#243

TITLE: The Vinland Map

DATE:  ca. 1440

AUTHOR: [unknown]

DESCRIPTION: This highly controversial map has only recently been uncovered (1957) and therefore has only a short history of scholarly analysis. This fact not withstanding, it may also claim, during this rather short period, to have undergone more intensive scrutiny and examination in both technical and academic terms than any other single cartographic document in history. Interest has been virtually international in scope and has covered every aspect pertinent to a document purported to be of seminal historical significance:  its historical context, linguistics, paleography, cartography, paper, ink, binding, “worming”, provenance or pedigree, etc. The choice of the name Vinland and the appearance of this Norse discovery prominently displayed on the map was what attracted such immense popular and scholarly attention. Two primary areas of controversy and concern initially swirled around this map: first is the claim of the map’s owner, Yale University, that it was the earliest and only cartographic documentation of the medieval Norse exploration of the northeast coast of America, Greenland and Iceland; and, second, the timing and nature of its discovery, without any known prior reference to it or its accompanying manuscript, either in contemporary or later literature. Therefore, no proverbial rock has been left unturned in subjecting these manuscripts to all of the state-of-the-art technology and worldwide scholarly debate. The result has been a polarization of many prominent authorities from many disciplines into three camps: the ‘believers’, the ‘nonbelievers’, and the ‘undecided’. Three publications that represent most of the pertinent arguments or analysis concerning these documents are: “The Strange Case of the Vinland Map, Geographical Journal; Washburn, W.E. (ed.), Proceedings of the Vinland Map Conference; Kristen A. Seaver, Maps, Myths, and Men, and Skelton, R.A., Marston, T.E., and Painter, G.D., The Vinland Map and the Tartar Relation.

This monograph, like others in the series, will only attempt to summarize significant aspects of the map itself, the purported literary and historical context of the map, and the arguments for its authenticity and major questions/doubts still pending or as yet unanswered.


THE MANUSCRIPT: First brought to the public’s attention in 1957 by an Italian bookseller, Enzo Ferrajoli from Barcelona, the document now known as the Vinland map was discovered bound in a thin manuscript text entitled Historia Tartarorum (now commonly referred to as the Tartar Relation). This manuscript text and map were copied about the year 1440 by an unknown scribe from earlier originals, since lost. The Tartar Relation itself was initially bound as part of a series of volumes containing 32 books of Vincent of Beauvais’ (1190-1264) Speculum Historiale [Mirror of History]. All indications (paper, binding, paleography, etc.) point to an Upper Rhineland (Basle?) source of origin for the present three-part manuscript. In its original form, then, this manuscript consisted of a volume that contained, first, the parchment leaves on which the Vinland map is drawn; then the four books (XXI to XXIV) of Vincent’s Speculum; and, finally, the Tartar Relation. The Vinland map and Tartar Relation had become physically separated from the 15th century Vincent text and were later re-bound together as a separate volume in their present 19th century (Spanish) binding. Only through extremely fortunate circumstances did the ultimate reunion with this particular copy of Vincent’s manuscript occur, also in 1957, in Connecticut.

        The Tartar Relation, in essence, is a shortened version of the more well-known text entitled Ystoria Mongolorum, which relates the mission of Friar John de Plano Carpini, sent by Pope Innocent IV to ‘the King and People of the Tartars’, which left Lyons in April 1245 and which was away for 30 months. Whereas Carpini’s Ystoria is not considered a rare text, no manuscript or printed version of the Tartar Relation has survived, save the one bound with the Vinland map. Clearly, Painter points out, any circulation that the Tartar Relation may have had in separate form was too limited, in view of the normal wastage of medieval manuscripts, to ensure its transmission to the present day.

The fate of the Speculum Historiale was very different, for Vincent’s work became a standard reference book on the shelves of monastic libraries and was constantly multiplied during the next two centuries in manuscript form. It is because the Tartar Relation, one, had the good fortune to become embodied in a manuscript of this popular work (possibly a substitute for, or an addition to, Books XXX-XXXII, which also contained an abridgement of Carpini’s own account) and, two, because, in general, a bulky manuscript like the Speculum Historiale had a better chance of physical survival than a slender one like Tartar Relation bound separately. Based upon various internal and external evidence, it is likely that the juxtaposition of the Speculum Historiale and Tartar Relation first occurred prior to the drafting of the Yale manuscript of 1440, but sometime after the 1255 date of the original production of the Speculum, so that the Yale manuscript is itself a copy of an earlier manuscript, now unknown, in which the Speculum Historiale, Tartar Relation and Vinland map were already conjoined. This foregoing explanation or scenario has been the one put forward by the ‘believers’.

According to these same scholars, the Tartar Relation text does have some significance in its own right as an independent primary source for information on Mongol history and legend not to be found in any other Western source. Additionally the Tartar Relation does act, partially, as one of the chief sources for some textual legends on the Vinland map with regards to Asia.

The associated edition of Vincent’s Speculum Historiale is not complete in itself. As part of Vincent’s encyclopedia of human knowledge entitled Speculum Majus, Speculum Historiale was included as a chronicle of world history from the time of man’s creation to the 13th century, in 32 sections or books. The Yale manuscript contains only Books XXI-XXIV, and comparative calculations indicate that 65 leaves are missing that could account for the table and text of Book XX (these four Books cover the history from 411 A.D. to 801 A.D.). It must be noted that the textual content of these Books show no relationship with either the Vinland map or the Tartar Relation, but, instead, are to be seen within the context of all 32 Books of the Speculum Historiale. Some authorities speculate that possibly the link or actual reference to the Vinland portion of the map could be supplied in the missing 65 leaves.




THE MAP: The map itself is drawn on a single sheet of vellum measuring 27.8 x 41 cm. (about 11 x 16 inches), folded down the center and trimmed to a rectangle corresponding in size to that of the other sheets in the codex, of which it forms a preliminary unsigned quire. The physical association of the map with the manuscript is demonstrated beyond question by three pairs of wormholes which penetrate its two leaves and are in precise register with those in the opening text leaves of the Speculum.   On the blank recto of the first leaf of the map, that is, on the first page of the quire, are written the words:   ”Delineation of the first, second, third part of the Speculum”.    The map is inscribed with 62 geographical names and seven legends.  The handwriting and ink appear to be similar to, if not identical with, those of the legends on the face of the map.  The outlines, names, and legends of the map are executed in a somewhat diluted brownish ink flecked with black. The handwriting is similar in character to that of the manuscript and shows the same idiosyncrasies in individual letters.’’ The map was therefore probably prepared by the scribe who copied the texts of the Speculum and the Tartar Relation.

That the map and the manuscript were juxtaposed within their binding from a very early date cannot be doubted. The physical analysis, together with the endorsement of the map, points with a high degree of probability to the further conclusions that the map was drawn immediately after the copying of the texts was completed, and in the same workshop or scriptorium, and that it was designed to illustrate the texts which it accompanied. These texts may have included, in addition to the surviving books (XXI-XXIV) of the Speculum and the Tartar Relation, other books of the Speculum conjectured to have formed the missing quires and a lost final volume of the original codex.

The association of the map with the texts is reinforced by paleographical examination, which has enabled the hands of the map, of its endorsement, and of the texts to be confidently attributed to one and the same scribe. Further evidence on their relationship and on its character must be sought in the content of the map.

The map depicts, in outline, the three parts of the medieval world: Europe, Africa, and Asia surrounded by ocean, with islands and island-groups in the east and west. Those to the east are named Insule Sub aquilone zamoyedorum and Postreme Insule; in the west are the Atlantic archipelagos, Desiderate insule, Beate isule fortune, Magnæ Insulæ Beati Brandani Branziliæ dictæ, and (to the north and northwest) isolanda Ibernica, Gronelada, Vinlanda Insula.  The map is drawn oriented with north to the top, that is, with names and legends written in horizontal east-west alignment. Whether the northerly orientation is deliberate or represents the most convenient method of arranging the elements of the map within the rectangular space at the cartographer’s disposal cannot be positively determined.  The influence exercised by the space to be filled upon the medieval cartographer’s choice of shape and orientation for his map is a factor which must often be taken into account, particularly in respect of maps in codices or on skins of vellum.

In the design of the Old World the map belongs to that class of circular or elliptical world maps in which, during the 14th and 15th centuries, new data were introduced into the traditional mappaemundi of Christian cosmology. The derivation of the map, in this respect, from a circular or oval prototype is betrayed by the general form of Europe, Africa, and Asia, which are rounded off (or beveled) at the four oblique cardinal points, although the artist had a rectangle to fill with his design.  The Catalan Atlas of 1375 (#235), drawn on six panels which, together, form a rectangle, also appears to have been based on a circular or oval model. In the northeast (that is, at the top right-hand corner, the map being oriented to the north), the coast of Asia is rounded off; and it has been conjectured that the southern half of the map, which would restore the proportions and framework of an ovoid, may have been lost or never executed.

Within the coastal outlines of the Vinland map hydrographic features (rivers and inland seas) are represented; orographic features are absent, and the five cities that are named (i.e., Rome, Jerusalem, Mecca, Cairo, and Alexandria) are not marked by symbols. The whole design is drawn in a coarse inked line, with evident generalization in some parts and considerable elaboration in others. In certain regions, such as East Asia, the precision of detail is doubtless illusory; in others, notably the Mediterranean coasts, Western Europe, and the British Isles, it is as plainly derived from experience, albeit possibly second-/third-hand.

Written in Latin on the face of the map are sixty-two geographical names and seven longer legends. The features named are seas and gulfs, islands and archipelagos, rivers, kingdoms, regions, peoples, and cities. The nomenclature is densest in Asia, where it is largely borrowed from the Tartar Relation or a similar text. The sources of all of the names and each of the legends are examined in great detail in Skelton, et al’s The Vinland Map and the Tartar Relation.  Although they are carefully written, some fragmentary forms and corruptions point to errors by the copyist or to his failure to read or understand what he transcribed. That his original was a map containing (in the manner of medieval cartography) drawings of monarchs with pavilions, cities, or standards is suggested by the frequent indication of kingdoms as Rex . . . or Imperium . . . (such drawings were a regular feature of mappaemundi and of the Catalan charts of the 14th and 15th centuries; the use of the word Rex . . . plainly implies an iconographic representation of a monarch.)

Before proceeding to analyze the geographical delineations of the map in detail, we may briefly survey the antecedent materials, cartographic and textual, to which comparative study of it must refer. It is, of course, not to be supposed that its anonymous maker had direct access to all surviving earlier works with which his shows any affinity in substance or design; but identification of common elements will help us to reconstruct the source or sources upon which he drew. Moreover, in the light thrown on the cartographer’s work-methods and professional personality by his treatment of sources which are to some extent known, we may visualize his mode of compilation or construction from materials which have not come down to us. We may even catch a glimpse of these materials, as they are reflected in the Vinland map, and of the channels by which they could have reached a workshop in Southern Europe (this assumes that the ascription of the manuscript to a scriptorium of the Upper Rhineland is valid).

As noted above, the representation of Europe, Africa, and Asia in the map plainly derives from a circular or oval prototype. Even when the world maps of the late Middle Ages, drawn for the most part in the scriptoria of monasteries, attempted a faithful delineation of known geographical facts (outlines of coasts, courses of rivers, location of places), they still respected the conventional pattern which Christian cosmography had in part inherited from the Romans, and, in part, created.  Until the second half of the 15th century, the habitable world continued to be represented as a circular disc surrounded by the ocean sea, with Jerusalem at the center and east (with the Earthly Paradise) to the top; and the symmetrical pattern of the T-O (or “wheel”) diagrams was still reflected in the more elaborate mappaemundi. Underlying their wealth of geographical detail and (in some cases) legendary lore, this was the model for the 13th century world maps of Hereford* and Ebstorf* (#224 and #226) , Fra Paolino Minorita and  Petrus Vesconte*  (ca. 1320, #228),  Ranulf Higden  (ca. 1350, #232),  Andrea Bianco* (1436, #241), and Giovanni Leardo* (1442-52/3, #242).  Note that the illustrations on the following page are all reduced to a common scale, based upon the north-south diameter; and that the Vesconte, Bianco, and Leardo world maps (indicated by the *) were originally drawn with East at the top and have been re-oriented here with North at the top for convenience of comparison.

Variations of this basic pattern were introduced to admit new geographical information, ideas, or new cartographic concepts.  From the early 14th century the representation of Asia and the Far East is amplified by information from the missions to the Tartars or (exceptionally) the travels of Marco Polo, and from the middle of the 15th century by the reports of Nicolo de’ Conti. The delineation of Africa and the Atlantic islands, in the 15th century, is progressively extended and improved by the Portuguese voyages of discovery and by the information of merchants and other travelers in the Sudan and Near East. The maps in the Byzantine manuscripts of Ptolemy, brought to Italy at the beginning of the century, introduced a radically different world picture, especially in regard to the Indian Ocean.

The circular form of the medieval world map, in the hands of some 14th and 15th century cartographers, is superseded by an oval or ovoid; and even in the 14th century rectangular world maps begin to appear, mainly under the influence of nautical cartography. The traditional orientation, with east to the top, came to be abandoned by more progressive cartographers, who drew their maps with north to the top (following the fashion of the chart makers) or south to the top (perhaps under the influence of Arab maps).

Most of these variations in the form and design of world maps were adapted from the practice of nautical charts and, in the 15th century, of the Ptolemaic maps.  Like the maps of Ptolemy, the portolan charts, or compass charts, were drawn with north to the top; and they were as a rule rectangular in shape. These were not documents of the study, but practical works whose design betrays their function, as aids to navigation by compass or wind-direction. Hence their orientation; hence, too, the network of rhumbs or bearing lines radiating from centers or windroses arranged on the circumference of a circle. The drawing of the coasts is characterized by geometrical forms (arcs for bays, pointed headlands, rectangular or crescent shapes for islands) and by the writing of names at right angles to the coastline. The “normal” portolan charts of the 14th and early 15th centuries embraced the shores of the Mediterranean, of the Black Sea, and of western Europe. Those of Italian authorship, with few exceptions, represented only features on or adjoining the coasts, with the more important rivers which carried trade inland; but charts drawn in the workshops of Catalonia and Majorca attempted a more detailed delineation of the interior. The world maps executed by chart makers such as Petrus Vesconte, Abraham Cresques, and Andrea Bianco, tended to reproduce features of design customary in nautical cartography, particularly the network of rhumb-lines.

If the Vinland map was drawn in the second quarter of the 15th century, and perhaps early in the last decade of that quarter, it would take its place after that of Andrea Bianco and would be contemporary with the output of Leardo, whose three maps are dated 1442, 1448, and 1452 or 1453. While the work of Leardo is considerably more sophisticated in compilation and more “learned” in its incorporation of varied geographical materials than that of Bianco, the world maps of both these Venetian cartographers plainly depend for their general design on models of the 14th century.

The textual sources on which the mapmaker could have drawn in this period may be roughly grouped as follows:


(a) Geographical treatises of classical authors (Strabo, Pliny, Ptolemy, Pomponius Mela) and of medieval writers (Vincent of Beauvais, John Sacrobosco, Pierre d’Ailly, the 14th century Libro del Conoscimiento ) .


(b) Travelers’ reports (e.g. for Asia, Marco Polo and missionaries or merchants before and after him, Nicolo de’ Conti; for Africa, the Portuguese voyages along the west coast, the journeys of merchants into the Sahara or up the Nile River).


(c) The texts contained in the manuscripts which accompany, or formerly accompanied, the Vinland map, namely the surviving books (XXI-XXIV) of Vincent’s Speculum Historiale, the conjecturally missing sections of the Speculum, the Tartar Relation of C. de Bridia describing Carpini’s mission, and perhaps other texts which contributed to delineations or legends on the map.


Of these three groups, the last was plainly that most accessible to the cartographer.  The degree to which he limited his compilation processes to this group of materials (so far as we can reconstruct them), or extended his search for geographical data beyond it, will help to determine the character of his map, as a representation of the known world, and the purpose for which it was drawn.

As previously noted, the outlines of the three continents form an ellipse or oval, the proportions between the longer horizontal axis and the vertical axis being about 2:1. Since the map is oriented with north to the top, the longer axis lies east-west, and the two greater arcs at top and bottom are formed by the north coasts of Europe and Asia and by the coasts of Africa respectively. The elliptical outline is interrupted, in its western quadrant, by the Atlantic Ocean and by the gulfs or seas of Western Europe, and in its eastern by a great gulf named Magnum mare Tartarorum; the curvilinear outline is however continued southeastward from Northern Asia by the coasts of the large islands at the outer edge of this gulf. The only parts of the design which fall outside the elliptical framework are the representations of Iceland, Greenland, and Vinland, in the west, and (less certainly) the outermost Atlantic islands and the northwest-pointing peninsular extension of Scandinavia.

It is not necessary to assume that the prototype followed by the cartographer was also oval in form. In fact his map has striking affinities of outline and nomenclature with the circular world map in Andrea Bianco’s atlas of 1436 (#241). The features common to both maps, and in some cases peculiar to them, are sufficiently numerous and marked (as their detailed analysis will demonstrate) to place it beyond reasonable doubt that the author of the Vinland map had under his eyes, if not Bianco’s world map, one which was very similar to it or which served as a common original for both maps. If this original was circular, the anonymous cartographer’s elongation of the outline to form an ellipse may be explained by his choice of a pattern into which elements not in the original, notably his delineation of Greenland and Vinland and his elaboration of the geography of Asia, could be conveniently fitted, perhaps also, or alternatively, by the need to fill the rectangular space provided by the opening of a codex. These two explanations, taken together, may account for a further modification probably made by the cartographer to his prototype. The northerly orientation of the map should perhaps be attributed to expediency rather than to the adoption of a specific cartographic model, for it enabled the names and legends to be written and read in the same sense as the texts which followed the map in the codex.

If the model for the Vinland map corresponded generally in form and content to Andrea Bianco’s world map, then the variations introduced by its author are not less significant than the general concordance. Apart from the change in orientation and the north-south compression of Eurasia and Africa associated with the adoption of the elliptical shape, they include the addition of Iceland, Greenland, and Vinland; a different delineation (not certainly deliberate) of the British Isles and of the Atlantic extremity of Scandinavia; a variation in the drawing of the Sinus Ethiopicus; revision of the design of East Asia within the framework of the original map; and the suppression of the iconography-pictures of cities, pavilions, and monarchs which is freely scattered over Bianco’s map, was presumably also in the prototype and (as noted above) has left its trace in the nomenclature of the Vinland  map.

Comparison of the geographical outlines of the Vinland map with those of Bianco suggests that its author, while generally following his model, was inclined to exaggerate prominent features, such as capes or peninsulas, and to elaborate, by fanciful “squiggles”, the drawing of a stretch of featureless coast. His personal style of drawing, save perhaps in the outlines of certain large islands, shows no sign of the idiosyncrasies of the draftsmen of the portolan charts, although these have left a clear mark on the execution of Bianco’s world map. Some apparent differences in the rendering of particular major regions in the Vinland map, which may be due to the use of a different cartographic prototype or simply to negligence by the copyist, are discussed in the detailed analysis which follows.


EUROPE:  With the reservations made in the preceding paragraph, the cartographer’s representation of the regions embraced by the “normal” portolan chart of the 15th century, the Mediterranean and Black Seas, Western Europe, and the Baltic, closely resembles that of Bianco in his world map, which reflects his own practice in chart making. The orientation and outline of the Mediterranean agree exactly in the two maps, although in the Vinland map it has a considerably greater extension in longitude, in proportion to the overall width of Eurasia. The distinctive shapes in which Bianco draws the Adriatic, Aegean, and Black Seas reappear in the Vinland map. The Peloponnese and the peninsula in the southwest of Asia Minor are treated with the anonymous cartographer’s customary exaggeration. The outline of Spain is depicted with slight variation from Bianco’s, the Atlantic coast trending NNW (instead of northerly) and the north coast being a little more arched. The remaining continental coasts of Western Europe are drawn precisely as by Bianco. This correspondence is particularly marked in the north-trending line from Flanders to the Skagerrak; in the flatheaded outline of Denmark (peculiar to Bianco and to the Catalan world maps); and in the orientation and shape of the Baltic, typical of the portolan charts and of world maps derived from them.

Scandinavia, as in all maps before the second quarter of the 16th century, lies east-west in both maps; but there is a conspicuous divergence in their treatment of its western end, which both cartographers extend into roughly the longitude of Ireland. Bianco shows Scandinavia as terminating in an indented coast projecting westward with a large unnamed island off-shore, divided from it by a strait; but the author of the Vinland  map has altered the island to a peninsula and the strait into a deep gulf by drawing an isthmus across the south end of the strait. This seems a more probable explanation of the feature than to suppose that it represents the gulf of the northern ocean supposed by medieval geographers to cut into the Scandinavian coast and drawn in various forms by cartographers of the 14th and 15th centuries, from Vesconte to Fra Mauro.  If the delineation by Bianco and that in the Vinland map rest on authentic information and not on the draftsman’s adherence to a cartographic model, it must be admitted that the source for it is unknown. A legend on Fra Mauro’s map (#249) indeed refers to the shipwreck of the Venetian Pietro Querini on the Norwegian coast in 1431, and it has been suggested that Fra Mauro obtained oral information on Scandinavia from Querini, who spent some time in Norway and Sweden and wrote an account of these countries.  If Querini’s evidence could have been used by Fra Mauro, it was doubtless no less available to Andrea Bianco (also a native of Venice), who is recorded to have assisted in the preparation of Fra Mauro’s world map and who included in his atlas of 1436 a special map of Scandinavia and the Baltic, showing knowledge of the Norwegian fisheries and an enriched nomenclature.

In its delineation of the British Isles, the Vinland map again diverges from that in Bianco’s world map. In both, Ireland has the same shape and coastal features, derived from the representation in contemporary Italian charts; and Bianco’s version of Great Britain also is that of the portolan chart makers, with the English coasts deeply indented by the Severn and Thames estuaries and the Wash, with a channel or strait separating England and Scotland, and with Scotland drawn as a rough square with little indentation.  The design of Great Britain in the Vinland map is markedly at variance with Bianco’s: the southwest peninsula is considerably extended, the Isle of Wight is laid down off the southeast coast (not, as by Bianco, in its correct position), the Thames estuary has almost disappeared, the channel between England and Scotland is not shown, and Scotland has a distinctly different outline, with a much more indented east coast and the north coast trending WSW-ENE (instead of W-E as in Bianco) to end in a conspicuous cape.

These differences seem too great to fall within the limits of the license in copying which the author of the Vinland map evidently allowed himself in those parts of his design which agree basically with Bianco’s rendering and may derive from a common prototype. In view of the novel elements in the northwest part of the map, we must reckon with the possibility, but no more, that its author found this version of the British Isles in a map of the North Atlantic which may have served him as a model for this part of his work and from which may stem not only his representations of Iceland, Greenland, and Vinland, but also his revisions of Scandinavia and Great Britain and of the islands between.

In the Vinland map, Europe is devoid of rivers, save for a very muddled representation of the hydrography of Eastern Europe. The lower course of the Danube is correctly drawn as falling into the Black Sea; but the copyist or compiler appears to have erroneously identified it with the Don (which debouches on the Sea of Azov), for the name Tanais is boldly written just above the river, with a legend about the Russians. The “Danube” is shown as rising just south of the Baltic and turning eastward in about the position of Poland; at this point it forks, and a branch flows in a general southeasterly direction to fall into the Aegean. On the source of this farrago, which is in marked contrast with the relatively correct river pattern drawn in Central Europe by Bianco and the chart makers, it is perhaps idle to speculate; it seems to involve a confusion of the Oder, the lower Danube, and the Struma.

The twelve names on the mainland of Europe are, with two exceptions, those of countries or states. The only European city named in the Vinland map is Rome, while Bianco’s world map marks only Paris.  In summary, with regards to the depiction of Europe, Bianco’s world map of 1436 shows little difference from those drawn by Fra Paolino and by Petrus Vesconte over a century earlier, except in its general proportions and some regional details.  Here the author of the Vinland map is following a well-established model.


AFRICA: The general shape and proportions of Africa, extending across the lower half of the Vinland map, also correspond to a type followed, with variation, in most circular world maps of the 14th and 15th centuries, and deriving ultimately from much earlier medieval and classical models.  The great arc of the southern coast, extending from Morocco to East Africa and bounded by the encircling ocean, is common to Fra Paolino (ca. 1320), Petrus Vesconte (#228), the Catalan-Estense map (#246), Andrea Bianco (1436, #241), Leardo (#242), and to Walsperger (1448, #245). That this representation shows an open seaway from the Atlantic to the Indian Ocean does not necessarily imply that the authors of these maps had any knowledge or conviction of the existence of such a passage or that they explicitly repudiated (if they knew) Ptolemy’s concept of a land-locked Indian Ocean (#119). Bianco’s atlas of 1436 indeed presents both versions, for his circular mappamundi is followed by a world map on the Ptolemaic model. In the Vinland map, as in the other circular maps mentioned, the form of Africa is dictated by the traditional cosmographical pattern of land environed by ocean and by the medieval trend toward symmetry and analogy in design. Thus geographers of that time could speak of a double Gades or a double Strait of Gibraltar, of the Pillars of Hercules in the west and the Pillars of Alexander in the east.

Alike in the general form of Africa (with one major variation) and in the detailed outlines of the continent, the Vinland map agrees with Bianco’s circular map of 1436 (which itself has, in this part, close affinities with the design of Petrus Vesconte). The northwest coast was by this date known as far as Cape Bojador, and this section is traced with precision in both maps. Beyond it the coast line, conventionally drawn, trends southeastward with two estuaries or bays similarly shown by both cartographers, although the anonymous map has a slight difference in the river pattern.

The hydrographic pattern of the African rivers in the Vinland map is a somewhat simplified version of that drawn by Bianco, with the Nile (unnamed) flowing northward from sources in southern Africa to its mouth on the Mediterranean and forking, a little below its springs, to flow westward to two mouths on the Atlantic; the western branch is named magnus [fluuius]. (Here again, Bianco’s representation closely follows that of Petrus Vesconte.)

The African nomenclature of the Vinland map, some fourteen names, is conventional, over half the forms corresponding to those of Bianco. Errors made by the anonymous cartographer in common with Bianco, or derived from their common prototype, are the transference of Sinicus mons [Mount Sinai] to the African side of the Red Sea and the location of Imperits Basora [Basra] in the eastern horn of Africa (Bianco also incorrectly places the Old Man of the Mountain (el ueio dala montagna; not in the Vinland map) in Africa instead of Asia).  The single African legend in the Vinland map relates to Prester John.

Africa is the continent in which we have noted some striking links between the Vinland map and Bianco’s world map of 1436. It also seems, although no doubt deceptively, to provide the latest terminus post quem for dating both. They have in common the precise tracing of the northwest coast as far south as Cape Bojador, and if they shared a common prototype, this (it might be supposed) could not have been executed before the voyage of Gil Eannes in 1434. Yet this section of coast had been laid down in very similar form on earlier maps; as Kimble puts it, “Cape Non ceased to be ‘Caput finis Africa’ about the middle of the 14th century”, and “the ocean coast as far as Cape Bojador (more correctly, as far as the cove on its southern side) was known and mapped from the time of the Pizzigani portolan chart (1367)’’.  No documentary record of any voyage south of Cape Bojador before 1434 survives; but the fact that it is marked and even named in earlier maps (as in the Catalan Atlas of 1375 buyetder) forbids us to assume that Bianco’s representation and that of the prototype followed in this part of the Vinland map must postdate 1434.   It is true that by 1448, when he drew his later chart, Bianco had obtained (no doubt when calling at Lagos or Lisbon on the Flanders voyage) intelligence of the Portuguese traverse of the African coast as far as Cape Verde, i.e. up to 1445; but the time lag between discovery and mapping was usually much greater.

The great advance in the knowledge which, from the second half of the 13th century, reached southern Europe about the interior of West Africa and the Sudan was reflected in many maps, from the information collected by merchants on the Saharan trade routes and in the markets of Northwest Africa. The wealth of detail for this region recorded by Carignano, the Pizzigani, and the Catalan cartographers is wholly absent from Bianco’s world map and from the Vinland map. The latter repeats Bianco’s anachronistic reference to the Beni-Marin and his erroneous location of two names; but these aberrations, which appear to be peculiar to Bianco, do not help in dating. Nor was the transference of the Prester John legend to Africa a novelty in the middle of the 15th century.  The other names in Africa all come from older stock, and the marked similarity of outline between the Vinland map and Bianco’s may go back to an earlier prototype now lost.


ASIA: If we are justified in supposing the cartographer’s prototype to have been circular, he, or the author of the immediate original copied by him, has adapted the shape of Asia, as of Africa, to the oval framework by vertical compression rather than lateral extension. Thus, in place of the steeply arched northern coast of Eurasia shown by Bianco, we have a flattened curve which abridges the north-south width of the land mass.      

It is in the outline of East Asia that the maker of the Vinland map introduces his most radical change in the representation of the tripartite world which we find in other surviving mappaemundi and particularly (in view of the affinities noted elsewhere) in that of Andrea Bianco. The prototype is, in this region, not wholly set aside for traces of it remain but rather adapted to admit a new geographical concept which, significantly enough, can be considered a gloss on the Tartar Relation. This concept is the Magnum mare Tartarorum with, lying beyond it and within the encircling ocean, three large islands which appear to derive from the cartographer’s interpretation of passages in C. de Bridia’s text.

This version of East Asian geography is found in no other extant map, and its relationship to the prototype followed for the rest of the Old World is best seen by comparison with Bianco’s delineation, which itself descends from an ancient tradition.  In the sector of his world map lying, by his orientation, between NE (G=greco) and ESE (below L=levante), Bianco draws the coastal outline as a curve interrupted by a semi-circular bay with a large island (tenplon chatai) and trending SE to a prominent cape (gog magog)  followed by a deep gulf and another cape (with the paradixo terrestro), after which the coast runs SW to join that of the island-studded sea which in maps of this type represents the Indian Ocean. The concept of the Caspian Sea (Hyrcanum mare) as one of the four gulfs of the outer ocean was inherited by medieval cartography from Strabo and Isidore of Seville. Rubruck was the first traveler to refute it: “It is not true what Isidore reporteth, namely that this sea is a bay or gulf coming forth of the ocean: for . . . it is environed on all sides with land”.  In the Vinland map, allowing for the flattening of the curve forming the northern edge of Eurasia, we find some evident correspondence with Bianco’s design. The outer coasts of the three large islands (Insule Sub aquilone and Postreme Insule) plainly repeat the outline and orientation of sections of Bianco’s coast; the channel dividing the mainland from the first, or most northerly, of these islands is laid down in the position of the semi-circular bay in Bianco; and the strait between the second and third islands coincides with the deep gulf in the east of Bianco’s map. The inner or western coasts of the three islands and the eastern coast of the mainland, fringing the Sea of the Tartars, have no counterpart in any known cartographic document, but are drawn with elaborate detail of capes and bays. Considering that this sea represents (so far as we know) the cartographer’s interpretation of a textual source, it may be suspected that the outline of its shores was seen by him in his mind’s eye and not in any map.




It is a striking fact, and one which perhaps does credit to his realism, that, in order to admit into his drawing of the Far East a representation derived from a new source under his hand, he has gone so far as to jettison the Earthly Paradise from the design.

The concentration of interest on the Greenland sector has led to the comparative neglect of the Asian section, which has topographical features at least as unusual. The most prominent of these is the Magnum mare Tartarorum [the Great Sea of the Tartars] set between the eastern shores of the mainland and the three large islands on the margin, and occupying an area approximately one-third of that of continental Asia. This great sea is connected in the north with the world ocean by a passage named as mare Occeanum Orientale [the eastern ocean sea]. Against the most northerly island is inscribed Insule Sub aquilone zamogedorum [Northern islands of the Samoyeds]; then in the center Magnum mare Tartarorum.  Between the two more southerly islands is Postreme Insule  [the furthest islands] and then, in the sea, a long legend: Tartari affirmant absq’ dubio q’ noua terra in extremis mundi partibus sic posita nec ultima terra nisi solummodo mare occeanum inuenitur [The Tartars state that without doubt a new land is situated in the extreme parts of the world and that no further land is found, only the ocean sea]. Finally in the southern entrance to the great sea is Mare Indicum [the sea of India] which is repeated further south, to the east of the African promontory. Five names are written on the mainland, near the coastline from north to south: Gogus [Gog, Magog is to the southwest)] Ayran [Bianco’s Airam], Moal [Turkish, Moghal], Zumoal [a nation of the Mongols], Terra Indica [Land of India, possibly the ‘third India’].

Of the five names and legends written on the sea, only one Tartari affirmant . . .  is found in or based on the Tartar Relation, and none are found in Bianco. Of the five in the second group, three are from the Tartar Relation and three from Bianco. As to the long legend from the Tartar Relation, it is to be noted that, significantly, dicta terra [the said land] has been altered to noua terra.

It should be further noted that the Vinland map is at variance with the Tartar Relation in several instances: e.g., in referring to eastern Asia, the Tartar Relation states that it is situated ‘at the very end of the world, and beyond it no land is found, but only the ocean sea’, while the Map shows the three large islands beyond. Again, the northernmost of these islands on the Map has the inscription, Insule Sub aquilone zamogedorum, while the text states that the Samoyeds are ‘poverty stricken men who dwell in forests’ on the mainland of Asia. The Tartar Relation also states that the Tartars have one city called ‘Caracaron’ (Karakorum) but this city does not appear on the Vinland map. It may be recalled here that there is nothing in the Tartar Relation referring to Greenland and Vinland. Some have concluded that, if authentic, the Vinland map was not drawn primarily to illustrate the Tartar Relation.




A reconstructed 12th-14th century Norse world map

after A.A. Björnbo in Cartographia Groenlandica


The remaining islands of Asia are drawn in the Vinland map very much as by Bianco, with some simplification and generalization, and may be taken to have been in the prototype. The three small islands in the Persian Gulf appear in both, though Bianco’s crescent outline for them (of portolan type) is not reproduced by the anonymous cartographer; the large archipelago depicted by Bianco (again in portolan style) in the Indian Ocean is reduced to four islands, and the two bigger oblong islands to the east of them are in both maps.  So too is the large island lying still further east in the ocean, roughly rectangular in shape, with its longer axis N-S, and named by Bianco ixole perlina; comparison with earlier maps of this type identifies the island as Taprobana [Sri Lanka, also a name occasionally applied to Sumatra].

The representation of the interior of Asia in the Vinland map is considerably curtailed by its author’s modifications of his prototype (assuming this to have been a circular mappamundi similar to Bianco’s): in the north, by vertical compression to fit the elliptical outline adopted; in the west, by the drawing of Europe on a larger scale; and in the east by the interpolation of the Sea of the Tartars. Instead of Bianco’s representation of the Arctic zones of Eurasia (with two zonal chords, delineations of skin-clad inhabitants and coniferous trees, and a descriptive legend), the Vinland map has only the two names frigida pars and Thule ultima. The elimination of East Asia by the western shoreline of the Sea of the Tartars has affected the distribution of place names in the Vinland map and its delineation of the hydrography. The four streams issuing from Eden, shown by Bianco as the headwaters of two rivers flowing west and falling into the Caspian Sea from the northeast and south, have disappeared from the Vinland map, in which we see only the two truncated rivers entering the Caspian from the east and south respectively.  A third Asian river (representing the Volga and the Kama) is drawn in Bianco’s map with a T-shaped course, rising from two headstreams and flowing south to a delta on the northwest Caspian. In the Vinland map the double headstreams are absent, and the river is traced from the Caspian to the ocean in northeast Asia, where it is given the name Tatartata fluuius; this delineation is not paralleled in any surviving map and doubtless springs from the cartographer’s misreading of the Tartar Relation. The general form of the Caspian Sea, as a rough oval, is similar in the two maps. The Aral Sea is not represented in either.

Within the restricted space allowed by his revision of the river-pattern and of the coastal outlines, the author of the Vinland map has grouped the majority of his names in two belts from north to south, on either side of the river which runs from the Caspian to the ocean. The nomenclature for Asia, with twenty-three names, is richer than that for the other two continents; some names come from the common stock found in other mappaemundi, but the greater number are associated with the information on the Tartars and Central Asia brought back by the Carpini mission. The location and arrangement of the names cannot, in general, be connected with Carpini’s itinerary (or any other itinerary order), nor with any systematic conception of Central Asian geography. They appear, rather, to be dictated by the cartographer’s need to lay down names where the design of the map allowed room for them. The four longer legends written in Asia or off its coasts are all related, by wording or substance, with the Tartar Relation.

For Asia the compiler of the Vinland map shows the same conservatism in his use of sources as for Africa; and, apart from the modifications introduced from his reading of the Tartar Relation, this part of the map could very well have been drawn over a century earlier. The cartographer’s neglect to use any information from Marco Polo or from the travelers in his footsteps, notably Odoric of Pordenone, is common to all maps before the Catalan Atlas of 1375 (in which East Asia is drawn entirely from Marco Polo) and to most maps of the first half of the 15th century.  The Asian geography of Fra Paolino and of Petrus Vesconte in the early 14th century, while it owes nothing to Marco Polo, betrays the influence of Rubruck in its nomenclature and in the corrected delineation of the Caspian Sea (although it retains the classical version of that sea, in duplicate).  Nothing is added to this stock of information either by Bianco or, apart from the data furnished by the Tartar Relation, by the author of the Vinland map. The latter, however, deserves credit for originality in his removal of the Earthly Paradise, an almost constant component of the mappamundi; for, as Kimble observes, “the vitality of the tradition was so great that this Garden of Delights, with its four westward flowing rivers, was still being located in the Far East long after the travels of Odoric and the Polos had demonstrated the impossibility of any such hydrographical anomaly, and the moral difficulties in the way of the identification of Cathay with Paradise”.


ATLANTIC ISLANDS: Both the “normal” portolan charts and the mappaemundi of the 14th and 15th centuries included the islands and archipelagos of the eastern North Atlantic, so far as they were known; and it is evident that they were in the prototype followed in the main by the author of the Vinland map for his representation of the tripartite world. His delineation of them, indeed, closely resembles that in Bianco’s world map, which is in turn a generalization, with nomenclature omitted, from the fourth and fifth charts (or fifth and sixth leaves) in his atlas of 1436. The affinity between the two world maps, in this respect, is so marked as to distinguish them from all other surviving 15th century maps and to confirm the hypothesis that one has been copied from the other or that both go back to a common model for their drawing of the Atlantic islands.  It is convenient (leaving Iceland, Greenland, and Vinland for separate consideration) to examine first the islands shown in the eastern Atlantic, from north to south.

To the north of the British Isles, the Vinland map marks two islands, presumably representing either the Orkneys and Shetlands or these two groups and the Faeroes. The two islands appear, in exactly the same relative positions, in Bianco’s world map, although they are absent from the charts of his atlas.

To the west of Ireland the Vinland map has an isolated island, also in Bianco; and to the southwest of England another, drawn by Bianco as a crescent. In Bianco’s fifth chart they are named, respectively, y de berzil [i.e. the legendary island of Brasil] and y de uentura.

Further out, and extending north-south from about the latitude of Brittany to about that of Cape Juby, Bianco’s world map shows a chain of about a dozen small islands, drawn in conventional portolan style. These islands, the Azores of 15th century cartography and the Madeira group, are represented in the Vinland map, in more generalized form and without Bianco’s characteristic geometrical outlines, by seven islands, having the same orientation and relative position as in Bianco’s map, and with the name Desiderate insule.  In the fourth chart of Bianco’s atlas, which depicts the chain of islands as in the world map, the names given to the “Azores” (north to south) are corbo marino, coruos, y de sanzorzi y de bentusta, y di colonbi, y de brasil, chapusa and lobo; and those of the Madeira group are porto santo,  y de madera, y dexerta.

Further south, the Vinland map lays down the Canaries as seven islands lying off Cape Bojador, with the name Beate lsule fortune.  The grouping differs slightly from that of Bianco, who arranges the seven islands, as in all charts which show them before the date of his atlas, in a line lying roughly ENE-WSW. It seems unlikely that the anonymous cartographer’s re-disposition of the islands, any more than the difference in their outlines from the chart-designs of Bianco, has any special significance. In the fourth chart of Bianco’s atlas, the islands are named (from east to west) y del fero, y de le palme, y de gomera, y de inferno, y de chanaria, p. santo, forte uentura, paruego( ?), p. sable, gracioxa, y de lancilotto, rocho.

In the ocean to the west of the Desiderate insule, the Vinland map has two large islands, irregularly rectangular in shape with the longer axis north-south; the name Magnæ Insulæ Beati Brandani Branziliæ dictæ is written between them. Their agreement in outline with the two large islands laid down in exactly the same positions at the western edge of Bianco’s world map is striking: in particular, the indentation of the east coast of the more northerly island and the peninsular form of its southern end, the squarish northern end of the other (and larger island) and its forked southern end, are common to both maps. These islands (unnamed in the two world maps) are Satanaxes and Antillia, which make their first appearance in a map of 1424 and have been the subject of extensive discussion by historians of cartography. In point of date, Bianco’s atlas of 1436 is the third known work to show the Antillia group, and the fourth chart of the atlas names the two major islands y de la man satanaxio and y de antillia. Since the outline given to these two islands both in the world map and in the fourth chart of Bianco’s atlas is easily distinguishable from that in any 15th century representation of them, the concordance with the Vinland map in this respect is significant. Here again we have plain testimony to the derivation of the Vinland map from a cartographic prototype, and to the character of this prototype.


ICELAND, GREENLAND, VINLAND: In the extreme northwest and west of the map are laid down three great islands, named respectively isolanda Ibernica, Gronelada, and Vinlandia Insula a Byarno re et leipho socijis, with a long legend on Bishop Eirik Gnupsson’s Vinland voyage above the last two. That they lie outside the oval framework of the map suggests that they were not in the model, apparently a circular or elliptical mappamundi, which the cartographer followed in his representation of Europe, Africa, and Asia.

The three islands are drawn in outline, in the same style as the coasts in the rest of the map; and there can be no doubt that the whole map, including this part of it, was drawn at the same time and by a single hand. For this part of the map there are no earlier or contemporary prototypes of kindred character for comparison, and indeed (except in respect of Iceland) no representations with much apparent analogy can be cited before the late 16th century.   Any attempt to divine the cartographer’s sources of information and their character can only proceed from analysis of his delineation, in the light of impressions formed about his working methods and style; from examination of the knowledge of lands in the northwest Atlantic available in Europe in the early 15th century, and of the channels by which it may have been transmitted; and from the scrutiny of later maps or texts for data that might have survived from the period in which the Vinland map was made. Here we have at once the most arresting feature and the most exacting problem presented by this singular map.

Iceland lies to the west and a little north of Scandinavia, and NNW of the British Isles. It is drawn as a rough rectangle, with a prominent west-pointing peninsula in the northwest, the EW axis being considerably longer than the N-S axis. Greenland, somewhat larger than Iceland, is dog-legged in shape, with its greatest extension from north to south. Its outline, on the east side, is deeply indented and in the form of a bow, the northeast coast trending generally NW-SE to the most easterly point, and the southeast coast trending NNE-SSW to a conspicuous southernmost promontory, in about the latitude of north Denmark; from this point the west coast runs due north, again with many bays, to an angle (opposite the easternmost point) after which it turns NW and is drawn in a smooth unaccidented line to its furthest north, turning east to form a short section lying WE. The approximation of the east coast and of the southern section of the west coast to the outline in modern maps leaps to the eye.   The delineation of Greenland as a large island is striking because the Danish cartographer Claudius Clavus portrayed Greenland on his maps, from 1427 onwards, as a peninsula of Europe, which became the usual 15th century interpretation.

Between Greenland and Vinland lies a channel slightly wider than that between Iceland and Greenland. The northernmost point of Vinland is shown in about the same latitude as the south coast of Iceland and somewhat lower than the north coast of Greenland; and its southernmost point in about the latitude of Brittany. Between these points Vinland is drawn as an elongated island, the greatest width being roughly a third of the overall length; the somewhat wavy details of the outline, if compared with this cartographer’s technique in other parts of his map, seem to be conventional rather than realistic. The island is divided into three great peninsulas by deep inlets penetrating the east coast and extending almost to the west coast. The more northerly inlet is a narrow channel trending ENE-SSW and terminating in a large lake; the more southerly and wider inlet lies roughly parallel to it. The name Vinlanda Insula . . . is written to the right of the northernmost peninsula.  No off-shore islands are shown round the coasts of Iceland, Greenland, or Vinland (the circle visible off the SE coast of Greenland is in fact a hole in the vellum, easily identified (like the other holes) by the square outline of the patch at the back of the vellum, showing through to the face).

The land depicted to the west of Greenland in the northwest Atlantic has the following legend (in translation): Island of Vinland discovered by Bjarni and Lief in company.  A longer legend describes their discovery and records a visit to Vinland in the last year of Pope Pascal (i.e., 1117-18) made by Eric, Bishop of Greenland.


NAMES & LEGENDS: Names and legends are transcribed ad litteram, with indication of line divisions; contractions are not extended. Translations of the legends only are given. To facilitate location of the name and legends on the original map, numbers have been added, keying them to the reproduction at the end of this section.


EUROPE

(a) The British Isles and other islands of NW Europe

1. Anglia/ terra/ insula = England.

[Isle of Wight, unnamed]

[Ushant, unnamed]

2.  Ibernia = Ireland.

3. Ierlanda insula: perhaps a variant of Irlanda, as a doublet of Ibernia; but more probably, in view of the position of the name, a corruption of one of the names current in 14th and 15th century cartography for islands or island groups north of Scotland.

[Two islands, unnamed, north of the British Isles. Residual from the representation described under the previous name, the large elliptical island being suppressed. Presumably intended for the Orkneys and Shetlands, or one of these groups and the Faeroes].


(b) Northern Europe

4. Dacia = Denmark. The form (for Dania) common in mediaeval cartography, and found in many charts and world maps.

[An island, unnamed, in the mouth of the Baltic]

[An island, unnamed, at the head of the Baltic]

5. Rex/Noruicorum = Norway.

6. Rex Suedorum = Sweden; placed this name south of the Baltic.

7. frigicla pars. Mediaeval world maps commonly show a pair of such legends, indicating the regions, outside the oikoumene, too cold or too hot for human habitation.


(c) Central and Eastern Europe

8. Apusia = Prussia.

9. Buyslaua = Breslau (Bratislava), where Carpini’s party stopped on the outward journey and was joined by Friar Benedict.

10. Rusij = Russia.

11. Rusij habent imperiu contiguum ex parte orietis mogalo tartaro / m. kan ex pane boreali habent mare frigidum et magnum flumen / quod medium montiii insularum q’ transit inter glacier borealis occeani / pro grediens [The Russians have their empire adjoining on the east that of the Mongols and Tartars of the Great Khan, to the north they have the frozen sea and a great river which passes through the midst of the mountains and islands, debouching amongst the ice of the northern ocean].

12. Tanais = river Don.


(d) Mediterranean countries

13. Rex/ francoru = France.

14. Hispanoru rex = Spain.

15. aben: plainly a corrupt and probably truncated form, which cannot be assimilated to any name on the south coast of Spain in other charts and maps of the 14th and 15th centuries.

16. vrbs/Roma = Rome.

17. Imperium/romanorum. Written in SE Europe, denoting the Eastern Empire.




(e) The Mediterranean

[The islands of Majorca and Minorca, Corsica and Sardinia, Sicily, Corfu, Cephalonia, Crete, Cyprus, and the Aegean Archipelago are drawn but unnamed]


AFRICA

(a) West of the Nile

18. Rex/Marr = Morocco. Although the second word is truncated, no trace of further letters can be seen in ultraviolet light.

19. Bela/rex The Beni Mann dynasty which ruled in Fez and Morocco in the 13th century and in Tlemcen until 1407.

20. Tunesis / rex = Tunis.

21. Maori = Mauri? The name is, however, placed too far inland and too far east for Mauretania, and this may be a corruption of another name in the prototype, e.g. the second element of Syrtis maioris (Ves Syrtes maiores).

22. Phazania = Fezzan. The Roman name, first used by Pliny; also in Ptolemy, but not in his maps. This name does not seem to occur in any other medieval map.

23. magnus [fl]uuius. The concept of the Western Nile, or “Nile of the Negroes”, represented in mediaeval cartography arose from the identification of the Niger, by some classical writers, as a western branch of the Nile and from subsequent confusion of the Niger, the Senegal, and the Rio do Ouro (south of C. Bojador). The name Magnus fluvius is not in any other map, but is implied in some texts, e.g. the Libro del Conoscirniento (ca. 1350): “The river [Nilus] forms two courses, the greater, flowing to the westward, called the Rio del Oro”; and Alvise Cadamosto (ca. 1468): “This river is said to be a branch of the river Nile . . . This river has many other very large branches, besides that of Senega, and they are great rivers on this coast of Ethiopia”.

24. Sinus Ethiopicus (apparently a scribal corruption of Sinus Ethiopicus).

25. Rex/ Soldanus = Egypt (Babilonia being the medieval name for Cairo).

26. Alexandria


(b) East of the Nile

27. Chaira (somewhat to the east of the river).

28. Sinicus/mons = Mount Sinai; placed on the African side of the southern end of the Red Sea.

29. Imperifi Basora = Basra; placed at the eastern tip of Africa.

30. Prestis Johannis Iste sunt terre populose ad meridiem prope sinum / australe posite sed diuerso ydiomatum in unum deum et in dnm ihm / xpm credunt ecclesias habent in quibus orare possunt dicunt[ur] [These are the populous lands of the Prester John situated in the south near the southern gulf. Although of diverse languages it is said that they believe in one God and in our Lord Jesus Christ and have churches in which they can pray].

31. Emibar superior = Zanzibar. The forms of the name are corruptions of the medieval name for the coastal region of East Africa known to the Arabs as Zanzibar

32. Ethiopi


ASIA

(a) South of the Black and Caspian Seas.

33. Mecca

34. Jerusalem

35. Samaca = Shamaka, in the eastern Caucasus, NW of Baku.

36. Nestoriani assidue processerunt usque ad terrain Kitay / ite reliqui filij israel quos dominus monuit transiuerunt / uersus montes hem modos quos superare non potuerunt [The Nestorians pressed on assiduously to the land of Cathay. The remaining children of Israel also, admonished by God, crossed toward the mountains of Hemmodi, which they could not surmount]. This legend, the first part of it seems to be distilled from references to the defeat of “Nestorians” by Genghis Khan and their diffusion in Asia. The second part of the legend relates to the medieval belief that the Ten Tribes of Israel who forsook the law of Moses and followed the Golden Calf were shut up by Alexander the Great in the Caspian mountains and were unable to cross his rampart. The “shut-up nations” were also identified with Gog and Magog and with the Tartars, who were held to be descended from the Ten Tribes.

37. Kemmodi/ montes/ Superiores/Excels. siue/Nimsini = the Himalayas.

38. Terra/Indica. This name is placed in the approximate position of India media of Andrea Bianco, who (like most medieval geographers) distinguished three Indias—minor, media, and superior.


(b) North of the Black and Caspian Seas

39. montes inferiores abrupti / In hanc terram primi fratres nostri ordinis iter faciendo ad tartaros / mogalos samogedos [et] indos transiuerunt nobisc per obedientiam / et subieccionem tam debitam q’ deuotam lnocentio sanctissimo Patri / nostro Pont. max. per totum occidentem et in reliqua parte usque ad mare occeanum orientate [Steep mountains, not very high. The first to cross into this land were brothers of our order, when journeying to the Tartars, Mongols, Samoyedes, and Indians, along with us, in obedience and submission to our most holy father Pope Innocent, given both in duty and in devotion, and through all the west and in the remaining part [of the land] as far as the eastern ocean sea].

40. lmperiu Tartarorum (written on the River Don).

41. magnus ka[n]. The form Magnus Canis is found in many world maps, usually following Marco Polo; the position of the name, between the Sea of Azov and the Volga, is curious, since Carpini only reached the camp of Kuyuk Khan after entering “the country of the Mongols, whom we call Tartars”, i.e. in Mongolia, far to the east. The cartographer has perhaps confused the Great Khan (Kuyuk) with Batu, Khan of Kipchak, whom the Carpini mission encountered on the Volga.

42. Tartaria mogalica. Presumably Carpini’s “country of the Mongols, whom we call Tartars”, i.e. Mongolia.

43. Zumoal. According to Carpini, one of the nations of the Mongols: “… Su-Mongal, or Water-Mongols, though they called themselves Tartars from a certain river which flows through their country and which is called Tatar (or Tartar)”.

44. Moal: Dr. Boyle writes: “Rubruck’s Moal represents the Turkish Moghal or Moghol, the native Mongolian word being Mongghol.”

45. Mogali

46. Ayran (NE of the Caspian) Perhaps Sairam in Turkestan, a station on the old highway, east of Chimkent and N.E. of Tashkent.

47. Magog.

48. Gogus. As early as the 13th century (e.g. in the Ebstorf mappamundi, Book II, #224) the cannibal nations of Gog and Magog, enclosed within the mountain rampart built by Alexander, were placed by Europeans in northern Asia. Hence their identification with the Tartars and their location by Marco Polo in Tenduc, with a probable reference to the Great Wall of China. “The theory that the Tartars were Gog and Magog led to the Rampart of Alexander being confounded with the Wall of China or being relegated to the extreme N.E. of Asia, as we find it in the Carta Catalana”; Gog and Magog are thus represented, usually behind Alexander’s wall, in NE Asia by most cartographers of the 14th and 15th centuries.

49. Kytanis = Cathay. The Khitai, who ruled in China for three centuries before the Mongol conquests under Ogedei and Kublai, “originated the name of Khitai, Khata or Cathay, by which for nearly 1,000 years China has been known to the nations of Inner Asia”. The name is misplaced to NW (instead of SE) of the land of Mongols.

50. Termacus Rex = Sarmatia

51. Tartartata fluuius. Carpini’s statement that “they called themselves Tartars from a certain river which flowed through their country” (see above, under Zumoal) reflects the opinion of other 13th century writers, such as Matthew Paris. The course of the river of the Tartars, as depicted in the Vinland Map, recalls Rubruck’s statement that the Etilia (i.e. Volga) flowed from Bulgaria Major, on the Middle Volga, southward, “emptying into a certain lake or sea . . . called Sea of Sirsan [?= the Mare Hyrcanum of 14th and 15th century world maps, i.e. the Caspian]”.

52. Thule ultima. In medieval cartography generally Thule is represented as an island north or NW of Great Britain; some writers identified it as Iceland. The Vinland Map’s location of the name, in the extreme north of Eurasia, places Thule (as Ptolemy and other classical authors did) under the Arctic Circle.


Inland Seas

[The Black Sea and Sea of Azov are drawn but not named]

[The Caspian Sea is drawn but not named. Carpini and his companions seem to have confused the Black, Caspian, and Aral Seas, the last of which was probably not visited by Rubruck and remained unknown to Marco Polo and to the mapmakers who followed him. The Vinland Map correctly distinguishes the Black and Caspian seas. Following an older model, it also represents the Caspian as a gulf of NE Asia;


Ocean and Islands

53. Terre non satis perscrutate/ posite sunt inter boreales glacies / ab iisdem abdite [Lands not sufficiently explored. They are placed among the northern ice and concealed by it].

54. mare Occeanum Orientale.

55. Magnum mare Tartarorum. The name and delineation probably embody the mapmaker’s interpretation of what he had read or been told of the Caspian Sea. The name Magnum Mare was applied by Carpini and Friar Benedict to the Black Sea while Rubruck called it Mare maius. Members of the Carpini party were somewhat confused about the courses of the rivers flowing into the two seas, supposing the Volga to enter the Black Sea.

56. Tartan i affirmant absq' dubio q’ / noua terra in extremis mundi partibus / sit posita nec ultima terra nisi solummodo / mare occeanum inuenitur [The Tartars affirm beyond doubt that a new land is situated in the outermost parts of the world, and beyond it no land is found but only the ocean sea]. The last phrase of the legend is inconsistent with the geographical ideas of the Mongols, contrasting with those of the Franks, as reported by Rubruck: “as to the ocean sea they [the Tartars] were quite unable to understand that it was endless, without bounds”.

57. Insule Sub aquilone zamogedorum. These islands, and the Postreme Insule, are associated with the cartographic concepts in the two preceding legends (see notes on Magnum mare Tartarorum and on Tartari a rmant . . .).

58. Postreme Inside.

59. mare Indicum

[Three small islands, unnamed, in the Persian Gulf]

[Six large islands, unnamed, in the Indian Ocean, with many more islands; = the Maldives and other archipelagos of the western Indian Ocean.]

[A larger oblong island, unnamed, to the east of the six islands = Ceylon.]


THE ATLANTIC AND ITS ISLANDS

60. Mare Occeanum. Twice written.

[An island, unnamed, west of Ireland = Island of Brasil. Thus, with characteristically circular shape, in many charts of the 14th and 15th centuries]

[An island, unnamed, SW of England. The usual name for this crescent-shaped island, in 14th and 15th century charts, is I. de Man (or Mam), sometimes Mayda]

61. Desiderate/insule. This is written in the center (between the fourth and fifth, counting from the north) of the chain of seven unnamed islands extending in a line N-S from the latitude of Brittany to that of C. Juby. These are the Azores, laid down in charts with this position and orientation from the middle of the 14th century to the end of the 15th, and the Madeira group. In many 15th century charts the chain has (usually written in larger lettering to the north of Madeira) the general name Insule Fortunate Sancti Brandani, or variants. The classical Insulae Fortunatae were the Canaries, the only group known in antiquity, and the association with St. Brendan arose in the Middle Ages. The general name Desiderate insule given in Vinland Map to these islands is not found in any other map; the only explanation we can hazard is that it may allude to the Portuguese attempts at discovery and colonization of the Azores from, probably, 1427 onward.

62. Beate fsule/ fortune. The Canaries, represented as seven unnamed islands.

63. Magnelnsulce/Beati Brandani/Branzilice/dictee. The name is placed westward of, and between, two large unnamed islands, to which it plainly refers. As the examples already cited show, the name Insulee Sancti Brandani (in variant forms) is commonly ascribed by chart makers to the Azores-Madeira chain. The Vinland Map is the earliest known map to move the name further out into the ocean and apply it to the Antillia group, the word magnce being added to justify the attribution and make a clear distinction from the smaller islands to the east. The alternative form Branzilio (or Branzilia), suggesting an association with the name of the legendary island of Brasil, is not found in any other surviving map. The name Brasil, in many variants, was generally applied by cartographers of the 14th and 15th centuries (a) to a circular island off the coast of Ireland, and (b) to one of the Azores, perhaps Terceira; the variant forms of the name include Brasil, Bersil, Brazir, Bracir, Brazilli.


ICELAND, GREENLAND, VINLAND

64. isolanda Ibernica = Irish Island. In no other map or text is the form Isolanda found, or the epithet Ibernica annexed to the name for Iceland. Medieval mapmakers, from the 10th century (Cottonian map, Book II, #210) onward called the Island or Ysland (v.l. Hislant), Islandia or Yslandia.

65. Gronelada. The Icelandic name Groenland, in variant forms (including the latinization Terra viridis), is used in all early textual sources. The name was introduced into cartography by Claudius Clavus (1427) as Gronlandia.

66. Vinlanda Insula a Byarno re pa et leipho socijs [Island of Vinland, discovered by Bjarni and Leif in company]. If the large island to which this name is applied represents the three lands, Helluland [Flagstone Land], Markland [Forest Land], and Vinland [Wine Land], discovered by Leif Eiriksson from north to south and divided here by the two deep inlets, either the name Vinlanda is here used collectively for all three, or it has been erroneously placed against the northernmost instead of the southernmost land. In the earliest written record of Vinland, that of Adam of Bremen (ca. A.D. 5070), it is referred to as an island; in the 12th-century Icelandic Geography, Helluland and Markland are said to be islands, and it is suggested that Vinland is “connected with Africa”; none of the other Norse sources referring to Vinland, from the earliest in the Islendinga bok of Ari Frode (ca. 1122-24), contains any indication that Vinland was thought to be an island. The Icelandic name was Vinland, variants of which are used in all the medieval texts.

The American landfall of Bjarni Herjolfsson in A.D. 985 or 986 and his sale of a ship to Leif Eiriksson in (probably) A.D. 1001 rest on the sole authority of the “Tale of the Greenlanders” in the 14th-century Flatey Book. This legend on Vinland Map, if it faithfully reproduces a genuine record, accordingly authenticates Bjarni’s association with the discovery of Vinland and adds the significant information that he sailed with Leif.

67. Volente deo post longu iter ab insula Gronelanda per meridiem ad / reliquas extremas partes occidentalis occeani mans iter facientes ad / austru inter glacies byarnus et leiphus erissonius socij terram nouam uberrima / videlicet vinifera inuenerunt quam Vinilanda [? or Vimlanda] insula appellauerunt. Henricus / Gronelande regionumq finitimaru sedis apostolicae episcopus legatus in hac terra / spaciosa vero et opulentissima in postmo anno p. ss. nrj, [=pontificis or patris sanctissimi nostri] Pascali accessit in nomine dei / omnipotetis longo ternpore mansit estiuo et brumali postea versus Gronelanda redit / ad orientem hiemale deinde humillima obedienci a superiori vo- / lutati processit [By God’s will, after a long voyage from the island of Greenland to the south toward the most distant remaining parts of the western ocean sea, sailing southward amidst the ice, the companions Bjarni and Leif Eiriksson discovered a new land, extremely fertile and even having vines, the which island they named Vinland. Eric [Henricus], legate of the Apostolic See and bishop of Greenland and the neighboring regions, arrived in this truly vast and very rich land, in the name of Almighty God, in the last year of our most blessed father Pascal, remained a long time in both summer and winter, and later returned northeastward toward Greenland and then proceeded [i.e. home to Europe ?] in most humble obedience to the will of his superiors.] Two historical events are here described: first, a voyage of discovery by Bjarni [no patronymic] and Leif Eiriksson “southward” from Greenland to Vinland; and second, a visit to Vinland by Bishop Eirik [Gnupsson] in a specified year, viz. A.D. 1117, his stay in the country, and his return.

The voyage of Bjarni and Leif is also referred to in the legend Vinlanda Insula . . . the additional information in the longer legend relates to the course (“southward amidst the ice”), to the character of the land discovered (“extremely fertile and even having vines”), and to the bestowal of the name Vinland by the discoverers.

The question “what kind of map is this?” the answer must be: a very simple map, simple both in intention and in execution. The links between the map and the surviving texts which accompany it strongly suggest that it was designed to illustrate C. de Bridia’s account of the Carpini mission (i.e., the Tartar Relation). They also prompt the suspicion that missing sections of the original codex may have been illustrated by the other novel part of the map, namely its representation of the lands of Norse discovery and settlement in the north and west of the Atlantic. It might be said that the dominant interest of the compiler or cartographer lay in the periphery of geographical knowledge, to which indeed the accompanying texts relate; and such a polarization of interest is exemplified in the themes of the seven legends on the map.

In finding cartographic expression for the geography of his texts, the maker of the map has practiced considerable economy of means. The design is reduced to its barest elements of coastal outlines and nomenclature. There is hardly a superfluous line in the drawing; any decorative detail which the compiler may have found in his map sources has been suppressed; no representation of topographical facts, such as relief, is attempted; and from the stock of names commonly found in world maps of the 14th and 15th centuries, even in small ones like that of Andrea Bianco, only a meager selection has been made. In a map of this form, drawn like the circular mappaemundi, on no systematic projection, we do not of course expect to find graduation for latitude and longitude, even if the quantitative cartography of Ptolemy had been known to its author. The only mark of modernization in design is the orientation of the map to the north.

Examination of the nomenclature has suggested that the Vinland map, in the form in which it has survived, is the product of a stage of compilation (the work of the author or cartographer) and a subsequent stage of copying or transcription (the work of a scribe who was perhaps not a cartographer). There is a decided incongruity between, on the one hand, the care and finish which characterize the writing of the names and legends, with their generally correct Latinity, and, on the other hand, the occurrence of onomastic errors which knowledge of current maps and geographical texts or reference to the prototype used by the compiler would have corrected. What other undetected changes or corruptions the copyist may have introduced into the final draft we cannot tell, since his original, the compiler’s preliminary draft, is lost. If Bianco’s world map be assumed to have resembled, in form and content, the model followed by the compiler for the tripartite world, we can however assess the performance of the final copyist by comparison of his work with Bianco’s map, so far as it takes us. He emerges from this test on the whole creditably, for the outlines of the two maps are (as we have seen) in general agreement. The chart-forms characteristic of Bianco’s style of drawing are not reproduced in the Vinland map; at what stage these disappeared we do not know, and they were not necessarily in the original model followed by the compiler. All the major divergences, in the geographical elements of the Vinland map, from the representation in Bianco can be traced to its compiler’s reading of the Tartar Relation or to changes forced upon him by the design adopted.

The process of simplification described above was presumably carried out in the compilation stage. If we are justified in supposing the scribe who made the surviving transcript of the map to have been ignorant or naive in matters of geography, the draft which he had before him for copying must have been the product of selection and combination already exercised by the compiler.




These considerations must govern our judgment of the date and place of origin to be ascribed to the map. The evidence, internal and external, which indicates that the manuscripts were produced in the Upper Rhineland in the second quarter of the 15th century can only apply to the map included in the codex.  If this is a more or less faithful copy from an earlier draft, any inferences on the date, place, and possible authorship of the preliminary draft, that is, of the compilation of the map, must be drawn from study of its content and of the source materials which went into it. This will furnish a terminus post quem for the making of the map and some evidence of the character and resources of the cartographer.

For the purpose of analysis the map divides itself into two distinct parts. In its representation of Europe, Africa, and Asia it can be referred to, and collated with, not only extant cartographic works of similar character and design, but also a text which is bound in the same volume and to which its content is clearly related. For its delineation of lands in the north and west Atlantic, the cartographic prototypes (if it had any) either have not survived or have been so transformed as to be difficult to identify; and if the codex once included a text relating to these lands, this too has now disappeared.

The major novelty of the Vinland map was that it appeared to show the full 360°.   It thus placed itself chronologically well before these later maps which showed or claimed to show the whole surface of the earth, the Florence Genoese  map (1457), the Toscanelli map (described in his 1474 letter), the Yale Martellus  world map of 1490, and the Behaim  globe of 1492 (# 248, #252, #256, and #258).  It was, therefore, of major interest to historians concerned with the growth of ideas about, and knowledge of, the Earth’s surface. There were controversies, it is true, about just how much could be read into the Map and there was, so far, no general agreement about its precise implications for the study of cosmography in the 15th century. Moreover, by showing islands liberally dispersed in the oceans east and west of the landmass, it opened up the question, some 30 years before it was discussed by Toscanelli, of the possibility of reaching Asia from Europe, not by a voyage of 16,000 to 20,000 km but by easy stages of 1,600 to 2,400 km which were within the capacity of oceangoing vessels at this time. The appearance of Vinland (of special interest to American scholars intent on enlarging the range of their own history), Greenland and Iceland raised the whole question of the level of navigational, geographical and cartographic technique in Scandinavia in the mid-15th century, on which the historians found the cartographers sharply divided but almost all disturbed in some degree by the nature of the Greenland outline which appeared on the Map. The depiction of Vinland, it was agreed, was notional rather than representational.

The Map was interesting to historians as apparent evidence that Norse voyages of the 11th and 12th centuries were known in the Upper Rhineland in the mid-15th century, and consequently that some continuity of knowledge existed between the early discovery of what we know as America and the rediscovery of western lands in the later 15th century. This information was limited in its scholarly impact by the failure of historians to find any other evidence of continuity or to discover that the evidence contained in the Map had ever been known to anyone concerned with exploration either before Columbus’s voyage or after. Finally, the inscriptions on Greenland and Vinland in the Map offered a few scraps of information which differed somewhat from what was commonly accepted. Bjarni, it was implied, had accompanied Lief on his first discovery of Vinland; Bishop Henricus, the Eirik of the annals, who was said there to have gone to look for Vinland, was stated to have found it, and at a different date. Subsequent scholarship has, however, tended to minimize the value of these novelties.

As a world map the Vinland map does not fit into the framework of medieval cartography as conceived in Western Europe. It bears no title or draftsman’s name, or date, and has been attributed to c. 1440 on the argument that it is in the same hand as the Tartar Relation, of which the Map is held to be an integral part. Some students have been reluctant to accept these propositions; the provenance of the Map had not been established, the nomenclature also presents difficulties, as does the representation of certain topographical features, in particular the accurate delineation of Greenland, a point heavily stressed by the editors of The Vinland Map and the Tartar Relation. Much argument has centered around the possibility that Norse voyagers might have circumnavigated and charted its coasts, or provided a written description of them.


SOURCES:  Analysis of the nomenclature and of its affinities with other maps or texts suggests some general remarks about the Vinland map and about its mode of compilation. In respect of toponymy, as of outline and design, the correspondences between this map and Bianco’s world map of 1436 are almost certainly too extensive to be explained by coincidence. It seems to be an inescapable inference that the author of the Vinland map (or of its immediate original) employed no eclectic method of selection and compilation from a variety of sources, but was content to draw on a single map, which must have been very like Bianco’s, for the majority of the names, as well as the outlines, in Europe, Africa, and part of Asia. For convenience of reference, this Bianco-type original, which has not survived, will be cited as O. The fact that, in regard to a few names or delineations, the Vinland  map seems to show affinities with charts in Bianco’s atlas of 1436, rather than with his world map, may suggest that O1 was of Bianco’s, or at any rate of Venetian authorship.

In those parts of the map in which (as noted above) the influence of O1 predominates, there are very few names which cannot be traced to it or to the common stock of toponymy found in contemporary cartography (and therefore perhaps in O1). Some of these anomalies (Aipusia, aben, Maori) are plainly the product of truncation or corruption in transcription, and indicate that the draftsman lacked the knowledge to correct his own errors in copying. In two other cases (Rex Marr, Bela . . . rex), it seems probable that names were never completed, perhaps because the draftsman could not read his model, or because he omitted a line containing the second half of a divided word. These instances suggest that the draftsman of the Vinland map, as we have it, may not have been its compiler, but that the map may have been copied from an immediate original or preliminary draft (having the same content) by a clerk or scribe who was no geographer and did not have access to the compilation materials. This hypothetical preliminary draft will be referred to as O2.

On this assumption, some other names (if they were not in O1) and all the legends (which can hardly have been in O1) must be attributed to the compiler of the map, i.e. the author of O2. Thus, in Europe, Ierlanda insula may perhaps arise from his misinterpretation of O1 or of some other map in which the names for Ireland and for the islands north of Scotland misled him; and Buyslava may come from the reports of the Carpini mission. In Africa, Phazania must have been taken by the author of O1 or from Pliny or Ptolemy; and magnus fluuius (if not a coinage of the cartographer) perhaps from a geographical text of the 14th or early 15th century. Sinus Ethiopicus could have been deduced from Ptolemy’s text; Andrea Bianco’s connection with Fra Mauro, in whose map this very name is found, and his conjectural association with O1 lend substance to the possibility that this name stood in O1, although corrupted in Bianco’s own world map. The transfer of Prester John from Asia to Africa was already made by Bianco and presumably in O1; but the compiler/author of O2, has enriched the plain name, found in Bianco, with details copied from the Tartar Relation, in which Prester John is recorded as an Asian king.

In this case the geography of the Tartar Relation has been corrected from that of O1. In Asia however, while a number of names and the basic geographical design derive from O1, the authority of the Tartar Relation of other Carpini information generally prevails in the toponymy. The degradation of names from this source points again to carelessness or ignorance in the copyist, although in one instance - Gogus, Magog - he, or the compiler of O2, has emended the debased form (moagog) in the Tartar Relation by reference to O1.   Unless we assume that the compiler of O2 was working from an earlier and more correct text of the Tartar Relation and that, if the same scribe copied the Tartar Relation and the Map, he failed to notice the discrepancy between the two. The only name which cannot be traced to either O1 or the Tartar Relation is Hemmodi (v.l. Kemmodi) montes, where a borrowing from a classical text (such as Pomponius Mela), in which the rendering of the initial aspirate was retained, may be suspected; the form in the Vinland map could hardly have been derived from Ptolemy’s. While the coupling of this name, in the Vinland map, with one from the Tartar Relation (Nimsini) may however mean that Hemmodi too came from a Carpini source, it is more likely that the cartographer was here trying to integrate his two sources.

Whether the novelties in the nomenclature of the Atlantic island groups were in O1 or were introduced by the compiler of O2 cannot be determined; the affinities between their delineation in the Vinland map and in surviving charts suggest that the names also may have been found by the compiler in maps which have not survived. The names for Iceland and Greenland may point to literary sources, perhaps of Norse origin (these names, however may have been in cartographic sources used by the compiler); so, with more certainty, do the name and legends relating to Vinland.

At each stage of derivation, from O1 to O2, and (less probably) from O2 to the Vinland map in its present form, there must have been a process of selection or thinning out of names. For Europe and Africa, Bianco’s world map has considerably more names than the Vinland map; in Asia the balance is redressed by the introduction of names from the Tartar Relation. In the absence of the prototype O1, we cannot say whether its author or the compiler of the Vinland map was responsible for introducing the few names in the Old World which must have come from classical or medieval literary sources and the nomenclature for the Atlantic islands. The names for Iceland, Greenland, and Vinland, with the legend on Vinland, must, like their delineations, be held not to have been in O1.

The question “what kind of map is this?” the answer must be: a very simple map, simple both in intention and in execution. The links between the map and the surviving texts which accompany it strongly suggest that it was designed to illustrate C. de Bridia’s account of the Carpini mission (i.e., the Tartar Relation). They also prompt the suspicion that missing sections of the original codex may have been illustrated by the other novel part of the map, namely its representation of the lands of Norse discovery and settlement in the north and west of the Atlantic. It might be said that the dominant interest of the compiler or cartographer lay in the periphery of geographical knowledge, to which indeed the accompanying texts relate; and such a polarization of interest is exemplified in the themes of the seven legends on the map.

In finding cartographic expression for the geography of his texts, the maker of the map has practiced considerable economy of means. The design is reduced to its barest elements of coastal outlines and nomenclature. There is hardly a superfluous line in the drawing; any decorative detail which the compiler may have found in his map sources has been suppressed; no representation of topographical facts, such as relief, is attempted; and from the stock of names commonly found in world maps of the 14th and 15th centuries, even in small ones like that of Andrea Bianco, only a meager selection has been made. In a map of this form, drawn like the circular mappaemundi, on no systematic projection, we do not of course expect to find graduation for latitude and longitude, even if the quantitative cartography of Ptolemy had been known to its author. The only mark of modernization in design is the orientation of the map to the north.

Examination of the nomenclature has suggested that the Vinland map, in the form in which it has survived, is the product of a stage of compilation (the work of the author or cartographer) and a subsequent stage of copying or transcription (the work of a scribe who was perhaps not a cartographer). There is a decided incongruity between, on the one hand, the care and finish which characterize the writing of the names and legends, with their generally correct Latinity, and, on the other hand, the occurrence of onomastic errors which knowledge of current maps and geographical texts or reference to the prototype used by the compiler would have corrected. What other undetected changes or corruptions the copyist may have introduced into the final draft we cannot tell, since his original, the compiler’s preliminary draft, is lost. If Bianco’s world map be assumed to have resembled, in form and content, the model followed by the compiler for the tripartite world, we can however assess the performance of the final copyist by comparison of his work with Bianco’s map, so far as it takes us. He emerges from this test on the whole creditably, for the outlines of the two maps are (as we have seen) in general agreement. The chart-forms characteristic of Bianco’s style of drawing are not reproduced in the Vinland map; at what stage these disappeared we do not know, and they were not necessarily in the original model followed by the compiler. All the major divergences, in the geographical elements of the Vinland map, from the representation in Bianco can be traced to its compiler’s reading of the Tartar Relation or to changes forced upon him by the design adopted.

The process of simplification described above was presumably carried out in the compilation stage. If we are justified in supposing the scribe who made the surviving transcript of the map to have been ignorant or naive in matters of geography, the draft which he had before him for copying must have been the product of selection and combination already exercised by the compiler.

These considerations must govern our judgment of the date and place of origin to be ascribed to the map. The evidence, internal and external, which indicates that the manuscripts were produced in the Upper Rhineland in the second quarter of the 15th century can only apply to the map included in the codex.  If this is a more or less faithful copy from an earlier draft, any inferences on the date, place, and possible authorship of the preliminary draft, that is, of the compilation of the map, must be drawn from study of its content and of the source materials which went into it. This will furnish a terminus post quem for the making of the map and some evidence of the character and resources of the cartographer.

For the purpose of analysis the map divides itself into two distinct parts. In its representation of Europe, Africa, and Asia it can be referred to, and collated with, not only extant cartographic works of similar character and design, but also a text which is bound in the same volume and to which its content is clearly related. For its delineation of lands in the north and west Atlantic, the cartographic prototypes (if it had any) either have not survived or have been so transformed as to be difficult to identify; and if the codex once included a text relating to these lands, this too has now disappeared.

The representation of the Atlantic, with Iceland, Greenland, and Vinland, was almost certainly not in the prototype used for the tripartite world, but was added to it by the cartographer from another source or other sources. The lucky accident that his sources for the Old World can be easily identified or reconstructed allows us to hazard some inferences about his treatment of his sources for the Atlantic part of his map. His apparent preference for the simple solution or the single source admits the possibility that the western part of his map also derives, in the main, from one prototype rather than that it combines features from several; it may have been modified by interpolation or correction from another source (as is the representation of Asia from the Tartar Relation), and this too must be taken into account. That the prototype, like that for the Old World, was a map also deserves consideration. This hypothesis indeed, while it must be tested by collation of other extant maps from which the prototype may be reconstructed, has (prima facie) some support both from the analogy of the cartographer’s treatment of the tripartite world and also from the uniformity of style which characterizes all parts of the drawing, alike in the east and in the west, in those parts where we know, and in those where we suspect, a cartographic model to have been followed. The historical statements about Vinland contained in the map, on the other hand, doubtless come from a textual source, as those in Asia and Africa can be shown to do.

The world picture of the 14th century, which was taken over into the mappaemundi of the next century, including the prototype used in the Vinland map, owed its general form and plan to geographical concepts of classical origin, confirmed and modified by the authority of the Christian Fathers. Patristic geography, as formulated in the Etymologiae of Isidore of Seville (7th century, #205), envisaged the habitable world as a disc, the orbis terrarum of the Romans encircled by the Ocean and divided into three unequal parts, Europe and Africa occupying one half and Asia the other half of the orbis, with the Earthly Paradise in the east. This theoretical and schematic construction did not necessarily imply belief in a “flat earth”, although it is uncertain whether Isidore himself admitted the sphericity of the earth.  The T-O map was simply a diagram designed to bring home to the reader certain basic geographical ideas in an easily apprehended form, and serving as a vehicle for biblical and legendary iconography. In the later Middle Ages the framework of the tripartite world was modified by elements from Arab cartography; thus the eastward horn of Africa, as a survival of the coast by which Ptolemy closed the Indian Ocean on the south, passed from Idrisi into European maps.

On this pattern were to be grafted geographical facts derived from experience and unknown to the creators of the model. New lands reported beyond the bounds of the known world were laid down as islands in the encircling ocean; rivers were moved about the map to accord with the progress of discovery; and a feature might be represented twice on a map, in the form handed down by tradition and in that recorded by a traveler.  The accommodation of new knowledge from land journeys in Africa and Asia during the 13th and 14th centuries and from oceanic discovery in the Atlantic during the 14th and 15th centuries confronted the mapmakers with “the task of pouring the new wine into the old skins”. That they were not daunted by this task is illustrated by the mappaemundi from Fra Paolino to Fra Mauro; and the Vinland map, or its prototype, exemplifies this harmonizing process in a peculiarly interesting way.

The geography of the Carpini mission of 1245-47, as illustrated in the Vinland map, is discussed in detail in The Vinland Map & Tartar Relation, however, here we may note only that the cartographer’s delineation of the nova terra in the east as islands in the ocean conforms to the convention of the Latin mappaemundi and to their authors’ usage in incorporating fresh matter of this kind into their design.

Lastly, it is evident that the geographical outlines of the map owe no direct debt to Ptolemy and that the few and faint traces of Ptolemaic nomenclature to be found in its toponymy are so ambiguous as to arouse the strongest doubt whether its author knew or used the Geographia.

The historians have been interested in the map’s claim to depict the whole surface of the Earth, notably the oceans to the east and west of the traditional landmass, as had not been done before. Usually an ancient or medieval “world map” would actually only display the oikoumene, or known world.  If the Vinland map was based, at some remove, on the Bianco map of 1436, as most cartographers agreed, this was one which showed the Earth as a single landmass, surrounded by only a relatively narrow band of ocean round the edges of its circular picture. The other part of the sphere, nearly half, according to the Ptolemaic convention which insisted the landmass accounted only for some 180° of longitude, was not illustrated. The major novelty of the Vinland map was that it appeared to show the full 360°.   It thus placed itself chronologically well before these later maps which showed or claimed to show the whole surface of the earth, the Florence Genoese  map (1457), the Toscanelli map (described in his 1474 letter), the Yale Martellus  world map of 1490, and the Behaim  globe of 1492 (#248, #252, #256, and #258).  It was, therefore, of major interest to historians concerned with the growth of ideas about, and knowledge of, the Earth’s surface. There were controversies, it is true, about just how much could be read into the Map and there was, so far, no general agreement about its precise implications for the study of cosmography in the fifteenth century. Moreover, by showing islands liberally dispersed in the oceans east and west of the landmass, it opened up the question, some 30 years before it was discussed by Toscanelli, of the possibility of reaching Asia from Europe, not by a voyage of 16,000 to 20,000 km but by easy stages of 1,600 to 2,400 km which were within the capacity of oceangoing vessels at this time. The appearance of Vinland (of special interest to American scholars intent on enlarging the range of their own history), Greenland and Iceland raised the whole question of the level of navigational, geographical and cartographic technique in Scandinavia in the mid-15th century, on which the historians found the cartographers sharply divided but almost all disturbed in some degree by the nature of the Greenland outline which appeared on the Map. The depiction of Vinland, it was agreed, was notional rather than representational.

The Map was interesting to historians as apparent evidence that Norse voyages of the 11th and 12th centuries were known in the Upper Rhineland in the mid-15th century, and consequently that some continuity of knowledge existed between the early discovery of what we know as America and the rediscovery of western lands in the later 15th century. This information was limited in its scholarly impact by the failure of historians to find any other evidence of continuity or to discover that the evidence contained in the Map had ever been known to anyone concerned with exploration either before Columbus’s voyage or after. Finally, the inscriptions on Greenland and Vinland in the Map offered a few scraps of information which differed somewhat from what was commonly accepted. Bjarni, it was implied, had accompanied Lief on his first discovery of Vinland; Bishop Henricus, the Eirik of the annals, who was said there to have gone to look for Vinland, was stated to have found it, and at a different date. Subsequent scholarship has, however, tended to minimize the value of these novelties.

As a world map the Vinland map does not fit into the framework of medieval cartography as conceived in Western Europe. It bears no title or draftsman’s name, or date, and has been attributed to c. 1440 on the argument that it is in the same hand as the Tartar Relation, of which the Map is held to be an integral part. Some students have been reluctant to accept these propositions; the provenance of the Map had not been established, the nomenclature also presents difficulties, as does the representation of certain topographical features, in particular the accurate delineation of Greenland, a point heavily stressed by the editors of The Vinland Map and the Tartar Relation. Much argument has centered around the possibility that Norse voyagers might have circumnavigated and charted its coasts, or provided a written description of them.


LOCATION:  Yale University, Beinecke Library New Haven, Connecticut

Reproductions:

Seaver, Kirsten A., Maps, Myths, and Men: The Story of the Vinland Map, 480 pp.

Skelton R. A. et al., Vinland Map and the Tartar relation, facsimile following p. 17; Plate VII names and legends keyed by numbers to the annotated list on pp. 128-141.

Washburn, Wilcomb E., Proceedings of the Vinland Map Conference (Chicago University Press), 1971 (after the table of contents).

Wallis, Helen, “The Vinland Map: fake, forgery or jeu d’esprit?” The Map Collector, Winter 1990, No. 53, 2-6.

Reproductions of details:

The Map Collector 1 (Dec. 1977), 16 (section showing the island of Vinlandia).

Bibliography:

Crone, Gerald R., Maps and their Makers. An Introduction to the history of Cartography, pp. 43-44.

Cahill, T. A., “The Vinland Map revisited: new compositional evidence on its ink and parchment,” Analytical Chemistry 59 (1987), pp. 829-833.

Edson, Evelyn, The World Map, 1300-1492, pp. 94, 210-211.

McGhee, Robert, “The Vinland Map: hoax or history?” The Beaver 67 (2) April-May 1987, pp. 37-44.

McCrone, Walter C., “The Vinland Map,” Analytical Chemistry 60 (1988), pp. 1009-1018.

Seaver, Kirsten A., Maps, Myths, and Men: The Story of the Vinland Map, Stanford University Press, 2004, 480pp .

Skelton, R. A., Marston, Thomas E. and Painter, George D., The Vinland Map and the Tartar Relation (New Haven and London: Yale University Press, 1965).

Towe, Kenneth M., “The Vinland Map: Still a Forgery,” Accounts of Chemical Research 23 (1990), pp. 84-87.

Washburn, Wilcomb E., Proceedings of the Vinland Map Conference (Chicago University Press, 1971).

Wallis, Helen et al., “The Strange Case of the Vinland Map: A Symposium,” Geographical Journal 140 (1974): pp. 183-214.

“That Vinland Map again,” The Map Collector 1 (Dec. 1977), 16.

Wallis, Helen, “The Vinland Map: fake, forgery or jeu d’esprit?” The Map Collector, Winter 1990, No. 53, pp. 2-6.

Witten, Lawrence, “Vinland Saga recalled,” Yale University Gazette 64 (1989), pp. 10-37.













The Vinland map detail: Vinland, Western Europe/Africa, Azores, Canaries










Translations of legends






The Skálholt Map, 1590 map by Sigurd Stefansson

Stefánsson attempted to plot the American locations mentioned in the Vinland Saga on a map of the North Atlantic. Stefánsson’s original is lost; this copy dates from 1669, and was included in description of Iceland by Biørn Jonsen of Skarsaa. The map mixes real, fictional and rumored geography. In its southeast corner, the map shows Irland and Britannia, and to the north of both the Orcades [Orkney Islands], Hetland [Shetland Islands], Feroe [Faroe Islands], Island [Iceland] and Frisland, a persistent phantom island

Kongelige Bibliothek, the Danish Royal Library.


Other maps displaying “Vinland”





Map attributed to the Danish theologian Hans Poulson Resen, 1605, manuscript in the Royal Library in Copenhagen, Denmark (detail below)



Detail of the 1605 Rosen map showing (from left-to-right) Helleland, Markland and Vinland


This map is a composite work in which Resen added delineations of his own to those made by an anonymous earlier cartographer. The original of this map, which Resen dated “September 1605,” right after the English pilot James Hall’s return from his first voyage, is still in the Royal Library in Copenhagen, but in such fragile condition that when J,K.V. Steenstrup wanted to use it to illustrate his 1886 article, he found it unsuitable for photographic reproduction. Instead, a scrupulously executed copy was made by W. Lynge. With skilled help from a clergyman and a librarian, Steenstrup was able to distinguish and read the map’s inscriptions, some of which had become almost invisible with age. In repeating these inscriptions the copyist, W. Lynge, was instructed to differentiate the two or three different handwritings found on the original, the most recent one being Resen’s own, according to Henrik Dupont at the Royal Library in Copenhagen. Of particular interest to Steenstrup, of course, was where Resen imagined that the Norse settlements in Greenland lay and where he thought Hall’s landing had taken place. The long legends (which Steenstrup considered the work of the map’s original creator before Resen added his own comments) indicated an east coast location for the Western and Eastern Settlements, but Resen reasoned, correctly, as it turned out that they must lie to the west of Greenland’s southernmost point.

With the benefit of hindsight we can see that Resen’s map reflects Dionysius Settle’s published account of his participation in Frobisher’s 1577 voyage and in other ways tries to fit in contemporary thinking about voyages beyond Greenland. For example, at the base of the Zeno-derived Estotiland (upper left-hand side), just west of the sound separating Helleland from Markland, the map pays tribute to Frobisher, Zeno, and Johannes Scolvus Polonus (1476), while the legend in the small sound separating Frysland from Gronland notes that John Scolvus sailed there in 1476 and that in 1507 Sebastian Cabot was stopped there by ice at 67 degrees north latitude.

The delineations and other geographical details demonstrate that Resen did not base his map on first-hand experience with northern seas or even with the countries to the north of Denmark, but drew on the Stefansson map or its progenitor, and on his wide reading. He shared the rapidly growing interest in his part of the world for saga accounts of Norse voyages to Iceland, Greenland, and Vinland, and he was clearly familiar with the story in the medieval Icelandic Annals about a small Greenlandic ship that had landed in Iceland in 1347 after being driven off course on its way home from a voyage to Markland in North America. In an obvious attempt to visualize the current that had produced this incident, the map depicts an arc of tree trunks drifting toward Iceland all the way from a region south of the Promontorio Vinlandia (shown at the tip of Terra Corterealis).

The Resen map reaches farther south on the American side than does the Stefansson map and calls its southernmost region Norumbega. This mythical wonderland, originating with Girolamo Verrazano’s modest location named oranbega on the North American coast of his 1525 map (Book IV, #347), had been considered the potential reward of travelers to Nuova Francia ever since Giacomo Gastaldi located Nurumberg on his 1548 map of North America (#399). What we do not know is whether Resen depicted this region as a continuation of Helleland, Markland, and Vinland because he thought that the name Norumbega reflected an ancient connection with the Norse, a notion that arose rather early and has been as slow to die as pity for the Norse Greenlanders. Put in colonial terms, Norumbega may have been considered as potentially useful as the three North American regions mentioned in the sagas.


Additional maps with Vinland displayed include those from Ranulf Higden’s Polychronicon (1350). These elliptical, or ark-shaped world maps are of schematic and cosmological design. They typically have 10 islands northwest of Europe. The second or third of these islands from the top [north] is called Svinlandia, Winiland, Wineland, or Widland in reference to a land of grape vines. None of the 21 known copies of Higden’s map spell these islands with an “F” at the beginning so there is no indication that the islands were confused with Finland. Furthermore, the placement and spelling are consistent with Vinlandia insula on the Yale Vinland Map of 1440. See also #232 Higden maps.







The Rudimentium Novitiorum, Lubeck 1475 & Hans Rust, Augsburg, ca. 1480 (#253)

At least two versions of a map in the anonymous Lubeck geography called Rudimentium Novitiorum show the Norse colony of Vinland northwest of Europe on an island that extends into the Atlantic Ocean. The two spellings, Vinl’ad, and Winl’ad, are variations of Vinland. Historians and geographers have claimed that Winlad on the Lubeck map is a misspelling for Finland. However, the spelling is consistent with contemporaneous historical and geographical references to Wineland and Vinland as well as the Higden maps which consistently use “W’ or “V” for Wineland. Furthermore, as the later Augusburg map distinctly has the spelling Vinland (where the ‘ represents “n”), we can be certain that use of a “V” for Vinland was not a mistake. Such maps confirm that Vinland was a well-known location to 15th century Europeans as thousands of copies of the Lubeck map were printed.