This is the accepted manuscript of the article published online in the Journal of Early
Modern History in 7 March 2019: https://brill.com/view/journals/jemh/23/1/articlep1_1.xml
Early Modern Nautical Charts and Maps:
Working Through Different Cartographic Paradigms
Joaquim Alves Gaspar & Henrique Leitão
Centro Interuniversitário de História das Ciências e da Tecnologia
Faculty of Sciences, University of Lisbon
Abstract
Of all the technical and scientific developments that made possible the European
maritime expansion, the nautical chart is perhaps the least studied and understood.
This fact is very surprising as it was with the information contained in those charts, and
later imported to geographical maps and atlases, that the newly discovered lands were
first shown to the European nations. There was, however, a deep incompatibility
between these two cartographic paradigms – the nautical charts and the geographical
maps – which remained unsolved throughout the sixteenth century and beyond,
despite the attempts to harmonize the technical principles of Ptolemy’s Geography
with the advances of nautical cartography. An eloquent symptom of such
incompatibility was the difference between what was understood as an accurate
depiction of the Earth, in the eyes of cosmographers and geographers, and what was
considered by the pilots as an accurate nautical chart. The misunderstandings around
these issues during the early modern period and the unsuccessful attempts at
reconciliation were, in great part, the cause for some polemics among cosmographers,
cartographers and pilots, such as the conflict in the Casa de Contratación around the
charts of Diego Gutiérrez, a fact not entirely understood by historians. At the core of
the difficulty lies the circumstance that only in the present day has the true nature of
the nautical chart, as a navigational tool, started to be clarified. How the differences
between geographical maps and nautical charts contributed to shape the History of
Cartography in various periods, and how they are related to conflicting scholarly
objectives and practices, is the subject of this essay. We will show, using the results of
cartometric analysis, that not only were those artifacts constructed using different
principles and with different purposes, but that they belonged to incompatible
cartographic paradigms, and we will argue for the relevance of this fact for the history
of science.
1
Introduction
Late-medieval and early-modern maps and nautical charts are complex artifacts that
have often challenged – and eluded – the understanding of scholars. Still, the historical
relevance of these documents is such that the results of their study often transcend
the boundaries of specialized fields, such as the History of Cartography or the History
of Navigation. It is the purpose of this essay to shed a new light onto the differences
between geographical maps and nautical charts historically, and to explain how these
differences contributed to shape not only the history of cartography but also to affect
the history of science at large.
The importance of focusing attention onto the subject is obvious: from the 15th, and
well into the 18th centuries, nautical cartography was the single most important source
of information for the construction of a global and geographically coherent depiction
of the world, most particularly the depiction collected during the sea voyages about
the coasts, islands and the shapes of landmasses abutting the oceans. The image of the
Earth that was conveyed during this period, as well as the discourse about the orbis
terrarum, was inevitably influenced by the origin of the information. The
consequences of importing geographical information from nautical charts into
geographical maps, more specifically those resulting from the fact that they belonged
to distinct cartographic paradigms, are deeper and historically more relevant than
previously thought. 1 In fact, although these two models have co-existed since the
Middle Ages, it was only in the early modern period that their inherent differences
became significant.
While in broad terms the differences between maps and charts are not new for
historians, their full historical significance and implications to knowledge-making
scientific practices of the time have not yet been fully explored. In fact, it was only in
recent years that techniques of cartometric analysis and numerical modelling,
especially designed to assess and simulate the geometry of old charts, were
successfully developed and tested. One of the most important conclusions that can be
drawn from the application of such techniques – this is indeed a crucial point to note –
is that these two cartographic models, that of the geographical map and that of the
nautical chart, were often incompatible with each other: not only their geneses,
evolution and use were associated to different social and professional circles and
practices but, more importantly, they were designed according to different principles
and exhibited different geometries.
The term ‘paradigm’ in the present context has no relation to the homonym Kuhnian concept that was
proposed in The structure of Scientific Revolution (Chicago: The University of Chicago Press, 1962). By
choosing this word, we aim to emphasize the relevance of aspects that are beyond the strict technical
matters of map construction, such as the professional context of chart making and the use of charts in
navigation. Throughout this text we will use the word ‘cartographic model’, instead of ‘cartographic
paradigm’, whenever we focus on the technical component only.
1
2
Secondly, we will show that to fully understand nautical charts historically, it is
imperative to know how navigation took place on board ships. Not only were those
charts constructed using the navigational data collected by the pilots by means of the
contemporary navigational methods, but also, they were used to support navigation
exactly the same way. 2 This can be stated in a somewhat metaphorical, albeit
expressive manner: one needs to look at these charts not only with the pilots’ eyes
but, more importantly, with the pilots’ hands. In other words, the full comprehension
of old nautical charts requires a full cognizance of the contemporaneous practice of
navigation, a fact that has not been sufficiently explored by historians. The intimate
connection between nautical cartography and navigation is an aspect that will become
evident in this essay. Hence, a brief, but necessary, detour into the technicalities of
early modern navigation might be useful.
Finally, we will note how pilots, cartographers, cosmographers and mathematicians of
the early modern period were, most of the time, incapable of grasping the intrinsic
differences between nautical charts and geographical maps, and how this difficulty
often led to misunderstanding and conflict. This article will make evident that, at the
core of the misunderstanding, lies the circumstance that information of geographical
interest was being acquired, structured, and displayed according to different processes
and with different purposes. Whereas for pilots, the nautical chart was an “instrument
to navigate” -- whose construction and use were to be understood in the scope of that
specific activity-- for geographers and other scholars, nautical charts were treated as
maps, that is, as attempts to represent the geography of the world. Thus, the very
same object was perceived in a distinct way by two professional groups.3
An important aspect of those conflicting views concerns the two different concepts of
accuracy that were at play. Whereas for geographers and learned cartographers, the
accuracy of a map was measured by the agreement between the latitudes and
longitudes represented on it and their correct values on the surface of the Earth (what
we call today absolute positional accuracy), for pilots and nautical cartographers, the
2
It is now consensual among most historians of cartography that pre-Mercator nautical charts were
constructed with compass courses, estimated distances and – from the beginning of the sixteenth century
on – observed latitudes. This kind of functional relation is not applicable to the present-day practice,
where the surveying methods used to acquire the information for the charts are usually distinct from the
navigational methods.
3
Our topic is therefore relevant to the classical debate around the role of artisans vs. scholars in early
modern Europe. For a recent and innovative statement of the issues at stake, see: Matteo Valleriani (ed.),
The Structures of Practical Knowledge (Dordrecht: Springer, 2017). See especially the introductory
article of Valleriani, “The Epistemology of Practical Knowledge”, pp. 1-19. See also Pamela H. Smith,
The Body of the Artisan: Art and Experience in the Scientific Revolution (Chicago: The University of
Chicago Press, 2004); Pamela O. Long, Artisan/Practitioners and the Rise of the New Sciences, 14001600 (Corvallis: Oregon State University Press, 2011); Lisa Roberts, Simon Schaffer and Peter Dear
(eds.), The Mindful Hand: Inquiry and Invention from the Late Renaissance to Early Industrialisation
(Amsterdam: Koninklijke Nederlandse Akademie van Wetenschappen, 2007); Pamela H. Smith, Amy R.
W. Meyers and Harold J. Cook (eds.), Ways of Making and Knowing: The Material Culture of Empirical
Knowledge (Ann Arbor: The University of Michigan Press, 2014).
3
accuracy of a chart was measured by its efficacy in the accomplishment of a specific
task –navigating a ship in a sea voyage. Thus, a chart was navigationally accurate when
the latitudes of the places, as well as the compass courses and distances between
them were accurate, and not necessarily when the shape of the landmasses matched
their “true” form. 4
What made this situation highly prone to ambiguity and conflict was the fact that a
specific object – the nautical chart – when transported from the realm of pilots to the
realm of geographers and learned cartographers, did not carry a clear (and, much less,
an explicit) explanation of its nature. Hence, the two professional groups interested in
geography - pilots and scholars - were not solely separated by social barriers but also
by a deeply contrasting interpretation of the meaning and content of a specific object,
and of the processes by which geographical information was acquired. 5
Several technical studies that support our work have been published in recent years,
but the full implications of these studies for the history of science – namely, the fact
that those two cartographic paradigms collided – have not been sufficiently explained
before. That is the purpose of this article. 6
Maps and charts
When a present-day topographical map and a nautical chart depicting the same region
are put side by side, the differences are evident: while in the topographical map the
information is concentrated inland, leaving the wet areas almost totally empty, in the
nautical chart the information is mostly located at sea and in the coastal areas, leaving
the dry land areas virtually blank. The reasons for this difference are related to the
distinct purposes of the two kinds of representations: while topographical maps are
intended to show all types of general information located on land, nautical charts are
constructed with the explicit goal of supporting marine navigation and display only the
features that can be useful for such purpose. 7 A subtler difference, noted by the late
American cartographer Arthur Robinson, concerns the way maps and charts are used:
4
We will see later in this article that this desideratum could only be met for some specific routes on the
chart.
5
This social barrier mostly applied to Iberian pilots and cosmographers and was usually not a feature in
Northern European countries. We are indebted to Sarah Tyacke for this observation.
6
For a general explanation on the use of cartometric methods of geometric analysis and numerical
modeling of old charts, see Joaquim Alves Gaspar, From the Portolan Chart of the Mediterranean to the
Latitude Chart of the Atlantic: Cartometric Analysis and Modeling. Unpublished doctoral dissertation
(Lisboa: Universidade Nova de Lisboa, 2010). For a discussion around the nature of early modern
nautical charts and its connection with the contemporaneous navigational methods, see Joaquim Alves
Gaspar and Henrique Leitão, ‘What is a nautical chart, really? Uncovering the geometry of early modern
nautical charts’, Journal of Cultural Heritage, 29 (2018), 1: 130-136.
7
Other less obvious differences, not perceived by the casual map reader, are the distinct vertical
references used to reckon depths and elevations, and the map projections adopted in the two types of
representations.
4
“maps are to be looked at while charts are to be worked on”. 8 Although Robinson’s
aphorism is wonderfully synthetic and expressive, it only focuses on the use of present
day maps and charts, and does not capture the whole spectrum of differences and its
historical significance. Looking into the history of maps and charts makes one realize
that, not only their geneses and evolution are distinct, but also their dissimilarities are
conspicuous, significant and historically relevant. While in general qualitative terms
this is not new for historians, the subject has never been addressed technically. When
browsing through the published volumes of the History of Cartography, most
especially those dedicated to the Middle Ages and Renaissance, it is notable how
terrestrial maps and nautical charts have been traditionally studied separately. 9
Moreover, the geometry and construction of old nautical charts are almost exclusively
addressed in a qualitative way, and the fundamental question of what happened when
information from nautical charts was imported to geographical maps is not discussed.
The result has been that some incorrect interpretations of those charts and maps have
been perpetuated in the historiography, as we will indicate later in this article. 10
The Renaissance chart of the Atlantic
Navigation in the European waters during the Middle Ages and early modernity was
usually made near the coast, using the available portolan-type charts and the
information registered in the pilots’ rutters. Sometimes it was necessary to move away
from the coast, to reach a distant island or cross a larger body of water, but these open
water tracks seldom took more than a few days. In those situations, the position of the
ship was determined using the information about the course steered and the
distanced sailed, estimated by the pilot. This method, designated in the present day as
dead reckoning, was known by the early modern Portuguese and Spanish pilots as the
ponto de fantasia (point of fantasy), a colorful name that expressed the uncertainty of
the estimation process (Fig. 1, left). Although considerable errors could be made with
the point of fantasy, they seldom represented a serious problem for navigation
8
Arthur Robinson, Joel Morrison, Philip Muehrcke, A. Kimerling and Stephen Guptill, Elements of
Cartography, Sixth Edition (New York: John Wiley & Sons, 1995), 15.
9
In the first and third volumes of the History of Cartography there are eleven chapters totally or partially
dedicated to the history of nautical charts. See J. B. Harley and David Woodward (ed.), The History of
Cartography, Volume One, Cartography in Prehistoric, Ancient and Medieval Europe and the
Mediterranean (Chicago: The University of Chicago Press, 1987); David Woodward (ed.), The History of
Cartography, Volume Three, Cartography in the European Renaissance, Parts 1 and 2 (Chicago: The
University of Chicago Press, 2007).
10
There is a single chapter in the History of Cartography where the mathematical aspects of chart
construction are addressed: John Snyder, ‘Map Projections in the Renaissance’, Volume Three, Part 1,
365-381. The author postulates that both the portolan chart and the Renaissance nautical chart are based
on the cylindrical equidistant (or equirectangular) projection. This is a repetition of an erroneous
interpretation that originated in the sixteenth century, propagated to the present day and was adopted by
respected authors in other important international publications.
5
because the correct position of the ship could be easily found as soon as the coast was
again in sight. 11
By mid-fifteenth century, Portuguese ships were already navigating away from the
coast, in long-distance voyages that often took several weeks or months. In these new
conditions, positioning the ship by dead reckoning was inadequate. As time elapsed
from the last known position, the accuracy of the point of fantasy steadily degraded to
the point of becoming almost useless. To face the new problem, a new navigational
method had to be found. The solution was the introduction of astronomical
navigation, made possible by the adaptation of the instruments of observation used by
astronomers on land – the astrolabe and the quadrant – and the development of very
simple procedures that could be effectively used by uninstructed pilots on board, to
observe the heavenly bodies and determine latitude. 12
The point of fantasy then gave place to the so-called ponto de esquadria (set point),
where the latitude became the prevailing element of navigational information. The
position of the ship started to be determined on the chart as the intersection between
the line representing the course steered and the parallel representing the latitude (Fig.
1, right). The superlative importance of this technical development cannot be over
emphasized, as the reliability of the new method permitted to establish regular
oceanic routes from Europe to India and beyond, for the centuries to come. We will
see next how the introduction of these techniques had a major impact on the
geometry of nautical charts.
11
The practice of navigation and the use of nautical charts in Europe, during the Middle Ages and
Renaissance, is addressed in various studies, most of them focused on the Mediterranean. See Eva Taylor,
The Haven-Finding Art: An History of Navigation from Odysseus to Captain Cook (London: Hollis &
Carter, 1956); David W. Waters, The Art of Navigation in England in Elizabethan and Early Stuart Times
(London: Hollis and Carter, 1958); Patrick Gautier-Dalché, ‘L’usage des cartes marines aux XIVe et XVe
siècles’, in Spazi, tempi, misure e percorsi nell’Europa dal Bassomedievo: aati del XXXII Convegno
storico internazionale, Todi (Spoleto: centro Italiano di studi basso Medievo – Academia Tudertina,
1996), 97-128; James Kelley, On Old Nautical Charts and Sailing Directions: Technical Essays (Melrose
Park, PA: Sometime Publishers, 1999); Piero Falcheta, ‘The Use of Portolan Charts in European
Navigation During the Middle Ages’, in Europa im Weltbild des Mittelalters: Kartographische Konzepte
(Berlin: Academic Verlag, 2008), 269-276; Eric Ash, ‘Navigation Techniques and Practice in the
Renaissance’, in Woodard & Harley (ed.), The History of Cartography, Volume Three: Cartography in
the Renaissance, Part I (Chicago & London: The University of Chicago Press, 2007), 509-527. Several
Portuguese and Spanish sources contain descriptions of navigational methods in the early modern period,
the oldest being the anonymous Portuguese manuscript of ca.1508 known as the Livro de Marinharia de
João de Lisboa. Among the best-known treatises explaining how to determine the position of the ship on
the chart are those of Alonso de Chaves, Quatri Partitu en cosmographia practica y por otro nombre
llamado espejo de navegantes (1520-38), chapter 2; Martín Cortés, Breue compendio de la sphera y de la
arte de nauegar (1551), chapter XIII; Rodrigo Zamorano, Compendio de la arte de navegar
(Sevilla,1581), chapters 20 and 21; and Francisco da Costa, Arte de Navegar (1596), chapter LIII.
12
A good introduction to the genesis of astronomical methods of navigation is Luís de Albuquerque,
‘Astronomical Navigation’, in Armando Cortesão, History of Portuguese Cartography (Coimbra:
Agrupamento de Estudos de Cartografia Antiga, Junta de Investigações do Ultramar, 1971), 221-442.
6
N
N
PD
d
PD
C
C
SP
PF
Figure 1 – Point of fantasy (left) and set point (right). The point of fantasy PF is the intersection
between the line representing the course, C, and the arc of circle representing the distance
sailed from the point of departure, PD. The set point SP (right) is the intersection between the
line representing the course and the parallel of latitude (the horizontal line).
During the fifteenth century, nautical charts used by Portuguese pilots in voyages of
exploration and trade adopted the same cartographic conventions as the portolan
chart of the Mediterranean. Whenever the value of magnetic declination was small,
these charts could still be used with the new astronomical methods of navigation. That
was indeed the situation in the Atlantic, north of Cape Vert, during the fifteenth
century. But as ships progressed to the south, the value of magnetic declination
markedly increased, and it eventually became obvious to pilots that the old
cartographic model was no longer compatible with the new navigational methods. The
incompatibility of the traditional charts with the set point method, under the effect of
magnetic declination, resulted from the fact that magnetic declination tilted the eastwest lines implicit in their geometry (parallels of latitude) relative to the east-west
direction (Fig. 2, left). On the contrary, a chart based on the set point method always
represented points with the same latitude on the same horizontal east-west line (Fig.
2, right). 13 A new and better chart had thus to be designed, this time based on the set
point method.
The Cantino planisphere, drawn by an anonymous Portuguese cartographer in 1502, is
usually considered the oldest extant nautical chart incorporating the latitudes of the
places (Fig. 3). With this chart, a new cartographic standard was established, adopted
by many other world maps of the beginning of the sixteenth century, such as the
Caverio planisphere (ca. 1504), the Hamy-King planisphere (ca.1504) and the two
printed world maps of Martin Waldseemüller (1507; 1516).
13
For a detailed explanation see Gaspar, From the Portolan Chart of the Mediterranean to the Latitude
Chart of the Atlantic (note 6), 13-21; ‘Blunders, Errors and Entanglements: Scrutinizing the Cantino
Planisphere with a Cartometric Eye’. Imago Mundi, 2012, 64, 2: 184-186.
7
N
N
PD
d
PD
C
C
C
C
m
m
PF
PF1
SP
SP1
Figure 2 – The influence of magnetic declination on the point of fantasy (left) and the set point
(right). PF1 and SP1 are, respectively, the point of fantasy and the set point, as affected by
magnetic declination . Notice how magnetic declination only affects the longitudinal position
of the set point, thus conserving the observed latitude .
Figure 3 – The Cantino planisphere (1502) depicts the world as it became known after the
exploration voyages in the end of the fifteenth century and beginning of the sixteenth to
Africa, Greenland, Newfoundland, Brazil and India. It is usually considered the first ‘latitude
chart’, that is, the first nautical chart representing places according to latitudes. Biblioteca
Estense Universitaria (C.G.A.2), Modena.
How accurate are those world maps? Or, more precisely, how well do the shapes of
the landmasses agree with the equivalent shapes on a modern map? Occasionally,
historians have tried to give an answer to this question by comparing the coastlines
depicted on them with the corresponding lines in a modern map. Such exercise is
equivalent to what is called today an assessment of absolute positional accuracy,
where the geographical positions shown on an old map are compared with the ones
accepted as true. The process is, however, theoretically meaningless unless the two
8
representations share the same map projection. 14 A possible way out of this difficulty
is to assume that the geometry of those early modern planispheres is identical – at
least approximately – to the one of the cylindrical equidistant projection centered at
the Equator (the so-called plate carré), where meridians and parallels form a square
grid. That was indeed the (erroneous) interpretation of most map historians until very
recently. While we can confidently assume that the parallels of latitude that are
implicit in those early modern planispheres are approximately oriented in the eastwest direction, which is a direct consequence of the use of the set point method, a
similar assumption cannot be made for the meridians, whose orientation reflects the
convergence of meridians and, as we will show, the effect of magnetic declination. 15
Figure 4 – Coastlines of Africa and Brazil in the Cantino planisphere (left), compared with the
corresponding outlines in a modern map (right). Notice the eastward displacement of Africa
and Brazil, and the longitudinal stretching of Africa. The projection of the modern map is the
equirectangular projection centered at the Equator, commonly known as square chart or plate
carrée.
Indeed, if we compare the representation of the South Atlantic in the Cantino
planisphere with the corresponding one in a modern plate carré map, the result is
unexpected (Fig. 4): Brazil and Africa appear both displaced to the east and the African
continent is markedly stretched in the east-west direction, making the Isthmus of Suez
look enormous, a feature that was replicated in all European cartography of the
14
A map projection is any systematic arrangement, on the plane, of the meridians and parallels defined
on a geometrical model of the Earth (a sphere or an ellipsoid). Different map projections exhibit different
geometries, depending on the geometrical or analytical process used to transform geographical
coordinates ( ) into plane coordinates (x, y). For example, in all cylindrical projections, meridians and
parallels are straight and perpendicular to each other, forming a mesh of rectangles. Other projections
may depict meridians and parallels as curved lives. Thus, comparing the geometries of an old chart with
the one a modern map, in order to assess its accuracy, only makes sense when they share the same map
projection.
15
This is easily perceived by the visual inspection of the geographical grid of meridians and parallels
implicit to those representations, which can be estimated using a sample of control points of known
latitudes and longitudes. For an explanation of the process see Gaspar, From the Portolan Chart of the
Mediterranean (note 6), 45-54.
9
sixteenth century and beyond. 16 An attractive reason that could be invoked for this
apparent distortion is the inadequacy of the surveying techniques of the time, when
longitude could not be determined accurately. An alternative explanation, proposed by
some authors, is that the eastward displacement of Brazil was purposeful, aiming to
fool the Spanish as where the dividing line of Tordesillas should pass. 17 None of these
reasons resists a detailed geometrical analysis.
N
N
SP
SP*
C
C
Cm
PD
Cm
PD
SP*
SP
Figure 5 – The effect on the set point of an eastward magnetic declination for a course in the
northeast and southeast quadrants (right). The graphs illustrate how the ship’s position SP was
marked on an ordinary chart using the compass (magnetic) course Cm, sailed from the point of
departure PD, and the observed latitude . N is the chart’s North, from where all courses are
reckoned. If magnetic declination were corrected for, the resulting set point SP* would lie to
the east or to the west of SP, depending on the quadrant of the course. Notice (right) how the
eastward displacement of SP (here representing the Cape of Good Hope), relative to its true
position SP*, is explained by the eastward magnetic declination in the South Atlantic, as
correctly perceived by João de Castro.
The main cause for the apparent displacement of Brazil and Africa on the charts of the
sixteenth century was correctly identified in 1538 by the nobleman and navigator João
de Castro (1500-1548), during a voyage from Lisbon to India. During this voyage,
Castro made systematic observations of magnetic declination with the purpose of
confirming (or disproving) the conjecture that its value varied linearly with longitude.
18 After describing its spatial variation along the route from Lisbon to Brazil and to the
16
A detailed cartometric analysis of the Cantino planisphere is in Gaspar, From the Portolan Chart of the
Mediterranean (note 6), 141-174; idem, ‘Blunders, Errors and Entanglements (note 11)’, 186.
17
Indeed, some Portuguese charts of the sixteenth century deliberately manipulated the coastline of Brazil
with diplomatic purposes. However, that is not the case of the Cantino planisphere. See Avelino Teixeira
da Mota, Reflexos do Tratado de Tordesilhas na Cartografia Náutica do Século XVI (Coimbra: Centro de
Estudos de Cartografia Antiga, Junta de Investigações do Ultramar, Separata LXXX, 1973), where the
idea that the positions of Newfoundland and Brazil on the Cantino planisphere were manipulated is
refuted.
18
During his voyage from Lisbon to India, João de Castro collected 37 daily values of magnetic
declination, which were determined along the route using the instrument designed by the mathematician
Pedro Nunes. These values were registered in Castro’s “Roteiro de Lisboa a Goa” (1538), together with
other navigational data. See João de Castro, ‘Roteiro de Lisboa a Goa’, in Luís de Albuquerque (eds.),
10
Cape of Good Hope (where the compasses pointed to the east of the geographic
north), and from the Cape of Good Hope to India (where the compasses pointed to the
west), Castro concluded that the eastward displacement of both Brazil and Africa on
the charts was caused by using uncorrected compass courses. Castro illustrated his
interpretation with examples, where the effect of magnetic declination on the position
of the ship, as determined using the set point method, is correctly explained (Fig. 5).19
A critical point to stress is that latitude charts were constructed, like portolan charts,
by transferring directly to the plane the uncorrected compass courses measured at
sea. That was done, not because pilots were unaware of the phenomenon of magnetic
declination but because that was the simplest thing to do. Correcting all charts for
magnetic declination would imply time-consuming and costly surveys, in order not
only to measure the true geographical directions between places but also to
determine the spatial distribution of magnetic declination. 20 Such an endeavor only
became truly necessary when the longitude problem was solved in the second half of
the eighteenth century, making it possible for the old latitude chart to be abandoned
and the Mercator projection to be fully adopted by mariners.
Coming back to the initial question about the accuracy of the early modern
planispheres, it seems clear that the modern criterion of absolute positional accuracy,
based on the latitudes and longitudes of the places, should not be applied here.
Knowing that charts were made for supporting navigation, an appropriate way of
assessing their navigational accuracy would be to evaluate to what extent they were
fit to that purpose, at the time they were used. More specifically, how accurate were
the latitudes, the magnetic courses and the distances between places? 21
Incidentally that was the mistake, not only of present-day historians, but also of the
cosmographers of the sixteenth century, as described next.
Obras Completas de D. João de Castro (Coimbra: Academia Internacional de Cultura Portuguesa, 1962),
Vol. I, 113-296.
19
Idem, 198-207. Castro’s text contains a remarkably clear discussion about the effect of the uncorrected
magnetic declination on the geometry of charts. A lengthy study of his ideas and contribution for the early
modern science, emphasizing his experimental work, is Reijer Hooykaas, ‘Science in manueline style.
The historical context of D. João de Castro’s works’, idem, Vol. IV, 231-426.
20
The earliest known cartographic representation of isogons (that is, lines of constant magnetic
declination) is in the unsigned chart of c.1585, attributed to the Portuguese cartographer Luís Teixeira,
representing the western margin of the Pacific Ocean, which was most likely part of a now lost
planisphere composed of four sheets of parchment. Our interpretation is those lines might have been used
during the late sixteenth century as an aid to navigation, rather than to correct magnetic courses. See
Joaquim Alves Gaspar and Henrique Leitão, ‘Luís Teixeira, c.1585: The Earliest Known Chart with
Isogonic Lines’, Imago Mundi, 70, 1, 2018, 221-238.
21
Two unrelated factors make such an assessment not easy to accomplish: the fact that only the specific
(often unknown) routes used to construct a certain chart were supposed to be represented accurately; and
the difficulty in knowing with sufficient accuracy the spatial distribution of magnetic declination in
ancient times. An assessment of the latitudes, courses and distances of five Portuguese charts of the 15th
and 16th century is in Gaspar, From the Portolan Chart of the Mediterranean (note 6), 108-119, 151-166.
11
The Lopo Homem incident
Around 1560, the Portuguese cartographer Lopo Homem, in a note addressed to the
king, harshly complained about the new official master chart that had been established
by the Cosmographer Major, the mathematician Pedro Nunes (1502-1578). 22
According to Lopo Homem, such chart had been prepared using the eclipses of the sun
and moon −to determine longitude− with the purpose of showing that from Lisbon to
India, and to the Moluccas, the real distances were shorter than the ones shown on
the charts. According to the cartographer, however, the charts made according to the
new pattern were “so wildly distant from all truth and navigational science that many
ships were lost and pilots were forced to buy their charts in Castile”. 23
Figure 6 – Excerpt of a nautical planisphere by Lopo Homem (1554). The black outline
represents the coastline of a modern plate carré map. Although the longitudinal stretching of
Africa was reduced, when compared with the Cantino planisphere, the eastward displacement
was not completely eliminated. This may be explained by the error, of about 10 degrees, made
in the determination of the longitude of Diu. Istituto e Museo di Storia della Scienza, Florence,
IMSS, n. 0946.
By the time this note was written, pilots were perfectly aware, as João de Castro was in
1538, that the longitudinal distance between Lisbon and India, as shown on
contemporary charts, was exaggerated. In his Tratado em defensam da carta de
marear (Treatise in Defense of the Nautical Chart), of 1537, the same Pedro Nunes had
already complained about the fact, which he attributed to the incompetence of the
pilots, “who represented as straight which had been sailed with so many detours”. 24
22
The royal master chart, or Padrão del Rey, was the official cartographic standard from which all charts
used by the pilots at sea should be based on. This chart was supposed to be kept up to date, concerning
the new geographical discoveries and surveys.
23
The undated letter of Lopo Homem is kept in the Bibliothèque nationale de France (ms. Coll. des Cinq
Cents de Colbert 298, fols 6 r-8 r) and was transcribed by Luís de Matos, Les Portugais en France au
XVIe siècle (Coimbra: Universidade de Coimbra, 1952), 318-322.
24
Pedro Nunes, Obras, Vol. I: Tratado da Sphera; Astronomici Introductorii de Spaere Epitome (Lisboa:
Fundação Calouste Gulbenkian, 2002 [First edition: Lisboa, 1537], 120-184 at 129.
12
Although Pedro Nunes was right about the exaggerated longitudinal distance between
Lisbon and India, as represented on the charts, he missed the point when he blamed
pilots for the apparent mistake. In 1547, he was appointed cosmographer major,
becoming responsible for the royal master chart. Then he finally had the opportunity
to correct what he considered to be a major imperfection, by ordering the longitude of
Diu (in India) to be determined by astronomical methods. The result was a new
cartographic standard in which the longitudinal distance between Lisbon and India was
partially reduced, as in the planisphere of the same Lopo Homem (1554) shown in Fig.
6.25
The examination of the Portuguese charts from 1554 onward makes clear that Nunes’
new master chart was not adopted by all. We have compared the longitudinal width of
Africa at the latitude of Cape Guardafui (in the easternmost tip of Africa) in a series of
nautical planispheres of the sixteenth and seventeenth centuries, and the results are
revealing. While the apparent longitude of Cape Guardafui remained more or less
constant from 1502 to 1628, the width of the Mediterranean increased during the
same period, making the longitudinal distance between its eastern margin and Cape
Guardafui decrease markedly. This means that the Mediterranean was artificially
stretched in the east-west direction, with the likely purpose of cosmetically reducing
the size of the Red Sea and, with it, of the Isthmus of Suez. 26 This fact suggests that
the new pattern introduced by Nunes failed after all, most probably owing to the
opposition of pilots and cartographers.
Coming back to the comments of Lopo Homem, why did the cartographer consider
Nunes’ standard “wildly distant from all truth and navigational science?” What kind of
errors would make the new master chart so unfit for navigation? Clearly, he was not
referring to the distances measured on them – particularly the one from Lisbon to
India – which were usually not to be trusted.27 The problem was, most likely, the
orientation of the African coastline, which no longer agreed with the uncorrected
compass courses. Although the allegations of Homem – that many ships using the new
charts were lost in their way to India and that pilots were forced to buy their charts in
Castile – are obvious exaggerations, the apparent failure of Nunes in imposing the new
standard demonstrates that the arguments of the cartographer had considerable
weight.
25
The astronomical observations ordered by Pedro Nunes in Diu were reported by the humanist and
historian Jerónimo Osório, in De Rebvs Emmanuelis Regis Lusitaniae (Olysippone: 1571), 424.
26
See Gaspar, ‘From the Portolan Chart to the Latitude Chart : The Silent Cartographic Revolution’,
Cartes et Géomatique, Bulletin du Comité Français de Cartographie, 216 (2013), 67-77.
27
It was shown before that it was not possible to conciliate, on a latitude chart, both the course and the
distance between places owing to the effect of magnetic declination A critical point to note is that the
courses between places was a much a more important piece of information for the pilots than distances
because it affected on a much larger degree the safety and effectiveness of navigation. For a more detailed
explanation see Gaspar, From the Portolan Chart of the Mediterranean (note 6), 13-21.
13
Who was then right and who was wrong in this dispute? Was it Pedro Nunes, who
used astronomical methods to determine the longitude of India and enforced the
result onto the official cartography? Or was it Lopo Homem, who insisted in keeping
the traditional representation and supported his position with practical arguments
concerning the safety of navigation? Because the requirement of Pedro Nunes (correct
longitudinal distance) could not be reconciled with the one of Lopo Homem (correct
compass courses), the answer to the question depends on the use of the charts. As
cosmographer major, Pedro Nunes was responsible for keeping the official master
chart and approving the nautical charts used in navigation. But charts were also made
for purposes not related to navigation, for example, for diplomatic purposes. This line
of reasoning suggests that Nunes may have been driven by the intention of applying
the Ptolemaic prescription of map construction, that is, the use of latitudes and
longitudes, to nautical charts. In other words, he may have intended to enforce the
use of a positionally accurate plate carré into nautical cartography.
A better explanation is suggested by an attentive reading of Nunes’ Treatise in Defense
of the Nautical Chart of 1537, where a detailed analysis of the geometry of the
contemporaneous charts is presented and the necessity of representing correctly the
distances in the east-west direction is emphasized. 28 However, nowhere in the text is
the influence of magnetic declination discussed, although Nunes was perfectly aware
of the phenomenon. This suggests that either he was convinced that courses were
corrected before being transferred to the charts, or alternatively, he did not consider
the effect to be geometrically relevant. Knowing that the eastward displacement and
stretching of Africa in the charts was mostly caused by the effect of magnetic
declination, this may explain why Nunes insisted that the distance to India should be
corrected: not to depict the peninsula according to its true longitude but to conserve
longitudinal distances. Once again, two distinct concepts of accuracy were at stake
here: the modern concept of absolute accuracy, applicable to the geographical
depictions of the Earth; and the concept of navigational accuracy, applicable to
nautical charts.
This episode also demonstrates that even when close contact between scholars
(cosmographers) and artisans (pilots and cartographers) was enforced, the full
understanding by the former of artisanal practices could be difficult to achieve.
Charts with multiple latitude scales
Another example of the clash between the geographical and the nautical paradigms in
early modern cartography is the dispute in the middle of the sixteenth century around
the making and use of charts with multiple latitude scales. This ingenious expedient,
28
Pedro Nunes, Obras, Vol. I: Tratado da Sphera, 129-133. An article containing a detailed discussion of
of the Tratado em Defesam da Carta de Marear (1537), and its reflex in Europe, is being prepared by the
present authors and will be shortly submitted for publication.
14
first used by Pedro Reinel ca. 1504 to represent Newfoundland, was intended to
conciliate the three elements of navigational information in the location of a given
region – the latitude, the compass course and the distance – in the presence of
magnetic declination. The idea was to locate that region on the chart according to a
compass course and a distance, measured from some origin (that is, using the method
of the point of fantasy), and add a secondary latitude scale only applicable to that
region. In the case of Pedro Reinel’s chart, the origin of the track used to put
Newfoundland on the chart was the Azores archipelago. The secondary latitude scale is
tilted as to indicate the direction of geographical north in the northwestern Atlantic
(Fig. 7).29
Figure 7 – Excerpt of a chart by Pedro Reinel (ca. 1504) with two different latitude scales. The
tilted scale near Newfoundland, indicating the direction of geographical north, was only
applicable to the region. Bayerish Staatsbibliothek, Munique (Cod Ican 132).
In 1539 Pedro de Medina (1493-1567), an influential humanist and author of the
celebrated Arte de Navegar (1545), arrived in Seville with a royal authorization for
making charts in the Casa de la Contratación. This authorization triggered a strong
conflict between the Piloto Mayor Sebastian Caboto (ca. 1474-1557), who was
responsible for the official master chart (the Padrón Real), and the cosmographer
Diego Gutiérrez (ca. 1485-1554), who held the effective monopoly of chart production
and sale. The problem was that the charts with multiple latitude scales produced by
Gutiérrez, with the approval of Caboto and the support of pilots, did not comply with
the Padrón Real. The dispute lasted for several years and involved other members of
29
The reason for this expedient was the fact that it was geometrically impossible, as referred before, to
conciliate the three elements of information, in the presence of magnetic declination. For a more detailed
explanation, see Gaspar, From the Portolan Chart of the Mediterranean (note 6), 118-119.
15
the Casa, including the cosmographer Alonso de Chaves (ca. 1492-1587). It was finally
resolved by a decree of Prince Philip (the son of Charles I of Spain), in 1545, who
ordered that all charts should comply with the royal master chart. In other words, the
use of multiple latitudes scales was banned from official cartography.
This well-known incident has been discussed by several authors, who focused on the
aspects related to the “scientific truth” and how it was assessed in the sixteenth
century. 30 But once again, what was really at stake here were two distinct and
irreconcilable concepts of cartographic accuracy: the geographical (absolute) and the
navigational. While most Spanish cosmographers considered that nautical charts had
to comply with the Ptolemaic geographical standards (that is, accurate latitudes and
longitudes), pilots were unable to reconcile them with the practice of navigation. The
problem was technically complex, and to consider that the reason rested on only one
of the sides of the argument would be an oversimplification. Charts with a secondary
latitude scale for Newfoundland continued to be produced, and it is curious to realize
that the only surviving Spanish chart of the kind was made by the same Diego
Gutiérrez in 1550, well after the royal order forbidding their production was issued
(Fig. 9). Also remarkable is the fact that the magnetic course and the distance
measured from the Azores to Cape Race (the southeastern tip of present-day Avalon
Peninsula in Newfoundland) in those charts are very accurate, a fact that eloquently
explains the support of the pilots.31
An alternative solution that would satisfy the cosmographers’ requirement would have
been to correct magnetic declination in all courses measured with the marine
compass. This would make positions determined with the point of fantasy and the set
point to be compatible, implying that the main latitude scale of the chart could be
used in the whole charted area. But Spanish pilots were not willing, or prepared, to
determine magnetic declination on board, and the process would imply costly changes
in navigation and nautical cartography. Not only pilots would have to start using
corrected compass courses, but also new surveys and new charts would have to be
Ursula Lamb, ‘Science by Litigation: A Cosmographic Feud’, Terrae Incognitae, 1, 1 (1969): 40-57.
See also Alison Sandman, ‘Mirroring the world: sea charts, navigation, and territorial claims in sixteenth
century Spain’, in Merchants and Marvels: Commerce, Science, and Art in Early Modern Europe, ed.
Pamela Smith and Paula Findlen (New York, Routledge, 2001), 83–108’; Antonio Sánchez, La espada, la
cruz y el Padrón, Madrid: CSIC (2013), 229-261; María Portuondo, Secret Science. Spanish
Cosmography and the New World (Chicago and London: The University of Chicago Press, 2009).
31
Five charts of the sixteenth century were analyzed: Pedro Reinel (ca. 1504); Lopo Homem (ca. 1550);
Diego Gutiérrez (1550), an anonymous Portuguese chart of the Atlantic kept in the Bibliothèque nationale
de France (ca.1560); and Sebastião Lopes (1583). In all of them the error in the magnetic course between
the Island of Terceira (Azores) and Cape Race (Newfoundland) is less than 2 degrees, the corresponding
error in the distance is less than 10.5 Castilian leagues (0.6 latitudinal degrees), and the latitude error of
Cape Race, as measured in the secondary scale of latitudes, is less than one degree. To estimate the value
of magnetic declination in the region, during the sixteenth and seventeenth century, the geomagnetical
model CALCS7K of Korte and Constable was used. See M. Korte & Catherine Constable, ‘Continuous
geomagnetic field models for the past 7 millenia: 2. CALS7K’. Geochemistry, Geophysics, Geosystems,
6, 1, 2005.
30
16
made to comply with the new standards. Browsing through the Iberian charts of the
sixteenth century quickly makes clear that such corrections were never made.
Figure 8 – Excerpt of the only surviving chart by Diego Gutiérrez (1550), with three different
latitude scales. Two of them are shown in the figure: one on the right, for Newfoundland,
which appears rotated; and the other on the left, for North America. The third latitude scale,
not shown in the figure was intended to be used in the western Atlantic and Caribbean Sea.
Some decades later, near the end of the sixteenth century, the instrument maker and
mathematician Andrés García de Céspedes (1560-1611) was appointed cosmographer
of the Casa de la Contratación with the mission of reforming navigation and
cartography. Contradicting the position of his predecessors, Céspedes clearly aligned
with the pilot’s points of view by asserting that because the primary purpose of charts
was to support navigation, they should be made in accordance to the navigational
practices and the pilots’ wishes.32 As a university-educated scholar, he knew perfectly
well that the ad-hoc projection used in nautical cartography, where uncorrected
compass courses, distances and latitudes were transferred directly to the plane of the
chart, was not geometrically consistent. However, no better alternative in which
rhumb lines were represented by straight segments could be found at the time.
Unfortunately, neither his improved Padrón Real nor any charts based on it survived to
the present day. 33
For a more detailed discussion on the solutions proposed by Céspedes see Alison Sandman, ‘An
Apologia for the Pilot’s Charts: Politics, Projections and Pilots’ Reports en Early Modern Spain’, Imago
Mundi, 56, 1 (2004): 7-22.
33
García de Céspedes’s Hydrographia, bounded to his Regimiento de Navegación (Madrid, 1606), is a
lengthy and rich treatise where the author explains his ideas about how to improve navigation and
nautical cartography in Spain. Although considerable emphasis is put on the aspects related to the
location of the Moluccas, whose political relevance was still felt at the time, the author makes a serious
32
17
The Mercator projection
The clash between Ptolemy’s cartographic prescriptions and the nautical chart reached
a dramatic climax with the work of Gerard Mercator (1512-1594). In 1569, the Flemish
mathematician and cartographer presented his map of the world with the revealing
title ‘’New and improved description of the Earth properly adjusted for use in
navigation’’. 34 In this map, meridians and parallels form a regular mesh of rectangles
where the spacing between parallels increases with latitude in such a way that rhumblines are represented by straight segments making true angles with the meridians. In
theory, this solution would enormously facilitate the planning and execution of
navigation because all courses between places – and not only a few, as in traditional
latitude charts – were supposed to be accurately represented. Moreover, Mercator’s
map was based on the latitudes and longitudes of the places, exactly as geographical
maps. Apparently, this fact alone would resolve the long-standing incompatibility
between Ptolemy’s paradigm and nautical charts.35
But was Mercator’s world map truly adjusted to navigation at the time it was
proposed? As a matter of fact, it was not, owing to its intrinsic incompatibility with the
navigational methods of the time, still based on compass courses. Only after the
longitude problem was solved and the distribution of magnetic declination was known
could the old model be abandoned for good and new charts constructed using the
Mercator projection. A relevant question that could be posed is whether the new
world map of Mercator – or rather, some larger scale nautical charts based on it –
could still be used for navigation, assuming that pilots were prepared to use corrected
courses. We do know that the graticule of meridians and parallels calculated by
Mercator, to serve as a reference for latitudes and longitudes, was sufficiently
effort to understand the pilot’s needs and to contribute to improving the information used to construct the
charts. The contributions of two authors to the subject are profusely cited and praised in the treatise:
Pedro Nunes’s Tratado en Defensam da Carta de Marear (1537), where the geometry of the
contemporary charts is discussed; and João de Castro’s Roteiro de Lisboa a Diu (1538), where the reason
for the apparent distortion of Africa on the charts is explained. For a technical discussion of Cespedes’s
work, see Víctor Navarro Brotóns, ‘Astronomia y cosmografia entre 1561 y 1625. Aspectos de la
actividad de los matemáticos y cosmógrafos Españoles y Portugueses’. Cromos, 3: 2 (2000), 362-368.
For a discussion of Céspedes’s contribution to the improvement of navigation and cartography, see
Alison Sandman, ‘An Apologia for the Pilot’s Charts: Politics, Projections and Pilot’s Reports in Early
Modern Spain. Imago Mundi, 56: 1 (2004), 7-22.
34
The original title reads: Nova et aucta orbis terrae description ad usum navigatium emendate
accomodata.
35
The purpose of Mercator appears to have been more ambitious than the use of his novel projection in
navigation. In the legend Inspectori Salutem, three objectives are enumerated: ‘Firstly, to spread on a
plane the surface of the sphere in such a way that the positions of places shall correspond on all sides with
each other both in so far as true direction and distance are concerned and as concerns correct longitudes
and latitudes […]; [secondly] to represent the positions and the dimensions of the lands, as well as the
distances of places, as much in conformity with very truth as it is possible so to do […]; [thirdly] to show
which are the parts of the universe which were known to the ancients and to what extent they knew them’.
For a transcription of the Latin text of Mercator’s legends and its translation into English see ‘Text and
translation of the legends of the original charts of the world’, Hydrographic Review, 9:2 (1932): 7-45 at
11.13. The Walther Ghim biography is in Gerardus Mercator, Atlas sive Cosmographicae Meditationes
de Frabica Mundi et Fabricati (Duisburg: 1595).
18
accurate for navigational purposes.36 The question is: how accurate are the
geographical coordinates of the places, as well as the directions between them? Two
types of errors affect the latitudes and longitudes of Mercator’s world map: those that
originated in an insufficient knowledge of the geography of the regions, such as in the
depiction of South America; and those associated with the distortions of the
cartographic sources used in the compilation. Only the second type of error is relevant
for the present discussion. 37
Figure 9 – Grid of meridians and parallels implicit in the Mercator world map of 1569 (left)
compared with its own geographical graticule (right). Notice the convergence of meridians and
parallels in the northern hemisphere and the irregularities of the supposedly straight and
parallel lines. Reproduced from Gaspar, ‘Revisiting the Mercator World Map of 1569’ (2016), p.
8.
Figure 9 illustrates the grid of meridians and parallels implicit in the Mercator world
map of 1569 (left), which was estimated using a sample of places with known latitudes
and longitudes, side by side with its own geographical graticule (right). If the
coordinates of the places were all exact the two grids would be coincident, which is
not the case. Several errors can be spotted by visual inspection of the map’s implicit
geographical grid (left): the convergence of meridians in the northern hemisphere,
which is not supposed to occur in a Mercator projection; errors in the latitudes,
especially in the northern Atlantic and the eastern Mediterranean, which are reflected
in the irregularity of the parallels; and the unequal spacing of meridians in various
regions, most especially in northern Africa and the Mediterranean. This irregularity is
better perceived by comparing the coastlines of Mercator’s world map with those in a
36
The method used by Mercator to calculate his projection was the object of two recent articles where it
was demonstrated, historically and numerically, that he has most likely used a table of rhumbs to
calculate the spacing of the parallels. See Joaquim Alves Gaspar & Henrique Leitão, ‘Squaring the Circle:
How Mercator Constructed His Projection in 1569’. Imago Mundi, 66, 1 (2013), 1-24; Henrique Leitão &
Joaquim Alves Gaspar, ‘Globes, Rhumb tables, and the Pre-History of the Mercator Projection’. Imago
Mundi, 66, 2 (2014), 180-195.
37
A detailed assessment of the accuracy of Mercator’s world map is in Joaquim Alves Gaspar,
‘Revisiting the Mercator World Map of 1569: an Assessment of Navigational Accuracy’, The Journal of
Navigation, 69 (2016), 1183-1196.
19
modern Mercator representation, as done before in this article with the Cantino
planisphere. This exercise makes clear that the eastward displacement and stretching
of Africa and Brazil, affecting the Cantino planisphere, are still present in Mercator’s
world map (Fig. 10).
Figure 10 – Excerpt of Mercator’s world map of 1569, depicting part of the Atlantic and Indian
Oceans, compared with the coastlines of a modern Mercator’s map (white lines). Notice the
longitudinal displacement and stretching of Africa and Brazil, identical to the one in the
traditional cartography of the sixteenth century, and the artificial enlargement of the
Mediterranean, shared by other planispheres of the sixteenth century.
How should the persistence of these distortions be interpreted, knowing that the
novel projection was intended to represent latitudes, longitudes and rhumb line
directions accurately? The discrepancy is easily explained by Mercator’s erroneous
assumption that both the latitudes and the longitudes were approximately correct in
contemporary cartography. This should not come as a surprise given the fact that
Mercator shared with his contemporaries the false idea that the meridians and
parallels implicit in the contemporary charts formed a square grid, a conviction that is
clearly stated in one of the Latin legends of the map: 38
“On the navigator’s charts the degrees of longitude, as the various parallels are
crossed successively towards the pole, become gradually larger with reference to
their length on the sphere, for they are throughout equal to the degrees on the
equator, whereas the degrees of latitude do not increase at all”.
See Gerard Mercator (1932), ‘Text and translations of the legends of the original chart of the world by
Gerhard Mercator issued in 1569’, Hydrographic Review, 9, 2, 7-45. For a discussion about the
misinterpretation of the geometry of the early modern nautical charts, see Gaspar, Blunders, Errors and
Entanglements (note 11), 186; idem, From the Portolan Chart of the Mediterranean (note 6), 33-34;
idem, ‘The Myth of the Square Chart’, e-Perimetron, 2, 2: 66-79.
38
20
Under this assumption, all that was needed to transfer the geographical information
from the available charts to the projection was to take note of the latitudes of the
places (as measured using the latitude scale of the charts), as well as of the apparent
longitudes, whose scale was supposed to vary linearly in the east-west direction.
Coming back to the question posed above: could Mercator’s world map still be used
for navigation at the time it was proposed, assuming that the pilots were prepared to
use corrected compass courses? The answer to this hypothetical question is no,
because the latitudes and longitudes of the places on the map, and hence the courses
connecting them, were not accurate enough for navigational purposes. To construct an
accurate Mercator’s representation, new global surveys were – once again – required,
this time based on the latitudes and longitudes of the places. 39
Again – as in the case of Nunes – not being fully aware of the artisanal context in which
nautical charts were constructed and used led Mercator to misinterpret the
information imported from them.
Final remarks
Pre-Mercator nautical charts are complex objects whose inner geometrical properties
are only now starting to be understood. Rather than attempts at a faithful
geographical depiction of the world, they were tools aimed to accomplish a specific
purpose – to support marine navigation – a fact that is often mentioned in
contemporary sources. Thus, they carry inside their internal geometry the imprint of
the activities of pilots on board, that is, a “signature” of their artisanal origin and of the
techniques used to navigate. Any hope for a better understanding of these
cartographic artefacts requires a full acquaintance with the contemporaneous
navigational practice. More specifically, any attempt at evaluating the accuracy of early
modern charts using modern concepts, or comparing them to modern representations,
should be made bearing in mind their purpose and the way they were used on board.
Ever since the Middle Ages, geographical maps and nautical charts have co-existed as
two distinct representations of the known world, but it was only in the first decades of
the sixteenth century that the confrontation between the two models became a
critical issue. Following its first printed edition in 1478 and the numerous other
editions in the next decades, Ptolemy’s Geography was widely disseminated and
studied in Europe. At the same period, because of the oceanic voyages of exploration,
39
Attempts at using the Mercator chart for navigation are known to have been made from the late
sixteenth century onward, by the Dutch and English. On these experiments, see Sarah Tyacke,
‘Chartmaking in England and Its Context, 1500-1600’, in David Woodward (ed.), The History of
Navigation, Volume III – Cartography in the European Renaissance (Chicago & London: The University
of Chicago Press, 2007), 1722-1753, at 1743-1745. For a general discussion on the resistance of pilots
against the adoption of the Mercator projection in navigation, see Mark Monmonier, Rhumb Lines and
Map Wars: A Social History of the Mercator Projection. Chicago & London (2004): The University of
Chicago Press.
21
nautical cartography was undergoing enormous changes, and became the single most
important source of geographical information on a planetary scale. Inevitably, scholars
compared and tried to harmonize Ptolemy’s data and cartographic prescriptions with
the image of the orbis terrarum conveyed by nautical cartography. This effort proved
to be extremely complex and a source of constant tension between those working
with maps and those working with charts. The clash brought into play many different
factors, such as authority, imperial ambitions, social and professional environments,
etc. But, at its core, lay a crucial and yet extremely subtle technical issue that was not
grasped by those protagonists: the fact that the two representations were impossible
to reconcile. The attempt of Gerard Mercator to construct a chart ad usum
navigantium, based on the latitudes and longitudes of the places, was a serious effort
to reconcile the two cartographic paradigms. We have seen, however, how his
achievement was much ahead of the time it was proposed.
In its broadest sense, the argument of this paper is about knowledge production in
early modern Europe. By focusing on cartography, it adds to the already abundant
literature on artisanal practices and their role in the shaping of science during that
period. More specifically, it shows how certain objects - nautical charts - were deeply
influenced by the practices of the artisans who used them, but, simultaneously, how
such influence was almost invisible to contemporary scholars and present-day
historians. Our argument puts in evidence the deep level of complexity in the
interactions between scholars and craft practitioners. Even when contact and
cooperation between these two groups was intense and stable – in fact, even when
this contact was enforced by legislation – and when the artifacts (instruments, charts,
etc.) were the same, deep incomprehension might arise.
One final consequence of our study is bringing to light the existence of a major debate
that took place in Europe around the clash between Ptolemy’s geographical paradigm
and nautical charts. It is no surprise that this debate has been almost unnoticed by
historians and that its nature was not correctly understood. In a general sense, this
may have to do with the difficulty in knowing, and fully integrating in the historical
discourse, the practices of artisans. But in a deeper sense, the reason is mostly
technical. Only today, with the use of digital cartometric methods of geometrical
analysis, can we fully understand the true nature of nautical charts, and therefore, the
technical reasons behind those misunderstandings, and conclude: first, that
historically, nautical charts can only be understood in the context of the specific
navigational methods they were intended to support; and second, that a nautical chart
should not be considered as a true geographical map but as an instrument for
navigation.
Acknowledgements: this project received funding from the European Research Council (ERC)
under the European Union’s Horizon 2020 research and innovation programme (grant
agreement 714033-Medea-Chart / ERC-2016-STG).
22