Cecil R. Humphery-Smith, AIH, FSA, FHS
The first statement to be made is that any understanding of Armory must be measured in the light of social and economic history. Most armorists, until recent years, have ignored this approach. To cite dramatic examples, simply to stress the point: Wagner misdescribes the Heralds' Roll for lack of palaeographical study, Denholm Young misuses the muster roll for Evesham for lack of reading heraldry with history. The text books ignore its importance almost entirely, and I was trained and graduated as a biochemist! That should provoke interest.
Study the feudal system in the context of seigneurial duties and obligations also imposed upon sons, and one can readily understand why, in representing one persona (the capital lord) only one coat of arms was required of these individuals and even the same arms on seals. Parage and toponymical relationships emphasise this point. The surviving analogy is in the heralds wearing their sovereign's arms on tabards.
Look at the small number of Chief Lords (around 200) who needed arms in (say) 1100, extrapolate to 1250 (calling that the beginnings of rolls of arms) and much becomes clearer.
Take a look at a quarterly coat used between 1250 and 1300 to find 18 different persons bearing it, Quarterly Or and Gules. Then there are 16 differenced with bends, labels, bordures, birds, mullets, escallops and variants upon the theme. Eight of the persons are known kin of blood. While other names occur to muddy interpretation, I suspect that some others are also related.
Perhaps it is necessary to distinguish pedantically between "cadency" and "differencing". By the latter I mean often substantial changes in the colours or design of a particular coat to distinguish one person from another who may be related by blood, feudal, or other ties or other associations; by the former I intend to refer to the distinctions between fraternal kin. Differencing followed a variety of methods in the Middle Ages.
a) change of tincture of field, or charge or both;
b) adding, excluding or changing an Ordinary;
c) adding, excluding or replacing a charge;
d) placing overall a label, bend, baston or border;
e) adding a canton to the dexter chief, at first plain metal, colour or fur, later bearing a single charge, but if illegitimacy attached, a canton of the mother's arms was usual.
It will be appreciated that in general it was not possible to ascertain automatically from a medieval shield the actual relationship of the bearer to the head of this family, but much more could be done to examine the rolls of arms for a solution to this important genealogical and historical problem.
Distinction has to be made between these and similar arms but also the surnames
derived from the estates they held or where they were born, consequently disguised the relationship. Conversely, different arms derived from the previous associations of the estates may have been borne by brothers or father and son, conversely suggesting no relationship.
We should also look at the use made of shields of arms at the end of the thirteenth century. There is some evidence that the more wealthy landowners might use more than one coat of arms. Examination of seals illustrates this and when compared with rolls of arms further matters for deeper study are revealed.
Of great importance is differencing by any clear system of cadency marks. They are, generally, single or sometimes multiple small charges added to the basic coat of arms and invariably of a significantly smaller size than ordinary charges.
The principal single charges of cadency difference or brisures found are label, crescent, mullet, martlet, annulet, fleur de lys, and rose. These were generally used within a single family and first appear in the last two decades of the fourteenth century in monumental heraldry in England, some forty years before they are admitted by the writers of text books. Occasionally they may be employed by others than family members , those enjoying feudal relationships, military or political allegiances.
The Label: This appears to derive from the costumes of the Middle Ages. The collar with tabs or points, often decorated, was fashionable among young gentry in the thirteenth century. It probably survives in the preachers' tabs worn, even until recent times, by certain clergy and university professors. The label is the earliest form to be used of a cadency mark in the strictest form. Normally there were three or five tabs on the label used on the paternal coat of arms by the eldest son, and this collar was discarded upon the death of the father.
The Crescent: It might be supposed that this mark derives from crusading times though there is no real evidence. Nevertheless it has two horns appropriate to a second son.
The Mullet is generally a star of five points and straight sides, artistic licence allowing for some variation in the regularity of angles. The rationale of this and other marks of cadency demands much further research in spite of the many papers that have dealt with the subject of such symbolism over the past two or three hundred years.
A deep study of British mediaeval rolls of arms provides the base for an account of the origins of English cadency and to the conclusions drawn here. To such sources must be added contemporary use of seals, geographical and genealogical interpretation.
Reference to the many textbooks which base their decision to reproduce opinions on the origins of cadency upon the school-manuals of the fourteenth and fifteenth centuries would be invidious. Those texts, in turn, depend upon such works as the Tractatus de Armis of Bado Aureo, the text of Bartolus Sassoferrato and similar fourteenth century writings. By the sixteenth century Gerard Leigh is stating categorically that the eldest son has a label of three points and - on the authority of someone called Honorius - one point for the father, the second for the mother and third for the son, and therefore five points for the grandson. Some writers attribute the origins of the English system to John Wrythe, Garter King of Arms about 1500 but something was obviously well established nearly a century and a half earlier. Upton has a parallel system to that of Johan Baptista Christyn with borders similar to the Scottish system formularised by Stodart. Gayre, in his well composed popular book on heraldic cadency, had not studied sources and Wrythe is too late for a system already established before Agincourt. Gayre does, however, make a good point, largely correct, that the English heralds knew that differencing was essential since each person must have his individual means of identification.
By the time of the later heralds' visitations we find John Bossewell listing the minor cadency differences: "note, that if there be any more than sixe brethren, the divise or signment of further difference, only appertaineth to ye King of Armes, especially when they visite severall Provinces: and not to the father of the children, to give them what difference he list, as some without authority doo alledge". We can however, compare the use of cadency difference marks by the family of Leventhorpe in the Great Cloister vault at Canterbury in two bays, the one about 1380 and second around 1410.
Upton and Christyn seem to be content that after the use of the label and the crescent as minor brisures almost any charge could then be used and it has been suggested that the French had invented the regular use of small charges in order to distinguish a seniority of cadency sometime in the fifteenth century but had abandoned them by his own time. This is arrant nonsense as rolls and seals on both sides of the channel show distinct cadency differencing in the thirteenth century.
They were certainly firmly established by the middle of the sixteenth century when Bossewell in 1572 has them accepted as a rule with the first son and heir to bear a label of three points, the second son a crescent, the third son a mullet, the fourth son a martlet, the fifth son an annulet, the sixth son a fleur-de-ls, the seventh son a rose, the eighth son a cross moline, the ninth son an octofoil (or possibly a primrose).
Sir George McKenzie suggests that uniformity should be very much studied in heraldry to avoid confusion and that our arms may be the more universally understood. He then goes on to argue against the use of minor brisures and was probably in favour of what became the Scottish system which, while complex, is far more precise to distinguish one member of the family from another. McKenzie was not alone, for John Gibbon writes "these differences above said will serve very well for the sons of Primary and Top branches of a Family; but when it comes to crescent upon crescent and upon that a mullet, etc., it becomes confusion and not distinction (as Guillim complains)".
Sir William Dugdale states "as these minute (brisures), they do not show the Time of the Descent; for we cannot know which of the crescent bearers are the uncle or nephew. And further, it is a very usual Matter for every new Riser at this day, that can find a Man of his Sirname that hath a Coat of Arms, presently to seal it by adding a Crescent or any other of these minor Differences which (states Dugdale) I seldom credit such kind of Differences nor Bearers unless it be by some other Testamony or Proof made manifest which cannot be counterfeited so well in the other Differences, except the Assumer should be thoroughly acquainted in the descent of him whose Line he seeks to intrude himself into" ....... "The so-called English system does not even effectively distinguish uncle from nephew and so even within one generation confusion of arms can be created and the arms cease to distinguish one armiger from another. It soon becomes seen as a family coat. Of course, this is only important when heraldry is being used for its proper purpose of identifying the individual in his acts, his deeds and particularly upon the field of battle, and in his tenure of land holdings the use of which he may signify by his seal. Quoting Sir Henry Spelman "a learned herald", Rideo igitur, et rejicio manutas istas icunculas quibus nec error defuit nec periculum" (Therefore I chuckle at and reject such minute figures in which there is no lack of error nor danger).
Once again the textbooks have misled us as to our origins and likely also as to actual practice since, even at the time of the heralds' visitations during the sixteenth and seventeenth centuries there is little evidence of the use of the textbook system in the differencing of arms, and the crescent becomes more common to distinguish one branch of a family from another - though, perhaps, as Dugdale suggests, simply because some newly rich family has become socially and territorially advanced and happens to have the same surname as one of the old armigerous gentry.
Fox-Davies made the rather silly suggestion that since most families of any count eventually married heiresses they could dispose of their minor cadency marks and pick up quarterings which for them would become sufficient means of differencing. It still raises the problem of the cadencies for children who themselves may need to use heraldry. Today, the assigning of a Grant of Arms by the English Kings contains the phrase that such armorial ensigns as are granted may be proper to be borne and used by him and his descendants with due and proper differences according to the Law of Arms. So the purpose is, as Guillim says that cadency differences are "extraordinary additaments whereby bearers of the same coat armour are distinguished from each other and the nearness to the original bearer is demonstrated."
Sir Anthony Wagner enlarges upon the point that in one stage in the development of heraldry it became the accepted belief that the plain ancestral coat came to be looked upon as belonging to the head of the house only while his near kin, brothers and younger sons and remoter kindred should difference or distinguish their arms often by fairly major changes. It can be shown that this was not entirely true in the Middle Ages.
We must be well aware of the interesting variations that are found in the coat of arms associated with the house of Mortimer, thoroughly studied by the late Commander A.W.B. Messenger and previously by Sidney Mason Collins. In more recent times Robert Norton has developed the work of Messenger. There are further notes by Tony Wilmott where he draws attention to Denholm Young's
conclusion that the St George's Roll may have been made for Roger Mortimer III by a seigneural herald. If it was then it was later recomposed as an ordered library copy as I have already noticed.
Based upon the medieval rolls of arms prior to 1320 we can demonstrate the introduction of these minor marks but then encounter problems of pedigree. In some rolls of arms father and son are mentioned together with such differences as the mullet, cinquefoil, label, martlet and annulet but the position of the son is rarely if ever mentioned. Argent two bars Gules and a label Azure for Nicholas FitzMartin (Glover's Roll (202), the label charged on each point with three bezants for William Martin in Collins Roll (134) and for another William Martin in Segar's (171) and Gwillim's (60) where there are five points to the label each charged with three bezants, show a beginning.
Whether there is any relationship between Morles, Anstabeter, Noliwers, Robert Melese, John Low, Roger Moles, William Aungebyn or Gilbert Brideshale none can tell. They may have been "brothers in arms", going to tournament, or more usually to war or Crusade with a shared identity and compact to care for each others domestic and feudal affairs if any one should be lost seriously injured or killed. Were they of one family, brothers in arms, relatives of different generations but recorded in the same listings or, more probably, members of the same military unit? Moeles is also shown in the rolls with a significant label. Clearly orthography plays a part in confusion. Moeles, Morles, Melese and Moles are probably identical. Territorial names disguise brethren as I have stated already. The paternal line of one family may bear the name of the major estate, but other estates within the domain may give their names to the sons who have been made responsible for them. Examination of instances in Kent have demonstrated this aspect very clearly.
I do want to make the point here, whatever may have been the later development of armory, that, at the outset, leaders bore arms for territorial or military identify. Their kin or principal followers bore similar or identical coats. Fathers and sons bore the same arms, only differenced when one needed to be distinguished by person or estate. Otherwise a son was representative of his father or grandfather and a difference was unnecessary. That is the general conclusion drawn from studying the rolls. Within the thirteenth century development of armory several knights might represent one military or territorial unit or otherwise be brethren in arms, as described above, and therefore bear identical or similar arms later adopted on an hereditary principal and then seen to be awkwardly confusing.
In the Parliamentary Roll of 1312 we find Giles Estlee bearing Argent a lion rampant Azure charged on the shoulder with a cinquefoil Gules and over all a label, but neither seals or rolls of arms record from whom he is being so significantly differenced except that in the same roll, Argent a lion rampant Gules charged on the shoulder with a cinquefoil Or appears for Nicholas Estlee (105) and in the Falkirk Roll (48) for Andrew Astlee we have the shoulder charged with a cinquefoil Argent. A similar coat where the lion is charged on the shoulder with a cinquefoil is surmounted by a bendlet Azure also in the Parliamentary Roll (829) for Richard of Echebaston who might well be an Astley using his territorial name, as might Robert Walkfare who has a mullet on the shoulder.
The Parliamentary Roll of about 1312 (CEMRA 44 (268)) has a fleur-de-lys on the shoulder of the lion of Giles Bruce or Breouse (346)) distinguishing him from another (160) and a similar mark is used in the house of Segrave (ANA II sub Lion).
Apart from the royal coats which have additional marks upon labels, the majority of other differences in the earliest rolls and seals seems to be by the use of the chief, the label, the bordure, and the bend with some use of the frank quarter or canton, much changing of colours and of charges around ordinaries. Ermine, a bend Gules is generally accepted for Apuldrefeld
(FW 290) but for Richard Barnacke, who may be related, there is a cinquefoil Argent which appears to be for difference in the upper chief in the second version of Collins Roll (609). I can find no occurrence of Gules a bend chequy in seals or rolls but Gules crusily Argent and a bend chequy Or and Azure a mullet Sable in the dexter chief appears in the Parliamentary Roll (567) for John Ornesby. Surely this is a mark of cadency or a difference similar to that used by De Vere to distinguish from Saye.
Azure a bend between six martlets Or and a label of five points Gules appears in St George's Roll of about 1285 (CEMRA 19 (266)) for Arnold Monteney. There are many references to the same family with changes of tincture to distinguish relationships but then there appears Azure, a bend between six martlets or a mullet Gules for John in the Parliamentary Roll (428) and in the First Dunstable Tournament Roll of 1308 (223) and with a mullet Vert on the bend for Thomas Monteney, Parliamentary Roll (1109). William de Tracy uses an Escallop in chief, (Parliamentary Roll 868 and Dunstable Tournament 225) on his arms of Or, two bends Gules. Whether there is any relationship between Thomas Mulsho and Thomas Kempton I cannot say, but one would expect that the two Roger Watevilles (Dunstable Tournament 100, Parliamentary Roll 425) and Robert and John (Dunstable Tournament 98, and Parliamentary Roll 423) are related quite closely since they were contemporaries.
Nearer to the system there is Michael Hartelawe (Parliamentary Roll 1011) William Hever (no doubt of the same family) (Dering Roll 35), Andrew Hartelaue (Parliamentary Roll 1012), Andrew Hartley (Dunstable Roll 46) and the contemporary arms of Ralph Gefoul in Parliamentary Roll (593).
The simple conclusion drawn from the evidence of the rolls is that while cadency differencing was clearly in use in the thirteenth century, the English system of cadency that was so forced upon the heraldic public during the seventeenth to nineteenth centuries was virtually unknown up to the 1320s and seems to have been largely the figment of the imagination of the textbook writers like so many causes of confusion in heraldry: was this part of a ploy to create a mystery or closed shop out of heraldry?
Studying the medieval seals (which will feature as a further item in the next Anglo-Norman Armory volume) in conjunction with records of landholding along with arms illustrated in the rolls the true origins of English cadency are best illustrated by pedigrees. Robert Glover's work in the British Museum Landsdown MS 872, fo.18-74, provides good examples. These sources show, as one would expect by comparison with armory in the rest of Europe, the clearer use of the changes of charges, changes of colours and application of labels, bends and borders or multiple charges, distinct from the minute brisures of modern English cadency.
Peter and Otis Grandison have Paly Argent and Gules, a bend Sable. Ralph and William change colours and add three eagles to the bend. Another Otis or Otho bore escallops on the bend. William Baron Grandison married the heiress of John Tregoz. Their first son was Peter who died in 1358. He bore the eagles. The second son Otho died in 1364. He bore escallops. John, the third son, Bishop of Exeter, died in 1369 and bore round buckles.
Baron Robert Marmion bore Vair a fess Gules and William his eldest son appears to have borne the same though perhaps the fess was Argent. The second son Robert of Tamworth made the red fess fretty Or and his youngest brother Geoffrey Marmion of Chakinden replaced the fess with three mascles Gules.
The D'Arcy coats based upon Argent three cinquefoils Gules descend from Norman D'Arcy of Nocton, Lincolnshire. His eldest son Philip bore the same coat. His second son John of Knath bore Azure crusily three cinquefoils Argent. The latter's first son bore the same and his second son added a bordure Or. Philip D'Arcy had three sons who bore arms and perhaps others who bore none at all. The eldest, grandson of Norman, also named Norman, bore the same arms as father and grandfather, even, apparently during grandfather's lifetime. The second son, John D'Arcy of Park bore Argent an escutcheon Sable within an orle of cinquefoils Gules.
The chronicler Froissart tells the story how during the French wars at the beginning of the fifteenth century Sir James d'Audeley, not the head of his house, brings his friends together along with his brother Sir Peter Audeley to witness his rewarding the loyalties of four of his esquires. The arms that he has composed for them are significant but clearly differ from elements of cadency. This can be seen from the clear examples of cadency development shown in the following shields taken from the rolls which I have catalogued in Anglo-Norman Armory Two from which they can be identified.
Though these are arranged in some order, I have not yet been able to construct the many possible pedigrees from thirteenth century sources which might help us to determine how those appearing in the rolls of arms and on seals were related so showing the basis of the development of the cadency system. What is clear is that any system was somewhat arbitrary. Even with the house of Nevil in which there was a clearer development what we see is the beginnings of the adoption of particular charges or marks to indicate cadency by the last quarter of the thirteenth century with a fuller standardisation developing during the fourteenth.
It is not until the evidence of the rolls of arms and the seals is seen in the light of genealogies properly constructed from the documentary evidence of land tenures and title that we see the system emerging. What I believe the evidence of heraldry thus linked to the genealogies suggests with very little doubt is that almost unexpected conclusion:- cadency differences were not used by the eldest son at all unless he had a separate identity in his own right as a territorial baron, held his own lands or military command. Then he might take a new coat of arms based upon his maternal or matrimonial inheritance or feudal allegiance. Otherwise, he represented his father in all things and used, if any, the same arms - just as a herald wears the tabard of arms of the sovereign. Sir Edward Coke in his Commentary upon Littleton (1628) recognised this "…the eldest (son) shall beare as a badge of his birthright, his father's armes without any differences for that as Littleton saith section 5 he is more worthy of blood but all the younger brethren shall give several differences". It is often suggested that this implies the use the evidence of seals and rolls after the father's demise but suggest otherwise. The younger sons are only armigerous and distinguished by cadency marks if they need to be for military purposes.
None reaches the octofoil and few even the annulet in the thirteenth century rolls. One might guess that there was a somewhat indiscriminate use of the marks of cadency; if the second and third sons were merchants or artisans unassociated with property or military life, perhaps the fourth bore a crescent for difference and the fifth one or more martlets, for example. Late into the fourteenth century older methods of differencing were still in use. - And as I once pointed out, at Hamsey in Sussex there are hatchments of the Bridger family in which the positioning of the claws of the three crabs around the chevron - upwards - downwards - to one side - to the other - in bend-opposing - combattant - is clearly a form of differencing within the family in the 17th century - don't get bitten by Cadency, it is a muddy subject.
is clearly not the same lecture that I presented in 1953, if any reader is old enough to remember.
Bartholus de Sassoferrato Tractatus de Insigniis et Armis (1358): "Heraldry in School Manuals of the Middle Ages" in The Coat of Arms, vol.VI (1960), I, pp.114-123, 163-169, 200-202; Evan John Jones Medieval Heraldry (1943) p.XXV;
Gerard Leigh The Accedence of Armorie (1562);
Sir Henry Spelman, Aspilogia, ed. Bysshe (1654) p.140; Nicholas Upton De Studio Militari (written ca 1441; first published 1486);
Johann Baptista Christyn Jurisprudentia Herioca;
Lt-Colonel Robert Gayre of Gayre and Nigg Heraldic Cadency (1961);
John Bossewell Workes of Armorie (1572) - the "Concord", fo.10;
Charles Boutell Heraldry Historical and Popular (1864);
Sir George McKenzie of Rosehaugh Heraldry in Scotland, Edinburgh 1680, p.72;
John Gibbon, Introductio ad Latinam Blasoniam (1682);
John Guillim, A Display of Heraldry (1679);
Sir William Dugdale, The Ancient Usage in Bearing of such Ensigns of Honour as are commonly called Arms (1682);
Arthur Charles Fox-Davies, A Complete Guide to Heraldry (1909);
Anthony Richard Wagner, (1946);
Commander A.W.B. Messenger Family History (1963) vol.I, No.5, pp.140-149;
Sidney Mason Collins The New England Historical and Genealogical Register, XCIX, p.271;
Robert Norton The Coat of Arms New Series, vol.4 (1980/1) pp.324, 325, 351, 379, 401-405 and vol.5 (1983) pp.190-192;
Tony Wilmott The Coat of Arms, New Series, vol.6, pp.107-112;
M. Denholm Young, History and Heraldry 1254-1380 (Oxford 1965) pp.90-95;
Anthony Richard Wagner A Catalogue of English Mediaeval Rolls of Arms (1950);
Cecil R. Humphery-Smith, Anglo Norman Armory (1972); Anglo-Norman Armory Two (1984);Some have commented on the choice of the appellation Anglo-Norman for this period. I agree that armorials which run into the early fourteenth century do not fit the dynastic frames set by pedantic historians, I explained my own ideas of the heraldic Anglo-Norman definition in the first volume of Anglo-Norman Armory.
Some have questioned the origins of Anglo-Norman Armory Two. I began transcribing original manuscript versions of the rolls of arms during the 1940's, joined by my wife two years before we were married in 1951 and afterwards by my late father who adopted my hobby on his retirement in 1955. With their help I produced the composite index and ordinary solely as a research aid conscious of the more exacting and precise work in progress for the New Papworth for which I was also working. The compilation made a deliberate if unqualified choice from the many sources that we had studied to produce as speedily as possible a simple working tool for research in the period. Many have already found it an invaluable guide. A third book in the series will comment in fuller detail.
Further related papers by myself have been, New Lights on Old Rolls in Communicaciones al XV Congreso internacional de las ciencas genealogica y heraldica (1982) pp.245-254; Thirteenth Century Cadency in Recueil du 11e Congrès International des sciences généalogique et héraldique (1972) pp.289-292; Family History 1, pp.140-149; Baronial Influences on Local Arms, XIIe Colloque internationale d'héraldique, Gröningen, 2001.
There is a great deal more to be said and I shall hope to put this on site before long.
ILLUSTRATED COPIES OF MY ARTICLES CAN BE OBTAINED AT £5.50 ($15.00) each, including postage. Payment can be made by secure means through www.ihgs.ac.uk/books. The title of the article should be given in full with reference "Heraldic A/c".
27 August 2003