[ISHMap-List] 'Cartographic innovations by the early portolan chartmakers'

Roel Nicolai rnicolai at xs4all.nl
Tue Jan 10 22:38:01 CET 2017



Thank you for your comments in reply to my post. I think this is a useful discussion and although we will not be able to come to an agreement on the origin of the charts, I would like to bring up a few more things (with apologies for being so verbose). I will start with your last remarks, which contain the most important comment regarding my work, and after that I will add replies on some specific issues you mentioned.


1.         A key comment you make is that my analysis method is not understood well enough and therefore not endorsed by many other researchers. That is only partly true, viz. where it concerns the mathematical aspects of that analysis and that was indeed very evident during the workshop in Lisbon. I discussed it at the time with Evangelos Livieratos and indicated I intended to write a few articles to explain cartometric analysis techniques and my results in ‘laymen’s terms’ in e-Perimetron. I have actually begun with the first article, at least with the analysis work. 


I used the term ‘partly true’ in the previous paragraph and I will explain in this paragraph what I consider to be the ‘partly-not-true’-aspect of the criticism that my methods are poorly understood. A significant part of my thesis and my book do not deal at all with cartometric analysis, although the latter is a distinguishing feature of my work.  Only two chapters of twelve plus some scattered sections in others contain fairly heavy mathematics; I cannot blame anyone without the right background for not understanding the details of that. However, the rest of my work can be understood easily enough by people without a mathematical background. It contains, for example, an analysis of medieval navigation methods and tools, of medieval ships and shipping, and of meteorological and oceanographic aspects. These are all relevant for the discussion on the origin of portolan charts and yet I have hardly seen any discussion on them.


I believe there is a far more fundamental layer of disagreement underlying, i.e. below the arguments that I just described. In citing my position you only mention my conclusions; not my arguments. I object to that because it makes it appear as if my position for a non-medieval origin of portolan charts is a ‘claim’ stated up-front. That is not the case. However, this does expose a fundamental difference in the approach to my work and to the work of many others, including yours. Most other research work considers the medieval origin as a departure point and seeks to corroborate that by finding supporting evidence or arguments. From the perspective of the scientific method that approach is less than desirable. As the 20th-century science philosopher Karl Popper pointed out, a hypothesis or theory can never be verified fully by focusing exclusively on supporting evidence. It easily leads to ‘confirmation bias’ (‘only seeing what one wants to see’). Popper therefore introduced a method aimed at trying to disprove a hypothesis, rather than to prove it, because a single piece of disproving evidence is much more powerful than multiple pieces of supporting evidence. This is indeed the approach I followed:  I approached the problem from the perspective of hypothesis testing and described the hypothesis of a medieval origin as having four aspects, which I test separately. That led to a conclusion that the medieval origin needs to be rejected, and not merely to a claim of a non-medieval origin, even less so to new hypothesis. Confirmation bias is pretty much everywhere in science. Sometimes scientists fall in love so much with a hypothesis or theory that they try to ‘protect’ it against contrary evidence, to prevent what Thomas Huxley (father of Aldous) called: “the great tragedy of science - the slaying of a beautiful hypothesis by an ugly fact.”


So, to sum up, by ignoring the contrary facts (or arguments) that my work has produced, you are opting for a lopsided discussion, which I am not prepared to conduct. The excuse that you do not understand my methods is too convenient, as not all my arguments depend on those methods. Moreover, you probably don’t understand either how your smartphone derives its GPS-based position, but I doubt that will lead you to mistrust that position. Also the results of much of my cartometric analysis are easy enough to understand even if you don’t understand the method I used to derive them. You might place word of caution with those results when you cite them, but, as I wrote earlier, to ignore them completely is unacceptable, that is, if you aim to practice science.


My methods have been checked and vetted by 2 mathematicians, an astronomer, 2 geodesists, a statistician, 2 map historians and a cartographer in the process running up to the award of my PhD. I hardly think they require the additional approval of mathematically-inclined map historians before they are deemed acceptable. But I do agree additional explanations in more or less ‘laymen’s terms’ would be very helpful. 

I am also convinced that mathematically-inclined and humanities-inclined people should not be treated as two separate groups. That would bring the risk of two separate discussions, with opposite conclusions. 

Now the specific comments.


2.         You have conceded the point, but still leave me puzzled on how you arrived at the idea that portolan charts show the length of the Mediterranean as too short.


3.         Regarding the Atlantic coastlines; these innovations are unsuitable for distinguishing between the hypothesis of a medieval and the one of a non-medieval origin, because both can provide an explanation for the phenomenon. Furthermore, by only providing your own argument in support of a medieval origin you make the discussion lopsided, which is reinforced by the tendentious formulation that it “… evidently contradicts it [Nicolai’s ‘claim’]”.  Your own arguments may be more plausible to you, but don’t forget that mine are more plausible to me. This subject is perhaps better left to a separate discussion, which, I am sure, both of us will be keen to have.


4.         Regarding my statement on “no scale-true maps before portolan charts”, you are right to point out that Ptolemy’s maps were designed to scale. I was talking about medieval maps, but I should have made that clear.


5.         Regarding the wind rose; I distinguish the eight-wind system (which indeed predates portolan charts) from the diagram on portolan charts, whereas you appear not to. The diagram is rather more than the eight-wind system in that it uses an ingenious way to generate 32 directions out of 16 points on a circle. So, while portolan charts cannot be said to have invented the eight-wind system, the wind rose diagram is entirely new.  At least, I have not come across any map with that same diagram.  


5.         Regarding my comment on the wind rose on the Roselli 1456 chart; you write that you didn’t say that Roselli’s innovation created 64 directions, but I cite from your note (11th entry in your bullet-pointed list in the chapter ‘The Innovations’):

“there is a systematic network of 32 wind or compass directions, based around one or two large hidden circles (for the doubling of that number see below under 1456 Petrus Roselli).”

Your mention 32 directions; doubling that number would yield 64 in my books. 

And, in the section ‘Can a ‘portolan chart’ be defined or merely recognised from its characteristics?’, third bullet point:

“… has a systematic 32- or 64-line network of compass directions (colour-coded in black, red and green)…”.

I think there is only one way to read your text and that is what I described in my previous post. If you feel it the wind rose only shows 32 unique directions, I suggest you modify the above descriptions accordingly.


Best regards,




From: Tony Campbell [mailto:tonycampbellockendon at gmail.com] 
Sent: Sunday, January 08, 2017 3:36 PM
To: ISHMap-List; Roel Nicolai
Subject: re: 'Cartographic innovations by the early portolan chartmakers'



I am grateful to you for a number of useful comments on my 'Cartographic innovations by the early portolan chartmakers' (http://www.maphistory.info/PortolanChartInnovations.html).  [For the full text of Roel's post see: http://lazarus.elte.hu/pipermail/ishm/2017-January/000536.html.]

This is precisely what I was hoping for, and it would be excellent if others contributed as well.

As a result I have made some additions and corrections to the online document:

1. I have corrected, at the beginning of 'The Innovations' section, the 'passed down as partial charts' comment

2. I have suppressed the comment about the length of the Mediterranean, which, as you say, Luis A. Robles Macias had already pointed out.

However, we do have continuing areas of disagreement and, without repeating what has already been said, I include a few brief comments here.

While I fully take your point about the need to distinguish hydrographic outlines from toponymy in a consideration of the charts' origins - and there are indications that they may have had separate developments anyway - I do not see how any kind of pre-medieval origin can be proposed for the Atlantic coastlines.  So, even if we agree that the toponymy is medieval, the Atlantic coasts, which were charted, up to about Bruges, either side of 1300, cannot be included in any hypothesis about the ancient coastlines of the Mediterranean and Black Sea.

You write: "Before the advent of portolan charts no known map was drawn to a consistent scale (neither was any map for a considerable period after that)".  What about Ptolemy?

You write: "The wind rose itself was a major innovation, because it didn’t exist before (the naming system of winds probably did, but not the diagram). It allowed directional measurements to be made across the chart, a novel usage of a map. The innovation to the wind rose by Petrus Roselli you mention is actually a minor one, but you misunderstand its meaning. Roselli indeed drew 16 more lines through the centre point of the wind rose from about 1456 onwards, but this does not increase the number of distinct directions to 64, as you claim. This number remains 32, as the new lines do not constitute new directions; they were already present in the (old) wind rose but do not run through the centre."

The portolan charts cannot claim to have invented the wind rose.  Ancient navigators had at least a mental compass, with 8 or 12 named wind directions, assumed to blow equidistantly from their respective points around a circle. Standard texts used those in contexts where we could cite a compass direction, and the House of the Winds in Athens physically demonstrates the 8-point version.  

As to the number of compass lines on the Roselli charts, we are not saying different things. I did not claim that the number of direction lines had been increased.  Indeed, in The History of Cartography (1987) I, p.396 (which I reference), I wrote that "in duplicating existing compass directions, the added lines serve no obvious practical function."

Your main concerns are of course cartometric ones.  First of all, it is incorrect to say that I 'reject' cartometric analysis. 

What I had hoped for from the Lisbon meeting last June was that it would attract a number of participants who looked at the charts from geometric and mathematical angles. The ensuing debate could then have demonstrated how much unanimity there was among those who were coming from a different direction to the traditional historians. Would different analytical techniques produce different results?  Would the necessary choices made by two different people working with the same technique produce an identical result?

Unfortunately, that did not really happen and what discussion there was did not help to convince the historians in the audience that the startling interpretation to which you felt your research led you was one we were forced to follow. You were well aware of that. 

If I might make a suggestion, which I hope is thought helpful.  Since I have commented, here and elsewhere, only on the implications of your claim for a pre-medieval original, and have said nothing about your methodology, my own views as to your cartometric evidence are of far less worth than would be those of your fellow practitioners.  What did not happen in Lisbon - perhaps from an understandable feeling among cartometric specialists that they would be heavily outnumbered by historians - could be achieved by a special meeting, or possibly a joint publication, devoted to cartometric analysis of the portolan charts.  

If, by that or some other means, your methods and results were endorsed by those well qualified to make such judgements, your case would be immeasurably strengthened.  But your work really needs that endorsement, which was not evident to me in Lisbon.

Good wishes,

Tony Campbell




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lazarus.elte.hu/pipermail/ishm/attachments/20170110/c7669658/attachment.html>

More information about the ISHM mailing list