,

&

Published in cooperation with the Center for American Places,Santa Fe, New Mexico, and Harrisonburg, Virginia

, &

Medieval European

Knowledge of America

Baltimore & London

Library of Congress Cataloging-in-Publication DataEnterline, James Robert.

Erikson, Eskimos, and Columbus : medieval European knowledge ofAmerica / James Robert Enterline.

p. cm.Includes bibliographical references and index.

ISBN ---X. AmericaDiscovery and explorationNorse. . Geography,

MedievalMaps. . Early maps. . Historical geographyMaps.. Nautical chartsHistory. . VikingsNorth

AmericaHistory. I. Title.E .E .dc

-

A catalog record for this book is available from the British Library.

James Robert EnterlineAll rights reserved. Published

Printed in the United States of America on acid-free paper

The Johns Hopkins University Press North Charles Street

Baltimore, Maryland -www.press.jhu.edu

To my beloved wife, Esther

List of Illustrations ixDirectory to the Chronological Survey xiiiPreface and Acknowledgments xvii

Front Map

. Introduction

I. Outstanding Misunderstandings

. Claudius Clavus

. The Inventio Fortunatae and Martin Behaim

. The Yale Vinland Map

II. The Chronological Survey

. Introduction to the Chronological Survey

.

. Early Scandinavian Geography

. Communication Links with Greenland

. The Unseen Bridge

.

. Late Greenland-Based Exploration

. Foundations of European Misunderstandings

. News Penetrates the Establishment

. Europes Westward Awakening

. Mastery of the Atlantic

.

. A New Continent Emerges

. An Old Continent Emerges

. The Misunderstandings Are Resolved

. Conclusion

Appendix: The Vinland Maps Ink Notes Selected Bibliography Facsimile Atlases and Reproductions Index

Front map North polar map . Native Eskimo map of Southampton Island and aerial survey map . Seward Peninsula, Kotzebue Sound area, Alaska . Seward Peninsula and Claudius Clavuss map

of Scandinavia . Sideways view of Foxe Basin and Southampton Island area . Canadian Arctic coast and Arctic Archipelago . Transformation of Vinland Map with Greenland scale corrected . Outline map of Bafn Island, west of Greenland . Hypothetical regional maps of Bafn Island . Incorrect placement of accurate regional maps to simulate

Vinilanda Insula . Geographic north compared with current magnetic north . World map of Claudius Ptolemy, .. . Anglo-Saxon Map, ca. .. . Reconstruction of Adam of Bremens conceptions, .. . Misattribution to Heinrich of Mainz, early s . Views of Eid pass from Greenland cathedral ruin . Svartenhuk neighborhood of Greenland . Psalter Map, ca. . Hereford map northwest region closeup, . Mariners chart by Giovanni Carignano, ca. . Scandinavia by Petrus Vesconte, . Vescontes map in Sanudos book, Oxford MS, ca. . Vescontes (or Paulinos) map in Sanudos book, Paris MS,

ca.

. The Northwest by Angelino Dalorto, . Norse runestone found in cairn on island of Kingigtorssuaq . Medici (Laurentian) marine chart, . Catalan map fragment in Istanbul, ca. . Pierre DAillys Seventh Figure, . Albertin de Virgas world map . Detail of Figure . Two world maps by Duminicus Ducier, , and a

annotation . North Atlantic nautical chart, . Claudius Clavuss map of the North . Reconstruction from the Vienna-Klosterneuburg coordinate table

Cosmography of Climates, . Reconstruction from the Vienna-Klosterneuburg coordinate table

New Cosmography, . Circular world map by Andrea Bianco, , to accompany

sailing charts . Vinland Map at Yale University, ca. . World map by Andrea Walsperger, . Vienna-Klosterneuburg Schyfkarte, before . Catalan map in Florence, northern portion, . Catalan map in Modena, ca. . Genoese map in Florence, . Detail of disk-style world map by Fra Mauro, . Nicholaus Germanuss map of the North in the Zamoiski Codex,

ca. . Detail of Europe from Nicholaus Germanuss world of

ca. . Scandinavia by Henricus Martellus, ca. . Northwest Europe by Henricus Martellus, ca. . World map by Nicholaus Germanus in Ulm Ptolemy, . Reconstruction of Martin Behaims globe, . Sketch map by Bartholomew Columbus, . Juan de la Cosas world map . Cantino map, . Giovanni Contarinis planisphere, . Old World hemisphere by Martin Waldseemller, . Johann Ruyschs planisphere, . Chart known as Kunstmann III, ca.

. World map from the Strassburg Ptolemy, . Gored map of Northern Hemisphere by Johann Schner, . Western Hemisphere from Johann Schners globe of . Sketch showing Great Arctic Strait, ca. . Western Hemisphere from Johann Schners globe of . Double cordiform planisphere by Orontius Finaeus, . Western Hemisphere from Paris Gilt Globe, . Stuttgart gores, s . Cordiform planisphere by Orontius Finaeus, . Scandinavia by Jacob Ziegler, . Scandinavia by Olaus Magnus, . The North by Nicolo Zeno, . North Pole and Greenland-Grocland area, Gerard Mercator, . Abraham Orteliuss map of Tartary, . Greenland and Vinland by Sigurdur Stefnsson, . Vinland by Hans Poulson Resen, . Mappamonde attributed to Christopher Columbus, ca.

Before Ptolemys Geographia Fortunate Isles Icelandic settlement Ibn Khorddhbeh Ottar of Norway Greenland-Vinland voyages Anglo-Saxon map Adam of Bremen Eric Gnupsson Icelandic Geography

Saxo Grammaticus Heinrich (not) of Mainz Ibn Said Kings Mirror Saga writing Historia Norwegiae Haldor/Arnold letter Psalter Map Hereford Map Nyaland and Duneyar Geographia Universalis Greenland trade Marco Polo

Rymbegla Carignano Vesconte/Sanudo/Paolino

Dalorto-Dulcert Kingigtorssuaq stone Gisle Oddsson Ivar Baardsson Markland voyage Medici atlas Inventio Fortunatae Zeno and Estotiland Catalan map in Istanbul Bjrn Einarsson Zenos Greenland

First Latin Ptolemy Pierre dAilly Albertin de Virga Pirate attack Duminicus Ducier North Atlantic chart Claudius Clavus Vienna-Klosterneuburg maps Andrea Bianco Vinland map at Yale La Sales Salade Andrea Walsperger Vienna-Klosterneuburg chart

Catalan map in Florence Catalan map in Modena Gunnbjrns Skerries Genoese map in Florence Fra Mauro Scolvus, Pining, and Pothurst Paolo Toscanelli Germanus and Martellus Columbus in Iceland Thloyd and Brasil Bergen-Greenland sailors Ferdinand van Olmen

Bristol voyages continue Bishop M. Knutsson Martin Behaims globe

Columbus in the Caribbean Hieronymus Mntzers letter Encounters with America

Juan de la Cosa Piri Reiss informer Azorean explorations Cantino map Giovanni Contarini Martin Waldseemller Johann Ruysch Sebastian Cabot voyage Kunstmann III map Strassburg Ptolemy Early Johann Schner Magellan and Schner-Finaeus series Ziegler and Magnus Nicolo Zeno the younger Gerard Mercator Abraham Ortelius Icelandic Vinland maps

thatColumbuss encounter with America was completely accidental.Those whoventured to think otherwise assumed that any prior knowledge he mighthave held about America would have sprung from Leif Eriksons contactwith Vinland. New evidence presented here suggests instead that Eskimogeographical information about a wider America made its way through theGreenland Norsemen into medieval European world maps. In Europe suchcontinental foreknowledge was not immediately correctly perceived, but itgradually drew Europes attention westward and may have contributed to thebirth of the Age of Discovery. This admittedly radical-sounding idea hasgrown step by step out of developments in recent decades.

On the eve of Columbus Day , Yale University announced the ac-quisition of its now famous but controversial Vinland map, presumed at thattime to depict Norse America in Canada. Yales press release called it thecartographic nd of the century, and the map engendered two parallel dis-putes. The lay world saw it as an attack on a long-honored hero, Columbus.The scholarly world saw it as a completely anomalous document that thrustnew problems into many branches of the study of history. Those disputeswere not resolved when Yale announced tests in showing that the mapsink appeared to contain twentieth-century pigments.The lay worlds uneaseregarding its hero continues regardless of that maps authenticity, for there isother evidence establishing beyond any doubt that the Norsemen did en-counter North America.While to some people, forgery seemed the obviousexplanation and offered resolution of the scholarly dispute, that was not theonly possible conclusion from the ink tests. This author in postulated anatural scenario for the maps contemporary history that gave an innocent

explanation of every known detail of the maps ink, based on modern pig-ment contamination, and reinstated a case for the maps credibility. Other re-searchers, in the s, came to similar conclusions, minimizing the impor-tance of the anachronistic ink pigments. Still others showed that the pigmentsmight not be anachronistic at all, that they could appear in nature with theobserved parameters. In Yale republished the map, the new press releasestating that it stands once again vindicated. Controversy nevertheless con-tinues on many fronts.

Even before the ink controversy, I was working at a resolution of thosedisputes, entertaining a possibility that the map could preserve genuine in-formation. However, if genuine it probably represents something other thanthe Canadian seaboard and something other than Leif Eriksons landfall. Iwill introduce evidence here to support that possibility. In answer to thescholarly problem of anomaly, I introduce evidence that the Yale VinlandMap is potentially just one member of a large group of pre-Columbian mapsall apparently recording Norse contact with America or native Americans.But the Norsemen didnt know they knew about America. In dozens of OldWorld maps the Arctic coast of Eurasia shows rich though incorrect detail;Europeans had never been to Arctic Eurasia, nor had they any cartographicknowledge of it. In the past, scholars have explained away these coastlines asfantasies. It occurred to me that many of these details do correspond exactlyto features on the Arctic coast of North America instead of Eurasia.

Cartographic correspondence has been discredited by its misuse in someresearch that reached sensational, implausible conclusions. My approach hasbeen more rigorous, conservative, and plausible. The cartographic examplesto be shown are not cryptic but very clear when viewed in the appropriateframework. Post-Columbian maps that contain apparently fanciful coastlineshave been increasingly understood by careful historical analysis, and I will dothe same for pre-Columbian ones. In addition to the maps, I have sought toidentify numerous travelers itineraries and geographical descriptions of pre-Columbian America, all providing mutual corroboration. While the YaleVinland Map suffered for several decades from its uncertain origins, theseother documents are of unquestioned provenance and have been cataloguedin world-class libraries for centuries. However, historians seem to have over-looked or misunderstood them because, as I will suggest, the maps apparentlyincorporate a previously unrecognized geographical distortion that wasunique to the pre-Columbian mind. That systematic distortion resultingfrom the misidentication of continents is analyzed herein.The thesis is elab-

orated in a chronological survey of geographical materials extending fromthe Norse Late Middle ages down through the post-Columbian Renaissance.

Did Columbus see these documents? Perhaps or perhaps not, but his con-ceptions of land in the west were inspired by scholars of the generations pre-ceding his, who did see them. There is a way that rufed feathers in the layworld may still be smoothed: I make an effort to see Columbus as more of arationally motivated, carefully researching proto-scientist than the way his bi-ographers have described himstrictly a luck-blessed, illogically motivatedadventurer.

This work is the promised companion volume to my earlier book, VikingAmerica. Neither is a prerequisite to the other, and each is written to be read-able by itself. Nevertheless, issues ancillary to one are taken up in the other.(A few specic matters in the earlier book are reconsidered in the presentone, as will become evident.) The prefatory comments made in Viking Amer-ica apply here also. The polar map on the endpapers of that volume is re-peated here at the beginning of the text. I refer to it throughout as the frontmap and recommend consulting it whenever the discussion enters unfa-miliar regions. The more detailed local maps in Figures , , and shouldalso be noted.

This is not a history book, at least not in the strict sense of the usual de-ductive methodology of history from written texts. It is a prehistory bookthat subjects maps and documents, as artifacts, to the inductive methods ofarchaeology. It is based on the belief that reliable knowledge of the past canbe culled from this artifactual record even if it is not spelled out in words.This work falls into place as part of the newly emerging cognitive archae-ology. I will not enter into any controversy between new and past schoolsbut will dip into both. In examining how information about America couldhave reached southern Europe, I have focused on the circumstances of thevarious documents creation and the lives of their creators. Known histori-cal details will be augmented at times by scenarios grounded in a structural-ists view of human nature. Interdisciplinary support ranging from psychol-ogy to physics will be brought to bear. However, the story so constructed isnot held out as proven truth. Instead it is a plausible theory to be tested againstindependent evidence. The day-to-day purpose of science is not the estab-lishment of universal, nal explanations but the proposal of testable theoriesthat can serve as stepping stones to still better theories. Nevertheless, themore nearly universal a hypothesis, the more testable it becomes and themore potential it has for spawning still further insights. The conclusions

reached here are thus a part of the sciences, subject to adjustment by new ev-idence or even to being overthrownor strengthened.

This book, as well as being divided into chapters, is also structured in twoparts. The second part is a wide-ranging chronological survey of documen-tary evidence, namely, the scores of documents supporting the theory thatNorse contacts in America had European repercussions.The attention to in-dividual documents there is necessarily limited. Part I examines in greaterdetail a smaller number of documents. A casual reader might, after complet-ing Part I, wish to read just the chapter summaries and the Conclusion.

in the research for this book deserves to be ac-knowledged. Obviously, none of these acknowledgments implies any of theseindividuals acceptance of the works validity, which may or may not be ex-pressed independently. David Woodward read the full manuscript twice inwidely separated versions, giving many useful suggestions and critiques. Ear-lier versions were also critiqued by Benjamin Olshin and the late VincentCassidy. Subtopics in Eskimo archaeology received constructive attentionfrom Robert McGhee and John MacDonald. Lunchtime conversations overthe years with Thomas Goldstein gave me ongoing encouragement, and hisenduring question of motivation for the discoveries was always in my mind.This research would never have been completed, let alone started, withoutAlice Hudson, and before her Gerard Alexander, at the New York Public Li-brary. Also importantly, several anonymous (to the author) publishers read-ers served to sharpen the arguments presented here. Last, and far from least,a highly intensive reading of a late version was given by Gregory McIntosh.

The greatest thanks to go my wife, Esther, to whom this book is dedi-cated. Many authors are thankful, as am I, for their wives emotional supportand encouragement, but she went further than that. She made it possible forme to work without distraction in hundreds of ways, and she devoted herprofessional career as a psychologist to our material support for several peri-ods during which I pursued nothing but this research.

,

&

Image not available.

Image not available.

Discovery consists of seeing what everybody has seenand thinking what nobody has thought.

Albert von Szent-Gyorgy

, the European discovery of Americawas purely a geographical event. It occurred, these people believe, when acertain lookout, weary after long sailing, cried, Land ho! He may have ut-tered those words in Spanish, Italian, Portuguese, Old Norse, Irish, or per-haps some other language. In fact, this scene arguably took place in severalsuch tongues, leaving pedants to disagree over which was the discovery.Should they assign importance to priority of occurrence or to continuity ofoccupation? Or should they invoke some entirely different criterion to saywhich statue may bear the inscription The Discoverer?

More than either pedantry or ethnic pride is at issue, however; aside fromthe ideal search for Truth, there is a practical matter involved. Most Ameri-can (and other) elementary schools use the discovery of America to instillcertain ideals and heroic models into childrens minds. The American adultswho landed on the Moon gauged their accomplishment by comparison withthis earlier discovery. If our schools are going to continue promoting Colum-bus as the most widely held cultural hero in the West, then we should un-derstand him with some certainty.

However, our schools should probably discontinue any such heroic pro-motion at all. This book holds that the very concept of a European discov-ery of America is clouded by a reciprocal concept of divulgence by nativeAmericans encountering the Norsemen. Ethnic priorities become secondaryto an intellectual development that ultimately involved all the peoples en-gaged in a protracted uncovering.

Serious thinkers already recognize that the discovery of America was amuch more complicated happening than merely a landfall. Indeed, it tooksome decades after , before most geographers realized that there was an

America distinct from Asia. Until that time Europeans, with infrequent ex-ceptions, presumed that any mainland on the other side of the ocean fromEurope must be Asia. The many explorers who were active in those decadesafter are the ones who really deserve to be called the discoverers ofAmerica in the sense of discovery pronounced by Szent-Gyorgy in the epi-graph to this book. Columbus, island hopping in the Caribbean, never setfoot on new world mainland until at least , probably , and that onlybriey and initially unknowingly. Even when he came to know that a newworld was involved, his state of knowledge was vague, theoretical, and basi-cally a misunderstanding that sprang from the traditional presumption ofantipodes.

These statements may already be unsettling to some who have been in-doctrinated with the heroic concept of Columbus. Those people may hearsuch statements as merely iconoclastic. Or they may come to his defense withthe thought,But such explorers would never have gone there at all if it werenot for Columbus. This thought may be so, but one must observe veryclosely whether it is summoned in defense of Columbus the man or Colum-bus the creed.

Nobody can ever deny that Columbus the man was important in the dis-covery of America. Columbuss demonstration of the possibility of crossingthe middle latitudes of the Ocean Sea remains a bold, epoch-making un-dertaking. But we also want to evaluate his situation rationally. First one mustmake a wholehearted attempt to see him as an ordinary human being, think-ing and reasoning in familiar ways and driven by familiar, fathomable desires.If he does not end up seeming larger than life, that is no criticism of the manbut a reevaluation of what we have done with our minds in the classroom.

Leif Erikson, too, was a discoverer of America. There is no longer anycontroversy among scholars about that. In his case the bias in teaching hadbeen in the opposite direction. The saga of Leif s voyage to Vinland ca. .. had always been introduced with an apologia, almost clandestinely, as arumor that very well might not be true. Teachers felt it should neverthelessbe mentioned for the sake of fairness and completeness. There are discov-ered ruins, now, to prove itthe Norsemen were in America.1 Furthermore,their culture was not the barbaric way of life from earlier centuries depictedin popular media; its technology was similar to that of many small towns andrural areas in medieval Europe.

Nevertheless, Leif s discovery was somehow premature; for some reasonit led to no obvious repercussions in Europe. Columbuss discovery wassomehow more timely.What does that word timely mean? A standard answer

is often couched in European political and economic developments of thelate fteenth century. Then-recent expeditions out into the ocean lookingfor islands are cited as important precursors. While such factors obviouslywere involved, they do not tell the whole story either. The Columbian dis-covery of America was not only an event in the expansion of Europe butalso an event in intellectual history, the history of ideas, indeed, the historyof science.That is the area in which I address the timeliness of the Columbiandiscovery.

The above-stated thought that later explorers would never have gone toAmerica had Columbus not done so may be wrong.The idea of sailing west-ward for Asia was not unique to Columbus. Scholars know that some of hiscontemporaries and predecessors entertained the same idea and that a few ofthem actively pursued nancial backing for a trip. The established attitudethat Columbus had to battle was not concerning the sphericity of the earth.Contrary to what our schools have taught children, those with a practicalconcern in the matter had known ever since Greek times that the earth hadto be round.2 But they also knew something about its size. Many inferredfrom astronomical measurements that the distance one would have to sailwestward from Europe to reach Asia was probably , miles or more.They knew that a ship would have to carry a years worth of provisions toattempt such a voyage safely. No crew or captain was willing to sail that longout of sight of land exposed to even the predictable dangers, and thereforethe voyage was considered impossible.This was the establishments argumentin Spain against Columbus, and they were right.

Yet somehow, during the fteenth century an attitude began to prolifer-ate holding that the earth was much smaller than indicated by astronomicalmeasurements. Columbus believed Asian shores to be only one-quarter theabove distance away, and similar arguments by various near contemporariesgave him courage. Even without Columbus, America would very likely havebeen discovered within a few years or decades. The project was in the air.

The question of timeliness now takes on a new interest. What does inthe airmean? Structuralist psychologists, led by Jean Piagets examples, havebegun to understand scientically the concept of a body of knowledgeandthe modes of transition of knowledge from one state to another.3 One ba-sic tenet is that people cannot assimilate or appreciate a new idea until theycan connect it logically to some contemporary established knowledgesomething already known.4 Thereby they accommodate to the new idea.Theunderlying question now seems even more confounding.Why did the com-pletely incorrect ideathat the shores of Asia lay only a few thousand miles

across the oceangain credence at this particular time? To what canonicalknowledge or gestalt or idea structure was it attached?

When the seekers after Asia in the West tried to enlist nancial backingfor a voyage, they needed evidence to support their theory. One argumentColumbus and his supporters used was that classical and medieval writings,including those of Aristotle, Ptolemy, and Marco Polo, contained geograph-ical references suggesting the near proximity of Asia. It seems possible how-ever, as Samuel Eliot Morison thought, that such authorities were merelysummoned up as afterthoughts5 employed in a wide-ranging attempt to but-tress an argument they already believed in.* Something unique to their ownera might better explain the timeliness of the ideas of Columbus and likethinkers. Something stronger than (erroneous) ancient authorities perhapsmade them invest years at royal courts and risk their lives at sea defending anargument about the distance to Asia that many contemporary theoreticiansrightly scoffed at.

Yet neither they nor later students of their ideas have been able quite sat-isfactorily to point out what carried them to such unjustied certainty. Mo-tivation for belief (above all, prot) there was aplenty, but for certainty, none.Some have concluded that Columbus was probably mad.6 Many writers con-centrating on Columbuss motivation have completely overlooked the mo-tivation of the others who proposed such voyages. Were they all mad? Theproblem has been so frustrating that some writers have turned to mysticismfor an explanation.7 Jean Piaget, the founder of structural psychology andscientic epistemology, would have allowed no such easy escape. He statedatly, The great man who seems to launch new movements is but a pointof intersection or of synthesis, of ideas that were elaborated by continuouscooperation. Even if he is opposed to current opinions he responds to un-derlying needs which he himself has not created.8 While structuralism hasbeen applied in certain inappropriate elds, and thereby discredited, currentresearch is conrming and extending Piagets concept.9

In Viking America I suggested (as others had before me) one stimulus thatcould have given rise to belief in the nearness of Asia: vague and misunder-stood but actual information concerning America.10 The present book

*Even Ptolemys Geographia, which was rst translated from Greek into Latin around , couldhave been available to Western scholars long before that if they had any interest in the subject.(See p. of Milton V. Anastos, Plethon, Strabo and Columbus,Annuaire de lInstitut de Philolo-gie et dHistoire Oriental et Slaves de lUniversit de Bruxelles, vol. [Brussels, ], ). Fur-thermore, if the Geographia were as exclusively important as some writers would have it be, itsinuence should have been felt two generations before Columbus.

shows evidence that residents of the Norse colonies on Greenland transmit-ted such information, in the form of maps, travelers descriptions, and livingnatives, to Europe during the century or so preceding Columbus. And in therecentness of those particular Norse contacts lies a possible explanation. Afourteenth-century resumption of westward exploration after centuries ofneglect could explain the timeliness of the southern European ideas of theproximity of Asia in the West. While Leif Eriksons eleventh-century con-tacts in Vinland ultimately resulted in abandonment of Vinland, those of theearly Renaissance Greenlanders might have led toward the European Ageof Discovery. And that resumption of the Norse exterior orientation, as weshall see, was itself a result of new contact with the native American peoplewe call the Eskimos.* The discovery of America would be an involved jointaccomplishment not solely attributable to any individual hero or ethnicgroup.

Numerous writers who surmised that Columbus had some kind of Norseinformation have been dismissed for giving no proof. The reader is speci-cally put on notice that I make no claims of nal proof here regardingColumbus personally. I leave it to others to draw conclusions in that regard,perhaps after further research. But the evidence that such information wasavailable in southern Europe in his time will be substantially increased, andwe will see some tentative evidence that Columbus had access to it. Histo-rian Samuel Eliot Morison, in his biography of Columbus, concluded thathe had an illogical mind. While asserting that Columbuss theory was du-bious, Morison said, One can well imagine him explaining it, his eyessparkling and his ruddy complexion aming.11 There were undoubtedlymany factors contributing to his motivation besides rational ones.12 Indeed,current writers on the subject have highlighted various aspects of that mo-tivation that are nowadays considered politically incorrect. But we will seethat a rational component to his motivation also becomes a possibility.

However, I do not think, as some media headlines suggested when VikingAmerica was published, that Columbus secretly followed Viking maps. Nordo I claim that Columbus necessarily engaged in any kind of major decep-tion (although others have shown evidence for such theories).13 The processof any information transfer from the Norse was likely much more subtle thanconscious deception. Structuralist theoreticians acknowledge the possibility

*Modern descendants prefer the term Inuit, or in Greenland Kalaalit, but we will be dealing witholder sources and will retain the term Eskimo. For the same reason we will use the term Green-lander to refer to the medieval Norsemen rather than the modern Inuit.

of unconscious response to vague inuences.14 The new historicism re-searchers nd many traces of unconscious behavior during the encounterwith the New World.15 Louis DeVorsey has shown how in post-Columbiantimes native American information sources silently entered into Europeanmaps.16 G. Malcolm Lewis showed that similar information led to systemat-ically misunderstood representations on European maps of America whichmotivated explorations even if they were ultimately dismissed as myths.17

Surely, that kind of vague inuence could have affected the reactions of Colum-bus and his predecessors and contemporaries to any Norse-transmitted fore-knowledge about America.

If the importance of the Norsemen and the Eskimos in European historyis elevated by this book, it is not meant to be at the expense of Columbus orhis bravery. Columbus also will be elevated, for he will become more un-derstandable as a person. The man becomes an intelligent, rational partici-pant in one of the greatest human accomplishments to date: the develop-ment of the scientic method. It is an auxiliary purpose of this book toexamine evidence in that direction.

It will be obvious that this book contains radical ideas and uses radical pro-cedures. In order to make a positive approach under these conditions, it willbe useful to review (and preview) some fundamental considerations: the cul-tural milieu of the interaction between the Greenland Norsemen and theEskimos, within both the European culture and the Eskimo culture; and theEuropean intellectual milieu in which the Norse Greenland colonies ma-tured, particularly with regard to geographical thought. This will introducehow maps looking like American lands could show up on medieval and Re-naissance maps of Eurasia.* Such an idea, however, will immediately raisedoubts in scholars who are skeptical about interpreting old maps by theirshapes.This leads nally in this chapter to a section on method, to assure thatwe are proceeding with due caution.

of the Norse contact with Amer-ica, in Viking America.18 I suggested that there was a renewed contact aftercenturies of hiatus during Early Renaissance times, involving Eskimo con-tacts. Some evidence suggests Norsemen made physical contacts westward

*The Renaissance is dened ofcially to have begun during the fourteenth century, but the sep-arate attitudes characterizing the Middle Ages and the Renaissance coexisted for centuries. Wemay consider the Middle Ages to have ended with Columbus.I will not maintain here, as I did in Viking America, that the Norsemen actually assimilated intothe Eskimos. That will be left as an open question, although it seems increasingly unlikely. Fur-

into the Arctic Archipelago (see north of Canada on the front map) and per-haps into North America itself. But the cultural contacts will be our maininterest here.The Norsemen lived in fully civilized communities that kept intouch with the European world.Their so-called Eastern Settlement was nearthe southern tip of Greenlands west coast, and the Western Settlement wasjust a little farther up the west coast. Their summer hunting area, Nordrse-tur, was above the Arctic Circle on the west coast.When fourteenth-centuryclimatic changes forced a decline in their agricultural and herding settlementson Greenland, the Norsemens lifestyle became more nomadic and they livedincreasingly by hunting. Written sources show that they soon encounteredthe Eskimos in Nordrsetur. (A brief eleventh-century encounter in Vinlandseems not to have been sustainedat least no evidence suggests it was.Southern Greenland itself was not occupied by living Eskimos when theNorsemen rst arrived.)

In the nineteenth century it was fashionable to believe that hostile inter-actions between the newly arriving Eskimos and the established Norsemencaused the eventual end of the Norsemen in Greenland and that there waslittle interaction otherwise. In the twentieth century, Gwyn Jones continuedthis idea, in concurrence with many European archaeologists. But FridtjofNansen and Helge Ingstad both rejected a fundamentally hostile interaction,and they have many followers in North America.19 Nansen and Ingstad bothconcluded that, even though individual incidents of hostility certainly didoccur, the overall interaction was generally neutral and sometimes positive.Danish archaeologists now concur in rejecting the extermination hypothe-sis.20 The thirteenth-century Eskimos that the Norsemen encountered wereengaged in the culmination of a migration across the Arctic from Alaska toGreenland. These Eskimos, members of the Thule culture, might have setexamples for the Norsemen on how to live in the high Arctic winters andwhere to nd hunting grounds.* In their turn, the Norsemen have beenthought to have set examples for the Thules in many technologies (like baleensaws and coopered tubs) that the Thules incorporated into their daily life.The result was construed to be the emergence of a new Eskimo culture, theInugsuk, recognized by the Scandinavian archaeologist Therkel Mathiassen

thermore, while the contacts were widespread, they were not necessarily frequent and probablyinvolved an element of accident.*The word culture in this usage is a modern anthropological construct used to differentiate onegroup with a common lifestyle history from another, for example the earlier Dorset culture Es-kimos. Each has a distinct spectrum of tools and artifacts in the archaeological record.The nameof the culture comes from the location where it was rst discovered by archaeologists.

and amplied by Erik Holtved and Junius Bird.21 Some archaeologists stillconsider any such cultural interaction between the Norse and Eskimos to becontroversial.22 Nevertheless, the Canadian archaeologist Robert McGhee,who is considered to be the dean of Canadian Arctic archaeology, sees ev-idence of positive cultural interaction. He believes it is likely that Eskimosand Norsemen had wide-ranging trade contacts and coexisted, sometimesamicably, for several generations in southern Greenland, leaving at least oneshared historical tale in the folk record of each.23 Norwegian linguist KnutBergsland has cited Norse loan words for domestic plants and animals in theGreenland Eskimo dialect.24 Danish National Museum archaeologist JetteArneborg has given theoretical arguments that we should also expect to ndcultural inuences in the opposite direction, from the Eskimos to the Norse-men.25 While no one has yet been able to demonstrate any in the archaeo-logical record, we nd some in this book.

I suggest that the Thule Eskimos occasionally shared their cumulative ge-ographical knowledge of America, gained during the recent migration, withthe Norsemen. There is no more pressing question about such a newly metpeople than, Where on Earth did you come from? Neither party has to bebilingual to pose the question or give an answer, as many explorers havedemonstrated.The Eskimosgeographical knowledge included the techniqueof drawing primitive mapscartograms (surprisingly not part of the Norsepractice). The idea of the map, and skill in its realization, have been part ofmany nomadic, so-called primitive, cultures because they assist survival. Themedieval Thules were a primarily coastal people living from sea mammalhunting, and they were acutely aware of every turn of the coastline.The pi-oneer anthropologist Franz Boas visited their descendants in the early s,before they had contact with modern Europeans. About them he said, If aman intends to visit a country little known to him, he has a map drawn inthe snow by someone well acquainted there, and these maps are so goodthat every point can be recognized.26 Figure shows an Eskimo map ofSouthampton Island (at the top of Hudson Baysee front map) that TherkelMathiassen collected on the Fifth Thule Expedition. Its only major short-coming is local variation of scale. Note a doubling of scale in the easternpeninsula. A similar doubling in the southeast coastline along Fisher Straithas caused a rotation of orientation of other features. More recently, Spinkand Moodie thoroughly surveyed the question of primitive Eskimo maps andconcluded that they were part of the indigenous culture. Certainly no onedenies that the relatively sophisticated geographical awareness that goes intocreating a map was part of the Eskimos culture.27 Part of the difculty in

recreating a properly scaled map undoubtedly arises because much of the in-formation was memorized in a linear itinerary-like fashion.28 It is knownthat Greenland Eskimos did carve maps on wood, which were apparently lin-early organized.29 The suggestion that Eskimos only learned mapmakingfrom post-Columbian contacts with Europeans is supported by no evidence

Figure . Modern aerial survey map of Southampton Island (a) compared with prim-itive Inuit Eskimo map (b)

Image not available.

and is motivated only by chauvinism about mapmaking as an advanced,civ-ilized technology. Agriculture-based societies have always looked down onthe hunter-gatherer mind as primitive,yet in reality the latter are just as ca-pable.30

Most of the maps we will examine more likely originated from Eskimoinformation than from Norse. Furthermore, they probably would have orig-inated from contacts more or less contemporaneous with the appearance inEurope of analogues of each map. In fact, many might have been the resultof European cartographers directly interviewing Eskimos brought to Europerather than transfer on paper.This is particularly true of some early maps re-sembling the Alaska area, where the Thule culture originated.* Eventuallythe maps showed features incompatible with an Eskimo origin, such as evi-dence of surveying involving latitudes. This would indicate that the Green-land Norsemen, who kept intellectual contact with Europeans, had alsolearned some notions of cartography from the Eskimos. Indeed, as theNorsemen eventually became nomadic hunters themselves when their cli-mate deteriorated, that cartographic skill became of vital importance to themalso. Such later maps may well have originated directly on permanent media.

The history, nature, even the very existence of any such cartography onAmerican shores in the Middle Ages must remain conjectural. Even the wordcartography may be inappropriate, for much of the geographical informationmay have been carried in the head, as it was by Boass Eskimos, and drawnon ephemeral media, if at all, or described verbally.31 More recently EdmundCarpenter described his observation of Eskimo hunters: When a man trav-els by sled into unfamiliar country, he continually looks back to see how thecountry will appear on his return.These brief glimpses, vividly recorded andfaithfully remembered, are enough so that he can nd his way back with ease.An elderly hunter may efciently guide a party through an area he has notseen since his youth, and then but once. Exactly this kind of performancewas also witnessed by anthropologist Hugh Brody. This is not a racist attri-bution of special powers but suggests a cultivation of perceptions that theenvironment requires for survival. Everyone has heard the statement that theEskimo language contains scores of special terms dealing with snow condi-tions.32 The language also contains a large special vocabulary devoted to ori-enteering and geographical terms. Robert Rundstrom has shown by struc-

*This is a retraction of a suggestion made in Viking America that the Norsemen themselves at-tained Alaska. The Eskimo origin interpretation is more plausible and has much evidence tosupport it.

tural analysis of the culture that the cartographic notions were just anotheraspect of a characteristic Eskimo mimicry of the environment, a mimicryalso noted by Carpenter.33 The notions were just as likely to have been pres-ent in the Middle Ages during the migration as more recently. Indeed, suchcartographic notions were probably a major contributor to the remarkablespeed of that migration.

In medieval Europe, however, the idea of the scaled map detailing localfeatures was seldom encountered in everyday life. Geometrical and geo-graphical information was transmitted in elaborate verbal descriptions sup-plemented at most by pictorial diagrams.34 Thus, even though I will con-tinue to use the words map and cartography, the reader must bear in mind thatthese will indicate broader concepts of information transfer as well as paperdrawings. The detailed history of any such transfer obviously cannot beshown. Indeed, if there were any hard-copy prototype source maps that wemight otherwise hope to have found in the historical record, they were prob-ably immediately discarded. Such has been the fate of the majority of me-dieval maps. Once incorporated into world maps by the few cartographerswho could accommodate to them, any localized cartographic materials hadno further purpose. P. D. A. Harvey stresses that before the popularization ofPtolemys chorography,* local maps, as opposed to world maps, were practi-cally unheard of in medieval Europe.

So, there is no hard direct evidenceand we should no longer expect tond anyof any cartographic inuence from the Eskimos to the Norsemenand Europeans. Therefore, we will be unable to provide a deterministic ordeductive history in this matter. But this does not necessarily limit us purelyto speculation. Nor does it necessarily require us to remain permanently inthe dark. We can still make some progress by proposing a formal hypothesisthat such transfer did in fact occur, and then examining the supporting evi-dence. There need not have been extensive contact, just enough to stir in-tellectual curiosity each time it occurred. Such a postulate is obviously notonly plausible but perhaps likely, even though it remains to be sustained. Inarchaeology, such hypotheses are frequently invoked to make cultural senseout of a large collection of objects and observations. In our case, the obser-vations are not eld archaeological but archival archaeological, the worldmaps and texts we shall look at. This postulate, we will see, makes it possibleto weave together a large number of odd, unexplained and previously unre-

*Ptolemys word for the mapping of features restricted to a locality sufciently small so that globalprojection considerations are unnecessary.

lated observations into a unied fabric, a hypothetico-deductive model.Thisis the essence of philosophy, even if the fabric remains to be tested. Passingthe test converts it to science.35

If this fabric withstands the upcoming test, it will suggest that there wasno specic rst European discovery of America as we think of it, not evenby the Norse; instead, there was a gradual Eskimo divulgence to Europe ofcontinental land existing just west of Greenland. Only much later wouldanyone conceive that this was not Asia.

were in the form of experiential, topo-logical descriptions (called cartograms by some writers) rather than rigor-ous grid-referenced datauntil Ptolemaic principles reemerged. Early Eu-ropean maps have a bad popular reputation that can be blamed on a specialaspect of the human brain. Our visual apparatus has evolved to put great em-phasis on recognition of the nely detailed features of familiar human facesthat make them familiar. That same equipment detects minor variations inthe shapes of maps. It is true that early cartograms departed from what is fa-miliar now, but most at least recognizably suggested reality, certainly in thefourteenth and fteenth centuries. At least they were based on information,right or wrong.

The cartographic depiction of Scandinavia as practiced in southern Eu-rope has a very curious history. Scandinavia had no native tradition of mak-ing maps, and for many centuries European maps depicted it quite minimally.(Nevertheless, Scandinavian scholars were quite aware of and kept up withsouthern European cartographical and cosmographical developments.)36Thenin the fourteenth and fteenth centuries many bizarre shapes suddenly ap-peared in the northwest corner of European maps, as we shall see.37 But thoughthese shapes had much variability in their unreality, they showed little or noevolutionary development toward recognizability during these two cen-turies. (A few maps could be construed to have some resemblance to Scan-dinavia, but few indeed and showing no effect on any evolution.) Meanwhile,other parts of the world map evolved recognizably. This atypical circum-stance concerning Scandinavia terminated abruptly early in the sixteenthcentury with the appearance of known Scandinavian cartographers. I willuncover at length a rational explanation of this development as a natural by-product of our hypothesis of Eskimo information divulgence.

The European world gradually lost contact with the Norse Greenlanders,and there is no record of any ofcial communications after the early s.Nevertheless there are many suggestions that Norsemen continued living

there and that Europeans continued going there for another century. Late-fteenth-century European fashions have been found in Greenland burials.Vague bits of information about the lands to the west, as we shall see, mighthave gotten through to Europe by various unofcial, unrecorded channels,both before and after . A similar thing happened with information fromunrecorded New World explorations after .38 What would have hap-pened to such information once it arrived in Europe? Part of the job of manycartographers, at least in the late Middle Ages, was to incorporate new datawhen creating a new map. Once in Europe any American data would havegiven rise to a classical situation in Piagetian epistemology: how were anysuch odd maps, descriptions, and geographical scraps to be understood in Eu-rope? They could not be connected logically to any established knowledgestructure. Mostly, I suggest, they would have been completely ignored anddiscarded rather than divulging the new continent.

How can something stare you in the face and yet be unseen and unmen-tioned? It happens all the time. According to the structuralists, an organismsimply cannot respond to a stimulus unless the stimulus has some rudimen-tary meaning or is somehow known to the organism.39 Infants exhibit thisphenomenon when rst presented with one object disappearing inside orbehind another and reappearing: they actively refuse to look at the anom-alous situation that seems to violate the permanence of objects.40 Later, inthe peek-a-boo stage, they are able to accommodate to this anomaly byconsidering it fascinating, humorous, and entertaining. Similarly, most Eu-ropean intellectuals would have simply rejected as preposterous evidence ofmainland so close in the west.

Nevertheless, a few geographers might have found enough hooks in theestablished state of knowledge to be able to assimilate the new information.One element of the established state of knowledge was, as previously stated,that no maps or information about the geography of Scandinavia existed.Authoritative European cartographic knowledge of Scandinavia was a vac-uum. Meanwhile, consider the hypothetical occasional maps of or informa-tion about American lands transmitted by Greenlanders that might havecome into the hands or ears of European cartographers.These mapsor theindividuals with the informationwould naturally have reached Europe byway of Scandinavia, so some cartographers seem to have jumped to the con-clusion that the cartographic information must therefore relate to the geog-raphy of Scandinavia! It turns out that the unrealistic outlines of Scandinaviawere fairly good likenesses of various American lands. The northwest cornerof the map of the Old World appears, on various occasions, to have assimi-

lated outlines with origins ranging separately from Labrador all the way alongthe Northwest Passage to Alaska, mirroring the route of the Thule migra-tion, under labels of Scandinavia. I will refer to such a mental telescopinghereafter as the provenance paradigm, indicating a presumption that anynovel geographical information may be associated with the locale of its im-mediate provenance.

Thomas Kuhn introduced a denition of paradigm as a framework of as-sumptions that guide the vision of a eld of studya canonical way of han-dling problems.41 He showed that within a given eld, practitioners tend toadhere unquestioningly to the established paradigms for long periods of time,giving them up only when a revolution in the eld establishes a new para-digm. My provenance paradigm may seem rather trivial compared to someKuhn studied, but it meets all his criteria in our restricted eld. Europeancartographers thought, If the data comes from Scandinavia, it must be dataabout something in or near Scandinavia.Why think any harder, especially ifit leads to the preposterous?

During the sixteenth century, learned Europe underwent a revolution inanother larger paradigm, the paradigm within which global geographicalthought took place.This new global geographical paradigm, which we prac-tice still today and which did not become rmly established until manydecades after Columbus, teaches that the globe actually contains a multi-plicity of distinct, unconnected, continents in opposite hemispheres. Beforethe sixteenth century, beginning in antiquity (with exceptions to be notedbelow), the established paradigm held that all the Earths mainland, exclud-ing islands, existed in a single interconnected continental cap, terra rma. Thevery meaning of the word continent originally was held together.That worldcontinent is known to us even today as the Old World, comprising inter-connected Europe, Asia, and Africa. It was known throughout the MiddleAges in Latin as Orbis Terrarum; in Old Norse, Heimskringla, the home cir-cle. The presumed circular outer perimeter was geometrically imposed bytheory as follows. Supposedly the globe was made up of a land sphere and awater sphere, mutually interpenetrating in the same global space. (Watersoaks into land and land sinks under water. There was obviously no conceptof the core.) Slightly off-centered from one another during Creation, theland world rose above the ocean surface but was still attainable by anchorchains if one was not too far at sea.42 In such a globe the boundary betweendry land and ocean is necessarily circular if the land sphere is comparativelysmooth, allowing for minor deviations such as bays and peninsulas. And, mostimportantly, in such a world there can be only one ocean and one terra. In

the Bible, God said, Let the waters under the heaven be gathered togetherunto one place, and let the dry land appear (Gen. :). I shall refer to thisoutlook henceforth as the one-ocean paradigm, reecting the originalsense of the one global ocean, Homers all-encircling Oceanus becomeglobal.

Alternatives to this paradigm, which arose from time to time throughouthistory, speculated about quadripartite worlds,multiple oceans, and antipodes.43

But there were pressures on most scholars to remain within the one-oceanparadigm. Such pressures have been inherent in the established paradigms ofall elds of scientic study and are probably related to fundamental psycho-logical processes.44 But this xation left its practitioners open to being mis-led by any potential contacts with America. According to the one-oceanparadigm, an American mainland would have had to be seen as Asian. Themost famous alleged practitioners of such a one-ocean assimilation of Amer-ica were, of course, Christopher Columbus and his followers, until the exis-tence of the New World was recognized.*

We will see evidence that Columbus was not the rst to be misled in thisway. Hypothetically, with the expansion of the Greenlandersknowledge, di-rect or indirect, of the continental Arctic coast and perhaps down the Cana-dian east coast, the continental nature of the land would become apparent.While European maps had for some time been assimilating local Americanplaces as part of Scandinavia under the provenance paradigm, they could notassimilate this continental aspect using that paradigm. So, the cartographershypothetically concluded, the preposterous was true: it must be eastern Asia.This was the only remaining alternative to postulation of a new continent,an even more radical concept. (Northern and eastern Asia were then as mucha cartographic vacuum as Scandinavia had been.) From the early fteenthcentury onward, southern European scholars increasingly seemed to try toaccommodate their concepts by addressing the question, How far westwardwould one have to sail to reach the continental land of Asia? During thatcentury, as we shall see, many European world maps displayed outlines ofnortheastern Asia that resembled the northeast corner of America.

The one-ocean paradigm, it must be realized, is a total global concept.North America would be seen as Eurasia and even, perhaps, South Americaas Africa (which is southerly from Eurasia and attached to it by a narrow isth-

*While Columbus never personally set foot on North America, Caribbean Indian stories aboutthe continent led him to believe he was near Asia. He sailed briey along Central America with-out thinking it was continental. After his later, brief encounter with South America, he thoughtof it as the theoretical Antipodes, not the ocean-cleaving New World we know.

mus).* The paradigm can invoke a view of eastern America as eastern Asia,but theoretically it could also invoke a view of western America as westernEurope. Thus, any data concerning the northwest part of America (Alaska)would have had to appear on Old World maps in northwestern Europe(Scandinavia). Since the result of this is graphically equivalent to the resultof the provenance paradigm, an ambiguity about the means of adaptationwould arise when Alaskan land appears (which it indeed does).When a mapdisplays such a conguration, I sidestep the ambiguity, not attributing it toeither paradigm in particular, just both collectively.

So, we are going to see evidence that there were two counterbalancingforces in effect for several centuries. On one hand, there were informationsources creating a possibility to divulge the existence of the American con-tinent. On the other, there were paradigms in operation that served to hideany such divulgence. I refer to the provenance paradigm and the one-oceanparadigm collectively as the divulgence-hiding paradigms. In Viking Amer-ica I used different terminology for these, labeling the one-ocean paradigmthe Grand Misunderstanding and the provenance paradigm the SmallerMisunderstanding.I will continue to use the term misunderstanding when re-ferring to specic instances and occurrences, but the paradigmatic conceptis much more useful for understanding the mental processes.

I attempt to demonstrate, for each European map studied, an Americanprototype area that the paradigms transformed into the resultant map. Thisapproach involves minute attention to the details of coastal and interior fea-tures that are preserved during the transformation. Indeed, the necessary un-hinging of ones visual and mental expectations may sometimes seem dis-concerting.

of the divulgence-hiding paradigms in Europe. When stories about the Es-kimos reached southern Europe, they apparently were misinterpreted as ap-plying to legendary peoples in northern Eurasia. The historical myth aboutsuch people of the North is closely associated with the myth of the Arim-phians and the Riphean Mountains, the great stone girdle which the Russpeople believed encircled the whole earth. While legend placed the Arim-phians south of the Riphean Mountains, a people called the Hyperboreans

*Some versions of Ptolemaic maps closed off the Indian Ocean in the east, which would haveprevented conceiving Africa as attainable from eastern Asia. However, this idea was far from uni-versal.

were placed just north of the Ripheans. Among others, the medieval histo-rian Adam of Bremen (ca. ) wrote of the most northerly people inthe world, the Hyperboreans. The name and myth of the Hyperboreansprobably originated in the ancient Greek worship of Apollo.45 As the mythevolved and the Hyperboreans became displaced northward in later centuries,the description of the Hyperboreans became more and more potentially ap-plicable to the American Eskimos.The Hyperboreans (and Arimphians) werealways emphatically described as a happy people.46 Such a description iseminently applicable to the Eskimos. The rst and most vivid descriptionthat all early modern writers on Eskimo life gave was of the Eskimo happi-ness and amiability. When Eskimos were eventually brought by Norsemeninto contact with Europeans and interviewed by learned men, the tempta-tion would have been strong to identify them as Hyperboreans.

Indeed, the standard medieval way of incorporating knowledge of newlyencountered peoples was to try to identify them with something already inthe traditional record. This is the approach described by structuralist theory,looking for hooks of attachment. In this way, the Mongols later became as-sociated with the legends of Gog and Magog, the enclosed nations it wasfeared world start the Armageddon. Another remarkable characteristic of thelegendary Hyperboreans was that when they attained old age, they threwthemselves into the sea from a cliff.This strongly resembles the Eskimo prac-tice of self-sacrice in old age.47 Numerous modern explorers habituallyused the term Hyperboreans to refer to the Eskimos they met, and to this daythe languages of the Arctic peoples are classed as Hyperborean languages. Acentral concept of the classical Hyperboreans was that they were worshipersof the Sun.While the Eskimos religion was primarily animistic and shaman-istic,48 they did have many superstitions and myths regarding the Sun thatwould have fed the mistaken identication. They engaged in elaborate fes-tivals of the winter solstice and believed in many practices intended to en-courage the quickened return of the Sun from the winter darkness.49

Columbus and his successors, operating within the one-ocean paradigm,identied native Americans as Indians. Those who were brought back toEurope learned European languages and related much about their homeland.The Norsemen are also known to have brought Eskimos into contact withEuropean civilization. Karlsefni brought back Skraeling natives from Vin-land. Circumstances suggest that these were Eskimos rather than Indians. AreFrodes early-twelfth-century Islendingabk described a Norse archaeologicalsurvey of Eskimo remains (presumably Dorset) in Greenland: in both theeast and west of the land they found dwellings and fragments of boats and

implements, from which they perceived that the people who lived there wereof the same kind as those of Vinland, whom the Norse Greenlanders callSkraelings.Fourteenth- and fteenth-century contacts were clearly with Es-kimos, identied by their kayaks, their latitude, and other distinctive details.In world traveler Bjrn Einarsson took two young Eskimos into hishousehold while in Greenland. If any Eskimos were ever taken physically toEurope, they would likely have been perceived as Hypereboreans.

An area of Alaska occupied by the western branch of the Thule Eskimosand their immediate predecessors, the Birnirk culture, is shown in Figure .50

It contains a feature concerning the Sun that the legend of the Hyperbore-ans emphasized, the transition from familiar days of daylight and darkness tothe phenomenon of midnight sun and extended days (and alternately nights).That transition, of course, occurs at the Arctic Circle. Until recently mostarchaeologists and anthropologists thought little about the Eskimos astro-nomical concepts, but new investigations by John MacDonald show that theEskimos were astronomically quite aware.51 In particular, they were aware ofthe transition at the Arctic Circle. No sophisticated global concepts areneeded simply to observe the difference in the Suns behavior on either sideof the transition. All that is needed is an awareness of the importance of theSun in daily activities, which they certainly had. In winter above the ArcticCircle during the Great Darkness, they fought a disturbed circadian sleepcycle by keeping time with the stars; the return of the spring sun was accu-rately anticipated and measured in quantitative stages.52 Right at the ArcticCircle on the day of the winter solstice, the Sun is hardly seen at all; but thenext day it resumes its usual appearances, providing a nontechnical means ofspecifying this location. Several Eskimo myths address exactly this phenom-enon.53 Being a primarily coastal people without sophisticated global con-cepts, the original Thules would not likely have had much experience orconcern with the fact that this transition occurs all along the Arctic Circlethrough the continental interior. To them, the transition would have beenassociated with the unique area along the Alaskan coast near Kotzebue Soundand Seward Peninsula. Later in this book we will see cartographic traces ofthis Alaskan area right at the Arctic Circle, but placed in Scandinavia, its ap-propriate place under the divulgence-hiding paradigms.

in a map that purports tobe of another place has always posed problems for historical cartographers,who accept as a truism that appearances can be deceiving.The phenomenonhas been referred to as wishful seeing. In other elds incorporating large

collections of data, the process of searching for novel connections has beenformalized as data mining or even data dredging.54 One famous map byPiri Reis contains shapes that have led Charles Hapgood to unlikely but spec-tacular conclusions about Antarctica that the scholarly world rejects.55 An-other group of maps has been taken by Pablo Gallez to represent the Straitof Magellan even before Magellan went there.56 Less spectacular and moreplausible explanations of these maps have been given.57 There are manyother cases of map interpretations leading researchers astray. An illustrativestudents game in the subject is to turn a recent map of some locality upsidedown and then to try to rationalize it as a medieval map of some otherplace.58 Even respected scholars in the eld have come up with multiple con-tradictory interpretations of the same unidentied coastline. In the smathematician Benoit Mandelbrot formalized the idea that coastlines in na-ture exhibit fractal self-similarity: Seacoast shapes are examples of highlyinvolved curves such that each of their portions canin a statistical sensebe considered a reduced-scale image of the whole.59 In particular he ob-serves, It is hard to tell small and big islands apart, unless one either recog-

Figure . Bering Strait, Seward Peninsula, Kotzebue Sound area of Alaska

Image not available.

nizes them or can read the scales. . . . many islands look very much likedistorted continents.60 The point is, one must exercise the greatest carewhen trying to interpret maps the way I propose we do.

How does one exercise such care? Exactly what should the methodologybe? In the s one answer to that question began to emerge as a move-ment within the community of historical cartographers reacting to Mandel-brots results: forswear any and all use of unidentied coastline interpreta-tions. One way to avoid trouble is simply to consider coastline similarities asnot evidence. This movement has gained wide adherence and has provideda secure area in which to do research on other aspects of historical cartogra-phy. Students in historical cartography are indoctrinated with this attitude,and it is often promoted with zeal. In fact, there is a widespread attitude thatany work that interprets unidentied coastlines must necessarily be wrong.

The doctrine of forswearing coastline interpretation does not yield anideal logical discriminator in the class of Occams razor.The highly respectedhistorical geographer J. Brian Harley declined to join in a blanket rejectionof topographic evidence.61 Mandelbrot himself stated, in a different but stillrelevant context, The eye has enormous powers of integration and dis-crimination. True, the eye sometimes sees spurious relationships which sta-tistical analysis later negates, but this problem arises mostly in areas of sciencewhere samples are very small.62 We will be looking at a large number ofsamples. It is quite possible that in looking away from some coastline simi-larity one may be throwing out the historical-insight baby with the doctri-naire bathwater. An example can be found in Alfred Wegeners theory ofcontinental drift. It was inspired by the matching shape of the coastlines ofAfrica and South America as well as other coastline correspondences.63

Might there be some more-discriminating methodological principle thanlooking away that could at the same time provide a barrier against gross er-rors while yet allowing strong coastline similarities as evidence?

The real trouble with Hapgoods and Gallezs work is not so much thatthey used map shapes as evidence but that they drove the evidence to mul-tiple implausible and unwarranted conclusions. Scientic research, as distin-guished from logical deduction, actually has few hard and fast rules, but itdoes have many rules of thumb that guide researchers toward logically ten-able results.The doctrine of disavowing shape correspondences is an attemptat such a rule of thumb. Besides disregarding that rule, Hapgood and Gallezviolated one rule that astrophysicist Richard McCray calls the tooth fairyrule, which states that a credible theory can invoke a mysterious, unknownagent (tooth-fairy) once, but only once.64 The above-mentioned authors

works contained a multitude of tooth fairies. In Wegeners case there wasa single mysterious, unknown agentthe driving force behind continentaldrift. That theory was subjected to huge controversy in its time, includingclaims of accidental shape correspondence. But it was grudgingly acceptedas correct when investigators discovered the corroborating evidence of sea-oor spreading a half-century later and developed the theory under the nameplate tectonics, and within that discipline, shape correspondence analyses arealive and well today.*

I will adopt a shape correspondence rule here that requires meeting vecriteria:

. Study only large, extended coastlines (not isolated minor wiggles).. Candidate coastal shapes must have multiple discernible features.. The features must be of appropriate proportions and relationships.. Only limited scale adjustments may be involved.. Unless very specically justied, rotations will not be allowed.

Undue attention to minor features in single sources has been the cause of alarge part of the mischief in the past; I stress overall shape and multiple cor-roboration. Only self-evident similarities will be used, avoiding as much aspossible any controversy over whether one shape really looks like another.Some attempt will nevertheless be made to use objective correspondencemethods that incorporate computerized matching. Furthermore, the passingof these criteria must lead to no implausible or mysterious agents or impli-cations whatsoever. The only unknown agent will be the postulate of trans-fer of Eskimo geographical information to Europe via the Norsemen. Thatplausible hypothesis is hardly a tooth fairyof the same magnitude that con-tinental drift was in Wegeners time. I endeavor to supply a plausible basis forevery other implication of a given interpretation.

Wegeners arguments were not based purely on coastline similarity. He alsofound nongraphical support in the distribution of related ora and fauna oneach side of the split as well as interior mountain range correspondences.Wewill look at nongraphical evidence here too, particularly in supporting tex-tual passages.

My approach will attempt to formalize the testing of the hypothesis bymodeling the structure of a hypothetico-deductive system. In this approach,

*The detailed mechanism of seaoor spreading remains a tooth fairy even now, but there isabsolutely no question that Wegener was right.

rather than the classical deductive paradigm used in historical research, theparadigm is that used in contemporary scientic research:

Hypothesis implies observation;observation is true;thus hypothesis is probably true.

The degree of condence in this probability is increased by increasing thenumber of conrmatory observations.This paradigm has sparked its share ofphilosophical controversy, but in the real world it has proven highly usefuland applicable.65 In a softened form, with strict mathematical implication re-placed by expectation, it has been widely used in the softer sciences, like ar-chaeology. In our situation, the observations will be the trivial case of in-stantiation, so I provide the largest possible number of independent instancesexpectable from and consistent with the hypothesis that Eskimo geographi-cal information about America passed through the Greenland Norsemen andinto Europe. At the same time, I hope not to encounter any contradictoryobservations and to be prepared to modify my hypothesis to accommodateany that may arise. I at times apply this paradigm in local arguments for aux-iliary hypotheses as well as overall. In the end, this paradigm is merely a for-malization of the familiar colloquial detectives observation that, whenyoure on the right track, things tend to fall into place.

Nevertheless, the reader must realize that, even with a formally groundedapproach, the ideas presented will also include speculations and conjectures.That continues to be the case with studies on continental drift,66 and hasbeen the case for most of the studies of the Norse encounter with America.Here, speculations will be clearly identiable as such, will stay close to evi-dence, and will be surrounded by conservative examination. And no com-plex structures will be built on speculative bases.

Certainly any conservative overview of the maps analyzed here must al-low that any of the correspondences of shape noted with North Americanland could be accidental.We shall see dozens of them, and a difcult problemis whether such correspondences are all likely to be accidental simultane-ously.When the correspondences all lie in the same area of the world, all aresubject to the same paradigms, and the selection of source maps is dictatedby the entire body of available evidence rather than being arbitrary, the prob-ability of accidental coincidence decreases substantially. We will see many ofthese correspondences corroborating one another, to build a circumstantialcase for correctness of the approach. Readers who can withhold a reexive

disapproval of all correspondence arguments a priori will see that the resultshere are plausible, do corroborate one another, and are corroborated by topo-logically separate (and separately weighted) interior features and other non-graphical evidence. Readers may then decide for themselves if accidental co-incidence seems, while possible, likely or not.

IOutstandingMisunderstandings

, made the rst Latin translationof the ancient Greek geographer Ptolemys work. In the second century ..,Ptolemy had perfected an advanced scheme of cartography from his prede-cessor Marinus of Tyre, one that allowed at maps to account for the curva-ture of the earth. Ptolemy also avoided the decay of accuracy over gen-erations of map copying by attaching tables of numerical coordinates forlatitude and longitude for each mapped location, an early example of the -delity of digital reproduction. His cartographic work, which was in additionto his theoretical treatises, may have included an atlas containing a world mapand many local maps, all with coordinate tables.1 The rekindling of interestin Ptolemy in the Renaissance ultimately would sweep away many of theproblems associated with medieval cartography. More and more Renaissancescholars began to read his work, and among the earlier to do so was theFrench cardinal, Gulielmus Filiastrus. Filiastrus had someone make him aLatin translation of the map captions in a Greek atlas presumed to have beendescended from Ptolemys. The Latinized maps were appended to a translation of Ptolemys theoretical work and handed over to Filiastrus in. Filiastrus dedicated the work to the pope and took the manuscripthome with him to Nancy, France. Following his death the next year, it re-mained there apparently without inuence on other cartographers until itsrediscovery in .

This Filiastrus Ptolemy of is the earliest known copy to have anymodications or additions to Ptolemys original maps and texts. Specically,besides the Latinization, it contains a map (Figure ) and a textual descrip-tion of the Scandinavian regions. Scandinavia had been vaguely known toPtolemy and even his predecessors Pliny and Mela, but not with enough de-

tail that he could include it in his local chorographic maps. Early in the co-ordinate tables listing the source of the map is made clear.2 Namely, whiledescribing in great detail the small islands between Denmark and Scandi-navia, the lists text includes the description: in the South, Salinghi, onwhich is located Salinga, the native home of Claudius Clavus Suarto, the sonof Niels Petrus Tucho and Margaret Ingrid Cicilia Osee Strango Viningh,chorographer of these present parts, at longitude, latitude.

Claudius Clavus Niger Cymbricus is the Latinized name of a Dane whomay locally have been known as Klaus Schwartz. He was an educated trav-eler who seems to have appeared in Rome around . Clavus apparentlywas a visitor at the Curia of Pope Martin V in . It is perhaps there thathis work became known to Cardinal Filiastrus. In his time, Clavus came tobe considered the leading authority on the geography of the Northin-deed, the only authority. While modern scholars continue such plaudits, itwill become apparent that Clavus was not actually trained as a geographer.In fact, I show that Clavus was not actually an authority on the North, theScandinavian geographical vacuum remaining until more than a centurylater, to be dispelled by one Jacob Ziegler and his contemporaries. Clavus leftno known record of the sources for his geographical ideas. Practically all thatis known of Clavus has been reconstructed from secondary references whocite Clavus as their authority.3 From these references it seems that Clavus didnot take his ideas of the North from a lone source. Rather, his ideas evolvedas he himself discovered more sources. I use the word discovered becauseClavus apparently belonged to that group of humanist scholars who, likePoggio, took great delight in searching libraries for forgotten books and ob-scure manuscripts. I suggest, as an auxiliary hypothesis, that it must have beensome such obscure information that Clavus, interpreting and lending undueauthority thereto, thrust into the vacuum of Scandinavian cartography. Theinformation was not necessarily particularly old, just esoteric. Support forthis hypothesis will be forthcoming.

Not only was Clavuss map the rst chorographic addition to Ptolemy, butit was the rst known map to claim to show Greenland graphically. (The to-ponym Greenland had appeared on earlier maps, but no outline.) Mind you,this earliest map to show Greenlands outline did not appear until a decadeafter Greenlands settlements supposed extinction. So, what was Clavusssource of data? Was it strictly conjectural? It would be a far stretch to thinkthe outlines on this map could have been derived from knowledge of eitherGreenland or Scandinavia itself (see Figure and front map). We shall seethat, while an element of conjecture was involved, there was probably an ac-

tual source of geographical data about some somewhere transmitted viathe Greenlanders.

The rst step of interpretation necessary is to note the unreal arc of landthat the map shows connecting Greenland to the Eurasian continent at thebase of Scandinavia. Other writers have attempted to explain this in variousways.To some this arc looks as though it could represent the edge of the po-lar ice pack. Nansen believed that this representation involved some mis-construction of the leftmost headland of the Medici Atlas of (Figure), perhaps one of Clavuss sources.4 But there is also a possible reality-basedway to explain Clavuss unreal arc (as well as the Medicis). In reality, north-ern Greenland does come into proximity with other land: the Arctic Ar-chipelago north of Canada, including Ellesmere Island. And this archipelagodoes come into close proximity with the continental Canadian coast. In fact,these proximities, bridged by winter ice, are exactly what made possible theThule migration into northern Greenland from Alaska. We will see belowthat Clavus had information about that migration. Any description of thatmigration would necessarily include some suggestion of vague connectionbetween northernmost Greenland, where the crossover occurred, and con-tinental land. But Greenland lay far to the west of Europe. Under the one-ocean paradigm, Clavus was evidently unwilling to consider it so far west asto spring from eastern Asia and was forced to promote the hypothetical arcconnecting Greenland to northeastern Eurasia somewhere beyond Scandi-navia.

Many writers have unquestioningly followed the modern Scandinavianscholar Axel Bjrnbos assertion that this conjectured arc was a fundamentalpart of all Scandinavian geographical theory. It was widespread, but I attempt,following Andrew Fossum,5 to show later that there were several other geo-graphical conceptions existent in the North.

Clavus himself admits indirectly, through his coordinate table, that the arcconnecting Greenland to Eurasia is essentially conjectural. Namely, he listsspecic coordinate data for every other place on the map except this arc. Af-ter describing three documented promontories of Greenland at North(by the left-hand scale), Clavuss list terminates. In his textual chorography,he gives the coordinates of those rst three headlands of Greenland and then,after the third, says: But from this headland an immense country is extendedeastward as far as outermost Russia. The pagan Karelians occupy the northof it, and their territory is extended past the North Pole to the eastern Seres[China], and therefore the pole which to us is in the north, is to them in thesouth at . Clavuss belief that the North Pole could be in the north for

some and in the south for others shows that he was not a trained geographer.He uncritically accepted an old gaffe of Marco Polos, apparently based on aconfusion between the North Pole and the Arctic Circle, in fact at approx-imately north latitude.*

Still more anomaly appears in Clavuss mention of the pagan Kareliansas occupying northern Greenland. The Karelians were actually a Finno-Ugrian people who inhabited the regions around the White Sea. In timespast they had indeed been quite resistant to attempts at conversion to Chris-tianity, but somewhat before Clavuss time they had been completely Chris-tianized.This pagan Kareliananomaly has already been explained by JosephFischer and other authors as an example of the provenance paradigm, with-out using that term, putting the Karelians in the Eskimos place.6 This is sim-ply a variant of the Eskimos identication as Hyperboreans.We will see fur-ther development of the idea that Clavus was aware of the Thule Eskimosand their migration, even though misidentifying them as Karelians.

An alternative interpretation regarding this Karelian anomaly has been de-veloped by linguists over the past century. As noted before, the GreenlandEskimos absorbed certain Norse loan words into their vocabulary.The Norseword for the Eskimos was Skraelings. Under the same pronunciation-shiftparadigms for the other loan words, Skraeling becomes something close toKareli in Eskimo,7 and it may be the origin of the Greenland Eskimos alter-nate name for themselves as Kalali instead of Inuit. In either case, Karelians asused by Clavus refers to Thule Eskimos.

The hard data in the rest of Clavuss coordinate table, as reected in themap, will provide us with our rst tangible evidence of the cartographic mis-understandings suggested. Let us focus our attention on the strange graphi-cal representation that Clavus identies as Scandinavia (refer to Figure ).The rst step of analysis for this or any other map we shall look at will be todeconstruct the map, using this word in a slightly different form of what wasadvocated by the historical geographer Brian Harley.8 He argued that mapsconstitute a cultural text akin to verbal texts in that they can be analyzed forfar more information than just their surface appearance. However, instead oflooking for political or social ideologies, we will be extracting alternativegraphical dualities that escaped the creators conscious mind. When a car-tographer puts together a new map, he has many things in his mind. The

*Apologists for Clavus tried to explain this away with convoluted arguments. Apparently theywere unaware that the idea came straight from Marco Polo (Chapter of The Book of Marco Polo,Road to Cathay).

foremost is usually some preexisting picture given by other cartographers. Al-most all new maps incorporate parts of previous maps, to which alterations oradditions are made. Attempting to unveil the rationale of those alterations andadditions can give a glimpse into the cartographers mind and less obvioussources. In reconstructing the map while analyzing the hypothetical inuenceson the cartographer, we will always start by focusing on the graphical essencethat was new in the map at its time and has no known source in reality.

To that end, we should not be focusing, in Clavuss map, on the knownlands of Denmark, England, and Ireland, which were all copied fromPtolemy.We should also not be focusing on the conjectural northern Green-land coastline arcing to Asia, for the very reason that it is conjectural and un-documented in the coordinate table. Let us also, for the time being, not fo-cus on the toponymic cities and landmark identications, which are notgraphical and which probably would have come from an originally separatedata resource.What remains that is cartographically novel, including the threequantitatively specied data points or promontories of Greenland from thecoordinate table, is shown in the top of Figure .This is the new cartographicor graphical text, whether hard copy or descriptive or imaginary, that Clavus,from whatever sources and for whatever reasons, tted into the existingchorography of northwesternmost Europe.

The bottom of the gure shows the accurate graphical chorography ofnorthwesternmost America, a detailed map of the Bering Strait and SewardPeninsula. The resemblance is striking, sufciently so to suggest an outra-geous hypothesis. It appears that this could be a candidate for a misunder-standing that placed Alaskas chorography in Scandinavia. Clavuss Gron-landia Provincia,the remaining three data points when the conjectural uppercoastline is removed, could be interpreted as the tip of the Chukchi Penin-sula of Siberia; his Islandia occupies the same position as the Diomede Is-lands; and his would-be Scandinavia nearly perfectly represents the actualSeward Peninsula.These Eskimo lands bear the same geographical interrela-tionships as the Scandinavian ones. As we proceed, we will see evidence tomake the above hypothesis seem less and less outrageous.

The area around Bering Strait has always been of great signicance in Es-kimo history. It and nearby areas are considered to be the birthplace of manyEskimo cultures, including the Thule.*9 Is it plausible to think that Thule

*While the name archaeologists give to a culture comes from the site of its rst modern un-earthing, this has no relation to the historical site of its rst development. That is determinedfrom extensive further excavations and analysis. Early in the history of Thule excavations the site

cartographic information about this area could have made its way to theNorsemen in Greenland and thence to Clavus in Europe? One guide to mak-ing this interpretation might be sought in the historic Eskimo geographical

of origin seemed to be Point Barrow on the north coast of Alaska. But Thule sites just as earlyor earlier have been found along the Bering Sea south of Seward Peninsula. This highlights theintense communication and mobility of the predecessor culture.The Thule explosion grew outof the Birnirk Culture, which had settlements in Seward Peninsula.

Figure . Portion of map purportedly of Scandinavia by Claudius Clavus (a) com-pared with Alaskas Seward Peninsula area (b)

Image not available.

interaction with early modern explorers. The historic Inuit Eskimos are di-rect cultural and genetic descendants of the Thules. Such early modern ex-plorers found authoritative Eskimo guidance widely available at many points.10

One of the early explorers in this Bering Strait area, Frederick Beechey, en-countered a party of local Eskimos, who of course shared little if any lan-guage with him and must have communicated primarily by gestures. Theywere in the act of drawing in the sand exactly such a chart of Seward Penin-sula and Bering Strait that included the Diomede Islands and the tip of theChukchi Peninsula:

On the rst visit to this party, they constructed a chart of the coast upon the sand,

of which I took very little notice at the time.Today, however, they renewed their

labor, and performed their work upon the sandy beach in a very ingenious and

intelligible manner. The coast line was rst marked out with a stick, and the dis-

tances regulated by the days journeys. The hills and ranges of mountains were

next shown by elevations of sand or stone, and the islands represented by heaps

of pebbles, their proportions being duly attended to. As the work proceeded,

some of the bystanders occasionally suggested alterations, and I removed one of

the Diomede Islands which was misplaced: this was at rst objected to by the hy-

drographer; but one of the party recollecting that the islands were seen in one

from Cape Prince of Wales conrmed its new position, and made the mistake

quite evident to the others, who seemed much surprised that we should have any

knowledge of such things.When the mountains and islands were erected, the vil-

lages and shing stations were marked by a number of sticks placed upright, in

imitation of those which are put up on the coast wherever these people x their

abode. In time we had a complete topographical plan of the coast from Point

Darby to Cape Krusenstern. . . .

. . . For East Cape they had no name, and they had no knowledge of any other

part of the Asiatic coast. Neither Schismareff Bay nor the inlet in the Bay of Good

Hope [both in Alaska] was delineated by them, though they were not ignorant

of the former when it was pointed out to them.11

So Beecheys Eskimos geographical knowledge covered exactly the areashown in the bottom of Figure , albeit with some imperfections. Rund-strom has noted that in Inuit cartography it is the activity rather than the -nal object that is important. This particular group was having an animateddiscussion as they proceeded, correcting one anothers contributions to thedrawing. They carried out their activity on more than one day. Clearly thisactivity constitutes a group learni