fol.86v: of Portolan charts and Trabizond

PLEASE NOTE: The folio is referred to in this article by the original foliation published on Yale University’s bibliotecapleyades site.  The Beinecke Library site now gives the same diagram the foliation  “folio 85v and 86r” (Image scan 1006231) – note added 24/02/2016


[14oo words]

al-Tusi's image of Gemini compared with figure in ivory dated to 1st-2ndC AD, Begram

al-Tusi’s image of Gemini compared with figure in ivory dated to 1st-2ndC AD, Begram

I should begin by pointing out that a medieval map or chart is never an isolated creation.  It is important to ask, always, what parallel sources are being assumed.

For the portolan-chart that parallel source was fairly monotonous in its basic form, as:

Trafalgar and Cabo de Santa Maria, levante sirocho, 3 leagues.

Cabo de São Vicente and Tavira, levante sirocho, 12 leagues [1]

A second level of technicality is suggested by a small parchment roll, written in Venetian dialect and commented upon by Marston. It may represent a commissioned voyage because the route apparently back-tracks and triangulates specific points. I mention it for its early suggestion of the commissioned survey, and because one of the points mentioned on the roll is Trebizond [Gk: Τραπεζοῦς, Lat: Trapezus; Turk: Trabzon], that same near which the  minimap’s route  enters Asia minor.

Marston writes:

Several places are mentioned frequently as if they were key spots in map-making. (These have been located on the Becharius portolan chart.) The first is Chavo San Angelo, the easternmost peninsula of Cape Matapan. The second is al Giro (on the roll; ‘Giro’ on the Becharius portolan chart). This is the easternmost point on the south shore of the Bosphorus, where it joins the Black Sea. The third is Trebizond  on the north coast of Asia Minor. Cyprus is also linked with two spots on the south shore of Asia Minor as well as Tripoli and Alexandria. perhaps our unidentified fortress was another such marker.

A third and essentially non-nautical text is that implied for the elegant and illuminated type of portolan-style maps such as the one in Cresques’ compendium. The form given its worldmap, and some its charts certainly do refer to the sea, its moons and tides, but the map’s informing texts – indicated by snippets attached to figures and places – are from the literary traditions.  Among those sources recognised in Cresques’ map are the Travels of Ibn Battuta,  Idrisi’s ‘Book of Roger’, wonderful Persian poems and the market-place tales of the Alf Layla wa Layla  ~  copies of which presumably all existed in the French king’s library before the end of the fourteenth century, since it would scarcely amuse the court to be presented with a short reference to a parallel text to which it had no access.

Apart from the minimap, which does evoke cartographic norms of the Mediterranean and Europe, the remainder of folio 86v is unlike any other among the portolan-chart, or portolan-style  maps. To judge from the manuscript’s botanical section at least, the map’s informing text is more likely a form of periplus than of portolan.

One obvious question  is why the minimap’s route should apparently begin in the north and enter Asia Minor by Trapezus/Trabizond. Apart, that is, from its apparently being a mapmaker’s  marker-point? Why not by Byzantium or even Caffa, since Caffa like Trapezus was a Genoese centre in medieval times, and both places had first been founded directly or indirectly from Miletos.

One reason may be that, in updating Ptolemy’s geographical and astronomical tables (for the two are interconnected), the Byzantine world which was bitterly criticising Ptolemy’s work by the fourteenth century, had developed alternatives for it.

One of the scholars involved in the updating was Gregory Choniades, who took to Trebizond the Zij-I Ilkhani by Nassir al-Din Al-Tousi (al-Tūsi), the most important astronomer from the observatory of Maraga, and whose zij had been composed  c.1270AD. Thereafter his tables were adapted by another Byzantine,  Chrysokokkes, to the standards required for Constantinople, and explained in that last scholar’s introduction to his own work, which is  generally known now as the ‘Syntaxis of the Persians’.

Another, and perhaps unexpected reason may be a mutual interest in matters pertaining to astronomy and its geometries felt by Genoese, Venetians and one must suppose to the equally noted mariners, and notorious pirates, of Dalmatia.

Their presence in the Black Sea, and their association with the Genoese, and/or the Venetians is known but not well understood, nor whether it was positively or negatively intended.  Nonetheless, it was strong, extending as far as Genoese trading centres about the Caspian, where we have no record of any Venetian establishment.  By the Caspian there is also evidence of  Byzantine objects, and of a Scandinavian presence during the Byzantine era.

This region provided a natural corridor for passage between the most heavily Persianised regions of the earlier centuries and Byzantium; as well as a route which continued directly to the west via the Genoese  – exactly the route, in fact, by which the Plague made the same journey.

If fol 86v [which I conclude is a map] should turn out to represent a map* be one which was devised somewhere between the Black Sea and the Caspian, one can only sympathise with scholars hoping to decipher its inscriptions.

*edited for clarity 28/04/2015. I have already analysed the folio, concluding that it is a map of the makers’ world and itinerary,  and had published the greater proportion of my analysis online by 2011.

It is one of the most ethnically and culturally diverse regions in the world, and one would expect here that in addition to the ongoing problems of the script and possible  encipherment,  researchers into the manuscript might have to look contend with often-changed place-names, languages, innumerable local dialects and various scripts adapted or adopted to represent others’ more complex sounds …and so on.

Goldschmidt [2] found, in a Franciscan monastery in Dalmatia and shortly before the first world war,  a portolan-chart of the Caspian Sea, inscribed clearly enough, and even with some clearly Italian terms.   Of it he writes, however that:

For a long time I could not make out what sea it represented. There were plenty of place-names along the coasts, and the writing was clear enough: Odatubissi, Charutebassi, Allamoy, Chiseliha, and so forth. Not a single name could I find that sounded familiar.

The place-names, he eventually found, were local ones, informed by languages that included Italian, and Greek, but also Russian, Persian, Tatar, and possibly Kipchak.  Goldschmidt gave as his opinion on authorship that:

About the author of this portolan[-chart] we cannot draw many inferences, but he may well have been a Dalmatian. .. I think we can assert that his spelling, his system of representing Tatar words phonetically in writing, corresponds to the conventions of the Venetian dialect, which would be the natural orthography for a Dalmatian. In Venetian spelling x would stand for our consonant “sh,” ch for k. When our pilot wrote Chexech I believe he meant us to pronounce it Keshek.

[added note – 9th November 2016] – “When speaking Persian, the Azerbaijanis pronounce k like ch. But when they come to ch they pronounce it ts.

quoted from Robert Byron, The Road to Oxiana (1937) p.66.

All goodwill to those working on the written text, in such a case. But if  the possibly-Dalmatian maker had adopted some variant version of the Roman script to convey the sound of Tatar, the result might well have defied Goldschmidt’s efforts entirely.

The origin of the first makers and users of ms Beinecke 408, and of the map (or chart) forming the greater part of fol.86v may be equally difficult to determine, despite the echoes of older maps, and of portolan, in the inset minimap.  Goldschmidt’s chart was more consistent and in a familiar genre, using a familiar script – and even having many recognizably Italian forms.

On the subject of the Voynich script, I can only say that I am relieved to be addressing the imagery, which is comparatively accessible – all I can contribute on the issue of the written text is an observation: namely, that it appears to me most like one of those derived from Aramaic script. That is, one of the various scripts that are acknowledged derivatives from that used for the official language used throughout the former eastern Roman empire while under Persian rule. Among these recognised derivatives are the scripts used for Sabaic minuscule (Arabia), Tibetan and Ubyk/Abkhaz (Georgia) – which gives an idea of the range over which they occur.

Points in which the map in fol.86v differs from the usual maps of earlier European origin, and again from the portolan style,  include negative ones, such as its not representing seas by parallel, waved lines  and not including Jerusalem.

I’ll have reason to mention the ”parallel waves” motif again, though, in relation to one of the astronomical diagrams.

And so forward..



‘levante sirocho’ is a wind-direction, equivalent to saying ‘East south-east’.
All the windnames and their magnetic equivalents are listed in the wiki article ‘Boxing the Compass’

difference between the portolan and the portolan chart (‘map’) is clearly explained online, in discussing  Michael of Rhodes.


1. Thomas E. Marston, ‘An aid to Medieval Portolan-chart making?’, The Yale University Library Gazette , Vol. 46, No. 4 (April 1972), pp. 244-246.

2. E. P. Goldschmidt and G. R. Crone, ‘The Lesina Portolan Chart of the Caspian Sea’, The Geographical Journal , Vol. 103, No. 6 (Jun., 1944), pp. 272-278.

as opening to further reading on the astronomical side see the article at Muslim Heritage.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s