Heraldry as a Systematic and International Language? About the Limitations of Blazonry in Describing Coats of Arms

New Haven, Yale, Beinecke 648, fo.2v

New Haven, Yale, Beinecke 648, fol. 2v (source)

Usually, this blog is dedicated to approaches to heraldic sources from the perspective of cultural history. To the contrary, this post will deal with one of the central subjects of heraldry as an auxiliary science: blazoning. However, I do think that this does preclude interesting insights also from the perspective of cultural studies, since I am convinced that heraldry as such can also be studied from this perspective.

As a matter of fact, I am about to write a paper for a conference in Nancy next month, where I will speak about our project to develop a technology that is meant to enable us to encode and thus digitally edit heraldic sources on a state-of-the-art level by using semantic web technologies. The main focus of this effort are, to begin with, historical sources, texts and illustrations from the twelfth to the sixteenth century.

The Idea

I will spare you the technical details for now, since we are going to elaborate on more explicitly in an upcoming paper. But to understand the question I will ask you later in the text, you nevertheless need to know the main idea at the background of this project.

At the moment, all blazonry, whether in books or in databases on the internet, is done by text. Any blazon is a combination of natural language words (even though there are attempts to standardise this), which is why they are dependent on an existing language (most of all French).

What we want to do is not to represent coats of arms and their components by parts of text, but by concepts. This means not to enter the description/blazon of a coat of arms, but the very idea of it. This sounds more complicated than it is—mostly. By using a special technology (semantic web technologies), we are able to represent the different parts of a coat of arms, their attributes and relations to each other by simple numbers and statements.

The great advantage of this, besides others, is that we are not bound to any natural language such as French, German or English in order to do so. It is as if you have the ideas of things represented in your computer, in the same way as if you had them depicted in an illustration, before giving them any name. Names you can add later on by declaring how those concepts should be called in the different languages, be it French, English or German. Thus, they can be retrieved from the system in any (heraldic) language, as you have taught the system the conceptual, non-textual information, which is always the same, and the according terms in those different languages. In addition, you can teach the system what properties different concepts may have or not, and specify them this way (e.g. that a lion may have different attitudes, that he may be armed in a different tinctute than gules, etc.).

The Problem

The problem is that such as system must be coherent, especially if you work with sources and blazons from different countries and in different languages. I am aware that there were several attempts to unify the blazon of different languages in order to establish concordances. The most prominent is perhaps the one by Baron Stalins for the International Heraldry Academy, published in 1952 as ‘Vocabulaire-Atlas Héraldique’, which later on was expanded by François Velde in his very useful ‘International Illustrated Glossary‘.

What I am  interested in, however, is the exact opposite of this.

I do not want to know how to harmonise international heraldic language and thus to overcome and remove any differences that may exist. What I am interested in are exactly those differences, the incompatibilities between different heraldic languages and traditions.

I know that heraldists like to understand heraldry and blazonry as a well ordered, highly formalised logical system. But to understand it better, e.g. in order to rebuild it in a computerised model as it is the case here, we have to know more about its actual limits. What is possible, what is not? Which difficulties occur if you dig a little deeper and go beyond the basic rules?

In my experiences, there are minor yet important differences in the way coats of arms are blazoned and even conceived:

  1. by different persons (having different personal backgrounds and styles),
  2. by different languages (based on different traditions in the use of heraldry),
  3. at different times (because the heraldic system is a historical one and is thus subject to change over time).

By Different Persons

I think this point may be obvious. Different people may very well blazon the same coat of arms very differently. But where do they start to disagree and to use different terms? And why? Are there any cases where people had different ideas of what is represented in a medieval orearly modern coat of arms and which term should be used to blazon it (e.g. in different editions of the same armorial)? Are there any charges and terms that can be easily confounded or that are ambiguous, so that you need more knowledge to be able to know what is meant? Thus, where are the limits of the clarity and unambiguousness of blazonry in a given language?

Scheibelreiter

Faden, Balken and Leiste in Scheibelreiter, Heraldik, p. 41.

Just to give you a small, German example: For the extent of horizontal ordinaries, Vaclav Filip, in his ‘Einführung in die Heraldik’ only gives two options: ‘Balken’ and ‘Leiste’ [1]. Georg Scheibelreiter offers three options: ‘Balken’, ‘Leiste’, ‘Faden’ [2]. The ‘Lexikon der Heraldik’ by Gert Oswald adds the ‘Strichbalken’ to this. [3] But if we look now at the ‘Handbuch der Heraldik’, written by Adolf Mathias Hildebrandt, we have again ‘Balken’, ‘Leiste’ and ‘Faden’, though ‘Faden’ is presented here as an very small vertical ordinary [4]. Which, if we look in the Wappenbilderordnung of the ‘Herold’ may represent a ‘Stab’ [5]. Heraldica.org finally gives ‘Balken’, ‘Querfaden’ and ‘Leiste’, depicting, in contradiction to the other examples cited, the “Leiste” as the smallest one.

Hildebrandt

Vertical and Horizontal ordinaries in Hildebrandt, Handbuch der Heraldik: Wappenfibel, p. 75.

Of course, these may be problems of modern heraldry. We are looking for more medieval and early modern examples. Where could the same coat of arms be legitimately described in quite different manners?

By Different Languages

Heraldry and the use of coats of arms developed quite differently in different countries. It seems plausible that the same holds true for their design and corresponding vocabulary. The question is, therefore, whether national or regional systems did indeed become incompatible. Are there elements of heraldic design in one country that are unknown in other countries, which as a result had no corresponding label in other languages? Where there things that mattered in one country and in one language that were irrelevant in another?

When you look for different types of bends on Heraldica.org, for instance, you find four English terms, ‘bend’, ‘bendlet’, ‘baston’ and ‘riband’, while there are only three terms in the German translation: ‘Schrägbalken’, ‘Schrägleiste’, ‘Schrägfaden’. It appears that there is no German equivalent for either baston or riband. The English language, in this case, is thus making a distinction that does not exist in German.

At Different Times

Heraldry is a historical system, and thus its components as well as its rules are subject to historical development and change. This means that the heraldic language, in its visual forms as well as its textual blazon, may become more complex or simpler as new conventions are established while others disappear, and there may be things that simply change their meaning. The term ‘sinople’, for instance, which initially stood for red, changed its meaning in the fourteenth century and began to refer to green [6]. The lions tongue as a mark of difference became relevant only in the 14th century[7]. Are there other terms that changed their meaning, or are there any conventions that changed their subject or their significance?

New Haven, Yale, Beinecke 648, fol. 6r (source)

New Haven, Yale, Beinecke 648, fol. 6r (source)

Your Experience is Needed!

To develop our system for future digital editions of heraldic sources, we require more clarity. While in your daily practice you will have become aware of differences that require some flexibility, a computer system cannot think this way. But you can teach it to. This is what we need to do, a little further into the process.

Consequently, already from the start we have to be aware of problems we will encounter, and we have to know about differences, ambiguities, and changes. As we all know, knowledge grows with practice. When you deal with historical coats of arms on a daily basis, you will know a lot about the problems I mentioned in this post. That is why we are asking you for your help! Do you know, from your experience, of any examples of the problems, differences and ambiguities outlined above? Where, in your opinion are the limits of traditional blazoning, especially if you want to do so in different languages?

[1] Václav Vok Filip, Einführung in die Heraldik, 2nd edition, Stuttgart 2011, p. 75.

[2] Georg Scheibelreiter, Heraldik, Wien et al. 2006, p. 41.

[3] Gert Oswald, Lexikon der Heraldik, Leipzig 1984.

[4] Matthias Hildebrandt, Handbuch der Heraldik: Wappenfibel. 19th edition, Hamburg 2007, p. 75.

[5] https://heraldik-wiki.de/wiki/Teilungen_(Hauptgruppe).

[6] See Gérard J. Brault, Early blazon: Heraldic Terminology in the Twelfth and Thirteenth Century with Special Reference to Arthurian Literature, Oxford, 2nd edition. 1998., p. 275.

[7] See Walter Leonhard, Die große Wappenkunst, München 1978.

If you have any ideas, thoughts or examples, please feel free to comment on this blogpost (be it in Englisch, French or German)! 

Cite the article as: Torsten Hiltmann, "Heraldry as a Systematic and International Language? About the Limitations of Blazonry in Describing Coats of Arms", in: Heraldica nova. Medieval Heraldry in social and cultural-historical perspectives (blog on Hypotheses.org), published: 25/05/2016, Internet: http://heraldica.hypotheses.org/4623.

Torsten Hiltmann

Torsten Hiltmann is Juniorprofessor for the High and Late Middle Ages and Auxiliary Sciences at the University of Münster. He is interested in new approaches to late medieval and early modern heraldry, the medieval notion of kingship and the methods and technologies of Digital Humanities. On hypotheses.org, he is maintaining, among others, the blog “Heraldica Nova”.

17 Responses

  1. Herbert Stoyan says:

    Do you know the Wappenbilderordnung of Arndt and Seeger? A book in german of 1996.
    Do you have an email address? I would like to send you an actual paper…

  2. My first comment: describing an armorial concept is much like blazoning in your favoured language. My second: there is only one armorial principle: contrast. All the rest can be derived from this – but the two are of less importance when the aim is to describe rather than create. My third: here we are only concerned with coats-of-arms, not crests or supporters or the like.
    As Torsten proposes, the three parameters: personal preferences, language and time, are primary criteria of distinction. On a more practical level, one may add the preferred level of detail (colours for arming, crowning, and languing are standard, therefore not necessary. Counting small charges or multiple partitions may not be necessary. Derivations from the ‘rules’ are mostly comfortable or unconcious shortcuts by painters or blazoners, it is easier to paint a secondary figure blue than the true colour when the brush is filled with blue – and remember people cannot spell, neither can they read handwriting when transcribing blazon to images or vice versa), major regional traditions (e.g. the French prefer the primacy of the figure for making ordinaries, others prefer the field), regional or universal synonyms (e.g. roses and cinquefoils, chiefs and per fess), semi-conscious variations (crown, double-tails, nowed, guardant – varies with sources, but are sometimes important), confounded readings or viewings (martlets and mullets / merlettes et molettes, coquilles or cocks, not to mention difficult charges like a rammer = pile hammer = paviour’s ram = hie = poteau de mer = mouton à pilonner = demoiselle = Meerpfahl = Wegwalze = Rammen); housemark- and Slavic totemlike figures are also difficult to describe. Size matters, whether a full page basilisk as in Grunenberg, or a smallish figure in a 5 x 5 table of arms. Styles are variable, and can be mixed, e.g. there are 3 different styles of lions in Berry, from poodles to standard lions rampant.
    For the level of describing concepts, one must decide on priorities and levels, and separate colour, figures, and placing. In my experience, it is important to differentiate between primary charges in two dimensions (e.g. a chevron), subcharges (between 3 escallops), brisures, and tertiary charges like powderings and charges on charges (e.g. mullets on crescent dx).
    Lastly, as mentioned in some of the commentaries: what do researchers want from a database of coats-of-arms? If they want details for an art historian, they will not get it. He/she would need the actual images, but they could get the reference, basic description and context. Armorists, historians and genealogists could get lists of similar arms and names for comparison.

  3. Karl Wilcox says:

    Torsten, I have a parser which reads British Blazons and converts them to a canonical form in XML, conforming to a schema called BlazonML which I have defined. (This is as an intermediate step towards actually drawing them – you can try this out for yourself at the website below). You can see the XML Schema here:- http://drawshield.net/BlazonML.xsd . I have also made some small steps towards to parsing French blazons to the same canonical form, which allows them to be compared. I also have some ideas about using this to develop a “metric” of similarity between blazons and several other ideas. I would be very interested to discuss these with you!

    • Dear Karl, this looks indeed very interesting in many ways! I’m looking forward to learn more about it. I will write you an email in the following days. Thanks for getting in touch!

  4. Hans Cappelen says:

    There are differences between the blazon as a “definition” of the theoretical identity of a coat of arms, and the blazon as a description of a certain reproduction of the arms. When writing “unnessary” I mean not necessary for defining the identity. A question is what you want to do with the new electronical blazoning system: is the purpose mainly to study blazons in a historical context, or is it (also?) to have a coherent, international system for blazoning old and new arms defining their identities.

  5. Hanno Wijsman says:

    Merci beaucoup pour ton billet intéressant, Torsten. Je n’ai pas de choses très concrètes à partager, mais je me permets de réfléchir un peu avec toi.

    La langue du blason est une langue naturelle. Comme tu le dis aussi, c’est un malentendu (bien répandu) de croire que c’est quelque chose de très systématique et formel. Au contraire, c’est une langue humaine avec toutes les possibilités, règles, exceptions et illogismes que toute langue humaine possède. Ceci dit, croire le contraire est bien compréhensible ; c’est, par exemple, ce que beaucoup de gens pensent aussi des langues de signes, utilisées par des sourds : que ce soit une langue artificielle et inventé d’une seule trait ; non, les langues de signes sont aussi de langues naturelles avec leur historique.

    Donc, en effet, comme tu le dis très bien, il y a dans la langue du blason des différences individuelles (chacun a ses tendances personnelles, son idiolecte), géographiques (par pays ou communauté linguistique) et historiques (par moment dans l’histoire). C’est la même chose pour toutes les langues humaines.

    La première conclusion est du coup décevante, car, justement, comme pour toute langue humaine, la langue du blason ne sera donc jamais simplifiable à un seul système logique qui englobe toutes les spécificités de chaque variante individuelle.
    Mais ma deuxième conclusion donne plus d’espoir, car, comme pour toutes les langues, au bout du compte et malgré toutes les variétés, on arrive très bien à communiquer et à se comprendre et la machine peut le faire aussi.

    La bonne nouvelle est que le français, l’allemand et l’anglais (etc.) sont là pour discuter d’absolument tout ce qu’on veut, ce qui complique sacrément l’affaire, tandis que le but des langues du blason est beaucoup plus restreint : décrire des images construites dans un espace restreinte et selon des habitudes restreintes.

    Décrire les armes par des concepts est un beau défi, mais il est essentiel d’intégrer d’emblée une souplesse qui englobera toutes les ambiguïtés et doubles sens. Les mêmes armoiries peuvent avoir des multiples expressions visuelles et des multiples expressions blasonnées. Et une seule expression visuelle ou blasonnée peut représenter plusieurs armoiries (‘homonymes’). Encore une fois : tout ça trouve des parallèles dans nos langues de communication courantes.

    • C’est bien que tu mets le point sur la question de la formalité (prétendu) de la langue de blason, Hanno. Oui, la langue de blason est une langue naturelle. Mais plus formalisée que des autres langues, et plus restreint dans son champ d’application. Compte tenu de ceci, elle devra donc normalement représenter un excellent cas de recherche pour les informaticiens et linguistes qui travaillent dans le domaine de NLP, je pense. Mais ton remarque me fait penser plus clairement le grand défi du projet même. C’est de trouver un bon équilibre entre ouverture pour les différences d’une part et les règles unifiant d’autre part. Trop des différences nuiront à l’intégrité des données. Trop de rigidité à la fonctionnalité.

  6. An interesting subject. Maybe it will be helpful to look at “related” computer programs of iconography, develop by art historians. A commonly used program is Iconoclass.

    • Thank you very much for your comment pointing to vocabularies like Iconclass. The problem is, that all the existing vocabularies and ontologies in the cultural heritage domain don’t’ offer much when it comes to heraldry. Mostly, you only can state that there is some heraldic information without any specification about its content. Getty AAT and, for textual editions, TEI only offer the possibility to state that there is a coat of arms. Iconclass goes a little bit further and gives you the possibility to specify whether the bearer of the coat of arms is male or female, on which support the coat of arms is depicted and whether it is accompanied by a name. But nothing as well on the heraldic information as such. That is the gap we want to fill.

  7. Michael Ramey says:

    Given that later heraldry will refine and redefine terms, certain devices (arms) will be a subset of earlier devices. The earlier devices might have been retired, but for a modern accounting, they will overlap. The only way to make them unique is to attach a time quality, and possibly a time & place quality, which is exactly what you want to avoid.

    You also run into the three bars vs. Barry problem. One instance of Barry might have “three bars” while another might have four. So you can describe an instance of a device, which may or may not be a “correct’ abstraction into a blazon.

    • Dear Michael, your comment hits the point. But instead to avoid time and space as influencing variables, we just want to include them. The idea is not to unify everything in a rigid system but to represent it in its diversity. Thus, your remark on the Barry is important. Here, for instance, you can see that a 1:1 translation from English to German isn’t possible since in English blazon you count the areas, in German blazon however the dividing lines. Thus, “barry of ten” in English would be “neunmal geteilt” in German. I wasn’t aware of this, so far. Thank you!

  8. Jovan says:

    You should consider the system developed by Zeljko Heimer, president of the Croatian Heraldic and Vexilollogic Society, as well as master of informatics sciences. His latest book deals with heraldic blazon as a formal language.

  9. Hans Cappelen says:

    One question is how detailed a blazon needs to be. Is the blazon to include only a few or many varying reproductions of a coat of arms? For an art-historian there will be many interesting details that is not of importance to a heraldist or a lawyer who is only wanting to identify the owner of the arms. The number of details is of significance to the freedom for the heraldic artist. Blazons are intimately connected with the heraldic traditions for stylizing the charges. When such traditions tell us how to stylize, it will not be necessary to include so many details in the blazon. In some blazoning I see a tendency to include details which should be unnecessary, e.g. tinctures on an animal’s tongue, claws etc.

    • Yes, this is just the problem I’m thinking of. Which details are important to identify a coat of arms correctly, which are only liberties of the heraldic artist? The point is, that those conventions are not fix but depend on time, space and context. This leads us to many interesting questions. How could two different traditions communicate to each other? Where there processes of exchange and entanglement? The challenge is to identify those differences. In order to study them, but also to make our machine work more properly, time-space sensitive, if I may say so.

    • Hanno Wijsman says:

      I agree with you, of course, that the pont of view is essential. But then, if you want to make a system that describes faithfully what we see, details should be included, I think. Why do you say that certain details “should be unnecessary” ? The image reproduced below from the manuscript 648 shows that even at the time (late 15th century, I guess) the “lampassé de gueules et onglé de sable” was thought important to blason. So who are we historians then to say that we should not blason those details ?

      • Yipp, that’s the interesting point, I was always wondering about blazonry. which of the details are meaningful, which aren’t? Is the position of the charge a little bit higher than usual already meaningful or a mistake by the painter? Or the number of elements when a coat of arms is semé of something? Or, regarding the lion, the position and form of his tail? Of course it would be best to encode everything we see. But nevertheless you have to define some limits or this kind of granularity. Because, if you want to compute those information afterwards, too many “meaningless” (for a given context) details would make it unmanageable. At the end, what we are looking for is the meaningful information within the picture in order to retrieve the code which is encoded in the picture, not the picture as such. And the question is, what is part of the code, what is changing depending on time, space and context. I see this more clearly now.

Leave a Reply

Your email address will not be published. Required fields are marked *