Beyond the myth of the Cilician Gates. The ancient road network of Central and Southern Cappadocia
p. 179-200
Abstract
Central and southern Cappadocia could very well have served as a major hub within the context of the ancient communication system of Anatolia. The whole district was, indeed, passed through by a series of routes, which effectively linked east and west, as well as south and north. The ‘southern’ road leading from Iconium/Konya to Podandos/Pozantı and the Cilician Gates, running across the Çakıt Suyu valley, ensured smooth and easy communication between the Anatolian plateau and the Mediterranean shores of Cilicia. The ‘northern’ highway, from Iconium/Konya to Colonia Archelais/Aksaray and Mazaka/Caesarea/ Kayseri, connected the inner land to the eastern boundary of Anatolia and especially to the Euphrates district. Another historically important road from Mazaka/Caesarea/Kayseri to Podandos/Pozantı and the Cilician Gates joined the former route to the latter, closing that sort of wide and ideal ‘road triangle’ – whose vertexes being Konya, Kayseri and Pozantı – which has really characterized that frontier territory, and which this paper tries to describe.
Full text
Introduction
1Most of the topographical researches published up to the present day which have dealt with the ancient road network of Cappadocia seem lacking in terms of a systematic and careful confrontation with the morphology of the landscape, and they do not always consider the actual practicability of the various routes being hypothesised. To paraphrase Luciano Bosio, possible networks of ancient roads were constructed without adding an adequate quantity of “work done by the feet” to the “work done by the mind”.
2What is more, the customary practice was to seek to place a toponym and then, in accordance with its position, to formulate hypotheses regarding the possible route of the road which could have led to the spot in question. It is unfortunate that, in the majority of cases, identifications of this kind were not supported by any convincing archaeological proof, nor even by evidence of a rather less concrete sort. Instead they were suggested solely on the basis of a toponomastic assonance between the ancient name as reported in the sources and that of the present-day village being proposed in the hypothesis. The upshot of all this was that, from time to time, roads that had a major relevance in strategic and military senses were put forward as having passed across land which cannot in any way be construed as being easily accessible, and in some instances the terrain would have been extremely dangerous, especially when it was being crossed by an army.
3The following analytic results, which are part of a wider PhD research conducted at the Universities of Padua (Italy) and Durham (UK), have involved both ‘deskwork’ and work more appropriately carried out in the field, with a systematic programme of surveys targeted towards a search for the so-called ‘earth truth’. It is thanks to this that the opportunity has arisen to put forward reconstructive hypotheses which are fairer and more reasonable. Indeed, what appears to emerge through this archaeological and topographic analysis is a road network that seems to be in keeping both with the archaeological data (classical and post-classical) and, most importantly of all, with the context of the territory and its specific morphological character, which certainly needs to play a key role in any statement of the options available in terms of possible routes.
4Central Anatolia was traversed by a series of roads which effectively linked east and west, south and north. Moreover, if we limit ourselves to a consideration of the principal roads mentioned mainly in Greek and Roman literary sources, it is possible to visualise a sort of ‘ideal triangle’, with one angle at Iconium/Konya, another at Mazaka/ Caesarea/Kayseri, and the third at Podandos/Pozantı (fig. 1).
Fig. 1: The ‘ideal’ road triangle in central Anatolia

The lines highlighted in white correspond to the road routes considered in this paper
5As far as this paper is concerned1, the focus will fall only on the Cappadocian sections of those roads.
Along the Uzun Yol
6As far as the northern side of this triangle is concerned – our particular attention here is on the Cappadocian stretch between Colonia Archelais/ Aksaray and Kayseri – it can be affirmed that this same road, which both Strabo2 and Pliny the Elder3 seem to refer to, was noteworthy for the unquestionable ease with which one could travel its whole length. Besides that, it has a certain historicity, linked with its tradition of being in constant and systematic use – perhaps from Achaemenid times right up to the present day4.
7This in fact is the Uzun Yol, the ‘Long Road’ (fig. 2) that represented the principal caravan route which, in the Seljuk period, not only linked Aksaray with Kayseri, but was also one of the three roads that formed the main axes of the road system within Anatolia in mediaeval times5. Its importance appears to be fully confirmed by the series of caravanserais which were dotted around the directrix: the Ağzıkara Han,6 the Öresin Han7, the Alay Han (fig. 3)8, the urban caravanserai at Nevşehir9, and the Sarı Han (fig. 4)10, five kilometres east of Avanos. Moreover, there was another caravanserai in the vicinity of Acıgöl11 and the Sünnetli Han (fig. 5) to the west of Kayseri, as recorded by the Arab historian Ibn Bībī12. This latter caravanserai might be placed, even simply on a toponymic basis, at the small peak of Sünnetli Tepe13, although it needs to be said that along its slopes it has been possible to identify only ceramics from the Iron Age and the Greek and Roman periods14.
Fig. 2: The Uzun Yol from Garsaura/ Aksaray to Ouénasa/Avanos

Fig. 3: The caravanserai of Alay Han

Fig. 4: The caravanserai of Sarı Han

Fig. 5: The Uzun Yol between Avanos and Mazaka

8Now, in an attempt to produce an on-the-ground reconstruction of the possible route taken by that road, it is clear that the caravanserais we have cited above need to be considered as important territorial markers. However, in this case at least, there are also other elements which might prove to be of use; these elements being certain traces revealed in a satellite image which, in effect, is a ‘photograph’ of that area of central Cappadocia as it was at the beginning of the 1960s.
9To be more precise, we are referring here to a Corona satellite image in which it is possible to distinguish (albeit with the necessary caution which the case requires) what is generally defined in the Anglophone archaeological community as a hollow way: that is to say, a “broad and shallow linear depression in the landscape, thought to be formed by the continuous passage of human and animal traffic”15. Indeed, as one can observe fairly clearly (fig. 6)16, the modern road, in the immediate vicinity of the caravanserai of Alay Han, appears – significantly – to be flanked, especially on its northern side, by a trace which seems to present the characteristics which are peculiar to that particular road typology; i.e. there are ‘shadow zones’ discernible which can, with justification,
10be associated with the very nature of the trace, ‘hollowed’ in relation to the land immediately surrounding it. Quite obviously it is not datable, if considered simply in isolation. However, the strict correlation which seems to exist between that road artery and the caravanserai brings one to the thought (at least) that they might possibly be contemporary, if not indeed to the belief that the road itself could be older and the han was constructed some time afterwards.
Fig. 6: Possible hollow way in the vicinity of the caravanserai of Alay Han

detail of the Corona satellite image DS009029040DV165, December 1961
11From our own point of view, what is of interest is the fact that, confronted with a territorial morphology which could offer an infinite number of variations of route, the various traces – from those visible on the Corona image, to the well-beaten road giving access today to Alay Han, and then to the modern highway which, significantly, connects Aksaray with Kayseri – all of them have followed more or less the same route. A route which may and indeed must be traditional precisely by virtue of the fact that roads are, in effect, ‘things of long duration’; things that are perpetuated in time and have no particular reason to be modified (if, as far as this area is concerned, we are permitted to exclude the last decades, as in that instance the change has been brought about by technology and the increases in speed which have resulted from it). In this particular situation it is by no means inappropriate to give due recognition to journeys made in times long past.
12It is not surprising, then, that the entire road is often flanked by mounds and höyüks17 and that architectonic elements which can be dated as being from the Byzantine era have in fact been reused and incorporated into the caravanserais to which we have just referred. Such is the case with the Öresin Han, where several similar elements have been identified as originating from an ecclesiastical building dating from between the 5th and 7th centuries AD.18
13It therefore seems superfluous to hypothesise that the ancient directrix (fig. 2) which led to Kayseri arrived at Acıgöl, proceeded via Tatlarin, and then continued in a long stretch as far as Gülşehir reaching the southern bank of the Halys/Kızılırmak, which it then followed as far as Avanos19. On the basis of the considerations that have been put forward above, it would be far more reasonable to take the view that there must have been a more rapid route which, passing through Acıgöl and Nevşehir, arrived at Avanos (the Greco-Roman and Byzantine Ouénasa)20 where it would have met the road descending from Ancyra/Ankara, which ran parallel with the southern bank of the Halys21.
14In particular, and perhaps in contrast with what occurs today, with the modern road crossing a geomorphologically complex tufaceous area which – maybe for this reason – makes it not particularly easy to travel upon, from Nevşehir the ancient directrix could have run further to the west, crossing the plain of Nar. Here, according to the Scottish-born geologist William John Hamilton22, underground settlements were found. We should also not overlook the fact that from here it would have been possible to reach Ouénasa with ease, passing close to such strong territorial marker as the burial mound at Çeç. This is an imposing funereal monument23 which, after comparing it with similar structures in Commagene, possibly dates from the 1st century BC24. However, what is of interest to our present discussion is the fact that at this tomb it was customary not only to venerate the deceased (who was most likely to have been a Cappadocian ruler), but also Zeus, to whom a sanctuary at Ouénasa was dedicated. The importance of the Ouénasa sanctuary was, as Strabo recounts25, second only to that of the temple at Comana26.
15From Avanos, according to Friedrich Hild’s hypothesis,27 once the main directrix had passed the Sarı Han and the Sünnetli Han, it would have reached the present-day village of Süksün, which stands some sixteen kilometres to the east.
16However, it is best, in our opinion, to make it clear that (up to the present time, at least) archaeological evidence has not been discovered which would lead us to think that the land was used for the passage of a road in ancient times. Moreover, a reconstruction of this sort would seemingly imply that there was some degree of similarity between the route followed by the old (suggested) road and that of the new one – a road which would have approached Mazaka from the west/north-west, passing by Boğazköprü and across an area of stagnant water which lay to the south of the village itself. This same area, as a matter of fact, is described in fairly plain terms in the travel diaries compiled by a number of British and French explorers of the 18th and 19th centuries:28 it was submerged by the waters of the lake which can be seen clearly and unambiguously on a pair of maps29 which date back to the first half of the 19th century (fig. 7). If one then takes also into account the commentary of Vitruvius, who records the presence of a lacus amplus...in itinere quod est inter Mazaka et Tyana/ “a wide lake... beside the road which connects Mazaka and Tyana”,30 it would be reasonable to imagine that this would have been a lacustrine area, or one covered by stagnant water, even in much earlier times.
Fig. 7: Central Cappadocia

detail of The Map of Asia Minor by John Arrowsmith (1844)
17In view of these considerations, one might now be inclined to think of a different directrix between Avanos and Mazaka (fig. 5). In particular, one could hypothesise a route which, beyond Sünnetli Han, curved slightly towards the south-east in the direction of the present-day village of Örenşehir, near to which, significantly, it would have been able to meet the road from Yeşilhisar31 which led to Mazaka. At Örenşehir, in fact, there is a territorial marker in the form of the ruins of a necropolis.
18Within this there is a rectangular tomb measuring twelve metres by seven metres which can be said to possess a certain architectural monumentality. Part of the elevation of the necropolis has also been preserved. The structure dates from the Roman era and it was sacked at some later stage, during the Byzantine period32. According to Hild, it “...gehörte zur Nekropole einer alten Stadt, die unweit östlich beim heutigen Viranşehir [nowadays Örenşehir] lag und auch in byzantinischer Zeit besiedelt war”. It is also of significance that this was the area in which Rott observed the foundations of a church33.
19From there, the road would have crossed the plain, running south of the marshy area34 and reached Karpuzsekisi, at the foot of the north-western slopes of the Erciyes Dağı. There a number of territorial and itinerary markers can be identified, which, as Starr reports on the basis of archaeological evidence, are clearly linked with a necropolis dating from the Byzantine era35.
20Then, as some travellers recall36, having passed Karpuzsekisi a similar route would have followed the little valley formed between the slopes of the Erciyes and those of the Yılanlı Dağ. This route would then have reached Mazaka/Caesarea from the south-west. Such an itinerary might be further (and, in our opinion, more clearly) confirmed by the fact that the ruins of the Greek and Roman city to which we refer have been identified as Eskişehir, which lies a few kilometres south-west of the modern centre of Kayseri37.
From Pozantı to Tyana
21Although they do not allow for an actual geographic contextualization of the topographic references they contain (they are too limited, besides being vague and generic), certain very short extracts by two authors of the Historia Augusta38 nevertheless play a key role in a credible reconstruction of the road network in this area of southern Cappadocia (fig. 8).
Fig. 8: From Pozantı northwards: a missing road?

22We are referring here, in particular, to an episode which relates to the life of Marcus Aurelius. During his return march towards Rome after his ‘peace-making’ military campaigns in the East (176 AD), Marcus Aurelius suffered the sudden loss of his wife Faustina, who, for several years, had accompanied him in all his expeditions. To render her a fitting memorial, the emperor decreed that the status of the village in which she had died (...in radicibus montis Tauri in vico Halalae… / “…At the foot of Mount Taurus, near the village of Halala…”)39 was to be elevated to that of a colony, and that a temple was to be built there in her honour. (Some time later the temple was re- dedicated to Elagabulus: …templum, quod ei sub Tauri radicibus fundaverat maritus…/ “…that temple which her husband had founded in her honour on the slopes of Mount Taurus”)40.
23Now, were it not for the fact that the name taken by this village after it had become a colony41 was recorded in both the Itninerarium Burdigalense and the Itinerarium Antonini42 as one of the stopping stations on the directrix that would have led from Tyana/Kemerhisar43 to the Cilician Gates, evidence of this sort would perhaps have been largely ignored. Or, at least, it would not have been at the centre of a discussion, which is still unresolved, whose origins can be traced back to the end of the 19th century. It is clear, therefore, that only a precise localisation of this colony will allow us to establish, with any rationality, the possible route to which the itineraries refer.
24Here, however, the problems really begin. The sole piece of topographic data recoverable from the sources is a very generic indication that Halala was at the foot of Mount Taurus.
25The first reference to this question can be found in the pages of William Ramsay’s Historical Geography of Asia Minor44. Without providing details of the ‘proofs’ on which his hypothesis is based, Ramsay initially opts to identify the colony with the present-day Başmakçı, a small village which is approximately 25 kilometres south-east of Tyana, at around the point where the Kozlica Deresi and the Kırkgeçit Deresi valleys merge.
26The same scholar then returned later to the question and, as we shall shortly see, proposed a new hypothesis. The binomial Faustinopolis-Başmakçı that he suggested has not only continued to enjoy a certain popularity, even in relatively recent years45, but it has also been held valid and worthy of consideration by the majority of those who have joined the discussion.
27From the late 1930s, in fact, a number of pieces of evidence of an epigraphic sort drawn from within the area relevant to the village of Başmakçı were used to corroborate and confirm the “...identification of Faustinopolis with the modern-day Pasmakci ...” which, at the same time, has allowed for an improvement of “...knowledge ...of the route taken by the ancient road which gave access to Cappadocia”. More specifically, confirmation in this sense has derived from the discovery of a milestone “...at Bağderesi, 6 km. north of Eminlik, on the Bor-Paşmakci road, which is believed to follow the traces of the ancient Tyana- Faustinopolis-Portae Ciliciae road...” and from “...another fragment of a milestone ...” at the village of Eminlik46.
28It should however be noted that one of Jacopi’s contemporaries, Louis Robert, had already expressed a certain amount of reserve as to the possibility of tracing the site of the ancient Faustinopolis purely on the basis of the two milestones discovered “entre Tyane et Paşmakci”47 which, moreover, in our point of view, did no more than point towards the existence of a road directrix between Tyana and Eminlik.
29However, it appears that not even Robert wished to deny the identification outright, and he put forward instead the idea that a favourable clue of a strictly epigraphic-onomastic kind might in some way be derived from an analysis of another two inscriptions discovered by Jacopi, one at Başmakçı and one at the neighbouring village of İmrahor48: “...il paraît frappant que les deux inscriptions mentionnent précisément deux personnages portant les tria nomina et ne tenant pas leur droit de cité romaine de quelque empereur; cela conviendrait très bien à la Colonia Faustiniana”49.
30To these epigraphic finds, then, we need to add another, found towards the middle of the 1960s in a field close to Başmakçı village, whose inscription bears a dedication to the emperor Marcus Aurelius Gordianus (Gordian III) from the decemviri and citizens of Col. Faustinopolit[a]|norum. Ballance, precisely on the basis of these elements, comments “that the city at Başmakçı was the Roman colony of Faustinopolis has long been suspected, and the following dedication merely makes the suspicion a certainty”50.
31It has to be said that the reference to the colony founded by Marcus Aurelius really seems quite clear in this case; however, other observations leave one undoubtedly perplexed. Indeed, considering the morphological characteristics of this area of southern Cappadocia, the only possible road communication between the colony and the area of the Cilician Gates would in fact have been through the Kırkgeçit Deresi valley (significantly, the river with the forty bends)51 (fig. 9). This most certainly would have involved a route that not only must have been winding and dangerous in the sense that it followed the river channel, but it would also have been a difficult road to negotiate because of the numerous and continuous variations in altitude.
Fig. 9: Stretches of the Kırkgeçit Deresi valley

32Such a theory conjures up visions of a sheep-track rather than a road intended for practical and efficient use by an army52.
33It was no coincidence that this was the very reason which led Ramsay to reconsider his initial hypothesis and suggest a different location for Faustinopolis: “Previously, it was supposed that the Roman Road from the Pylae Ciliciae to Tyana turned off towards N. N. W. at Takhta -Keupreu, passing close to Pashmakji, near which are some ruins. In Hist. Geogr. p. 351 ff., Faustinopolis- Halala was placed on this path somewhere near Pashmakji. But this path was not a Roman Road: it is only practicable for horses”53. Specifically, having first oriented and placed the colony in the neighbourhood of Ulukışla54, Ramsay later held that on the evidence of a series of milestones that he himself discovered in the area, it would probably be more correct to place it between the villages of Beyağıl and Porsuk55.
34However, even if, in the 1990s, renewed interest in these studies of Anatolian topography sought to confirm the hypothesis of a more logical location of the colony of Faustina along the Çakıt Suyu valley close to Porsuk56, in more recent years a new and not entirely convincing theory has been put forward which, once again, appears to be based solely on epigraphic evidence. On the one hand recognition is given to the real and practical difficulty of movement within the Kırkgeçit Deresi valley, which, above all, “...n’était pas viable en hiver, ni pour un convoi lourd même pendant la belle saison...” and then there is the consequent admission that the main directrix must have run through the valley of Çakıt Suyu before then curving towards the north in the direction of Tyana and crossing either the Çaykavak pass or the nearer pass of Avluk.
35Yet despite all this, there is still an insistence that Faustinopolis needs to be identified as Başmakçı57.
36A proposal of this kind, based on the idea that between the Cilician Gates and Tyana two separate main road arteries might have co-existed, does not appear to resolve the question in any way whatsoever, given that in these terms not only would the directrix mentioned in the ancient itinerary sources – upon which, we must stress, Marcus Aurelius’s army marched when returning from the East – have to be judged thoroughly impractical both winter and summer alike, but it would also mean that a Roman colony was kept detached from the ‘normal’ road network of this area. Such a resolution actually leaves us somewhat perplexed.
37In order to move towards a solution which, at the very least, will be plausible, there would seem to be no harm in approaching the problem from a diametrically opposed starting point to the one traditionally adopted. Rather than seeking first to locate the colony and then to trace the road which connected it, it would be more appropriate, in this specific case, to reverse the order of the two tasks. It would be better if, first of all, and with the aid of literary, epigraphic and archaeological sources, we were to try to identify the various possible routes for major roads within the area, whilst evaluating, with the evidence obtained through a systematic and essential reconnaissance of the land, the actual possibility of whether a road could have followed a particular line. Then, at a subsequent stage, and after consideration of all the data at our disposal, we would find ourselves in a stronger position to put forward a hypothesis as to where the colony was likely to have been situated.
38An analysis of this type would reasonably lead us to the conclusion that the only possible alternative after leaving the plain of Pozantı would be to follow the natural corridor provided by the Çakıt Suyu valley, which offers a comfortable route, and, furthermore, one which had a definite tradition behind it. From there (fig. 10) the road would have been able to follow the course of the river, most likely keeping to the left bank, as is borne out by two milestones, one attributed to Caracalla and the other to Constantinus and Licinius, found a few kilometres north of Pozantı58.
Fig. 10: The road linking the Çakıt Suyu valley and Tyana

The stars roughly mark the places where the milestones were found
39Still keeping to the left bank of the river, after having passed the Kırkgeçit Deresi59 and the ‘tricky’ stretch between the narrow and sheer walls of a gorge60, the road would have reached Çiftehan, whose thermal springs, still in use today, might in some way back up the report of the Revd. E.J. Davis in relation to the presence of a “bath of Roman construction …with vaulted roofs of masonry … somewhat ruined”61. Quite apart from that, we can also call to mind the toponym of Aquis calidis applied to a place in that area and registered in the Tabula Peutingeriana which, as Miller suggests, would have been located between Tyana and Paduando (that is to say Kemerhisar and Pozantı)62.
40After this, the directrix would have turned towards the west. It would be reasonable to suppose that it continued along the left bank of the Çakıt Suyu – i.e. the bank on the sunny side – where some habitual use which we might define ‘traditional’ if not exactly ancient can be attested by means of a series of traces of a road which have been observed about 100 metres to the north of the modern highway and more or less in parallel with it63. These traces stretch for a total distance of some seven kilometres between the modern-day villages of Hasangazi and Porsuk64 (fig. 11)
Fig. 11: Traces of the Nineteenth century road along the Çakıt Suyu valley

Google Earth
41Significantly, it is precisely in the neighbourhood of the village of Porsuk that an archaeological excavation which is still in progress, conducted by a French team, has brought to light one of the most important sites in the Tyana area and in southern Cappadocia as a whole. It is here, in fact, that we find Zeyve Höyük, on a wide tabular plateau at the extreme east of the Tapor Dağı mountain chain which serves as a divide between the Çakıt Suyu and the Kılan Deresi. The site enjoyed a continuity of life from the Hittite period until roughly the Imperial Roman era65. It takes a certain importance upon itself not only because of the archaeological evidence it yields, but above all (in our view) on account of its topographic location. Because of its high position vis-à-vis the surrounding plain, the site conclusively dominates the valley just as it did in antiquity, and it must also have played a role in governing the access and the control of traffic to and from the Cilician Gates66. Furthermore, it would have fulfilled an identical role with regard to the Bolgar Maden silver mines, which are approximately 8 km (as the crow flies) south of Zeyve67. We also need to bear in mind that there would certainly have been a byway which led off from the principal directrix running along the Çakıt Suyu valley. This minor road68 would have followed the watercourse which irrigates the south side of the höyük and then curved to the south-west in the direction of Cybistra/ Ereğli, passing by the prominent Hittite rock-relief monument at Ivriz.
42A little further on, between Porsuk and Bayağıl, there are the archeological ruins, dating from the Byzantine era, of the fortress which the Arab itinerary sources refer to as aç-Çafçaf69.
43Still in this same area, and apart from the pre-protohistoric site and the Byzantine and Arab settlements, if one keeps in mind (a) the finding of a fragment of moulded entablature70, (b) various epigraphic fragments dating from the 2nd and 3rd centuries AD71, (c) numerous fragments of ceramics which can be dated from the Roman period up to the 5th and 6th centuries AD72 and (d) the foundations of a Byzantine church73, it will then perhaps be possible not only to comprehend the importance of the area even better, but also appreciate its systematic and continuous frequentation and use over the thread of time. Then, in order to define the key role which the area held in terms of roads and itineraries even more precisely, it should also be remembered that, close to the village of Porsuk, William Ramsay discovered no fewer than three milestones74. The first one, attributed to Elagabalus and dating from the first year of that emperor’s reign (218 AD) records a distance of XXXVI miles to P[ylis] – with reference, therefore, to a directrix leading in the direction of the Cilician Gates. The second, which is of uncertain date, bears the inscription Faustinianae col[oniae] | XXIII a Tyan[is] this referring to a directrix towards Tyana. The third, which dates from the time of Gordian III, unfortunately has no indication of distances or places of reference75.
44Having pondered these several considerations, and returning now to the question from which our deliberations started, it would not seem too hazardous to propose that Faustinopolis should be located precisely in that very same area.
45With the present state of archaeological research in this sector of southern Cappadocia, it is impossible to propose with any certainty a location for that Roman colonia which would be any more precise and accurate. However, given the cultural and strategic importance that Zeyve Höyük appears to have commanded since very ancient times, one might suggest that it was located on that exact site. Since only a small part of the extensive area covered by the site has been excavated so far, and since the presence of a necropolis dating with some caution from the 3rd century/late antiquity has been discovered on its eastern boundary, one might perhaps wonder about the presence of a settlement (contemporary or maybe earlier) in that central part of the höyük which has not yet been the subject of systematic research. One could also add that the absence of any archaeological proof, from the Roman period, of any architectural ‘monumentality’ which the site would have had on account of its status as a colonia does not seem to be an adequate reason for denying that an identification such as this is correct. Indeed, an investigation of the sort we have indicated could stem from the discovery of that fragment of moulded entablature to which we have referred above – even though at the moment it is one single example; but we also should not neglect the phenomenon of the systematic re-use of antique material within the foundations of more modern buildings, or enclosed within their walls – a phenomenon which is particularly widespread in that part of Anatolia76. It is for reasons such as these that research in the nearby villages, both populated and abandoned, might perhaps prove to be of a certain interest. We should also emphasise that the indications of distances in Roman miles between Podando/mansio Opodando/Pozantı and Faustinopolim/civitas Faustinopoli which appear both in the Itinerarium Antonini and in the Itinerarium Burdigalense favour a location of this nature, and it would also appear compatible with the distances revealed along the stretch of road which we have just endeavoured to describe77.
46Perhaps even more interesting are the observations that on one of the milestones discovered at Porsuk, not only is the toponym of the Roman colonia explicitly cited78, but also that at the point where the milestone was placed a directrix may reasonably be thought to have branched off which should have reached Tyana after XXIII milia passum. In fact a road which, from the area of Porsuk, would have arrived in Kemerhisar after following a route of approximately 35 km does exist. We refer here to a stretch of road suggested some years ago by Michel Coindoz79 and Thomas Drew-Bear80. In contrast with the modern road which runs to Kemerhisar via the Çaykavak pass (along which nothing of particular archaeological significance has been found up to the present time), it must have kept slightly further to the east and reached the Kozlica Deresi valley, crossing the col between the slopes of Avluk Tepe and İlhan Dağı (fig. 10).
47Specifically, the possible ancient road to Tyana might have curved towards the north, thereby breaking away from the main artery – significantly – at the höyük of Zeyve, following a route which perhaps was not wildly different from that of the present-day road which leads to the artificial basin which is to be found approximately two kilometres further to the north81. Furthermore, in this same area, Coindoz revealed the significant presence of a series of “tumuli funéraires”, which on a numismatic basis may be dated to the 2nd century AD82. It would have then been able to proceed to the north-west along the valley floor, which nowadays is submerged under the waters of that same basin, passing close to the Hacı Davut Tepesi, where Coindoz was able to unearth an abundance of ceramic work, part of which dates from the Hellenistic-Roman period. Then, passing by “quelques vestiges de constructions récentes” – according to local tradition, these are the remains of a caravanserai – the road would have continued north, still on a relatively easy route even if the context is decidedly more mountainous, and climbed the slopes of İlhan Dağı (fig. 12). It is reasonable to suppose that it would have gone on by following the curves of the level, reached the pass and, from there, descended along the north side of that same mountain range which dips rather more gently towards the Kozlica Deresi plain83.
Fig. 12: At the foot of the Avluk Tepe and İlhan Dağı. The possible road towards Tyana

48From there on, the possible route followed by the road may be suggested by the milestones which have been found at Eminlik, on the left bank of the Kozlica Deresi, and at Kavuklu, which is about 5.5 km further to the north84.
49The road would finally have reached Tyana, and it is reasonable to think that it would have passed close to the site of Salmanlı Çiftlik. Evidence of this is suggested by other milestones85.
50Although this stretch of road between Porsuk and Eminlik presents a number of difficulties as to the route followed – difficulties which are obviously linked with the morphological characteristics pertaining in that part of southern Cappadocia – what we have here is a route which would have provided for far easier travel than the route which is traditionally considered to have been the preferred via of communication between Tyana and the mansio Opodando (i.e. the road that extends along the valley of the Kırkgeçit Deresi). Some confirmation of this can be derived from a piece of evidence linked with the trade in ceramic items from Avanos. The caravanners who exported their celebrated wares to Cilicia recalled that of the directrices most frequently used to reach the Çakıt Suyu valley from Kemerhisar, this route, along with the main road which climbs the Çaykavak pass, was the most popular. Significantly, they further recalled that “la vallée de Kırkgeçit Deresi était ... impraticable lors de printemps suivant un hiver trop enneigé”86.
51At this point, therefore, considering that there was this alternative route which was practicable, certainly easier to negotiate, safer, and less exposed to possible ambushes or other attacks, it seems possible to exclude the road passing through Başmakçı from the list of principal roads which made up the possible road network of southern Cappadocia. More logically one might consider the existence of a road system formed by two directrices; one running from east to west (Cilician Gates – mansio Opodando/Podando/Pozantı – Faustinopolis/Porsuk – Cybistra/Ereğli – Iconium/ Konya) and one from north to south (Caesarea/ Kayseri – Tyana/Kemerhisar – Faustinopolis/Porsuk), which appear, fascinatingly, to have converged in that strategic area along the Çakıt Suyu valley where, over the course of the centuries, an important pre-protohistoric settlement (Zeyve Höyük/Porsuk), a Roman colony (Faustinopolis) and a Byzantine fortress (aç-Çafçaf) would have all followed one another87. This, then, was a strategic area in every sense of the term, and one which could very well have served as a major hub or pivotal point within the road network of southern Cappadocia.
52On the other hand, a piece of epigraphic evidence known as the Itinerarium Cappadociae also appears to offer some confirmation of a preferential status accorded to these same road directrices88.
53Indeed, the itinerary described there89 seems to correspond with the route we have sought to propose: from Tarsus to Tyana, via Panhormos (Pozantı), Aquae Calidae (Çiftehan) and Tynna, which has been identified with some certainty as being located on the site at Zeyve90.
54Furthermore, Christol and Drew-Bear have recently proposed that the toponym mutation Caena – which the Itinerarium Burdigalense91 locates between mansio Opodando and civitas Faustinopoli – should be considered as a “déformation de Calida” (i.e. Aquis calidis/Aquae calidae) and that it should therefore be associated with Çiftehan92.
55If one were to accept this proposal, one might recognise the road which ran along the Çakit Suyu valley as being part of that ancient itinerary (which is in contrast with what is traditionally proposed)93. One might also add (with all the necessary caution) that, in general, the distances – both ancient and modern – along the stretch of road between civitas Faustinopoli and mansio Opodando would be thoroughly compatible94.
56For the sake of completeness, it also needs to be said that once a reconstruction of the road network such as this is accepted, there is still a problem ‘of miles’ between Faustinopolis and Tyana, since the distance recorded in both the Itinerarium Antonini and the Itinerarium Burdigalense is incompatible with the 35 kilometres (more or less) which separate Porsuk/Zeyve from Kemerhisar95.
57On the other hand, however, when everything said up until now is considered, the principal, preferred, most comfortable and least difficult route – also for the transit of an army – can only be found in the directrix Porsuk - Avluk Tepe/İlhan Dağı - Eminlik - Kavuklu - Kemerhisar.
58At this point, then, how can the case be resolved? As seems to have been suggested by Cuntz (among others) in the critical apparatus96, the problem might be due to there having been an incorrect transcription of some sort in the Roman numerals. After all, it would be fairly straightforward – but also facile, perhaps – to arrive at a total of XXIII Roman miles (that is to say, the number recorded on one of the milestones at Porsuk) if one were to surmise that a figure X had been lost at the beginning of the mileage indicated in the Itinerarium Burdigalense (thus enabling the smooth transfer from XII to XXII); and in the case of the Itinerarium Antonini to suppose that a fairly common error had been made in that a figure X had been transformed into a figure V so that the distance became XVIII miles instead of XXIII.
59Alternatively, however, one might think along with Cuntz97 – albeit with all the caution which the case requires and in the knowledge that a ‘heavy’ hypothesis is involved – that what has been lost is some intermediate stopping place. This theory in fact would be much more in keeping with a possible stretch of road which needed to cross a mountainous district and along which just one single stopping place in very nearly thirty-five miles would have been too demanding. In this sense one could think, even in hypothetical terms, that a new mansio needs to be inserted in the area of Eminlik/ Kavuklu – where, as will surely be recalled, a number of milestones were found. That could significantly have made an excellent stopping place, since not only was it on the plain; it was also (and above all) plentifully supplied with water.
...and from Tyana to Caesarea, passing by the fortress of the black camel
60Having reached Tyana, the road would then have continued to the area occupied by the present town of Niğde (fig. 13). With regard to the initial stretch of this road, the site of Köşk Höyük, some four and a half kilometres north-east of Kemerhisar itself, would undoubtedly have been an important territorial marker which must have had a significant influence on the route taken in ancient times.
Fig. 13: Towards Caesarea passing by the Fortress of the Black Camel

61This site has a certain relevance both from an archaeological point of view (in that it provides evidence of settlement from the late VIIth or early VIth millennium BC through to the Byzantine era)98, and also topographically (since it was initially chosen because of its proximity to a spring of natural water which can still be seen at the foot of the slopes of Köşk). During the Roman period this spring was monumentalised and served as a source for the hydraulic system operating in Tyana. Furthermore, recent investigations have unearthed the ruins of a small church (possibly a chapel) with two apses. This building dates from around the 13th century AD, and it may offer further confirmation of the strategic importance of the site.
62Possibly keeping to the eastern side of the Niğde plain, it would have been possible to reach Andabalis/mansio Andavilis99, known today as Yeniköy/Aktaş100. There, the church of Constantine and St. Helena, built significantly where the road leading to Colonia Archelais/Aksaray branched off, formed an important road marker. This church can reasonably be dated as being of the 6th century AD101.
63As to the following stage, the directrix would have arrived, as some travellers recall, at Mysty102 or “Misli… a small village, almost underground”103, which we can identify in the vicinity of Konaklı, approximately 27 kilometres north-east of Niğde104. From there the road led on to the caravanserai (marked on Arrowsmith’s map as “ruined khan” – fig. 7) which there must have been in the vicinity of Edrye Kieui105, which we can perhaps identify as the Edikli of today106.
64In order to arrive at Caesarea/Kayseri, rather than following the route of the modern main road which crosses the Araplı pass and then descends into the plain, our road would have been more likely to have crossed the longitudinal valley that can be entered in the vicinity of the town of Araplı, and, twisting and turning along the banks of a water course, it would then have come to Kyzistra, nowadays Yeşilhisar. Some confirmation of this can in fact be derived from the position of Zengibar Kalesi, the fortress of the Black Camel as Ainsworth recalls107, which is situated precisely at the point where at least two road arteries (i.e. the one that extends along the Mavrucan Deresi valley, connecting the Yeşilhisar plain with Derinkuyu108, and the one coming from Tyana) would have formed a junction.
65As far as the strategic importance of that fortress is concerned, some further clues might perhaps be found in one of the attractive and lively miniatures (fig. 14) drawn by the historian and calligrapher Maṭrāḳçī Naṣūḥ in his history of Suleiman I’s campaign for the conquest of Iraq (1534-36)109. Thanks to the captions included, it has been possible to recognise Kayseri (at the foot of the image) and Yeşilhisar (at the top). It might also be possible to recognise the fortress at the summit of a mountain which rises isolated from the surrounding plain (shown in the immediate vicinity of Yeşilhisar) as the kale of Zengibar. Considering the strategic and military nature of Matrākçī’s opus which, as has been suggested, was also intended to be of some importance in the planning and organisation of successive military campaigns (and not only those of Suleiman)110, one might perhaps, with some element of justification, consider the fortress as a marker of strength within the area; one which controlled not just the territory but also – and possibly above all – one or more key roads.
Fig. 14: Zengibar Kalesi in a miniature by Maṭrāḳçī Naṣūḥ

Yuraydın 1976
66Going northwards from Yeşilhisar the directrix would have kept always to the west side of the plain, as Esme Scott Stevenson records: “On the opposite shore [of our road], in a cleft on the slopes of Argaeus [there is a] small village built entirely of black stones, which is called Sheik-Shâdân”111. This can easily be recognised as the present-day Şeyhşaban, which is built on the eastern side of the valley.
67In contrast with the modern highway which passes through the centre of İncesu before heading north, the old road may have taken a slightly different route. As seems to be hinted by the road traces found by Frederick Starr “…a few miles east of Incesu…”, the latter road would gradually have moved across from the western to the eastern side of the plain, until it reached the above mentioned directrix linking Avanos and Kayseri.
Closing the triangle: Cybistra, a junction at the foot of the Taurus
68As far as the ‘southern road’ is concerned, after leaving Iconium/Konya it arrived at Cybistra/Ereğli112, from where there would have been a road which led to Porsuk and to the directrix we have already mentioned which curved from there towards Podandos and the Cilician Gates, thereby connecting the Anatolian uplands with the Mediterranean coastal strip (fig. 15). This appears to have been the route of Xenophon’s army on its march towards the East113; Cicero would have passed this way en route to Tarsus after his appointment as governor of Cilicia114; and Baldwin and Tancred chose to march that way with their band of soldiers during the First Crusade115.
Fig. 15: The alternative routes from Cybistra to Porsuk

69As regards the Cappadocian stretch between Cybistra and Porsuk, there is at present no archaeological data available which might enable us to put forward a concrete hypothesis concerning its route. The only useful piece of information we have which might assist us in reconstructing it could be recoverable, in Hild’s view, from what is reported by the 9th century Persian geographer Ibn Ḫurdāḍbih. Ḫurdāḍbih records the existence of a road artery which, setting out from the Cilician Gates, would have reached aç-Çafçaf, and from there Hirakla/Ereğli, passing by Wâdi-’Tarfâ, a toponym which means the “valley of the tamarisks”116.
70Since the plain east of Ereğli is a flat and predominantly arid area (“…eine mehr oder minder vegetationslose Steppe…”)117, Hild proposes (even with some reserve) that the main road should have passed to the south of that, where the land is richer in water. In particular, he suggests that, after proceeding east along the mountain slopes of the Bolkar Dağları, the road would have run through the so-called Kamerettin pass (1,650 metres above sea level) at the western extremity of the Kılan Deresi valley; then it may have curved towards the north- east, entering the valley of the Çakıt Suyu just off the site of Zeyve Höyük/Porsuk.
71Now, we have to admit that for anyone to propose a route which would have been rough, uneven, and close to the northern slopes of the Taurus; a road which would have involved considerable variations in height (around 700 metres) along its length, seems far from acceptable from our point of view, since there was a far more practical and comfortable alternative in the wide valley of Çakıt Suyu and the plain which stretches to the east of Ereğli. The whole suggestion becomes even less attractive when one considers that a hypothesis of this sort has been formed solely on the basis of a toponym which, when all is said and done, has clearly been misunderstood, since tamerisks (one is reminded of that celebrated poem by D’Annunzio)118 are shrub-like trees which can also survive quite happily in a very dry environment.
72In this sense, it seems more reasonable to think that there may have been only a byway along the Kılan Deresi valley and that the main route would not have differed too greatly from the modern one and from that used by a couple of English travellers of the Nineteenth century. Such a route would have always kept to the plain in order to arrive at Ulukışla (which is about nine kilometres to the west of Porsuk) passing by Çayhan119.
73What appears to emerge at the end of this archaeological and topographic analysis is a road network which really allowed for efficient connections within the context of central Anatolia.
74This analysis reveals, above all, a road system which appears to have remained constant and had no substantial alterations over the course of the centuries precisely on account of the fact that roads are actualities which perpetuate themselves over time and which find no reason to undergo modification.
75A continuity of this kind can, with good reason, be attributed to the strategic importance of that borderland linking East and West in both military and economic senses.
Bibliography
DOI are automaticaly added to bibliographic references by Bilbo, OpenEdition’s bibliographic annotation tool. These bibliographic references can be downloaded in APA, Chicago or MLA formats.
Format
- APA
- Chicago
- MLA
Abadie-Reynal, C., “Porsuk. Rapport sur la campagne de fouilles de 1989. Chantier est”, Syria LXIX, 1992, 349-377.
Abadie-Reynal, C., “La céramique du haut-empire a Porsuk”, in C. Abadie-Reynal (éd.), Les Céramiques en Anatolie aux époques hellénistique et romaine. Actes de la Table Ronde d’Istanbul (22-24 mai 1996), Paris, 2003, 101-109.
Ainsworth, W.F., Travels and Researches in Asia Minor, Mesopotamia, Chaldea, and Armenia, I-II, Londres, 1842.
Balance, M.H., “Derbe and Faustinopolis”, AS XIV, 1964, 139-145.
10.2307/3642469:Beck, A. / Philip, G. / Abdulkarim M. / Donaghue D., “Evaluation of Corona and Ikonos high resolution satellite imagery for archaeological prospection in western Syria”, Antiquity 81, 2007, 161-175.
10.1017/S0003598X00094916:Berges, D. / Nollé, J., Tyana. Archäologisch-historische Untersuchungen zum südwestlichen Kappadokien, I-II, Bonn, 2000.
Beyer, D., “From the Bronze Age to the Iron Age at Zeyve Höyük/Porsuk: a Temporary Review”, in L. d’Alfonso / M.E. Balza / C. Mora (éds.), Geo-archaeological Activities in Southern Cappadocia –Turkey. Proceedings of the Meeting held at Pavia (20.11.2008), Milan, 2010, 97-109.
Beyer, D. / Chalier, I. / Kirner, F. / Patrier, J. / Tibet, A., “Zeyve Höyük – Porsuk : Rapport préliminaire de la campagne 2011”, Anatolia Antiqua XX, 2012, 177-203.
10.4000/anatoliaantiqua.332:Blaizot, F., “L’ensemble funéraire tardo-antique de Porsuk : approche archéo-anthropologique (Ulukışla, Cappadoce méridionale, Turquie). Résultats préliminaires”, Anatolia Antiqua VII, 1999, 179-218.
Burnichon, P.J., “À travers le Taurus. De Césarée de Cappadoce à Adana. Souvenirs de voyage”, Études religieuses, philosophiques, historiques et littéraires LXI, Paris, 476-499.
Cassia, M., Cappadocia romana. Strutture urbane e strutture agrarie alla periferia dell’Impero, Catane, 2004.
Childs, W.J., Across Asia Minor on Foot, Edinbourg/ Londres, 1917.
Christol, M. / Drew-Bear, T., “L’aménagement de la Via Tauri sous les Sévères”, Anatolia Antiqua XVII, 2009, 239- 254.
10.3406/anata.2009.1283:Coindoz, M., “Recherches archéologiques dans la région d’Avanos (Cappadoce) : le tumulus du Çeç”, Anatolica XII, 1985, 1-28.
Coindoz, M., “Le site de Porsuk et le voies de communication entre la Tyanitide et le Portes Ciliciennes”, in B. Le Guen-Pollet / O. Pelon (éds.), La Cappadoce méridionale jusqu’à la fin de l’époque romaine. État des recherches, Actes du Colloque d›Istanbul (13-14 Avril 1987), Paris, 1991, 77-90.
Coindoz-Kleiman, C. / Coindoz, M., “Avanos. Un village cappadocien menacé”, Dossiers Histoire et Archeologie 121 (Novembre), 1987, 80-93.
Davis, E.J., Life in Asiatic Turkey. A Journal of Travel in Cilicia (Pedias and Trachœa), Isauria and Parts of Lycaonia and Cappadocia, Londres, 1879.
De Goeje, M.J., Kitâb al-Masâlik wa’l-Mamâlik (Liber Viarum et Regnorum), auctore Abu’l-Kasim Obaidallah Ibn Abdallah Ibn Khordâdhbeh, Leida, 1889.
Deniz, B., “Ağzıkara Han”, in H. Acun (éd.), Anadolu Selçuklu Dönemi Kervansarayları, Ankara, 2007, 321-345.
Deniz, B., “Alay Han”, in H. Acun (éd.), Anadolu Selçuklu Dönemi Kervansarayları, Ankara, 2007, 51-75.
Dikilitaş, G. / Açikgöz, F., “Activity of Conservative Restoration of the Constantine-Helena Church in the Ancient Site of Andabalis”, in L. d’Alfonso / M.E. Balza /
C. Mora (éds.), Geo-archaeological Activities in Southern Cappadocia –Turkey. Proceedings of the Meeting held at Pavia (20.11.2008), Milan, 2010, 129-135.
Drew-Bear, T., “Inscriptions de Cappadoce”, De Anatolia Antiqua I, 1991, 130-149.
Duda, H.W., Die Seltschukengeschichte des Ibn Bībī, Copenhague, 1959.
Equini Schneider, E. / Morselli, C. / Spanu, M. / Vismara, M., “Varia Cappadocica”, ArchClass XLIX, 1997, 101-209.
Erdmann, K., Das Anatolische Karavansaray des 13. Jahrunderts, I-II, Berlin, 1961.
Forrer, E.O., “Kilikien zur Zeit des Hatti-Reiches”, Klio 30, 1937, 135-186.
10.1524/klio.1937.30.30.135:French, D.H., Roman Roads and Milestones of Asia Minor. Fasc.1: The Pilgrim’s Road, Oxford, 1981.
French, D.H., Roman Roads and Milestones of Asia Minor. Fasc.2: An Interim Catalogue of Milestones. Part I, Oxford, 1988.
French, D.H., “Pre- and early Roman Roads of Asia Minor. The Persian Road”, Iran XXXVI, 15-43.
10.2307/4299973:French, D. H., Roman Roads and Milestones of Asia Minor. vol. 3 Milestones. Fasc. 3.3 Cappadocia [British Institute at Ankara Electronic Monograph, n. 3], 2012.
Hamilton, W.J., Researches in Asia Minor, Pontus, and Armenia with some account of their antiquities and geology, I-II, Londres, 1842.
Harper, R.P., “Podandus and the Via Tauri”, AS XX, 1970, 149-153.
10.2307/3642593:Hild, F., Das Byzantinische Strassensystem in Kappadokien, Vienne, 1977.
Hild, F. / Restle, M., Tabula Imperii Byzantini 2. Kappadokien, Vienne, 1981.
Jacopi, G., Esplorazioni e Studi in Paflagonia e Cappadocia, Rome, 1938.
Johnston, N.J., “The Urban World of the Matraki Manuscript”, JNES 30, n. 3 (Jul., 1971), 159-176.
10.1086/372116:Karaçağ, A., “Avanos Sarı Han”, in H. Acun (éd.), Anadolu Selçuklu Dönemi Kervansarayları, Ankara, 2007, 211-234.
Lebreton, S., “Les Portes de Cilicie”, in A. Gangloff (éd.), Lieux de mémoire dans l’Orient grec à l’époque impériale. Actes du Colloque International (Lausanne, 8-11 avril 2011), forthcoming.
Lucas, P., Voyage Du Sieur Paul Lucas, Fait Par Ordre Du Roy Dans La Grèce, L’Asie Mineure, La Macédoine et L’Afrique, I-II, Paris, 1712.
Mandel, G., I caravanserragli turchi, Bergame, 1988.
Maraval, P., Grégoire de Nysse. Lettres. Introduction, texte critique, traduction, notes et index, Paris, 1990.
Miller, K., Itineraria Romana, Stuttgart, 1916.
Özergin, M.K., “Anadolu’da Selçuklu kervansarayları”, Tarih Dergisi XV, 1965, 141-170.
Öztan, A., “Archaeological Investigations at Kösk Höyük, Niğde”, in L. d’Alfonso / M.E. Balza / C. Mora (éds.), Geo- archaeological Activities in Southern Cappadocia –Turkey. Proceedings of the Meeting held at Pavia (20.11.2008), Milan, 2010, 83-96.
Pelon, O., “Six campagnes de fouilles à Porsuk (Turquie méridionale) de 1969 à 1977”, CRAI 122, n. 2, 1978, 347-359.
10.3406/crai.1978.13479:Pelon, O., “La fouille de Porsuk-Ulukişla”, in J. L. Bacqué- Grammont / M. Amandry / A. Davesne (éds.), Anatolie antique. Fouilles françaises en Turquie. Catalogue de l’exposition (1er décembre 1989-16 avril 1990), Paris/ İstanbul, 1989, 14-19.
Pelon, O. / Kuzucuoğlu, C., “Le site de Porsuk et les mines de Bulgarmaden”, Pallas 50, 1999, 419-435.
10.3406/palla.1999.1567:Ramsay, W.M., The Historical Geography of Asia Minor, Londres, 1890.
10.1017/CBO9780511709753:Ramsay, W.M., “Cilicia, Tarsus and the Great Taurus Pass”, GJ 22, n. 4, 1903, 357-410.
10.2307/1775456:Ramsay, W.M., “Lycaonia”, Jahreshefte des Österreichischen Archäologischen Institutes in Wien VII, 1904, 57-132.
Robert, L., Hellenica: XVII, Épitaphes de Cappadoce, in RPh, 1939, 210-214.
Rosada, G. / Lachin, M.T., “... civitas Tyana, inde fuit Apollonius magus... (ItBurdig, 577,7-558,1)”, in L. d’Alfonso / M.E. Balza / C. Mora (éds.), Geo-archaeological Activities in Southern Cappadocia –Turkey. Proceedings of the Meeting held at Pavia (20.11.2008), Milan, 2010, 111-127.
Rott, H., Kleinasiatische Denkmäler aus Pisidien, Pamphylien, Kappadokien und Lykien, Leipzig, 1908.
Scott-Stevenson, M.E.G., Our Ride through Asia Minor, Londres, 1881.
Spanu, M., “La Cilicia nella Tabula Peutingeriana”, in
C. Marangio / G. Laudizi (éds.), Palaià filìa. Studi in onore di Giovanni Uggeri, Galatina (Lecce), 2009, 635-652.
Starr, S.F., The Ancient Roads of Asia Minor: 1961 Expedition, New Haven (Connecticut), 1962.
Taeschner, F., “The Itinerary of the First Persian Campaign of Sultan Suleyman, 1534-36”, Imago Mundi XII, 53-55.
Talbot Rice, M., The Seljuks in Asia Minor, Philadelphie, 1961.
Taylor, C., Roads & Tracks of Britain, Londres/Toronto/ Melbourne, 1979.
Tekocak, M., “Aksaray İli Merkez, Ortaköy, Sarıyahşi, Ağaçören ve Eskil İlçeleri Yüzey Araştırmaları: 2010”, 29. Araştırma Sonuçları Toplantısı 2, 2012, 83-116.
Texier, C., Asie Mineure. Description géographique, historique et archéologique des provinces et des villes de la Chersonnèse d’Asie, Paris, 1862.
Thierry, N., “Avanos-Vénasa – Cappadoce”, in H. Ahrweiler (éd.), Geographica Byzantina, 1981, 119-129.
10.4000/books.psorbonne.1977:Thierry, N., La Cappadoce de l’antiquité au moyen âge, Turnhout (Anvers), 2002.
10.3406/mefr.1998.3661:Turchetto, J. “ ‘... e fino a Mazaka... passando per Soandos e Sadakora...’ (Strabo, XIV, 2, 29). La koinè odós straboniana e la possibile viabilità della Cappadocia centrale”, Journal of Ancient Topography 22 (2012), 2013, 83-94.
Ur, J.A., “CORONA Satellite Photography and Ancient Road Networks: A Northern Mesopotamian Case Study”, Antiquity 77, 2003, 102-115.
10.1017/S0003598X00061391:Ur, J.A., “Google Earth and Archaeology”, Society for American Archaeology. Archaeological record 6, n. 3, 2006, 35-38.
Wilkinson, T.J., “Linear hollows in the Jazira, upper Mesopotamia”, Antiquity 67, 1993, 548-562.
10.1017/S0003598X00045750:Wilkinson, T.J., Archaeological landscapes of the Near East, Tucson, 2003.
10.2307/j.ctv1jf2ddx:Wilkinson, T.J., “Ancient Near Eastern Route Systems: From the Ground Up”, ArchAtlas, version 4.1, http://www.archatlas.org/workshop/TWilkinson07.php
Wilkinson, T.J. / French, C. / Ur, J.A. / Semple, M., 2010, “The Geoarchaeology of Route Systems in Northern Syria”, Geoarchaeology: An International Journal 25, n. 6, 2007, 745-771.
10.1002/gea.20331:Williams, F., “Xenophon’s Dana and the Passage of Cyrus’ Army over the Taurus Mountains”, Historia XLV, 1996, 284-314.
Yuraydın, H.G., Beyān-ı Menāzil-i Sefer-i ‘Irāḳeyn-i Sulṭān Süleymān Hān, Ankara, 1976.
Appendix
Abréviations
Anatolica = Anatolica : annuaire international pour les civilisations de l’Asie antérieure. Istanbul : Institut historique et archéologique néerlandais.
Antiquity = Antiquity : A Quarterly Review of World Archaeology.
ArchClass = Archeologia Classica. Roma : L’Erma di Bretschneider.
AS= Anatolian Studies: Journal of the British Institute of Archaeology at Ankara. London : British Institute of Archaeology at Ankara.
CRAI = Comptes rendus / Académie des inscriptions et belles-lettres. Paris: de Boccard.
GJ= The Geographical Journal. London: Royal Geographic Society. Oxford: Blackwell.
Historia= Historia: Zeitschrift für Alte Geschichte= Revue d’Histoire Ancienne. Stuttgart: Steiner.
Iran= Iran: The Journal of the British Institute of Persian Studies. London: British Academy, British Institute of Persian Studies.
JNES= Journal of Near Eastern studies. Chicago (Ill.): University of Chicago Pr.
Klio= Klio: Beiträge zur Alten Geschichte. Berlin : Akademie Verl.
Pallas = Pallas : Revue d’études antiques. Toulouse : Pr. Universitaires du Mirail.
RPh = Revue de philologie, de littérature et d’histoire anciennes.
Syria = Syria : revue d’art oriental et d’archéologie. Beyrouth : Institut français d’archéologie du Proche- Orient.
Footnotes
1 I would like to thank Mr C.F.J. Jones for the translation of this paper. All the photographs and the maps shown in the article were taken and made by the author.
2 Strabo XIV, 2, 29.
3 Plin. Nat. hist., II, 112.
4 For a more detailed description of this road cf. Turchetto 2013, with earlier bibliography.
5 Cf. Mandel 1988, X-XI and, in general terms, about the Seljuks in Asia Minor, cf. Talbot Rice 1961.
6 Cf. Erdmann 1961, 97-102; Özergin 1965, 152; Hild 1977, 71; Mandel 1988, 36-39; Deniz 2007a, 321-345.
7 Cf. Erdmann 1961, 167-168; Özergin 1965, 158; Hild 1977, 71; Mandel 1988, 46-47.
8 Cf. Erdmann 1961, 81-83; Özergin 1965, 145; Hild 1977, 71; Mandel 1988, 16-17; Deniz 2007b, 51-75.
9 Cf. Mandel 1988, 116.
10 Cf. Erdmann 1961, 130-135; Özergin 1965, 161; Hild 1977, 80; Mandel 1988, 52-55; Karaçağ 2007, 211-234.
11 Özergin 1965, 163.
12 Cf. Duda 1959 for a translation from the Arabic of Ibn Bībī, and, specifically, 92 and 329 as regards the question of Sünnetli Han.
13 Hild 1977, 71, 80, with bibliography.
14 French 1998, 18.
15 Ur 2003, 102. In general, re hollow ways, cf. Taylor 1979; Wilkinson 1993; Wilkinson 2003; Beck et al. 2007; Wilkinson 2007; Wilkinson et al. 2010.
16 The image has deliberately been darkened in order to better reveal the less marked (and consequently less visible) traces on the ground.
17 Cf. Tekocak 2012.
18 Cf. Hild 1977, 71. A funerary inscription (in Greek) was found in the immediate vicinity of the Ağzıkara Han as well (Equini Schneider et al. 1997, 196-198).
19 Hild 1977, 79-81. In support of a reconstruction of this kind of the route taken by the road, which is dependent largely on the premise that, according to the author, Nevşehir was no more than an insignificant village (“unbedeutendes Dorf”), it is also reported that north of Acıgöl there was a bridge with three arches (of which there is no trace whatsoever today).
20 For the identification of antique toponyms associated with the present-day town of Avanos, cf. Thierry 1981. On the site in general, cf. Hild/Restle 1981, 302; Cassia 2004, 210-211. According to P. Maraval the Ouánota cited by Gregory of Nissa in a letter sent to Adelphius to thank him for the hospitality he received at his villa, which must have been in that area (Greg. Nyss. Ep., 20) can also be associated with Ouénasa /Avanos (Maraval 1990, 259, note 3; 343).
21 This road is attested in ItAnt, 205, 7 - 206, 7, 29 (Cuntz): Item ab Ancyra / per Nisam Cae / saream / m. p. CXCVIII / Gorbeus m. p. XXIIII / Orsologiaco m. p. XVIII / Aspona m. p. XX / Parnasso m. p. XXII / Nisa m. p. XXIIII / Asiana m. p. XXXII / Saccasena m. p. XXVIII / Caesarea m. p. XXX.
22 Hamilton 1842, II, 250.
23 Of conical form, it has a circumference of around 300 metres at its base, and a height of approximately 30 metres.
24 Cf. Coindoz 1985 and Coindoz-Kleiman/Coindoz 1987.
25 Strabo XII, 2, 5.
26 Cf. Thierry 1981.
27 Hild 1977, 80-81
28 Cf., among the others, Hamilton 1842, II, 256-257 (“…we continued N.E., crossing the valley obliquely from left to right, towards the N.W. point of Mount Argaeus …At half-past nine we passed round the N.W. point of the mountain, and entered an extensive plain stretching far the N.E.; in the center of it is a marshy lake or Sas, full of reeds and rushes, and much frequented at all seasons of the year by wild fowl of every kind…”); Texier 1862, 548 (“La route de Césarée à Ingèsou suit les pentes du mont Argée, et à douze kilomètres de cette ville rencontre le grand marais de Salzik engendré par les deux rivières; de nombreux troupeaux y paissent pendant l’été, mais il est impraticable quand la fonte des neiges grossit les rivières; c’est dans ce marais que le Mélas rejoint le Sarimsak... Les pentes occidentales du mont Argée sont moins abruptes que celles du nord...”); Scott Stevenson 1881, 188 (“…I believe it to be the commencement of the marsh called Salzik, which extends more or less all the way between Injehsu and Kaisariyeh, and is formed by the waters of two rivers, the Melas or Kara Sou and the Saremsak, at their junction on the way to join the river Kizil- Irmak….”).
29 Cf., among other sources, John Arrowsmith: The Map of Asia Minor (1844) and Joseph Grassl: Specialkarte der Asiatische Turkey (published 1860).
30 Vitr. De arch., VIII, 3, 9. However, it cannot entirely be excluded that this reference was intended to apply to another marshy area to the south-west of the Erciyes.
31 Re that road vide infra.
32 Hild 1977, 118; Thierry 2002, 37-38.
33 Rott 1908, 203.
34 “…we continued, between the lake on our left and the roots of Argaeus on our right .…” (Hamilton 1842, II, 257).
35 Starr 1962, 65.
36 “…we entered a narrow plain which separates the lofty ridge of Yelanli Dagh from the mountain [Argaeus/Erciyes].… ,” (Hamilton 1842, II, 257); “…we passed the foot of Yilanli Dagh – mountain of snakes, so named not without reason – … skirting always the base of Argaeus…” (Childs 1917, 206).
37 Cf. Cassia 2004, 169.
38 Iul. Capitol. M. Aur., XXVI, 4-7; Ael. Spart. Ant. Carac., XI, 6-7.
39 Iul. Capitol. M. Aur., XXVI, 4.
40 Ael. Spart. Ant. Carac., XI, 6.
41 Despite the fact that no literary sources exist which specify the name given to Halala after its elevation to a colony by Marcus Aurelius, most scholars agree that it may have been Colonia Faustiniana (as an inscription suggests – CIL, III, 12213 – see, among others, Drew-Bear 1991, 134). Faustinopolis, on the other hand, represents a sort of “...forme ‘vulgaire’ de basse époque, qui ne doit son existence qu’à l’analogie et à la perte de la culture littéraire classique...” (Christol/Drew-Bear 2009, 251, note 70). Among scholars, however, the use of the toponym Faustinopolis is widespread (cf. Lebreton forthcoming, among others).
42 ItAnt, 145, 1-4, 20 (Cuntz): Andabalis m.p. XVI / Tiana m.p. XVI / Faustinopolim m.p. XVIII / Podando m.p. XXVI; ItBurdig, 577, 7 - 578, 4, 93 (Cuntz): civitas Thyana mil. XVIII /inde fuit Apollonius magus / civitas Faustinopoli mil. XII / mutatio Caena mil. XIII / mansio Opodando mil. XII.
43 Re Tyana, vide, most recently, Rosada/Lachin 2010.
44 Ramsay 1890, 346 (“Faustinopolis must be near Pashmakji...”).
45 Cf, among others, French 1981, 19 (“From Kemerhisar southwards... The road then runs straight to Başmakçı, the ancient Faustinopolis”); Coindoz 1991, 83 (“La vallée du Kırkgeçit Deresi... Nous l’avons parcourue pour partie au dèpart d’Halala/Faustinopolis- Başmakçı).
46 Jacopi 1938, 32-33 (translated from the original Italian).
47 Robert 1939, 211.
48 Jacopi 1938, 32-33.
49 Robert 1939, 212-213.
50 Ballance 1964, 141.
51 With regard to this directrix, the route that Ballance envisaged in his hypothesis – a position which was also accepted by subsequent scholars – was as follows: “It climbs steadily to a pass at 1,625 m., 3-4 km. north of Eminlik, and then drops rather steeply into the plain between Eminlik and Başmakçı. Below Başmakçı, it follows the very steep-sided valley of the Kırk Geçit down to the latter›s confluence with the Çakıt at Tahta Köprü, 5 km. below Çiftehan...” (Ballance 1964, 142). The route drawn in fig. 9, along the Kırkgeçit Deresi valley, is only a suggestion and does not follow any actual modern road.
52 Here it is quite extraordinary that this same opinion has also been adopted by other scholars who, in spite of all the evidence to the contrary, adhere to the supposition that the Tyana-Cilician Gates road directrix ran through this valley. Ballance himself, for example, states that “this valley is said to be impracticable for wheeled traffic” (Ballance 1964, 142), whereas David French, who appears to be a lot more aware than the others with regard to the actual inconsistency inherent in a solution of this sort, remarks that “whatever surprise may be expressed, the Roman road does, indeed, run down the valley (or rather, gorge) of the Kırkgeçit” (French 1981, 19). The commentary to be found in Coindoz 1991, 83 is also highly significant: “cette voie de communication... n’avait, semble-t-il, jamais été explorée sans doute à cause des réelles difficultés de circulation qu’elle présente aujourd’hui. La chaussée n’étant plus entretenue, elle est très souvent effondrée et les berges sont parfois trop étroites pour être utilisées. La marche y est donc malaisée et il faut à l’occasion évoluer dans les eaux de la rivière”.
53 Ramsay 1904, 112
54 Ramsay 1903, 396 (“The view was long held that this path up the Takhta-Keupreu water was the line of the Roman road to Tyana, and my ‘Historical Geography’ (p. 346) was written under that misapprehension. The only change needed in p. 346 is to read Ulu-Kishla in place of Pashmakji”).
55 Ramsay 1904, 111 (“At the southern base of the peak of Loulon stand the villages Bey-Aghyl and Porsukh: the ancient village Halala lay probably between the road and these villages”).
56 Williams 1996, 293-296.
57 Christol/Drew-Bear 2009, 249 ff.
58 Cf., most recently, French 2012, 274-276, nn. 166-167, with earlier bibliography.
59 At the confluence of the Çakıt Suyu and Kırkgeçit Deresi another milestone was found. It has been attributed to the Emperors Macrinus and Diadumenianus (cf. Christol/Drew-Bear 2009, 244-247 and, most recently, French 2012, 273-274, n. 165, with earlier bibliography).
60 The stretch between Pozantı and Çiftehan would in effect have been a ‘tricky’ stretch of the road on account of the morphological characteristics of that area. Despite that, it would have been a compulsory route between the two towns simply because no possible practical alternative routes exist. Some of the travellers’ descriptions clearly show the difficulties encountered in negotiating that stretch: “A partir de Tata-Keupru [in the vicinity of Çiftehan], nous avons suivi la vallée, ou plutôt le couloir étroit où la rivière s’est frayé un passage entre deux falaises escarpées. Après une heure ou deux de marche dans ce défilé, nous atteignons le han de Ak-Keupru (Pont blanc) [in the vicinity of Pozantı] …” (Burnichon 1894, 672); “…at Chifte guard-house, [the road] took to the mountain-side and climbed steeply by zigzags for five or six hundred feet from one level stretch to another…” (Childs 1917, 275).
61 Davis 1879, 218-219.
62 TabPeut, IX, 2; Miller 1916, 664. Re this kind of identification, cf. also Christol/Drew-Bear 2009, 249.
63 The traces referred to here are visible thanks to Google Earth. The Corona satellite images, which date from December 1961, do not allow for any of the traces to be viewed because of a thick layer of snow. For the use of Google Earth for archaeological purposes, cf. Ur 2006.
64 Here the writer wishes to express his thanks to Aksel Tibet of the Institut Français d’Études Anatoliennes (Istanbul) who pointed out that these traces would have related to the route of the nineteenth-century road which crossed the Çakıt Suyu valley.
65 A great deal has been published on the subject of the Zeyve Höyük/Porsuk site. Here, in relation to the excavations in the various sectors which make up the site, let it suffice to mention, among the most recent, Beyer 2010 and Beyer et al. 2012, where a bibliography of earlier literature can be found. For the necropolis (dating with some caution from the 3rd century/late antiquity) on the eastern fringes of the höyük, vide Blaizot 1999. See also the paper by Lebreton in this volume.
66 Regarding this aspect cf. Pelon 1978, 347-349; Coindoz 1981, among others.
67 Cf. Pelon/Kuzucuoğlu 1999.
68 Vide infra.
69 Ibn urdā bih (in De Goeje 1889), 73. Re this fortress, vide also Hild/Restle 1981, 268.
70 “Lors d’une promenade sur le plateau qui se trouve au sud du hüyük [Porsuk], de l’autre côté du Darboğaz Cay... j’ai aussi découvert, probablement exhumé par des fouilles clandestines, un fragment d’entablement – première témoignage de l’existence d’une architecture monumentale tardive sur le site de Porsuk. Il s’agit d’un bloc de corniche décorée de modillons, datant probablement de l’époque romaine, et plus particulièrement sans doute du Haut-Empire” (Abadie-Reynal 1992, 377).
71 Cf. Pelon 1989, 14-19; Drew-Bear 1991, 141-142, 145-147.
72 Cf. Abadie-Reynal 1992, 376.
73 Forrer 1937, 234-235.
74 These three milestones were found in “a cemetery by the roadside ... The group of milestones stood together at this point, close to where the roads [i.e. the modern road which runs along the valley of Çakıt Suyu and the one which leads off and turns northwards in the direction of Tyana] forked.” (Ramsay 1903, 401- 403). Cf. also Ramsay 1904, 111-113 and, most recently, French 2012, 235-237, nn. 133(A), 133(B), 133(C), with earlier bibliography.
75 It is interesting to wonder whether this third milestone might possibly have referred to the directrix which led west in the direction of Cybistra/Ereğli.
76 Clear examples of this sort can be found in the neighbouring town of Kemerhisar.
77 The number of Roman miles indicated in the itineraries are, respectively, 26 and 25, whilst the “modern” distance is approximately 36 kilometres.
78 The hypothesis that this could be interpreted as “un milliaire de la colonie de Faustiniana sur le territoire de la cité, car il n’est pas nécessaire que la voie principale traversât la ville qui donnait son nom à l’étape” is essentially dictated by the fact that, as already mentioned, according to Christol and Drew-Bear, Faustinopolis has to be identified as Başmakçı (Christol/Drew-Bear 2009, 249).
79 Coindoz 1991, 81-83.
80 Drew-Bear 1991, 134-135; Christol/Drew-Bear 2009, 252-253.
81 As is explained on a notice alongside the small artificial dyke, the basin formed part of the Çakıt Erozyon Kontrolu Projesi which was undertaken in 2005 by the T.C. Çevre ve Orman Bakanlığı (Turkish Ministry of the Environment and Forests).
82 Coindoz 1991, 82. Today nothing remains of those funerary structures, as is also the case with other relics of an archaeological sort which were observed by Coindoz and about which we shall speak later.
83 As far as this last stretch is concerned, in our considered view it is not possible to reconstruct in detail the exact route which the road followed. This is because the area does not present any particular obstacles of a morphological kind which might incline one towards one alternative or another.
84 Besides the milestones we have already referred to, three others were discovered in more recent years. In general, re all those milestones cf., most recently, French 2012, 270-273, nn. 163(A), 163(B), 163(C), 164(A), 164(B), with earlier bibliography.
85 Re those milestones (which can be referred to Gordianus III and Constantine) cf., most recently, French 2012, 267-270, nn. 162(A), 162(B) and 162(C), with earlier bibliography.
86 Re this evidence, cf. Coindoz 1991, 80.
87 This is not to suggest that these three settlements coincided topographically or that one was simply built on the former site of another. What is intended is that they must have stood – even if at different points – specifically within this very area.
88 CIL, VI, 5076.
89 Re the (unknown) chronology of this inscription and the related issues, cf. Berges/Nollé 2000, II, 325-326.
90 As regards the identification of Tynna, cf. Miller 1916, 664; Forrer 1937, 149; Pelon 1978, 349; Williams 1996, 299, among others.
91 ItBurdig, 578, 3, 93 (Cuntz).
92 Christol/Drew-Bear 2009, 250.
93 Cf. Ballance 1964, 142 and fig. 2; French 1981, 123 and Map 6. In both cases the positioning of Caena is wholly generic and is not supported by archaeological proof of any kind.
94 Indeed, the thirteen Roman miles recorded between civitas Faustinopoli and mutatio Caena are reasonably compatible with the approximate distance of eighteen kilometres which separate Porsuk and Çiftehan, whilst the twelve Roman miles as far as mansio Opodando coincide almost exactly with the sixteen kilometres needed from Çiftehan to Pozantı.
95 In any case it should be pointed out that even if one were to follow the stretch of road passing along the Kırkgeçit Deresi valley, as some have suggested, the distance would in no way be compatible with the miles between Faustinopolis and Podando as recorded in the two ancient itineraries.
96 ItAnt, 145, 3, 20: “fere XXIII.”
97 ItBurdig, 578, 2, 93: “fere XXII. Una statio videtur excidisse”.
98 Cf, most recently, Öztan 2010.
99 ItAnt, 145, 1, 20 (Cuntz); ItBurdig, 577, 5-6, 93 (Cuntz).
100 Cf. especially Hild/Restle 1981, 238.
101 For more information about the church, cf. (most recently) Dikilitaş/Açıkgöz 2010.
102 Lucas 1712, I, 182.
103 Hamilton 1842, II, 295.
104 Re this identification, cf. Hild/Restle 1981, 238.
105 Hamilton 1842, II, 295.
106 Re this identification, cf. Hild/Restle 1981, 173.
107 Ainsworth 1842, I, 209.
108 Re that directrix, vide Turchetto 2013, 84-87.
109 Re Ma rā çī Na ū vide Taeschner 1956; Yuraydın 1976.
110 On these aspects and also on the overall importance of Matrākçī Nasūh’s miniatures, cf. Johnston 1971, among others.
111 Scott Stevenson 1881, 182.
112 This place is generally identified in the ruins of Tont Kalesi, near the present-day village of Gökçeyazı, about 10 kilometres south-east of Ereğli (cf., in particular, Hild/Restle 1981, 188-190, where there is an extensive bibliography; Spanu 2009, 646).
113 Xen. Anab., I, 2, 19-23, 25.
114 Cic. Ad Fam., III, 6, 6; III, 7, 4; XV, 1, 2-3; XV, 2, 1-2; XV, 3, 1; XV, 4, 2-4; Ad Att., V, 18, 1; V, 20, 1-3; VI, 4, 1.
115 Gest. Franc., X, 5.
116 Ibn Ḫurdāḍbih (in De Goeje 1889), 73.
117 Hild 1977, 63.
118 “Piove su le tamerici / salmastre ed arse”, G. d’Annunzio, La pioggia nel pineto, vv. 10-11 (from Alcyone, 1904).
119 Cf. Ainsworth 1842, II, 71, who writes: “At a distance of twelve miles from Eregli …a small village called Kayan”; and also Davis 1879, 233 ff.: “At 10.30 a.m. we came to the village of Tchaian … . We left Tchaian at 1.45 p.m… reached the first branch of the river of Eregli at 5.45 p.m. and, after crossing two other branches, one a considerable stream, entered the town.”
Author
University of Padua
jacopo.turchetto@libero.it
Only the text can be used under the OpenEdition Books License license. Other elements (illustrations, attached files) are “All rights reserved”, unless otherwise stated.
Le Mort dans la ville
Pratiques, contextes et impacts des inhumations intra-muros en Anatolie, du début de l'Age du Bronze à l'époque romaine
Olivier Henry (ed.)
2013
La Cappadoce méridionale de la Préhistoire à l'époque byzantine
3e Rencontres d'archéologie de IFEA, Istanbul 8-9 novembre 2012
Aksel Tibet, Olivier Henry and Dominique Beyer (ed.)
2012
Hittitology today: Studies on Hittite and Neo-Hittite Anatolia in Honor of Emmanuel Laroche’s 100th Birthday
5e Rencontres d'archéologie de l'IFEA, Istanbul 21-22 novembre 2014
Alice Mouton (ed.)
2017