Armorials as evidence of coats-of-arms

Steen Clemmensen, october 2007; based on the presentation The Armorials  - Groups and relations, at the 26th International congress of heraldry and  genealogy, Brugge 2004.

Most armorists use a small selection of standard reference works, e.g. Siebmacher, Burke, Papworth and Rietstap, as their primary source identification of older coats of arms – many without questioning their reliability. A few professionals take the opposite view and hold e.g. Burke’s General Armoury as nearly worthless as evidence. The truth might lay somewhere in the middle, but that is another discussion. One objection is that they rarely state their sources, so the user has little evidence to base his arguments on. But what may one use instead? Seals and funeral monuments may be available, but they are either uncoloured or restored, and descriptions of the restoration process are harder to get at than most other things. So one would turn to the armorials and line up as many as one can manage.

That is a commendable practice. It is almost too trivial to repeat that blazoned or painted armorials are our main source to the colours used in medieval coats of arms. In one form or another, this statement is a key message in nearly any basic introduction to heraldry. But using armorials is not without risks, and those risk are seldom discussed. What might look as proof could well turn out to be circular conclusions or flights of fantasy. Most armorist will known that citing both Burke and Papworth will not bolster their arguments, as Papworth is merely Burke as an ordinary supplemented with some citations from unnamed armorials. Citing multiple armorials may well present the same type of non-evidence.

 Ever since Paul Adam, some 50 years ago, noted the similarity of the Normandy sections in the Wijnberghen and Toison d’or armorials, armorists have recognised that a segment of an armorial might reappear over a period of several hundred years. What is less acknowledged is how widespread this might be.

While the early armorials, as we know them, are nearly all short original collations, by the year 1400 long, composite editions, not collations, dominate the surviving manuscripts. Of course, many segments are contemporary collations, but not the armorial as a whole. It is a little disturbing that of the less than 350 medieval armorials which have survived, possibly 10% belongs to a very small number of groups of clones. Three examples of such groups are based on the armorials Urfé (URF), Toison d’or (ETO, Golden Fleece), and Navarre

As an example, look at the Bergshammar, now in the Riksarkiv in Stockholm. It was manufactured in the Bruxelles area around 1450. The number of segments in it may vary according to the criteria used. Their number will increase if the criteria are very detailed, but here it is put at 66, with some having several clearly delineated subsegments. Jan Raneke, who edited the armorial in 1975, recognised 3 sources, Gelre, Toison d’or and Lyncenich – alias Gymnich, though he is incorrect in the details.  Most of the 66 segments of Bergshammar come from these 3 armorials. Only 16% of the 3400 items are from presently unknown sources.

BHM   Segment                       Items       Source

34        Polen                                 79        ETO 17*

35        Berry                               134     

36        Milan                                 22       

37        Lothringen                         15       LYN 59

38        Savoie                              22       GEL 39 + ?

39        Namur                             24        LYN 74

40        Normandie                     101        ETO 11= WIN 3

41        Bretagne                          79       

42        Hildesheim                       12        GEL 12

 The integration of the sources is the most complex that I have encountered, but in many ways typical of a 15th century armorial – and they are unlikely to be the result of later rebinding. Batches of segments are taken from one source, but not in progressing order. Then the process is repeated from a second source, and so on.

 Most of the Toison d’or is here, and a large part of Gelre too, while the Lyncenich is a source of lesser import. The Wijnberghen-Normandy segment is recopied from Toison d’or. And this really means from the Toison d’or manuscript, now in the Bibliothèque de l’Arsenal in Paris. That the artisan must have had the actual Toison d’or and the actual Gelre before him, is evident from the way the items are organised, though unlikely, it could have come from lost exact copies.

An example of how a segment (from Toison d’or, ETO) may be copied so as to look as an independent armorial is given on page 83 in the Paix de Arras, available on this website.

 The next example is from the armorial Berry (BER), attributed to Gilles de Bouvier, roy d’armes Berry.  It has largely unexplored relations to several other armorials, but I will look at one particular segment, the Ile-de-France, personal domain of the King, and as such it should be of central importance to his chief herald. But it is not. It is actually a reduced copy of a fragment of the Navarre – which is in blazon only. And this casts a light on the problems of moving from blazon to paint.


NAV               Branch                              BER

donor                                         acceptor


916           Soissons                                 

917           Coucy                                       268

918           Coucy  & bend                        

919           Berlaimont                               

920           Clermont                                  

921           Clermont  ch. escallop            

922           Clermont & label                      

923           Clermont & label                      

924           Clermont & label                      

925           Moreuil                                      269

926           Moreuil & bend                        

927           Plessis                                     

928           Hangest                                   

929           Hangest                                    270

930           Hangest & label                       

931           Raineval                                   271

932           Raineval ch. escutcheon        

933           Beauvais                                  272

934           Poix                                         

935           Cramailles                                273

                                                        (new  274  inserted)

936           Montigny                                   275

His intention was obviously to make a display of families rather than the display of persons favoured by the Navarre herald. In this process, he did make a number of mistakes, including the most common: replacing a mullet with a martlet. Making one of the mistakes is easier to understand reading it in French rather than in English. In this case he needed to adjust the rest of the blazon in order to make it understandable. So crusily becomes a crescent :

                      NAV:1015  lermite de caumont  de gueules croixetes d’or à .iij. molettes du même

                      BER:298   le sr de cormont de gueules au croissant d’or acc. de 3 merlettes du même

 It will come as a surprise if there is a Lord of Cormont with this coat of arms. Of the several Cormont’s arms known, neither comes remotely near the present one. Fillièvres (NAV:1017) – Siliènes (BER:299) is another examples of this substitution.

 BER   text                        Branch                            NAV  text

267       philippe d'artois                      Artois                                           

268       ceulx de vervins                       Coucy                             917        le sire de cursy

269       ceulx de moreul                       Moreuil                           925        le sire de moirieul

270       ceulx de genli                          Hangest                          929        m aubert de hangest

271       ceulx de raineval                     Raineval                         931        le sire de raineval

272       ceulx des boves                      Beauvais                        933        le castelain de beauvès

273       ceulx de cramailles                 Cramailles                     935        le sire de cramailles

274       ceulx de paillart                       Sourdon                                       

275       ceulx de montigny                   Montigny                         936        le sire de montegny

transmission and change of name from NAV to BER


Even in the short fragment of BER:297-311, there are more mistakes.  A few are probably here, because he was not able to read the whole text, and had to guess either the name and/or the coat of arms, a reduced Quieret (NAV:1016) becomes Rinzy (BER:300) and a reduced Beauval (NAV:1018) metamorphoses into Breteval (BER:301). Another, not uncommon, mistake, is misreading the lines, and thus doubling the name. The unwary will now have a Gourlay (BER:309; actually BER:311 = NAV:1040) with the arms of de la Motte (NAV:1038), the genealogists can have a field day, and the purists may speculate on the change from Argent a cross moline Sable to Azure a cross moline Or.  However, not all changes are mistakes. Some are simple changes of title as the incumbent was contemporarily better known by a different lordship.

 Preliminary alignments have been made of another ’French’ group, named from its eldest member, Urfé, which includes the Prinsault in addition to the armorials named above. The group exhibits most of the features discussed above, but sheds an additional light on the process of inheritance. The LeBlancq, manufactured in Lille around 1570, appears to be a copy of an Urfé-clone with URF segments in irregular order: 1, 21, 18, 19, 17, 13, 8, 16, .. , with a tail of a large part of the Bellenville (BEL segments: 37-43, 14, 18-20, 29, 21, .. ) and with a small part of a Lyncenich clone (Austrians) inserted. The English segment of LBQ appears to be a merger of Urfé and Lyncenich. It might be far-fetched, but perhaps the owners of the 3 armorials were contemporary bibliophiles in that provincial capital.  The latest member of the group, SIC, a copy made around 1640, is unlikely to have much to do with the famous herald,  Jean Courtois dit Sicile. It is most likely a late copy of an Urfé-clone with a prologue, English segment and a list of knights added from a Toison d’or clone.

 In summary, most late medieval armorials are composite works, edited from whatever sources available. These might be short contemporary collations, but will often turn out to be reuse of older material. As such, they will exhibit faulty transcriptions, confounded blazons and wrong attributions. And for many coats of arms the actual colours and figures provided are uncertain. This implies that apparently independent armorials might be derived from a single source – and that the evidence they provide is not evidence, but is just a circular argument. Such reuse has continued from the late Middle Ages into our time – and is to be found in most of the standard reference works. Though the references are generally trustworthy, any family mentioned in an armorial needs independent verification before its use of a coat of arms should be accepted.


 1.        P. Adam-Even et L. Jéquier, Un armorial francais du XIIIe siècle. L’armorial Wijnberghen, Archives Héraldiques Suisses, 65 (1951):49-62, 101-110; 66 (1952): 28-36, 64-68, 103-111; 68 (1954): 55-80. [WIN].

2.        Michel Pastoureau et Michel Popoff, Grand armorial équestre de la Toison d’or, fac-similé du manuscrit no 4790, propriété de la Bibliothèque nationale de France, conservé par la Bibliothèque de l’Arsenal, I-II (Saint-Jorioz: éditions du Gui, 2001). [ETO].

3.        Paris, BnF,, armorial dit du héraut Vermandois. [VER].  -  Paris, BnF,, armorial dit du héraut Navarre. [NAV].  -  Emmanuel du Boos, Armorial de Gilles le Bouvier, héraut Berry, d’après BnF, (Paris: éditions du Léopard d’or, 1995). [BER].  -  Christiane van den Bergen-Pantens, Gelre, B.R.15652-56, (Leuven: Jan van Helmont, 2001). [GEL], incl. reprint of P.Adam-Even: L’armorial universel du Heraut Gelre, from Archives Héraldiques Suisses, 75 (1961): 48-85; 76 (1962): 68-73; 77 (1963): 63-79; 78 (1964): 75-80; 79 (1965): 70-82; 81 (1967): 72-83; 82 (1968): 70-83 .  -  Bruxelles, KBR, ms.II.6567, armorial Lyncenich alias Gymnich. [LYN].  -  Jan Raneke, Bergshammervapenboken – en medeltidsheraldisk studie. (Lund. 1975). [BHM].  -  Nancy, Bibl.Municipale, ms.1727, armorial Nicolas de Lutzelbourg. [NLU].  -  London, British Library, Add.11542:94r-106r, armorial de la Paix d’Arras. [APA/a]; Paris, BnF, [APA/b].  -  Paris, BnF,, armorial Clémery Européen. [CLE].  -  Paris, BnF,, armorial Urfé. [URF].  -  Paris, BnF, + BA, ms.4910, armorial dit du héraut Sicile. [SIC].  -  Paris, BA, ms.4150, armorial dit du héraut Charolais. [CHA].  -  Paris, BnF,, Un provincial d’armoyries .. Le Blancq. [LBQ].  -  Michel Pastoureau et Michel Popoff, Armorial Bellenville – BnF, (St.Jorioz: éditions du Gui, 2004); Léon Jéquier, L’armorial Bellenville, Cahiers d’Héraldique, 5 (1983). [BEL]. Transcriptions of several  manuscripts were graciously provided by Emmanuel de Boos and Michel Popoff.

4.        Paris, BnF,, armorial Clément Prinsault. [PRT].