HOME (main menu)

TONY CAMPBELL articles

PORTOLAN CHARTS menu

INDEX

complete SITEMAP

What the site is ABOUT

WHAT'S NEW

Data collection
Errors and omissions are inevitable in a study of this scope, particularly in relation to toponymy. The nearly 3,000 names in the current iteration of the Excel spreadsheet (which will continue to be modified) have been gathered from several hundred charts and atlases. The data collection has also been done in various ways at different times over the past 30 years. The initial work in the 1980s was the first of its kind and there was no model available. Returning to the subject - post-retirement and prompted by the ground-breaking toponymic work in the 2007 Ramon Pujades volume, which uniquely presented an almost complete set of images for use on a desk-top computer - I extended the date range up to at least 1600 and went back to look for some of the less usual names. Even restricted to the period covered by Pujades (up to 1469) there were nearly 100 available charts and atlases. With an average of perhaps 1,200 mainland names on each, that brings into play somewhere in the region of 120,000 toponyms.

Clearly nobody could grapple with such a herculean challenge as that. The necessary shortcuts I adopted make it inevitable that there would be omissions, especially among the rarer names encountered in recent years and therefore not checked generally, as had been done initially for all the 'significant' names. There will certainly be errors, too, whether from faulty transcription (by myself or others) or the kind of mistakes it is all too easy to make when manipulating a large Excel spreadsheet. I hope these will be pointed out to me so that I can amend what will always be a 'work in progress'.

Textual and cartographic sources for toponymy
One important area that has not yet been examined in sufficient detail is the relation between text, portolan chart and map. How distinct was the world of written text from that of cartography, the scriptorium from the chartmaker's atelier or the mapmaker's desk? Or were those frontiers, if they existed at all, permeable ones? What is needed is a comprehensive listing, extracted, on the one hand, from the toponymically-rich environment of medieval geographical, travel and pilgrimage literature, and, on the other, from the full spectrum of medieval maps. Such a resource could supplement the data from the two 13th-century maritime guides, the 'Liber de existencia riveriarum' and 'Lo compasso de navegare', as well as this Excel census of portolan chart names. Patrick Gautier Dalché's careful examination of the accounts from the various crusades might serve as the model for the textual component of such an enterprise. The relevant literature from the 12th to 15th centuries would be large but a certain proportion will already have been published. The most appropriate vehicle for such toponymic accumulation and comparison could well be Pelagios Commons. Founded in 2011, this project is working to provide freely accessible online links to the place-names of the ancient and medieval worlds by means of the scanning and indexing of textual and cartographic documents. On this see further, Rainer Simon, Elton Barker, Leif Isaksen & Pau de Soto Cañamares, 'Linking Early Geospatial Documents, One Place at a Time: Annotation of Geographic Documents with Recogito', e-Perimetron, 10, no.2 (2015): 49-59.

Oral transmission
In their most dynamic element, namely their toponymy, the portolan charts need to be understood as the product of orality rather than textual transmission.. This is counter-intuitive when that is often the only part of the chart that can be 'read'. Yet this contention follows logically from the fact that no early chartmaker, as far as we know, was a sailor. The first one to double up in that way was Albertin de Virga (active from the late 14th century) and almost all those who imitated him in that way were fellow Venetians. From the outset, then, fresh information about place-names must have been brought to the chartmaker rather than being personally sought out by him from the local inhabitants, unless he was simply copying from an existing chart. Versions of the early navigation guides, the portolani, might have been available to Vesconte, but there is scant evidence that he would have tried to extract names from one of those and then research where to place each on his chart. Nor could those manuals, or any other textual source of which we are aware, have provided Vesconte with the bulk of the almost 500 toponyms for which his productions provide the earliest sighting on a portolan chart.

So, if the names did not initially come to the charts via a literary route they must have reached the chartmaker orally, in the way described by Fancesco Beccari in 1403, who tells us it was 'masters, ship-owners, skippers and pilots' who were his informants. The immense implications of this for portolan chart history - as discussed in G.3. The staged introduction and repetition of new names - have not, as far as I know, been considered before (though I might have missed that). The documented pattern of introduced toponyms shows numerous piece-meal additions, presumably provided by dozens of individual sailors after many voyages (for name totals, see the blue column in Table A of Black and red names considered together).

If each new name involved oral transmission from a local source, via a sailor informant to a specific chartmaker, how likely is it that the combined results would provide consistency of language or dialect? How would 'quality' be assured, in the sense of accuracy of pronunciation (whose?) and correct positioning on the chart? How many insignificant places or features were chosen, how many of greater importance never reported? And what might constitute 'significance' anyway, to a non-travelling chartmaker who had to take on trust what he was told? In the light of all that, what could, or should, we expect from portolan chart toponymy? It is not clear how any of this could be tested but if a comparative evaluation could be carried out into the place-names on the charts, whose incidence is now far better documented than it was, those historians (whom we should encourage to use this toponymic information) would have a clearer idea of the appropriate ways in which this unique resource could be used.

The pelagi and the charts' initial construction
Future research could with advantage also be directed at the light the Carte Pisane can throw on the portolan charts' formative period, which still remains shrouded in a speculative mist. One factor that could be considered is the likelihood that different techniques might have been used for surveying the various areas. A crucial consideration in the case of the charts' core, the Mediterranean, must be the pelagi, the open-sea crossings of up to 1000 km or even more whose distances are already stated in the early 13th-century 'Liber', with another contrasting series provided at the end of that century in 'Lo compasso' (for comparison between them see Gautier Dalché, 1995, pp.205-19). A GIS exercise to test out the accuracy of those measurements is surely called for. In any event, the fact that some headlands recur frequently in the descriptions of the pelagi point to a potential series of overlapping triangles. Even if those did not form part of the initial construction they could surely have helped to correct its errors.

Such adjustments, which might have been possible in the enclosed Mediterranean, with its large, well-spaced islands, would not have been feasible in the open Atlantic Ocean, particularly in the Bay of Biscay. It would be inappropriate, therefore, to assume analogies between the compilation of the already sophisticated Mediterranean outlines on the Carte Pisane and the tentative attempts to plot the long European Atlantic coastline between the time of the Carte Pisane (c.1290) and that of the 1313 Vesconte atlas.