Dynasties, heraldry and emblematics

[Solo]

Well this one is my garden. If you have any point to discuss like flag entries you think would work better or dynasties you'd like me to add to the list, this is where to debate and ask.

I'm also gonna give some details how I do things regarding heraldry and emblematics in general because there's an overall pattern I follow to keep things consistent and coherent through the timeline.

DYNASTIES :

- Completed works are : french and norman dynasties (including subcultures, occitan, francoprovençal and breton). Aragonese and catalan dynasties.

- Work in progress right now are german dynasties. It's a bout one third complete I'd say. There are hundreds of them so it's a bit longer than any other.

MAP FLAGS ENTRIES :

- About 1700/2200 done. France, England, Scotland, Germany should be completely covered.
- Crusader heraldry is completely covered.
- Work in progress is : Spain and Portugal, about 90% done. Improving Eastern Europe and russian principalities (not even halfway done).
- Areas lacking : Scandinavia. If you have ideas feel free to give them.

[Solo]

This is a copy of my post about the colours and the random generator in coat_of_arms.txt from the Patrum Scuta thread.

So about the colours. This post may be of some interest to anyone fancying heraldry so I'll explain things a little bit.

The method is a statistical one named héraldique comparée, developped Michel Pastoureau back in the eighties within the field of history of mentalities and states that studying patterns in the construction and evolution of emblems informs the historian on many aspects of the medieval society. It was the backbone of his theory (and book) on the rise of colour blue from the bold association of the image of the grieving virgin mary and the prestige of the french monarchy allowing it to overcome its status as banned colour from the roman code (there were even serious studies before that questionning the very ability for the romans to see the colour physically) to the nowadays colour of Europe and most prefered colour of europeans.

Now what he had to do before he could study the colour blue or the banner of the kingdom of burgundy is gather a lot of material to compile a database of known frequencies in heraldry. This is what interests us since he studied 12 000 emblems in all kind of sources (seals, rolls, paintings, etc) and I doubt anybody will manage to throw himself in that amount of work in a near future. This will be useful both for colours and the construction of the emblems but right now what interests me are colours.

Since the timeframe covers all the medieval era, I can directly use his most generic results covering the middle ages in all of europe (before 1500) and this is where things get interesting. Looking at frequencies we learn that not all colours are equal in their use. In fact only five of them are commonly used and cover 95% of the cases.
- The two metals are rather equally distributed (42% for gold, 51% for silver) close enough at least to be kept at equal chances without making things unrealistic.
- The three common other colours are dominated by gueules (red), with is declining during all the timeframe (before 1200 it accounts for two thirds of the total) but still appear on 53% of the shields with a slight particularity : 4% of the cases are associations with furs that are not even represented in the game (you'll see 4% isn't that marginal afterall).
- The second most common is sable, or black. It accounts for 24% of the cases, one quarter (more common with gold, you can probably read the influence of the imperial arms in that).
- The third colour is azur or blue which only has a 20% frequency despites the fact that it's the rising colour through all of the timeframe. Awfully rare before 1200 (less than 5% of the arms designed before that date) it gets popularized with the french dynastic (burgundy, vermandois, anjou, dauphiné etc) and political (bar, bourgogne) alliances.
- Sinople, the heraldic green accounts for 2% total of all medieval arms. It's about as common as the two furs, hermine and vair.
- All other cases make up for 3% of the total. Being a summary of association, most of these are untraditionnal (never say prohibited, heraldic laws have never been set in stone) combination. They may be intentionnal to separate from the common flock as it is in the arms of the kingdom of Jerusalem or simple matter of taste. It points how uncommon the association between black and red (think Albania) or similar cases are. The case of purple can also be included here, the colour being exceedingly rare and not purple at all (as stated in an earlier post, its the mix of all other colours on the palette of the artist, so not really pretty most of the time).

So having read that you guess that using a code that allows sinople (green) or purple to be as common as gueules (red) isn't something that will give you a realistic result. It may gives more variety but is it something you really want when it didn't exist is the question. In the middle ages everybody was fine with arms looking similar in every manner because imitation also fuels prestige. The count of Savoie adopts the imperial banner because it reminds everyone he was an imperial officer in a region far from the heart of the empire where authority is dimmer. That patterns repeats itself every where when authority or prestige has to be reinforced in some way (think of the counts of Barcelone adopting the burgundian banner when they became counts of Provence). It's all intentionnal and almost never random. Trying to differenciate people that tried to present themselves as similar in some way is fighting against the mecanics of the medieval society.

Back to the point, coding those frequencies is a rather simple task. By repeating one colour in the code, you get it twice in the choice, which if you push things a little bit allows you to repeat colours in a manner that will recreate semi accurate frequencies. It's only a matter of how precise we want to be, allowing a bit more sinople than it really was is no big deal if it's five times less common than gueules (we can't account for differences between regions and periods which would vary much more than the variation here). This is how I coded things without repeating the colours too many times (14 colours is about the same as vanilla) :

color = {
{ 255 206 11 } #0 Or
{ 255 255 255 } #1 Argent
{ 170 10 10 } #2 Gueules (51% freq.)
{ 10 10 10 } #3 Sable (24% freq.)
{ 42 94 137 } #4 Azur (20% freq.)
{ 10 116 54 } #5 Sinople (2% freq.)
{ 170 10 10 } #6 Gueules
{ 170 10 10 } #7 Gueules
{ 170 10 10 } #8 Gueules
{ 170 10 10 } #9 Gueules
{ 170 10 10 } #10 Gueules
{ 42 94 137 } #11 Azur
{ 10 10 10 } #12 Sable
{ 10 10 10 } #13 Sable

}

banned_colors =
{
{ 0 1 } #0 métaux
{ 0 1 } #1 métaux
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #2 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #3 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #4 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #5 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #6 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #7 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #8 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #9 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #10 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #11 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #12 couleurs
{ 2 3 4 5 6 7 8 9 10 11 12 13 } #13 couleurs

}

I banned non metallic colours from associating with each other because it's not very representative (happens more locally than anything, for ex in german speaking Switzerland). Sinople is a bit more common but by experience it's at an acceptable level. I only regret not being able to implement furs but otherwise the result are very good as it is. The field colour has a tendency to be metallic but I guess it's a flaw in the system (can be corrected by repeating and inverting patterns but is it worth it for anything other than common designs).

[Solo]

This post is a copy from Patrum Scuta too and is part of the same discussion

Yup stats on the patterns themselves are the third and most complex part. I had to go into details for colours even if the logic is very simple so it's done by advance for the rest. If you have looked at Brian Timms recreation of those medieval rolls, you certainly noticed a complete roll can be done with a rather limited number of design. Creating a set of random emblems out of thin air for such a number of important people is a bit like looking at rolls like the Bellenville, Gelre or Wijnbergen : you have princes and their vassals featured in clusters (which sadly we can't recreate) with very repetitive constructions from one to another. Right now the set look much more like what you would find at the end of an almanach on heraldry, featuring all known patterns in a linear order, with equal weight.

Not that those rare designs don't exist, in fact if you look at some localized rolls like the zurich roll you will find some funny and exotic charges like musical instruments, tools, people and such things but chances are, if you choose randomly any important noble from any random region in Europe, his arms will be something very common most of the time. Rare cases are almost always canting arms or inherited protoheraldic dynastic emblems (sometimes both at the same time, like the dauphins in Viennois, Forez and Auvergne) and are known enough to us that they don't feel akward like a pair of cissors, or a lute would be.

In fact the higher you go in medieval society the more repetitive emblems are, with crosses (obvious religious signification, imperial banner), lions or eagles (both prestigious animals in the bestiary associated with nobility or authority), bends (banner of Lotharingia), fesses (very common in feudal heraldry with example like Saxony, Luxembourg, Hungary), pals (banner of Burgundy), well you get the idea. Alliances, pretentions and imitation does the rest spreading those patterns downward into smaller and smaller heraldic groups (thats how heraldry gets more and more complicated, while the emblem on top of each of those pyramids is always very simple in its design). Perhaps I'll make an illustration to show how it works on a large scale with the burgundian and bosonide opposition and their long lasting influence over most emblems in the Barcelona-Dijon-Pisa triangle.

Colour codes are not important, those are just my usual colours that's all. From experience it's better not to decide on them before finishing the whole template (some concept may work better with less saturation than others). The only colour I never change from those is gold but that's only because it's what works with my gilding texture overlay (not sure I'll need that in the CK2 context).

[Solo]

Las part of the discussion, regarding the designs themselves

First of all there are four kinds of emblems in heraldry (as studied in medieval heraldry, not in the blazon which rules are superficial and useless to historians) : Animals (lions, eagles, etc), geometrical divisions (barry, bendy, lozengy, full crosses), geometrical charges (bend, fesse, pal) and other charges (stars, keys, castles, crosses and the like). The balance between those changes a lot depending on geography and chronology : while animals are more common in early heraldry (more than half before 1200) their frequency decreases regularly to fall under 20% after 1500. On a more global scale, such as the extensive study made by Pastoureau, you will find that those balance each other much better. While designs of the geometrical superfamily (geometrical divisions and charges) are a bit more frequent, the other two kinds are usually equally spread out.

I took a simple course and chose to even things out. There are so much things we can't account for (and one is the way the game choose to assign an emblem to individuals or holding), that blindly following percentages wouldn't mean anything in the end (plus we can correct things other way). So I chose to divide the 240 emblems equally and got four programs of 60 emblems to fill with the most common designs. I then took my old databases and searched for the statistics on designs and divided again those 60 in a set number of designs according to their respective usual weight in heraldry. Those are western centric, since the weight of the craddle of heraldry - Northern France, Southern England, Low Countries, Western Germany - in the known material is almost overwhelming (although personnal results for occitan and north italian emblems are included in this one for example).

So, things we can't account for since they are numerous, deserve a few notes. First, there's the evolution of tastes during the timeline. Emblems in early heraldry are very different from emblems in late medieval heraldry. They have more secondary charges, which also means a greater number of colours. This is either due to alliances, since heraldry is an individual system - people can choose to wear the same but it's an individual choice to do the same as your next of kin, change them completely or use a "brisure" of any kind - or simply due to the fact that the medieval mind hates blanks and have a tendency to fill them over time (stars, martlets, shells, mullets, besants or others depending on the local tastes, fill that role in heraldry). They are also much more consistent in their construction. Early emblems have a tendency to vary a lot in the number and organization of their charges : a pally can be three pals on the next seal, a lion can be passant, inverted, then rampant and be the same rampant regular lion in the first place (this is why the arms of Richard Lionheart were probably simply an undefined rampant lion if you ever read about the wild theories on his seals).

Another thing that can't be represented in those is the meaning. That's the main and most interesting feature of heraldry as a whole. Nothing is ever random in heraldry. Charges and colours are choosen with great care, because that emblem, to the rest of the society is like an identity card which can only be decoded through context (and that's where things get more complicated for us centuries after). That meaning can be political and represent a choice made according to your political and matrimonial alliances (or pretentions) spreading a combination of colour (colours are both emblems and symbols) or a given design (geometrical or not) through various groups of individuals.

The meaning can also be religious or symbolic. An emblem can be a symbol just as much as a symbol can be an emblem (that's exactly the case when crusading rulers wear crosses of all kinds and transform them later into emblems). The french fleur de lis is at first a christian symbol adopted by Philippe Auguste that later become an emblem of the capetians, but also a symbol of prestige and authority (the late arms of the kings of Bosnia for example).

Lets take the arms of Savoy again as an example (Gules a cross silver, a white cross on a red field if you want). That's the imperial banner obviously, a proto heraldic emblems inherited from the carolingian rule. It represents the count of Savoy as a vassal of emperor, an imperial officer (at the time they are designed) and a christian prince. Now place them in the context of south eastern france, where the counts of Savoy and dauphin de Viennois are facing each other, with intertwined holdings everywhere in the regions and living less than 50km apart (Montmélian and Grenoble at each side of the Grésivaudan valley) and the same pretentions on ruling the alps from french to italian side. The dauphins chose an emblem that features the capetian colours (third degree capetians), which clearly state their opposition to the imperial rule and their preference for the french side. They also wear a christian emblem, the dauphins which in medieval bestiary is a fish, the first christian emblem (that would be obvious to medieval people even if it's not to us nowadays).

One last meaning that can't be taken into account that is not to be ignored totally are the canting arms. In that case the arms simply make a play on the holder's name. It can be related to colours (Rougé in brittany wore Gules), charges (Lucy for lucius in latin, a fish we call brochet in french), or even a division (Maugiron wore an irregular gironny, as for "mal gironné" a bad gironny). Those are rather common and many akward charges encountered are often associated to this meaning. What this means is a family named Leusse or Leuczon (those are real french ones) encountered in the game would be likely to wear a fish of some sort but they won't because it's random, while you will find the fish in the arms of another random family but the chances to get the name and the meaning to match is even more unlikely than what it is with other kinds of meanings.

So, knowing all that we still want to make things look like a scrambled (because names, meanings, alliances are all messed up compared to any real historical or even fictionnal counter parts) roll of arms, in which we'll encounter common designs more often than rare ones. There was a lot of awfully rare ones in the original set which means many things are gone. They are sometimes replaced by repetitions of common ones : you need more lions than you need dragons, because the former represents 50% of all animals usually encountered in heraldry while the latter accounts for only 1% of all the bestiary. It also affects subdivision of one specific type and its variants : shields with one or three bends are much more commons than shields with bars (inverted bends which are highly uncommon in fact, while they were equally frequent in the original pattern file) or even two, four or five bends (two fesses are common, two bends are rare, that's imitation at work).

In the bestiary of medieval heraldry you will encounter mostly rampant lions, regular ones. The lion itself represents half of all animals and in most cases it's featured in the rampant regular style (facing or inverted are very rare, but passant of all kind have been included). The two other common animals are eagles, always single headed (two headed is 15th c.), and the martlet, incredibly frequent in french and english shields (the latter because of the norman invasion of course). The five other classifications of the bestiary represents only one quarter of all cases and are equally divided between leopards, quadrupeds (deers, bears, wolves), chimeric animals (dragons, griffins, panthers), birds (ravens, falcons and wings), and fishes (pikes, bars and dolphins). If it's not named specifically in that list, it's just too rare to be featured in the sample.

In the second part we have somme common designs like crosses (as long as they are not divisions), fleurs de lis (higly stylized), castles and towers (very close in design, a tower is really just a castle with one tower while the castle has three), écus (shields), quintefoils (medieval roses, stylized too) or besants/annelets (plain/empty circles) which account together for half the cases encountered. You can also find crescents (as much a christian symbol as it is a muslim one in medieval heraldry), stars (usually six, eigth or sixteen pointed), keys (a religious symbol commonly used as emblem), military related objects and weapons (gonfanon, drinking horns, swords, axes, escarboucle). That's already 90% of all cases (I think a detailed list would be boring, that's the more diverse category of all).

Third part is made up of geometrical divisions of the field. They are usually compared/combined with the last category (geometrical charges, which by opposition have a number and can be repeated like one bend, three bend, rather than a bendy of two colours). There you find three very common kinds : the cross (20% or so with all variants), barry (fascé/burelé, that's 15%) and chief (chef, 14% or so). The rest amount for half the cases spread among a long list of designs such as (by frequency ranging from 7% to 1%) : saltire (sautoir), checky (échiqueté), paly (palé), bendy (bandé), fretty (fretté), semy (semé), lozengy (losangé), quarterly (écartelé), vairy (vairé), per pale (parti), gironny (gironné), plain (not sure about the english, term, that's only the colour like Albret).

And in the last category we find a more limited number of designs. Two kinds account for 80% of all cases by themselves : that's fesses and bends which are often in their most simple form (one fesse/bend, regular) but can sometimes be repeated (two fesses and three bends/bend coticed are common designs). Then you have chevrons and pals (one or three in both cases) and a bit less frequent (and mostly in french/english heraldry), the gemelles (jumelles).

If you managed to read all that, you probably start to see how repetitive heraldry is in reality. Out of maybe 200 different designs in the original pattern sample, we downgrade things to maybe half of that, a hundred different constructions. On the other hand I tried to include many secondary charges to these (a third colour would have made a difference here but thats what we get for now), which are an important aspect of heraldry and have frequencies of their own (the most common one is the martlet which was completely missing from the sample, but shells, stars, mullets or besants are common too). Some specific designs are repeated over and over with minor differences in number like martlets for gemelles, shells and stars in chief, alérions, shells or stars on a bend and so on.

[firelordzuko]

I don't believe there's a way to tie the likelihood of certain charges and tinctures to the culture of the dynasty, is there? (For example, a Spanish dynasty having an increased chance of castles or Lions Purpure, or Russian dynasties using unheraldic emblems, )

In any case, I love what you have done to the heraldry of minor nobles in the game; I found most of the Vanilla ones absolutely ghastly and I approve of the addition of marks of difference. I assume it is impossible to increase the number of designs? Obviously including a wider range of furs, semées and bestiary would upset the relations otherwise, but if possible, wonders could be done to increase the variety. Even early medieaval heraldry wasn't half as boring as you make it to be, especially amongst the lower nobility and from the 13th Century (or so), with the rise of tourneys, people started caring about duplicate arms. One example that springs to mind is Scrope v. Grosvenor in 1389 when two English knights discovered they both used Azure a bend Or undifferenced. Obviously, complete avoidance of duplicate arms would be impossible, even within the same de jure kingdom, but increasing the pool of designs to choose from would make it much less obvious.
There are also some very strange choices in your collection, particularly the saltire engrailed, which I doubt is that likely, and perhaps a lack of interesting things to do with inescutcheons. There might also be a greater variety of ordinaries, in particular common variations on bend and fess and their respective divisions (wavy, embattled, indented ...) and complex charges / ordinaries (chequy and fusilly in particular had some commonality, especially in Central Europe).

[Solo]

Sadly no the engine only works with religion_groups so I can't even design an orthodox set or a scandinavian set. This is also why the set is designed to be an all around solution. As it is, it is biased towards western heraldry simply because the amount of known arms of all rolls feature mostly arms from the craddle of heraldry (northern France, southern England, western Germany and the lowlands). Of course french/norman heraldry seems out of place in eastern europe as much as eastern european designs would in the opposite case.

One way to work around this is to include as many historical dynasties as possible. And that's really where vanilla is utterly bad because it only features dynastic entries without arms or anything. When you have hundreds of dynasties with their correct emblems they will randomly pop everywhere in your realm and add that authentical feeling that's currently missing in most places, with canting arms, furs and local use of colours and designs (breton heraldry is a very good example).

This is really the only way to counter the lack of variety without having everyone and their grandmother using absurd designs. In any case furs can't be used properly in the engine, they just didn't think of it probably. So randomly created can't use them unless they draw from the third set which is horribly rare (because again of the engine that doesn't work properly).

It is really utterly boring or interesting depending on how you see things. But there's no way to make it work like it should because it's not repetitive at all, it's arranged in heraldic group which is completely different than just being repetitive. It often starts with imitation which is really the most important aspect of early heraldry, then splits according to local mode patterns. One good example is England and Denmark, which is not a coincidence. They form an heraldic group with Brunswick, Holstein and all other subcomponents which origin is the Poitiers dynasty and its lion. The lion becomes a leopard in Richard's arms then gets imitated by other rulers (like Otto of Brunswick) who then arrange it over time like Denmark (the semy) then become themselves the head of new heraldric groups because of vassals or neighbours imitating them in their turn. It's a ever repeating process that makes common designs pop everywhere in Europe.

Of course there are differences between periods and areas. Early french, and specifically norman heraldy is more than repetitive, it's the purest there is because it's almost only arranged by heraldic groups and nobody choose his arms without considering what its liege or cousin will think of it. It's where the brisure find its origin, as a way to distinguish relatives or allies (at first it's not only dynastic but also political and has absolutely no rule whatsoever) and at the same time still claiming your appartenance to a defined group of people using the same design.

Really cadet branches with optionnal predefined design possibilities (you simply code a pool of arms and names if any appeared) would have been a marvel but it's too much to expect from Paradox. Absent that, making sure the game has hundreds of choices to draw from is the best workaround I could come up with. It would help tremendously if they gave us the ability to restrict from using random design in any given culture but well I don't see that happening (and I'm really not the best person to make any suggestion seeing I have zero credit with the devs).

I can't really change the template anymore without discarding the thousands of dynasties already coded in the files (change the template and everything changes). Another problem is that I can only use this file to add more templates and if I do it will restrict even more the random occurences from other templates (the priority goes to the first template specified so with 240 it takes about 99% of all designs right now), completely removing them from the game probably. Otherwise I would have long made another template file to add more variants, it's just the game would not use them in more than 0.5% of all cases (and prevent using those in the multicolour template at the same time). I'm not sure if I explain clearly the problem but it's a design flaw that could easily be fixed by the devs adding a weight factor for each entry.

On a side note : the saltire and the cross engrailed or indented (it's really the same, as is the cross of losanges) are fairly common in french/norman/breton heraldry (by that I mean it's the most common variant of both designs).

[firelordzuko]

Thanks for the quick reply. Ah well, it works well enough, I suppose, and all in all the heraldry really is flavour only.

On a side note : the saltire and the cross engrailed or indented (it's really the same, as is the cross of losanges) are fairly common in french/norman/breton heraldry (by that I mean it's the most common variant of both designs).

Not quite the same, at least in English blazon -- though you might be right in French. A line engrailed consists of circular arcs meeting in points and is opposite to invected (so, like the bordure on the arms of Berry), while indented is a straight zig-zag line (Franconia, some Scottish quarterings). I have to admit I have never seen more than half a dozen saltires / crosses engrailed and none indented, most of them on Ermine (some variation on Ermine a saltire / cross gules surrounded by this and that, Lennox?), though I'll take your word for it. I have no idea what you mean by cross of losanges, though (I know the Cross of Lorraine, of course, which really seems out of place; and I know enough Italian to guess that losange is French for lozenge, but no further.) ...

[Solo]

Yes lozenge is the spelling in modern english sorry.

It is really the same, in primitive heraldry. There's no difference at first and progressively become three different designs : a bend engrailed, indented or a bend of lozenges is really at first only a difference in the artist interpretation of the design. Just like rampant and passant lions don't exist until the end of the XIIth century but become two separate designs from that point, before giving birth to many variants, like the leopards (which is not the original symbolic meaning since it has a negative image in medieval symbolism as opposed to the lion).

In the design I originally decided to stop at early heraldry (ends ca 1280-1300) because I couldn't imagine the aspect of late medieval heraldry (from ca 1300 to 1550), like the cimer and quartering, working in a 1066 without making things look utterly ridiculous. So there's an emphasis on the early aspects of emblematics, an unusual weight of proto and primitive heraldry and interpretations that come with itn compared to what would be the map of Europe in let's say 1400. I still had to mitigate things a little, still giving claws and tongues to lions and eagles because it's a game and there's no need to go overboard on accuracy either and it didn't hurt consistency in the first place.

So I (usually) respect the difference between engrailed, indented or a arrangement of lozenges but I'm not making a big case of it either because I know deep down that before the mid XIIIth century it would have been the artist's choice only and not an established tradition yet. This also explains why I sometimes arrange things a bit differently. There's a good recurring example to illustrate that, that I often see and have to correct because it's a complete misinterpretation : the eagle of Navarra. Much too often I read that the eagle facing right is a distinctive feature of the emblem which is a total misunderstanding of primitive heraldry in which lions or eagles can face both sides without any particular signification (later it would but many regular lions were facing the opposite side on the earliest seals). It's really no different than Richard's lions being arranged several different ways (including one facing right in 1189 iirc) or the pals of Barcelona varying in numbers from one seal to the other.

[firelordzuko]

I see. To be entirely accurate, having any sort of heraldry in 1066 is of course incorrect (just imagine the 867 start date, Charles the Bald bearing Azure semée-de-lys!), so it should at least be kept rather plain. I wouldn't call it primitive, of course, early heraldry in its simplicity is far more beautiful and refined than the paper heraldry of the 16th to 19th Centuries. Not for nought modern heraldry is based on the design principles of the earliest heralds (though the College of Arms seems to be fond of vaguely post-modernist aesthetics).
I was unaware that crests (cimiers) are a late mediaeval innovation. Surely, when arms moved from seals to shields and thus became ... well, arms, there were already primitive crests (of the crest-shield variety, see Frisia or Navarre). Plus tourneys and some degree of associated pageantry existed throughout the formative period of heraldry, which is also the main part of the game. (Not that crests could be depicted in the game engine, mind me.)

[Solo]

The early starting date will of course be something odd making many aspects of the game feel completely out of context. Worse than that, imagine the new pagan king of France after conquering it (surely this kind of things are meant to happen and I doubt Paradox will change their emblem system for TOG), using those traditionnal arms of France which are the most elaborate and innovative (in the 1180-1190's context) christian emblem of its time.

The first cimiers, which is the very first form of what is later named a coat of arms (the term itself is a translation/adaptation of a XVth c. french term), appeared in Germany right at the time when other important changes appear in heraldry (claws and tongues for animals in french/norman heraldry, apparition of quartering in spanish heraldry) during the last quarter of the XIIIth century. The addition of those changes justify the split between early and late medieval heraldry as different periods. I use the term primitive heraldry in the same academic way, following the definition of periods made by Pastoureau :

- Pre-heraldry is anything before heraldry and not specially linked to it. Greek or Roman emblems (as long as they're not just used as symbols of course) for example are pre-heraldic emblems.

- Proto-heraldry is anything attested before heraldry and directly linked to it (or retroactively interpreted as a previous proto-heraldic emblem most of the time). The bosonid cross (Provence-Toulouse-Pisa), the burgundian pals (Grandson-Nice-Faucigny-Barcelona), the lotharingian bends (Bourgogne-Lorraine-Champagne-Alsace) or the ottonian eagle are all good examples but there are many many others taking many different forms (the capetian blue, the dynastic surname of the dauphins, the coins of the counts of Saint-Pol, and even the three besants of Boulogne on the tapisserie de Bayeux).

- Primitive heraldry defines the period between the birth heraldry and the time when consistent rules and traditions begin to appear. So it's between 1133 and 1200. Emblems are usually very simple and bicolour. Blue is still completely excluded as it was not part of the roman code of colour used in emblematics and symbolism, in liturgy or on the battlefield. The number of pieces, their aspect or their orientation can vary a lot from one seal to another without having a specific meaning.

- Early heraldry is the following period (until ca 1300) when heraldry still isn't formatted or strongly codified, and remains in its purest form. Some traditions start to emerge and the style is more consistent (both very progressive, you can still find the old form of vair at the end of that period for example). It's also the time when heraldry move from an individual to a more dynastic based system (it's not a feudal system, you can choose to continue using the arms of the previous ruler if you feel it is convenient but nothing says a different ruler for England or France wouldn't have used his dynastic arms and replaced the old ones).

- Late medieval heraldry (until 1550) is more consistent as a period. Although it's less codified than people usually imagine but is progressively more and more restrictive than the earlier periods for sure. Gules was the dominant colour of primitive and early medieval heraldry, blue is the dominant one from now on. With the expansion (both social and geographical) of heraldry comes more complex and imaginative designs to the point that it's pretty easy to tell when (and often where) an emblem was designed simply based on its construction and colours. Heraldry is now a dynastic and feudal based system (rules enforcing traditions) and quartering start to replace both the tradition of merging dynastic arms and the creation of new heraldic groups. As a result at the end of that period heraldry is far less creative and much more codified,

[Ixor_Drakar]

I know the German dynasties are wip so I'll just point out House Wettin and Askanien have the same coa. I know the Askanien represents their hold over Anhalt and the Wettin would later have a similar coa based on controlling the electorate. I'm no expert so I can't suggest a better alternative for Wettin except perhaps one pertaining to their control of the March of Lusatia which doesn't exist in game.

[Solo]

The Askanier would be the one changing, the possible alternative being the arms they used for Aschersleben. The similarity between successive rulers of a same title was completely intentional, and even more frequent in imperial lands than anywhere else.

[ahog]

I can send you the flags I use for the danish provinces, if I knew how to upload files. Some of them are probably a bit anachronistic, i.e. from the 14th/15th century, but they're the CoA's of the main cities in each province.

I just pulled the pictures from the web and then into gimp, so I'm sure you can make some that are more fitting your style.

[Solo]

Actually I'm not missing much in Denmark. Vestervig/Viborg and the Eastern Scania county.

It's the rest of Scandinavia that is problematic since most of its heraldry was designed in the XVIth and XVIIth centuries.

[Cèsar de Quart]

Solo, I was in the Musée Languédocien in Montpellier this morning and I saw coinage with those strange arms of Montpellier-Aragon (Montpellier, au chef d'Aragon). In the ledge it said they were coined under James I. But anyway it reinforces the not-so-personal character of Catalan heraldry (Spanish in general), where people change arms depending on their estates, family and marriages, more than just lineage and inheritance (or taste).

It would not be strange to see James I using Montpellier au chef Aragon in Montpellier and Aragon full in other Aragonese territories. But still, they could be later, maybe his son James, king of Mallorca.

Should we change the arms of Mallorca into these? After all, the purple or blue bend was something made up in later rolls, never seen during the existance of the Kings of Mallorca themselves.

[Solo]

Remember I've already shown you those arms as painted in Pont-Saint-Esprit in the house of the Piolenc family, a while ago in the PS thread.

As a coincidence, I was actually wondering yesterday about Mallorca and the county/duchy as I'm now polishing things and trying to eliminate redudant arms I've sometimes given to different rankings for lack of a quick solution. There's no kingdom title (although we could make one really) so the duchy could use those arms, but I have no idea for the county itself if not the aragonese arms with the bend.

EDIT : Still have the picture on my imageshack :

[Ixor_Drakar]

While messing around in southern Italy I noticed Tarsia and Catanzaro have the same county COA.

[Cèsar de Quart]

Remember I've already shown you those arms as painted in Pont-Saint-Esprit in the house of the Piolenc family, a while ago in the PS thread.

As a coincidence, I was actually wondering yesterday about Mallorca and the county/duchy as I'm now polishing things and trying to eliminate redudant arms I've sometimes given to different rankings for lack of a quick solution. There's no kingdom title (although we could make one really) so the duchy could use those arms, but I have no idea for the county itself if not the aragonese arms with the bend.

EDIT : Still have the picture on my imageshack :

Yes, although the colours on these arms are a bit strange. I am not sure if the bezant for Montpellier is altered or if it's a mistake.

For the county, either use the Aragon with the bend, or use the arms of the city of Mallorca, the citadel of Ciutat de Mallorca (nowadays called Palma).

[Cèsar de Quart]

I am finding it increasingly difficult to play as anyone in Southern France while everything is in French. It feels utterly, unavoidably wrong.

We already discussed it and you're in the winning side. However, I'd like to be given the choice. As I said when we discussed it, as much as you're right, it still feels wrong.

The problem is that you're purging all trace of Occitan from the files, and I honestly don't have the time to go through them and re-occitanise the mod. Maybe I'd do it once, but I would not for a second patch, and so on and so forth. Dynasty names, onomastics, toponymia, it's just too many files to be something "I could easily change myself".

I'm not asking for you to back down. I'm just saying that if you could offer an Occitan pack of files, which is easier for you to do, since it's you who's changing them...

All those Aimerys and Guillaumes...

[Solo]

It's a huge work and it's not only occitan people.

The problem that is specific to France is that - when it comes to written documents, the only reliable testimony you can use - you can consider occitan and franco-provençal to be dead languages since the XVIth century at least (probably earlier, official documents in the XIVth century Viennois are all written in french already for example).

Now most cities have occitan names but contrary to other countries or areas where tradition can be traced back, just like you can find arabic names of almost all cities in Iberia without wondering if someone made them up in the last ten years; in France you have to be very very cautious about what you read.

- Many cities changed names during the early modern era, and even more at the revolution meaning half the cities in France don't have their medieval names but a modern more complex or completely different name. This is a process that still goes on and renaming still occured in the XXth century to accomodate a local politician or the postal services (thousands of renamed cities).

The reason the city have been given occitan names by city councils, is simply tourism. If you think for one second that politician will care about historian and historical accuracy let me stop you right there, they don't. The city of Montélimar choose not to keep its true occitan name for something that sounds more like the modern name of the city, as did another city not far that is also named countless time in medieval documents. In both case I know city councillors and advised them to keep the true name but they wanted something more "occitan sounding" so it is "Péré Lato" instead of "Peyrelada" and "Montéléimar" instead on "Montelh".

My point is, you basically can't use any of what you can find easily on the internet (wikipedia or not, it's not the problem and wikipedia works fine for plenty of things even in linguistics). What that means is a huge work to translate everything after looking in depth at the etimology of the toponym to trace back the changes that occured after our timeline. In the case of France they are often countless, some cities having been renamed a dozen times for various reasons : the lords of the place (Montfort-L'Amaury, La Roche-Guyon, La Roche-Bernard, Montélimar, etc), honouring a ducal or royal city (like renamed before, during and after the revolution like Bar-le-Duc, or Villeneuve-le-Roi), because it was a castle (Trie-le-Château) or an abbey (countless of something-l'Abbaye) or just because some new activity was established there like mining (L'Argentière) or even tourism (Alba-la-Romaine).

I know this is slightly off topic but this is my point : finding the occitan name for Alba-la-Romaine, Grenoble or l'Isle-Jourdain probably takes a couple minutes using wikipedia or the city website. But what kind of value does it have when the people from those cities in the middle ages called the cities Aps, Gratianopolis and Insula ? If it has to be done it must be done correctly and in that case it's a huge amount of work probably too much for a game as much of a perfectionist as I am.

There's a huge gap in occitan and franco-provençal due to the lack of written acts after the conquest which begins when the language is still changing a lot. In some case you can find the medieval endonyms but we're speaking hundreds of holdings to find with only a few documents to go through.

I've said I want to do a gascon culture which is the true separate subgroup (if it is even considered onen which is an ongoing debate) of occitan and the most resistant area to acculturation by far (which explains the success of the anglo-norman kings there). When I find the time to work on that I may do a huge adaptation of occitan and franco-provençal at the same time but I really can tell it will be pharaonic and will require a lot of help to back translate the french names without too much extrapolation (there will be half of it in some area, that's a sure thing).

[Solo]

One more problem is the bug I encountered with the county names and the map legend. If this isn't solved, the counties everywhere on the map will have to be permanently labeled whether they can change culture or not. That seriously threatens any serious work you can do to make the mod more realistic regarding culture changes.

[Cèsar de Quart]

Yes, I know. And I know it would be a pharaonic task (in which, as I said when we discussed it, I'm more than glad to help you, and that, since this is a game, a bit of plausible imagination would not be out of place), but I was advocating more for the Vanilla broken Occitan as a placeholder.

The exonym in counties, well... I don't think it's that important, except in places in which it makes little sense like Muslim Spain or the Crusader Levant (Tell-Bashir or Turbessel? Well, damn).

So why don't we make a compromise. If not all of them can be changed, then let's just change some of them. The ones that are more plausible to actually change.

- The Crusader Levant. (Levantine for Frankish/Norman/Occitan)
- Muslim Spain. (Andalusian for several Peninsular languages)
- Bohemia, Poland and Silesia. (Polish for German or Czech)

Maybe Greece and Anatolia (Greek-Turkish-"Crusader") and some parts of England (the Welsh Hen Ogledd corresponding to Cumbria and Northumbria. Gododdin, Rheged, etc).

***

I know you dislike compromises, but for the sake of fun, I think it's a good way out of this problem.

[Solo]

- About occitan names

I've modded in the occitan toponyms this week end but as expected it doesn't appear ingame since it's controled by the king of France and the system works by top liege only (another great PI idea ruined by poor implementation).

The character names are on the way, I need to do the women still (which is the most diverse list ever and very hard to translate from latin, french or english, complete opposite of english women given names for example). Of course the forms are only a possibility since there's an alternate spelling, or several ones, for almost all of those names depending on the area and the timeframe.

I'll make a specific gascon list at the same time although I don't know when we'll implement the culture ingame. On the other hand I've decided to let all francoprovençal use french modern forms. It's really too much work and there isn't the same kind of reliable sources I can use as a base list (like the song of the crusade or the chronicles of Montpellier). It'd just be a huge guesswork if I did.

- About 1.10 and TOG. A few very good news regarding emblems.

We can use culture arguments in the random emblem generator instead of just culture groups. Practically it means I can now add a specific orthodox pattern set for example. This should make things more authentic in the eastern roman empire at the very least. I could also cook something for certain specific heresies like the cathars who rejected the use of the traditionnal christian symbols and emblems (cross or keys).

We can also specify separate pagan (maybe more, I need to test, would be great to be able to do the same with muslims and tengri) emblems in landed_titles.txt by coding them. That means I can make pagan titles match the dynastic entries and have for example Lithuania use its traditionnal Gediminai emblem, Mecklenburg have its boar head or Pommerania its griffin. Restricting the system to religion when it's really more cultural than anything else was a mistake they did, but at least know there's a workaround that should do the trick.

[Cèsar de Quart]

- About occitan names

I've modded in the occitan toponyms this week end but as expected it doesn't appear ingame since it's controled by the king of France and the system works by top liege only (another great PI idea ruined by poor implementation).

Like the Vanilla random heraldry system. Great idea, good concep, lazy product.

The character names are on the way, I need to do the women still (which is the most diverse list ever and very hard to translate from latin, french or english, complete opposite of english women given names for example). Of course the forms are only a possibility since there's an alternate spelling, or several ones, for almost all of those names depending on the area and the timeframe.

Great! By the way, I have a list of Catalan names which include names from the early XIth Century, some as fancy as "Lloboslleó", translated Lupursoleo in Latin (sometimes spelled "Lloborselló"), Guadall, Rostany, Rampó, Mir/Miró, Berà, Isembard, Isemir and the "Ermen" family (Ermenter, Ermemir, Ermensal, Ermenguer, Ermerany, Ermenat...) and the "Sig" family (Sifré/Segifré, Sindreu, Sicard, Siumar, Sigemir, Segimon, Sigrau...), while leaving out some later Catalan names, especially Francis (Francesc or Ffrancesch, in old ortography, a name that no one bore until Saint Francis got famous) and some XVth Century Arthurian names like Lançalot, Galvany, Artús...

or a very common one that I see in some modded lists of Catalan names: Jordi. As common as it may be today and as popular as it got among the nobility in the XVIth Century, before it became the patron of chivalry, it was a slave's name, a name that belonged to Greek and Slavic slaves.

Some women's name get weirder. Saurina, Timbor, Jacina, Romea, Aurembiaix...

Another thing is the patronymics. I know that you can get into some "Joan Joan", "Pere Pere" or worse, "Ramon-Berenguer Berenguer-Ramon", but it feels like a waste not to use this system in what was an overwelmingly patronymical collection of onomastics. Almost everyone in Medieval Catalonia had a composite name that evolved from being the given name + the father's name, into fossilised double names. Every dynasty got their own fossilied double names: The Montcadas had tens of Guillem Ramons and Ramon Guillems, the House of Barcelona fathered a lot of Ramon Berenguers, the Rocabertins used Felip Dalmau and Bernat Hug, the Empúries issued quite some Ponç Hugs and Hug Ponç's... the Cardonas used Ramon Folch to such extent that Folch became fossilied as a part of their own "surname", the dynasty being known now as the Folc de Cardona.

The only dynasty that didn't use patronymics extensively was the House of Urgell. Because every single count of Urgell was called Ermengol up to Ermengol X in the mid XIIIth Century. In game we'd have Ermengol Ermengol. It's a price I'll pay to have Bernat Hug, son of Hug Bernat, son of Bernat Ramon, son of Ramon Borrell... but will you?

By the way, real life did have some weird things the game would portray correctly: Introducing the crazy Ferran Joan Ramon Folc de Cardona, son of Joan Ramon Folc III de Cardona, great-great-grandson of Ramon Folc IV de Cardona. Some serious patronymics here!

I'll post my list of names in a while.

I'll make a specific gascon list at the same time although I don't know when we'll implement the culture ingame. On the other hand I've decided to let all francoprovençal use french modern forms. It's really too much work and there isn't the same kind of reliable sources I can use as a base list (like the song of the crusade or the chronicles of Montpellier). It'd just be a huge guesswork if I did.

Pity, but fair enough... you're already making a good effort with the Occitan and Gascon names.

- About 1.10 and TOG. A few very good news regarding emblems.

We can use culture arguments in the random emblem generator instead of just culture groups. Practically it means I can now add a specific orthodox pattern set for example. This should make things more authentic in the eastern roman empire at the very least. I could also cook something for certain specific heresies like the cathars who rejected the use of the traditionnal christian symbols and emblems (cross or keys).

We can also specify separate pagan (maybe more, I need to test, would be great to be able to do the same with muslims and tengri) emblems in landed_titles.txt by coding them. That means I can make pagan titles match the dynastic entries and have for example Lithuania use its traditionnal Gediminai emblem, Mecklenburg have its boar head or Pommerania its griffin. Restricting the system to religion when it's really more cultural than anything else was a mistake they did, but at least know there's a workaround that should do the trick.

Orthodox patterns for Byzantium!! You're probably rejoicing and making some tests already.

[Solo]

I have a list of catalan names somewhere. I love making statistics and parsing genealogies but I'm less comfortable with the difference between catalan and occitan spelling. Basically in a list we want the common names, those you'll encounter at least 1% of the time if there's enough of them. That's because unlike previous Paradox titles, you can simulate frequencies and all entries have the same weight, so Jean will be as common as Gargamel is (thank god PI gave me such a good example that I can use it in almost any conversation lol).

Also and that's debatable now that there's the earlier start date (tbh we haven't debated whether the mod will focus only on 1066+ or not when it comes to flavour like emblematics and naming patterns) but it's better to focus on the 12th and 13th century in general. If a name becomes antiquated after 1050 there's no need to use it in the list, the engine already have a nice feature for those with the parent name frequencies, making them less and less common over time (if theres any trace of it in the history files). The same way you don't want to use anything specific to the period when cultures became less specific towards naming, usually during the 14th century and foreign names started to become more and more frequent (exactly your point with Francesc).

Saurina I believe could be a diminutive or an adaptation of Saure, a name you would encounter in northern Provence (Saure de Mévouillon, Saura de Mevolhon from the cadet branch of the Nice-Orange dynasty is one example). I wouldnt be surprised if it was the direct adaptation of the hebraic Sarah in occitan. Aurembiaix I remember encountering several times. Timbor remind me of Tiburgis (very common, in the same dynastic group as above) which spelling was Tibort in occitan vernacular.

I've used a ton of double names for occitan in the first list and kept the same in vernacular except for the couple francoprovençal or gascon ones like Bernat-Ato. The patronym glitch only happens if there's cultural conversion with the iberian group I think so it's a bit secondary (and there's a lot of strange things or names popping in that case anyway).

History has proven to capable to produce strange things on its own anyway, Aimar Azemar (Adhemarius Adhemarii) for example is the name of a viscount of Marseille and his close relatives were named Giraud (his brother, who married a Dragonette de Mondragon, daughter of Dragonet de Mondragon), Giraudet (son of the previous) and Giraudonet (son of the previous). Transmitting legacy through the names is one of the most distinctive traits of the medieval provençal culture.

There's one example that best illustrates that (I don't know if it's unique but never heard of anything similar anywhere else). The family is named Auger and lives in alpine Provence, near Gap. They inherited lands from two other families named Reynier (Raynier) and Oze (Osa) in the 12th century (iirc). The tradition they established for themselves is as follows : the first born would always be named Guillaume Auger (Guillem Auger) as he is the head of the family. The second son would be named Pierre Reynier. The third son would always be named Francon d'Oze. Only there's a trick, which is a nightmare for genealogists : if the said Guillaume Auger would come to die without heir, the second son would then assume his place and take the name Guillaume Auger in his turn (Francon d'Oze would also become Pierre Reynier at the same time). Of course that makes a lot of Guillaume Auger in the acts and it continued like that until the XVth century.

[Sponsored content]