzantine travelers: known birthplaces compared with residences (cf. Map 8.1) Birthplace Locations* No. % Italy Sicily 9 Rome 1 Reggio (Calabria) 1 Asia Minor (1 each) Amneia Amotion area Kyzikos Elrenopolis in Isauria Synada Europe Greece 4 Athens 1 Thessalonica 2 "Graecia" 1 "Moesia" (i.e. the lower Danube} 1 Dalmatia 1 Constantinople Georgia Kaxeti 1 Caliphate (1 each) Jerusalem Alexandria Total 11 39 18 21 14 7 28 100 Residence % n 26 19 44 0 100 24 41 0 Notes: "Numbers in this column specify birthplaces. 0 Two more travelers resided either in Sicily or, more probably, Constantinople: Anons. 237-8. 214 21,5 if E ^ E m v> a a-- .s, 2 LU . v H £ CP ' Vi ■i i si- reside in Constantinople as to have been born there, so far as we can tell. The implication is manifest: already in the eighth and ninth centuries, the capital was a magnet for immigration from the provinces.6 Even more striking, nearby Bvzantine Italy naturally supplied many travelers to the early medieval west, but distant Asia Minor was not far behind. Overall, Constantinopolitans' share outpaced that of all the Byzantine territories between the Frankish west and the Bosporus combined. In fact, the capital's dominance is even more pronounced when we consider that several of the Asian residents came from the greater capital region, for instance, the see of Chalcedon and the monasteries of Bithynia. Two groups of travelers merit a closer look. The first group is particularly interesting for their places of origin. Though they were predominantly of Hellenic culture, these travelers came from territories conquered by the forces of Islam, and add an important dimension to our investigation. The issue of origin is equally important for the second group: the large Greek community established at Rome offers its own problems and opportunities. About one in seven eastern travelers 1 have uncovered set out from the Arab world (49:14 percent of 340). They were mostly Christian and Hellenic. That so many appear from research geared to Byzantine and Latin source materials - presumably the least rewarding for the Arab world - challenges the conventional wisdom. Could levels of communications between the Christian and Muslim shores of the Mediterranean really have been so low in the early Middle Ages if so many moved across it? The importance of Jerusalem reflects the fact that the sample is overwhelmingly Christian. But the other main regions of the Caliphate are represented equally, from Iraq to Tunisia. Nearly half of these travelers (twenty-two: 45 percent) came to the westas envoys, chiefly from the patriarchate of Jerusalem (twelve) and also from the ecclesiastical authorities of Alexandria and Africa (three). Others, of course, represented the Islamic authorities of the Caliphate (seven), from Damascus, Baghdad and Fostat. Surprisingly, immigrants (seven), in addition to pilgrims (seven) - the two are not always distinct-make up a fair share of the rest. And, contrary to conventional wisdom that Muslims did not leave Islamic territory to trade, we find a few merchants (six), including some certain cases of Muslims, operating at least as far north as Rome. Other travelers on business of some sort (two) and prisoners or slaves (two) add a final contingent.7 We have already observed a persistent flow of western pilgrims to the Holy Land. It might be explained away as a triumph of religious devotion over economic realities. But that does not explain why easterners should travel from the Holy 6 See Pros. 7 The reasons for travel of the remaining three are uncertain. 217 table 8.4 Travelers' points of departure in the Arab world Place Number Percentage Palestine Jerusalem 1 6 Ascalon 1 18 37 Mesopotamia "Babylonia" 1 Baghdad 4 Constantia 2 7 14 Egypt Alexandria 7 7 14 Syria Damascus 4 Antioch 1 6 12 Africa Fostat 1 7 14 Uncertain 4 8 Total 49 100 Land to the west, and why they should be such an appreciable fraction of the total group of eastern travelers. Was travel involving Jerusalem borne along by deeper, less visible currents? The Holy City was the starting point for a third, but only a third, of the eastern travelers. Another, less obvious characteristic of some other eastern travelers further undermines the notion of interrupted intercourse, at least between Byzantium and the Arab world. Beyond the travelers who came directly from the Arab world to the west, another substantial group of eastern voyagers (28: 8 percent) also traveled to the Arab world at other times. In all, nearly a quarter of all eastern travelers also came from or went to the Caliphate (77:23 percent of all eastern travelers). They traveled, naturally, as envoys, but also as slaves or prisoners, and as pilgrims. Although some of the eastern travelers with Arab connections made only a brief sojourn in the Caliphate, others lived there much longer: a dozen years or more is not uncommon.8 To these we should add westerners resident in the Arab world who returned to the west for short periods, for example, western religious permanently attached to the church of the Holy Land who returned to the west as emissaries of the east. Migration, voluntary or not, between the Byzantine empire and the Caliphate is unmistakable. Most immigrated to the Byzantine empire or Italy; some actually moved back and forth.9 Most of the long-term residents of the Caliphate display a blurred, complex cultural identity. Two who had been born in the Byzantine empire and transported to the Caliphate as slave boys kept their knowledge of Greek.10 Almost all the others seem also to have spoken Greek. Even Zachary, the most clearly Arab among the group, shows a complex cultural personality.11 This former Muslim converted to Christianity and resided in Rome for at least nineteen years. There he served the papal court in an intimate and respected capacity as a physician, before defecting anew to his old religion and homeland. The story of this long-term Arab resident of Rome underscores the complexities and importance of Rome in tracking Mediterranean travelers. The Eternal City offers rich but somewhat problematic evidence. Its political and religious significance allied with its geographic situation on the edge of the Byzantine sphere of influence to make it a focal point for Byzantine travelers to the west; it was the single most frequent destination of our eastern travelers (see below). There they will have found people who spoke their languages and shared their culture. If the best-documented human groups, the monastic communities, are any indicator, the thriving and numerous Greek community dates from the seventh century. For the less well-to-do, no small encouragement came from the Roman custom of feeding pilgrims.12 They came as pilgrims. But they came also, perhaps especially, as immigrants and refugees from the turmoils of the end of empire, mainly from Syria and Palestine, the regions most affected by those seventh-century upheavals. At least in the early decades there is reason to suspect an influential number ofnative E.g., Theopemptus, Stauracius, Theodoms 7 and his wife, Anon. 207, Basilius 4, and Abu Aaron all to adulthood; Gregorius 5, twelve years; Elias 1, c. forty-two years, to name only those we could classify as immigrants. See Pros. Gregorius 5; Elias 1; Hilarion; the pilgrim-merchant-would-be ambassador Basilius 4 shunted around the Mediterranean, but showed no sign of returning permanently to the Arab world as far as we can follow him. See Pros. EIiasi;Leontius. Zachartas 4. J hat pilgrims received food at the expense of the Holy See is explicitly documented in the mid-7th C: Martin 1, Epištola, JE 2080, PL, 12C..600D-601A. Cf., e.g., Pope Leo Ill's construction of a bath for pilgrims as well as a pilgrim hostel, endowed with lands to supply food to pilgrims from "distantregions": Liber pant., Duchesne, 2.28.3-10. Archaeology may shed more light on the infrastructure of pilgrim hospitality in the early Middle Ages in Rome in particular. See for the early medieval textual evidence a few indications in Schmugge 1983. 218 219 Greek-speakers among the lay elite of the duchy of Rome.15 Was the Greek-speaking population of early medieval Rome an isolated, entrenched community of old immigrant families? Or did it continue to attract travelers of broader horizons? We know nothing about how quickly early immigrant families assimilated into the local culture, but their children, at least in the first few generations, probably helped to people the dozen or so Greek monasteries which sprang up at Rome from the seventh to the ninth centuries.14 Local Roman recruitment has in fact been judged better attested than the other obvious source, the Hellenic regions of southern Italy and Sicily. It has furthermore been suggested that the numbers of native Greek-speakers born in Rome dwindled in the course of the ninth century and that this, joined with a lessening stream of travelers from abroad, helps explain the disappearance of many Greek monasteries. Other Greek-speakers came from Constantinople, the more distant provinces of the Byzantine empire, and the Melkite churches of the Caliphate.15 A lost Greek inscription discovered on the Tuscan coast has cast a new shaft of light on the problem. This tomb marker commemorated three generations of a family of high officials in the early eighth-century Roman church. Several of the individuals are known from other sources, but their family connection and identity as Greek-speakers is new. It is plausible that the family came from Syria; they will have immigrated no later than 710, since by that date the grandfather may well have held a high post and accompanied Pope Constantine to Constantinople.16 The choice of Greek for their tomb strongly suggests that the family remained hel-lenophone down to the third generation, assuming that the grandfather had immigrated, and was not himself born in Rome or elsewhere in Byzantine Italy. Since their provenance remains only probable, they figure among the quarter of Byzantine residents of Rome whose geography I have classified as unknown. But it is certainly legitimate to suspect that they were locals. Closer scrutiny of individuals adds some nuance to the overall picture at Rome. Nearly a third of all the eastern travelers I have identified were more or less 13 See Sansterre 19S3,1: 20-1. Note that the latest evidence adduced there and in the relevant footnotes is from the first half of the 8th C. See ibid., 1:47 and 2:103-4^388 for the slight evidence available on theschola Graecorum. This schola seems to have come into existence in the 9th C. and, in any case, was then insufficient to maintain Greek monastic vocations at the earlier level. 14 Sansterre 1988, 704-7. 15 Sansterre 1983,1: 39-47, has conscientiously collected the evidence. 16 Inscripfionesgrauat 14, no. 2263, with Cosentino 1996, 507-16. Cosentino's observation that the uniqueness of the name of Mamalus 1 in the Roman church reinforces the identification holds equally for Moschus: see Llewellyn 1981, 367. More onomastic research mightadd further clarity to the question of the family's provenance. For the other family members, see Pros. "Agatha," "Anastasius 1," "Marouse"and "Sergius 2." long-term religious residents of Rome, or went there on pilgrimage.17 Is there any reason to believe that within this subset of eastern religious visitors or residents of the city, many came from further afield than Rome itself? It is possible at least to circumscribe the unknown and specify what is knowable. For a little over a quarter of this special group (27 of 102: 27 percent) we have no geographic data other than their presence at Rome. For over a third we know cither the birthplace (two), last residence before Rome (twenty-two) or both (fifteen).18 Although the Byzantine empire supplied more residents (Table 8.5), the Caliphate sent a respectable share of Rome's religious (Table 8.6). The trackable religious residents and pilgrims in Rome confirm the traditional view, that Byzantine southern Italy supplied a substantial share of the Greek monks of Rome. Italo-Byzantines of all descriptions constitute over a quarter of the entire group.19 Adding the six males from the Tuscan inscription, on the hypothesis that they were all born locally, would boost the local share of certain or possible Italo-Byzantines to nearly half (44 percent: 15 of 34). Rome itself occurs as a likely birthplace only twice, and both cases come late in our period. One might suspect that its very banality might militate against the mention of Roman birth in Roman sources, so that the numbers of "locals" are understated, and many of the Greek monks of unspecified origin might in fact have been Romans.20 Although the numbers are small, it is also worth noting that the Sicilian residents antedate the Arab conquest of the island, while the Calabrian ones show up in the late ninth century, a pattern which accurately reflects the changing weight of the two neighboring regions within Byzantine Italy.21 17 102 of 340 (30 percent). i have included various refugees, whose motives are generally recognized as religious in whole or in part; immigrants with religious connections which justify assimilating them to pilgrims (e.g. the Syrian bishop Theopemptus 1 and his daughter who settled in Rome in the early 8th C). Among the more fuzzy cases, I have accepted the suggestion that Harun ibn Yahya was a pilgrim. 18 Thirty-nine individuals; 38 percentof the overall group of eastern pilgrims or religious resident at Rome; forty cases of provenance, since Basilius 4 made two trips to Rome, from differing residences. '9 Including Methodius (by virtue of his birthplace, not his residence) would yield ten of thirty-nine individuals - 26 percent. 20 Greek monks at Rome named Benedictus or Martinus (q.v.) would be obvious candidates for local birth, if only the possibility of their having taken a local religious name (cf. e.g., the case of the Constantinopolitan Petrus 7) could be excluded. Among Greek monks, Roman saints' cults (and, probably, the names they imply) gained ground in the later 8th and gth C: cf. Sansterre 1983,1:158-9. 21 As usual, Table 8.5 is organized in descending order of magnitude, and within sets, chronologically. From about 827, one would expect the Arab invasion of Sicily to have spurred renewed migration to Rome. In fact, archaeological evidence may link the monastery of St. Basil in Scala Mortuorum with Sicily or southern Italy: Sansterre 1988, 707-8ni6. 220 221 TABLE 8.5 Religious travelers to Rome arriving from the Byzantine empire: known residences and/or birthplaces 30 Byzantine of 40 known provenances; 75 percent; see Maps 8. and 7.3. Name Residence Birthplace Constantinople (1 3 of 30: 43 percent of Byzantine provenances) Methodius 1 Constantinople Sicily Basilius 3 Constantinople Anons. 194-7 Constantinople lacobus Constantinople Petrus 7 Constantinople (Samarra) Basilius 4 (second trip) Constantinople Theognostus 2 Constantinople Anon. 243 Constantinople Anon, 273 Constantinople Blasius Constantinople Amotion sly (9 of 30: 30 percent of Byzantine provenances) Sergius 1 Sicily Palermo Stephanus 1 Sicily Sicily Anon. 75 Sicily Anon. 214 Reggio (C.)? Daniel 2 Reggio (C.) Taormina Elias 1 Reggio (C.) Sicily Elias 2 Reggio (C.) Reggio ohannes 22 Rome Rome Anon. 308 Rome Rome? Asia (5 of 30: 17 percent of Byzantine provenances) Eustratius Kyzikos? Hilarion Bithynia Isaac Bithynia Gregorius 9 Eirenopolis Anon. 244 Bithynia Europe (3 of 30: 10 percent of Byzantine provenances) Zacharias 1 Greece? lohannesl2 Monemvasia Anon. 21 3 Thessalonica? Kyzikos Georgia Georgia Isauria Athens table s.G Travelers arriving at Rome from the Caliphate (io of 40 known cases; Map 8.1) Given in chronological order Name Residence Birthplace Theopemptus 1 Mesopotamia Constantia Anon. 6 Mesopotamia Constantia Gregorius 5 Jerusalem Crete Anon. 246 Jerusalem Anon. 262 Jerusalem Basil i us 4 (first trip) Jerusalem Anon. 245 Alexandria Anon. 264 Alexandria Anon. 263 Antioch Harun ibn Yahya Ascalon Italo-Byzantines were an important component of Rome's Byzantine religious population. Yet by these lights, they were not the overwhelmingly dominant one: as many Greek monks came from the admittedly large and diverse Caliphate as certainly came from Italy (Table 8.6). Constantinople looms large, since Constantinopolitans outstrip Italians. Their weight increases further still when we consider the greater capital region. This preponderance probably reflects the Imperial City's unique position in the cultural and religious life of the empire, as well as its preeminence within the communications network. Outside Italy, the empire's closer European provinces furnished only a few monks: other factors can outweigh geographic proximity. Another quarter of the Roman religious travelers of known provenance came from the Arab world (Table 8.6). That Jerusalem numerically heads the list of religious travelers from the Arab world arises, in part, from the natural links between the two greatest centers of Christian long-distance pilgrimage in the early Middle Ages. It also reflects enduring cultural and personal contacts which had originated in the seventh-century Levantine flight to Italy, the final stages of which we probably grasp directly in the immigration of Theopemptus 1 and his daughter.22 Whatever else is true, it is certain that almost a third (31 out of 102: 30 percent) of individual eastern pilgrims and the like attested at or en route to Rome did not come to Rome from Byzantine Italy. Even in the unlikely event that all 22 See in general Sansterre 1983,1: 9-51; on the enduring links with Syria and Palestine, some more material in McCormick 1998b, 36-8. 222 those whose provenance is uncertain came from the Eternal City or its neighboring regions, Greek-speaking Rome does not look like an entrenched, isolated society. The cosmopolitan character of the eastern religious at Rome is only strengthened by the group's broader geographic exposure. Broader geographic exposure includes, in addition to the cases of known origin or provenance, travel subsequent to the individuals' attestation in Rome. In this respect, two thirds of the easterners in Rome for religious reasons show connections beyond Rome to the wider Byzantine and Mediterranean world (66: 65 percent). Thirteen of them are attested in more than one place outside Rome. Constantinople tops the list of places which they knew (29), followed by Asia Minor (17), Greece (14) and the Caliphate (11), other places in Italy (5) and, far behind, Georgia, Bulgaria, Frankland {2 each), and Sardinia (1). The conclusion is clear. We can learn something about the travel experiences of most of the documented Byzantine individuals or small groups who were in Rome for religious purposes. Their personal movements span a broad geographic range, running from Bulgaria to Africa and from Sardinia to Samarra. The Greek population of Rome was a very cosmopolitan group indeed. Although many travelers had been to Sicily or Calabria, mainland Greece and the Caliphate, including Jerusalem and Africa, were also well represented. A second, methodological conclusion follows as well. The fact that, even by the broadest definition, about a third of this group remains of unascertained geographic horizons, and that at least a quarter of those whose provenances we know certainly came there from Byzantine Italy, means that we must be cautious about concluding that every Graecus mentioned in a western source came from the eastern Mediterranean. For the rest of our study, this means that individuals' documented trips or movements are surer gauges of communications than foreign individuals lacking such evidence. 3. Social profile Most of the officially sponsored travelers and pilgrims belonged to the elite. This is obvious in the case of almost all ambassadors, officials and exiles, and apparent in a fair number of the pilgrims. Wealth is, and was, the most obvious source of social status. In a way quite different from Carolingian Europe, however, the state formally structured the middle Byzantine aristocracy so that institutional affiliation provides a second important element for describing this set of travelers.23 Some (six) of those travelers whose family is recorded in hagiography are explicidy identified as having wealthy parents.24 But most envoys, pilgrims, 23 See e.g. Kazhdan and McCormick 1997, 24 See Pros. "Blasius"; "Ellas 2"; "Gregorius r72- 9"; "Hilarion"; "Iacobus"; "Methodius 1." officials, and exiles offer no direct information on their economic status. We might be tempted to suspect that many poorly documented travelers, particularly low-ranking imperial officials, or monks were less well off than the exalted social personages who otherwise crowd our catalogue of travelers.25 Yet the very fact of their travel and associations often suggests significant disposable wealth, and at least one Byzantine monk (Anon. 275) is positively identified as wealthy, paradoxical though that might appear. A few eastern travelers can be detected financing their movements as they went, along the lines of the Anglo-Saxon Willibald. Gregory Dekapolite, though he was born to wealth, rented himself as a "debt-servant" to another monk to pay part of his way toward Rome.26 The unspecified operations of pilgrim-businessman-ambassador Basil of Jerusalem maintained him in Constantinople for twenty months, until he ran out of money.27 Basil's travels around the Mediterranean and the length of time that he sustained himself in Constantinople, not to mention his priestly status, show that he was not from the lowest stratum of society. Yet along with the other imposter legates of the eastern patriarchates, the bishops and officials assembled at the Council of Constantinople of 869-70 judged that he deserved forgiveness for cooperating with Photius' plots, since he was "poor and foreign."28 One saint alone among this set of travelers certainly came from a less than distinguished milieu.29 A soldier in an elite regiment like the pilgrim Peter probably stood above the bottom of the economic ladder, at least before his capture by the Arabs. So far as we can tell, most of these travelers came from well-to-do backgrounds; among them, at least, the lowest reaches of the economic spectrum are not very low. Religious status was an essential component of social identity. For the total group of eastern travelers, religious outnumber laypeople.30 But among the different types of travelers who make up this group, the picture varies considerably. Laymen outweigh churchmen among ambassadors; marital candidates and officials are all lay.31 Pilgrims, on the other hand, are almost entirely religious. 25 E-g- "Daniel 1" or any of the many monks and travelers lacking other data. 26 Pros. "Gregorius 9." 27 Pios. "Basilius^.." Blaise is perhaps another case of self-financing travel. Despite his wealthy background, he kept busy weaving purple cloth and doing calligraphy in the monastery of St. Caesarius at Rome. 28 "FItcoxous • ■ • Kai l^vous": Concilium Constantinopolitanum a. 869, Actagraeca, act. 9, Mansi 16.397B; cf. Anastasius'version, ibid., 156E: "pauperes etperegrini." 29 GregoryAkritas: see Ch. 7.3. 30 186 religious (55 percent); 146 laymen (43 percent including non-Christians); four who entered the church in connection with their travel (1 percent); four of uncertain status (1 percent). The pattern holds for the combined group of envoys, state-sponsored and religious travelers on whom this chapter particularly focuses: 154 religious (57 percent); in laymen (41 percent); four changes in status; and three uncertain. 31 Envoys: eighty-one lay (55 percent) to sixty-four religious (43 percent), with three uncertain. 224 225 Eastern pilgrims differ somewhat from contemporary westerners with respect to position within the church, for there are a priest, a few deacons, and even some bishops among them."52 Missionaries were mostly religious.33 Church-wise, pilgrims and envoys differ distinctively. Pilgrims were overwhelmingly monks; a minority (8 percent) held any ecclesiastical rank that we know of. The envoys who were churchmen, on the other hand, typically held dignities ranging from deacon to bishop. Bishops and metropolitans alone count for a third (34 percent) of envoys who were churchmen. Monks, though present, seem far less numerous.34 Whether they were envoys, officials, missionaries, or brides, insofar as we can see them, lay travelers were generally of high status. Just how high emerges from their state dignities. One in seven of this group of lay travelers occupied the very summit of the hierarchy of dignities: they were imperial princesses, or patricians like Theodosius Baboutzikos (Figure 8.1).35 In fact, almost half of the lay travelers we can discern ranked in the upper ranges of the state aristocracy.36 That the loftiest dignities are disproportionately high and more numerous than lower ones implies one thing: the sources tend to mention explicitly only the highest-ranking members of groups. Such Byzantines were no less accustomed to escorts than westerners (cf. Ch. 6.3). No patrician or protospatharios traveled without a staff or retinue which would have included officers holding lower dignities. Subordinates usually appear because they were entrusted with some specific task mentioned by our records.37 That so few individuals can be identified as staff personnel confirms the selectivity of the sources' identification of travelers 32 A total of 92 pilgrims, of whom 82 (89 percent) were clergy, and only 10 laymen, of whom four took religious vows in the course of their pilgrimage. At the time of travel, three were bishops, one was a priest, three were deacons. There are other priests, but it is unclear when they were ordained. 33 Missionaries: eight clergy; one bosilikos nnthropos who is presumably a layman. In its technical sense, the term usually designates military men; it is perhaps being used here in the broader sense ofsenior associates of the emperor: cf. A. Kazhdan, ODB i: 266 and Philotheus, Kletoralorjicm, 215.6-8, where the phrasing seems to exclude the churchmen mentioned in the same breath. 34 Deacons: three; priests, including two abbots and some patriarchal officials: nine; archbishop, metropolitans and bishops: twenty-two; monks: eighteen, plus two more abbots. To put the monastic envoys in social perspective, there is one further case ofa probable monkwho was a former patrician: Pros. "Michael 1." 3 5 Twelve patricians and three imperial princesses, of 104 (14 percent) laymen traveling on official business. Patricians and the other ranks are discussed by Oikonomides 1972, 294-8 and Winkelmann 1985, 29-68 who emphasizes (68) that aside from the very highestdignities, office retained a determinant influence in this, the formative period of the middle Byzantine state aristocracy. 36 Thirty-nine held the rank of spatharios or higher. See Pros, for details. 37 E.g. Stephanus 4. 226 X Figure 8.1, "(Mother of God, help Thy servant) /+ Theodosius, Patrician, Imperial Protospatharios and Chartoularios of the Vestiarion." The lead seal bearing this Greek legend was once attached to a document issued by Theodosius Baboutzikos, a close relative of Byzantine emperor Theophilus. This side bears his name and title. Theodosius headed the "imperial" Vestiarion, a bureau which stored precious objects, ingots, and cash, and whose arsenal supplied war materiel for sea and land expeditions. In 840, Theodosius went on a lengthy embassy to Venice and the Frankish empire to negotiate alliances and military support (R455). Archaeologists discovered this seal at Haithabu in 1966; it shows that Theodosius was in communication with the Jutland emporium. Could he have been recruiting Viking mercenaries for the beleaguered Byzantine forces? Courtesy of the Archäologisches Landesmuseum der Christian-Albrechts-Universität, Schloss Gottorf, Schleswig. and delineates their distortion.38 Even among the types of travelers most likely to merit mention, it was mainly only those of the very highest status who were in fact mentioned, as in the west. One can begin to imagine the extent to which our informants ignored lower-status travelers of less-favored types. 4. Under way So, the bulk of eastern Christian travelers moved on their own initiative, as pilgrims, or on state business. A few combined elements of both types, traveling as envoys of the Christian church of Palestine. All eastern pilgrims aimed for Rome, 38 In addition to preceding note, Anons. 249-50 and 279-80. 227 easterners heading west under way most were religious, and most of these churchmen were monks. Afair number of churchmen also traveled on state business, but laymen were more numerous there. A final subgroup of Christian travelers partially overlaps with both of these broad categories, and partially escapes them. Whether pilgrims, envoys of some sort, or traveling on some other purpose, forty-nine travelers to the west started out from the former Byzantine territories of the Caliphate. About three quarters of them were Christian.30 What more do these individuals and small groups tel] us about the broader issues of travel and communications? Western travelers showed a rather wide age span and some relatively advanced ages in a health regime which did not make for old bones. The canonical rules on allowable ages of ordination seem to have been followed, at least for priests and bishops. In Byzantium, the situation differed. Notwithstanding scholars' frequent presumption that Byzantine churchmen's careers respected the canons about legitimate age - thirty for priests and bishops, twenty-five for deacons -there is real room for doubt on this score, at least for the eighth and ninth centuries.40 Our data deepen that doubt for priestly ordinations at Constantinople and in Georgia.41 For Byzantine career patterns, clearly documented ages must be our guide. On the other hand, since the ninth-century Roman church seems to have respected canonical norms for priests and bishops, the numerous Byzantine pilgrims who received their ordination in Rome shed some light on the age of travelers. The prosopographical evidence from Jerusalem equally suggests that its patriarch also kept to the canons, so hagiopolitan ordinations can be used with greater confidence.42 Overall, age can be addressed with varying degrees of precision for about forty-five (13 percent) of all eastern travelers. For once, Byzantine data are superior to western. For some, we know only that they were "young" or "old," and we have already seen that for one biographer at least, forty-five qualifies as "good old 39 12were not, ifHarun ibn Yahya is reckoned a Christian. 40 For the ages defined by the Quinisext council, see above, Ch. 6.3, My sample entirely confirms the skepticism ofV. Grumel, 1934: 352. A fairly clear example conies from the V. Iocmnis Psichaitae (EHG 89 6), 3,108.8-13: Patriarch Tarasius had wanted to consecrate two brothers as priests, but their abbot finally persuaded him to observe the canonical ages and consecrate the younger one as a deacon, and the older one as a priest. That an abbot had to prevail upon the patriarch to observe the canonical age is telling. 41 The career ofMichael of Synada suggests observance, since he was between thirty-eight and forty-one when he became priest and bishop; Joseph the Hymnographer, on the other hand, would have been between 18 and 25, while the priestly ordination of Constantine-Cyril certainly disregarded the canonical age: see Ch. 7.2. In Georgia, Hilarion seems to have been twenty-three or twenty-nine years old. See Pros, for details. 42 All three cases we can tes t suggest pries tly ordination between thirty-three and thirty-eight/forty-one: Michael 3, Theodoras 5; Theophanes 1. age. "43 But for another forty-five trips involving twenty-six individuals, we can be much more specific, as Table 8.7 reveals. The travelers' age span is, again, very wide. Table 8.8 lays out the relative frequency of different ages among them. As we might have expected, more than half the travelers were in the prime of life, in their 20s, 30s, and 40s. Three women seem to have been in their teens; for two, marriage motivated the travel and explains the age. Pilgrims, immigrants and ambassadors run the age spectrum; the only envoy who may have been twenty was an exceptional case, since Joseph the Hymnographer was charged with evading imperial security and conveying a dissident message to Rome. Constantine-Cyril looks a little on the young side when he was sent on an embassy to Samarra, but he was probably not the leader of the delegation. Young people under thirty comprise a third ofthe known cases, suggesting that the world on the move was largely a younger one. Yet what surprises again is just how late in life some Byzantines undertook lengthy trips within the Mediterranean basin. Elias the Younger, an inveterate traveler, was eighty when he answered the imperial summons and headed for Constantinople. His death on the road occasions little wonder. But nearly a fifth of these trips occurred when the travelers were in their fifties or even their sixties. This finding entirely confirms the scarcer - and independent - evidence for western travelers who headed across the Mediterranean at ages which, by early medieval standards, appear advanced. The age of some eastern and western travelers also hints, in its own way, that the travel infrastructures spanning the early medieval Mediterranean were less rudimentary than is sometimes assumed. Looking at the easterners' broader geographic horizons strengthens the impression of long-distance geographic mobility. For such travel seems to have been a fairly typical element in the life experience of many. More than one in twenty eastern travelers made multiple trips across the breadth of the Mediterranean world, and they did so across the entire period under review.44 Two even attempted the trip thrice. With few exceptions, these men traveled on official business of some sort.45 The sole pilgrim who went to Rome twice was also acting as a merchant; he was due to make a third trip under patriarchal and imperial auspices.46 The predominance of multiple trips by officials and ambassadors sheds light on the inner workings of a medieval government, for it makes it likely that the Byzantine 43 E.g. "young": Anon. 6, Daniel 2, Drogus; "old": Anon. 273; "Good old age": see above Ch. 7.3, on Gregory Dekapolite. Por this hagiographica] topos, Browning 1981, 123. 44 21/340; 6 percent. See Table 8.9. 45 Theognostus went firstas a refugee and private representative ofthe deposed Patriarch Ignatius; his second trip was as the restored patriarch's ambassador. Epiphanius and Euphemianus traveled as Theodore Studite's private emissaries to Rome. 46 Basilius4. 228 229 TABLE 8.7 Age at which easterners undertook travels Age Name Journey Comment 9/26 Stephanus 1 Sicily-Rome * c. 12? Anna Constantinople-Pavia c. 12? Evanthia Constantinople-Benevento 12 Elias 1 Sicily-Africa slave, first time 1 5 or less Joseph 1 Sicily-Greece C. 15 Elias 1 Sicily-Africa slave, second time 18/30 Joseph 1 Constantinople-Rome c. 19 Clemens Byzantium-Moravia * * c. 19 Elias 2 Taormina-Rome 19/23 Sergius 1 Palermo-Constantinople * 20 + Methodius 1 Sicily-Constantinople * 23/9 Hilarion Georgia-Jerusalem 24/9 Constantine 3 Constantinople-Samarra 24/5 + Methodius 1 Constantinople-Rome * 25/6 + Naum Constantinople-Moravia * 26 Gregorius 5 Seleucia-Jerusalem 30 + Anon. 293 Moravia-Constantinople imperial summons* 30 + Basilius 4 Rome-Constantinople 30 + Basilius 5 Rome-Constantinople * 30 + Methodius 1 Rome-Constantinople * 30 + Petrus 12 Rome-Constantinople * 30 + Zosimas Rome-Constantinople * 30/6 Hilarion Jerusalem-Georgia-Constantinople 33/5 Theophanes1 Jerusalem-Constantinople-Rome 34/6 Gregorius 9 Ephesus-Rome 35/6 Constantinus 6 Constantinople-Moravia 36 + Methodius 3 Constantinople-Moravia 36-8 Theodorus 5 Jerusalem-Constantinople-Rome 37 Elias 2 Reggio-Patras 38 Gregorius 5 Jerusalem-Rome 40s-early 50s? Hilarion Constantinople-Rome 42/8 Clemens Moravia-Bulgaria 42/55 Georgius 8 Palestine-Africa-Spain * * * 48/9? Naum Moravia-Bulgaria 48/52 Michael 3 jerusalem-Constantinople-Rome 54/7 + Methodius 3 Moravia-Constantinople imperial summons c, 55 Elias 1 Africa-Jerusalem c. 57 Michael 2 Constantinople-Salz first trip c. 57 Elias 1 Africa-Italy c. 57/8 Gregorius 5 Rome-Constantinople c. 58 Elias 1 Taormina-Greece c. 62/8 Elias 1 Reggio-Rome 65 + Gaudiosus Messina-Constantinople 66 + Michael 2 Constantinople-Aachen second trip 80 Elias 1 Regglo-Constantinople imperial summons; dies Notes: *Based on Roman canonical age. **AssumingClementand Naum were ordained at Rome in 867; advanced age appears to be confirmed by the fact that they were not sold as slaves; cf. below, pp. 2 5 of. ***Assuming his Spanish biographer understood the term "adolescent" in the same way as Isidore of Seville. TABLE S.8 Ages of travelers broken down by life decade Life decade Number Percentage 80s 1 2 60s 3 7 50s 6 13 40s 5 11 30s 14 31 20s 6 13 Teens 9 20 Under ten? 1 2 Total 45 100 TABLE 8.9 Eastern travelers making multiple journeys Name Date of first journey Destination Purpose Number of journeys made Constantinus 1 before 708 Rome; Constantinople Sergius 1 before 708 Constantinople Gaudiosus after 721 Italy; Constantinople Eutychius 726/8 Constantinople lohanne5 7 752 Rome, Pavia Georgius 5 756 Rome, Marseilles, Benevento Constantinus 4 784 Rome Michael 2 803 Franks Paulus 5 808 Venice Epiphanius 3 808/9 Rome Arsaphius 810 Aachen Euphemianus 817-18 Rome Theodorus 6 824 Franks Zacharias 3 843/7 Rome Basilius 4 855/7 Rome Lazarus 2 855/8 Rome Theognostus 2 861/2 Rome Methodius 3 863 Moravia, Constantinople Petn-is11 866 Rome Nicetas4 869 Bari Ali c. 898? Lucca; Samarra pilgrimage-immigration 2? immigration 2? immigration-official business 2 official 2? diplomacy 3 diplomacy 2? diplomacy 2 diplomacy 2 official 2? diplomacy 2 diplomacy 2 diplomacy 2 diplomacy 2 diplomacy 3 pilgrim/merchant 2 diplomacy 2 envoy/refugee 2 missionary 2 diplomacy 2 official 2 slave; envoy 2 231 administration - quite unlike its Frankish counterpart - attached importance to regional expertise and experience in its foreign missions. The embattled abbot of Stoudios may have been lifting a page from state practice when he twice dispatched the same individuals as his private representatives to the pope. It is also exceedingly interesting to see, in the first quarter of the ninth century, individuals repeating such lengthy travel who were not professional travelers and who lacked the transport resources of the Byzantine state. Still more interesting in this respect is the appearance in the next generation of a freelance traveler, Basilius 4, who made the trip from east to west twice under his own auspices. Might this be yet another indication that travel between east and west was growing more common? Certainly there is no hint in the telling of Basil's story at the Council of Constantinople that his long-distance voyages were any more unusual at that date than those of the "10,000" homonymous compatriots the Roman monk Peter claimed. This makes us wonder how typical travel may have been in the lives of those travelers we can detect. Most eastern travelers (230: 68 percent) appear in the sources only in connection with one trip, so it is impossible to judge whether this trip was unique in their life experience. Nonetheless, of the remaining 32 percent about whom we know more, it is striking that three out of four seem to have had some kind of documented travel experience beyond their main trip to the west.47 Outside the empire, we have already seen that a good number of the broader group of travelers had been in the Arab world, independently of their western voyage. Fewer travelers (seven) visited the other main, extra-Byzantine region, the Slavic world north and west of Constantinople, and all are confined to the ninth century. Before their mission to the Danube Slavs, Constantine and Methodius had been to Cherson and beyond. Clement, Naum, and, apparently, Agatho, bishop of "Moravoi," all traveled in the upper Danube and among the Balkan Slavs. Where in the west were the travelers going? That envoys constitute the single largest group means that courts attracted the lion's share; north of the Alps, the predominance of courts was even more complete. Geographically speaking, Rome attracted the most: of 340, more than half (179: 53 percent) had Rome as their main destination; a number of others, for instance, Byzantine envoys to the Lombard or Frankish kings, had Rome as a secondary destination. Interrupted trips are of obvious interest. Thirteen or fourteen trips involving eighteen or nineteen eastern travelers seem never to have been completed. They mostly involved ambassadors or envoys of some sort. Whether or notone traveler's 47 Twenty-seven of the better-documented travelers offer no evidence of other travel, i.e. 25 percentofthe latter group. Three quarters do suggest some further kind of travel. This takes into account the data on birthplace and residence discussed above. TABLE 8.10 Eastern travelers: main destinations Region Number Percentage Italy (including Istria) 237 70 Europe, N. of Alps 75 22 Other 28 8 Arab world 18 Constantinople 2 Dalmatia 4 Sardinia 3 Uncertain 1 Total 340 100 trip really was interrupted is unclear, and three or four travelers died in the course of their trip.48 Only three of the twelve aborted trips reflect difficulties with the infrastructure of travel. Elias the Younger's first voyage into slavery was circumvented by the Byzantine navy; Peter, metropolitan of Sardes, en route to the papal court at Rome, drowned in the Adriatic when the new Byzantine warship in which he was sailing sank (a second dromon conveying the other ambassadors survived) ,49 The nature of the accident was presumably banal in ninth-century shipping. Joseph the Hymnographer's interrupted voyage westward is more interesting because it is a rare case among the 340 eastern travelers which directly supports the conventional wisdom that an Arab menace disrupted long-distance communications. While he was attempting to make a clandestine trip to Rome, Arab pirates fell upon his ship, captured him and held him as prisoner in Crete until he was ransomed (R436). Like the capture of a merchant skipper who had set sail from Rome, Joseph's case signals the critical repercussions on the travel infrastructure of the Arab conquest of Crete (R729). The other interrupted trips show no connection to structural difficulties of transport. The four travelers who accompanied Michael Syncellus as ambassadors of the church of Jerusalem to Rome were supposedly arrested by the iconoclast authorities in Constantinople en route west. More careful scrutiny has shown that they interrupted their westward voyage in the capital before iconoclasm was restored, that is, they lingered voluntarily at Constantinople, and they exposed 48 Although scholars have assumed that Constantinus4 never reached Rome, other explanations of the peculiar circumstances of his trip are possible; see Pros. For the deaths, see below. 49 R431; the same event interrupted the profitable voyage of the African slave merchant Anon. 218; R573. Donnolo had the good fortune to escape from the convoy of slaves taking him to the Arab world: R804. themselves to criticism for doing so.50 In any event, other travel to the west is unambiguously documented in precisely these years, so that the hagiopolitans' decision to delay en route cannot be attributed to the infrastructure. Epiphanius never reached Rome on his second trip there, because he was arrested by the imperial police for attempting to smuggle anti-imperial documents to the pope.51 The case of the remaining six individuals (i.e., two trips) underscores the point. Imperial command halted the legation led by Theodore of Laodicea and Zachary of Chalcedon, when the embassy's sponsors were toppled from power.52 In other words, there is explicit evidence of structural impediments to communications in the movements of our 340 travelers, but it is confined to the Arab conquest of Crete. The known mortality of eastern travelers to the west was also low. Of the three or four deaths (4 of 340:1.2 percent) en route, thatofTheodosiusBaboutzikosis uncertain; Peter of Sardes, we have already seen, perished in a shipwreck. The cause of Lazarus' death is unknown, although he was not a young man at the time of this, his second embassy to Rome, twenty-five years after his first voyage. He was in any case not lost at sea, since his body was returned to Constantinople for burial. Elias the Younger, who was over eighty when he died on the road at Thessalonica, presumably succumbed to the afflictions of old age; he had had a strong premonition that he would never return home.53 Although in both cases the numbers are low, easterners appear to have suffered less mortality than westerners (1.2 percentas opposed to 3 percentwestern-ers). One underlying reason for the difference may lie in the distinctively lethal disease environment that northerners identified in the Mediterranean world. Natives of the region would naturally display stronger resistance to the local infections; since the bulk of eastern travelers never went beyond Italy, they remained in their native disease pool, unlike the transalpine travelers who ventured into the Mediterranean.54 Even assuming the worst for the uncertain cases, we are left with a handful (six) of interrupted travels that can be ascribed to disruptions of the infrastructure.55 They scarcely justify the conclusion that the decay of shipping, collapsed infrastructures, or enemy attack precluded Mediterranean travel. In this the eastern travelers' experience, again, reinforces the inferences drawn from the western wanderers. 50 See Pros. "Michael 3." 51 Pros. "Epiphanius 3." 52 R5 66. The projected voyage westward of a Byzantine bride also came to grief, presumably for political reasons: Pros. "Anon. 227." 53 Pros. "Theodosius3"; "Lazarus 2"; "Elias 1." 54 See above, Ch. 6.4; on migration and mortality, see the reflections ofCurtin 1989, xiii-xvi. 55 Counting the uncertain cases of Constantinus 2, Theodosius 3, Lazarus 2, as well as the certain cases of Joseph and Peter ofSardes, and addingto themElias 1 and the slave merchant Anon. 216. 234 In the eighth and ninth centuries, detailed accounts of individual travelers' routes that specify the key stops along the way are rare. This has left the essential question of change or continuity in route structures largely unanswerable. The 240 eastern travelers help to overcome this difficulty, for the cumulative force of their travels' scattered detail points to changing routes. Thus, eighth-century travelers sailed to the west via Sicily (e.g. R206). Late in the ninth century, some easterners still sailed through the strait of Messina and around southern Greece into the Aegean (R729). But about 830, we first hear of a pilgrim approaching the west through Corinth (R422). A generation later, one pilgrim followed the Danube west toward Rome, while another traveled from Jerusalem to Rome via Tripoli in Africa (R622; R519). And we have already seen that the Adriatic drowned a Byzantine ambassador en route to Rome. Do these new coordinates imply new routes? Synchronous parallels in the movements of western travelers suggest, once again, that we are seeing changes in reality, not simultaneous changes in the pattern of Greek and Latin documentation. This then is the second main group of travelers. Most were Byzantines, at least by culture. Whatever their language, one in seven set out for the west from the Caliphate. Even more striking is what our travelers' movements suggest about communications with the Caliphate, since over one in five of the whole group had been in the Arab world at some time in their life. Half of all easterners show up in Rome. The resulting Greek settlement in Rome was quite cosmopolitan. One wonders whether it is characteristic of an underdeveloped economy with a powerful capital that relations with neighboring regions should be about as intensive as with the capital. This is certainly the lesson suggested by a quite independent phenomenon, the geographic patterns of official documents' circulation inside the Byzantine empire.56 The convergence may hint at deeply buried connections between economic realities and administrative ones. At the same time, well over half of the Greek residents of Rome are of undetermined origin. Many, perhaps most of them, could conceivably have come from Byzantine Italy or even Rome itself.57 These men and women are exceptional in many regards, beginning with the extent to which they are documented. The several hundred travelers nevertheless As reconstituted by surviving deposits of lead seals; in a given province about 10 percentof seals come from the capital, and another 10 percent from provinces neighboring the one where the seal was found: CheynetandMorrisson 1990,116; cf. 111 and 113. 57 Note that the hypothetical total of certain Italo-Greeks (nine) + uncertain cases (sixty-two) comes to 71 percent of the total of 100 monks documented at Rome, which is not far from the 80 percent of seals which originate in the province of discovery, according to Cheynet and Morrisson 1990. 235 did not invent their own travel infrastructures. From the aggregate patterns of their individual movements, we can begin to recover the routes, means and rhythms of long-distance travel in the early medieval Mediterranean. Even more importantly, they begin to reveal for the first time how these deeper structures changed over time. But first we need to take a closer look at some more meagerly documented or simply smaller groups of travelers: traders, slaves, defectors, and their ilk. Traders, slaves, and exiles Of their own free will, early medieval people crossed the Mediterranean in order to trade or to raid as well as to negotiate or pray. Some sought careers in the ancient capital on the Bosporus; others immigrated to Europe. Others still traveled under dire compulsion, as exiles, refugees, hostages, or as slaves. In most cases, their numbers probably were much greater than the high, mighty, and holy who dominate surviving sources. But people of low, suspect, or ambiguous social status appear more rarely and less completely in the early medieval written record. A fair number are of uncertain geographic origin - shippers and traders, for instance, attract only passing mention in sources treating the more important people who traveled aboard their vessels. Where was the home berth of the six slavers that Bernard the Frank saw anchored in the harbor of Taranto in 867? We may surmise, but we do not really know; nor can we do more than speculate on the ethnicity of the merchant skipper with whom St. Blaise sailed from Rome to Greece. The smaller numbers of these kinds of travelers make it more practical to treat together easterners and westerners. We will begin, again, with a few thumbnail sketches and meet some merchants, a slave, and a refugee. We will also test the anecdotes against the broader group of like travelers, combining this time the scarcer data of east and west. Next we will cast a glance at the travelers who are so few, or so low, as to be almost invisible: the immigrants, the sailors, fishermen, and wanderers who could not pass themselves off as pilgrims. Finally, we will turn to a rich source of information about all aspects of the early Middle Ages which has passed almost entirely unnoticed: early medieval fiction. Imaginary travelers shed real light on early medieval travel and communications. 1. Traders, slaves, and politicos Traders In this era, merchants were travelers almost by definition. Alas, theycountamong those ofwhom the sources speak the least, so far as specific individuals traveling 237