8,5:4 Socratu

The little horizontal line which seems to join "Trallis" to Ad Vicum has has been irritating scholars of Asia Minor since the 19th century. It carries a label and distance which are partly illegible. The readings are debated. Talbert reads from it at left So[-?-]tv (the word is divided by the join between the two parchment sheets) and at the right IX (TPPlace2349).

The problem seems to have some of its cause in the disorder of the line below, which connects Hierapoli and Lavdicivm pilycvm (Laodicea ad Lycum). That Trallis and its onward figure XXVIII do not belong in this line has meeting mentioned in emendation 8:5,2 above. That the distance VI belongs between the twin cities is evident from the fact that Hierapolis and Laodicea ad Lycum are indeed just 6 Roman miles apart. What then are we to make of the number XV after Hierapoli and the strange little horizontal line? Not to mention the peculiar upwards hook at the page-edge before the line reaches Lavdicivm, perhaps a scribe's slip of the pen in a crevice where erasure was too difficult.
I at first entertained the idea that the chart-maker might here have intended to draw a clone, or second copy, of the highway from Hierapolis to Laodicea, with this variant proceeding onwards via Colossae and Sanaos (RE, DARE) to Apamea. That would make the little horizontal line a link in a greater chain. However the mileages given on the chart are insufficient to cover that great distance, and this idea had to be rejected.

Scholarship in the 19th century concurred in reading the little label as "socratu". Ramsay 1890 initially declared this to be a station belonging on a major route which is "the great road from Ephesus to the east ... given in a fragmentary way in the Peutinger Table", "the backbone of the Roman road system" (49). Radet then proposed its identification as Motella (530, see DARE and Trismegistos), attributing it to a miscopying. Ramsay 1897 was won over, but rejected Radet's explanation, instead proposing the Greek Atyochorion (DARE) could be "corrupted readily into s-cor-atu or socratu" when it was latinized (580). In either case the two writers agree about the locality, with Ramsay writing: "Considering that the village lies on the south bank of the Maeander, and that the road passes along the north bank, as M. Radet rightly recognizes, this interpretation seems indisputable" (see also Ruge, RE III A, 1, 784, and Ramsay's 1899 repetition of the argument in a review of Radet). The location they agree about is amid rugged hills.

Miller also adopts the reading Socratu and places the site somewhat higher up the Maeander's course, as a ford or bridge in the vicinity of modern Dayılar, while contending that the road approaches it from the south via Colossae. He argues that the distances before and after "Trallis" (XV and XXVIII) would accurately trace out this distance (727, Figures 228 and 233). This is not plausible as we take 28 Roman miles to be the Tralles-Magnesia distance (see above). In any case the notion that one place-name has completely overwritten another without trace of the collision requires corroborating or contextual evidence to be compelling.

The only thing I am therefore certain of is that the small line does not denote the highway through Colossae. Nor can So[-?-]tv denote Sanaos, despite the words' similarity of initial and length. What is to be done? I tend to the Radet/Ramsay solution, which supposes an itinerary along a footpath or mule-path through a village sited between the main roads. My emendation lengthens the little line to make plain that it departs from the main route near Hierapolis, that after 15 Roman miles the itinerist reached So[-?-]tv and that after a further 9 miles the traveller must have either regained a main road near Peltai (Pella) or arrived at a dead end. The distances stated are really too small to cover such a large landscape.

I welcome more investigation, a fresh reading of the scans, or pointers to any research  on the topic which I may have overlooked.