Cymdeithas Hanes Teuluoedd Clwyd
Clwyd Family History Society
Clwyd FHS


Patronymic Paranoia
by Dafydd Hayes

(From Hel Achau, Issue 36, April 1992)


This article is based on a talk given at one of the Society's monthly meetings about the mysteries of the old Welsh system of personal names and I gave it the rather whimsical title of 'Patronymic Paranoia'. The Concise Oxford Dictionary defines paranoia as, amongst other things, 'mental derangement with delusions of grandeur' and an 'abnormal tendency to suspect and mistrust'. In the context of patronymics and Welsh names in general, both definitions are very apt. What I want to show you is how family names developed in Wales and how you can then work the process in reverse to get at your Welsh ancestors.

We are all so accustomed to having a family name, which we inherit at birth and which is, to all intents and purposes, stamped right through us just like Blackpool rock. I dislike the expression that so-and-so was born Joe Bloggs. I am not aware that babies these days are born with a name stamped on some part of their anatomy and Joe would have inherited the family name Bloggs probably from his father and was given the name Joe by his parents. It is only the female of the species who can easily get rid of the family name and that by becoming the property of another! I want you to forget the idea that surnames are something fixed and constant. All family names had to start somewhere and, if you have Welsh ancestry, you will quite likely be able to find just when that surname was formed. You are most unlikely to be able to do that with an English family name.

In case you are not acquainted with the Patronymic system of naming, I would suggest that you try a simple written exercise. The only requirement is that you know you own name and that of your father and grandfather. You should start by writing down your own name (use only one name, the one you are usually known by). To the right of that, leaving a space, write down your father's given name (again only one name). The men should now put 'ap' between the two names and women should put 'ferch', 'ap' meaning 'son of' and 'ferch' meaning 'daughter of' in this context. There we have the basis of the Patronymic system. Everybody was known as the son or daughter of their father. As you can imagine there could be many people with the same name. So what's new in Wales? Back to the exercise. Again to the right of what you have written, leaving a space, you should write your grandfather's given name, that is your father's father. This time both men and women fill the gap with 'ap'. Now you have a three-generation patronymic. So who needs a surname. To continue the process you simply add further fathers' names inserting 'ap' between each name.

That is all there is to patronymics! Well perhaps there is a bit more. You had your own name and all you needed for identification was your father's name plus, possibly, your grandfather's. There was no particular stigma attached to being illegitimate in Wales, at least not in Mediaeval times, and so you would not be unduly concerned if your patronymic was not that of your half brothers and sisters. The legal system in Wales before the coming of the Normans needed something more than your short patronymic. As the family unit, the tribe if you like, was responsible for the doings of every member of that family, it was important that you knew your ancestry – in extreme cases to nine generations. Generally five or seven were enough. Can you rattle off seven generations? Five? In Mediaeval Wales to know your pedigree was of great importance, affecting everything from the ownership of land to the payment of fines.

Let us take a look at an early Welsh pedigree, not in its original form, even if it was ever written down at the time, which is most unlikely. You can forget about Domesday Book and 1066 and all that. These pedigrees take off from that time and go back, if not to the year dot, then to something very close to it. This is where that first definition of paranoia comes in and we start suffering from delusions of grandeur. Anyone with an ounce of Celtic romanticism would like a pedigree going back to the Welsh Princes and beyond but, alas, for most of us it is very difficult to prove, even if we know in our heart of hearts that we are descended from them!

This example is taken from an 8 volume masterpiece called Welsh Genealogies A.D. 300-1400 by P.C. Bartrum who has made a life's work of collecting Welsh pedigrees from various sources, cross checking and indexing them. The second set, covering the period 1400-1500, is another 10 volumes. I have never met him but this man is my hero! Some of these later volumes have pedigrees extending into the mid-sixteenth century and it is just possible to connect through into the earliest parish registers. When you are reading the pedigree, notice particularly the sound of the names because, as you go back through the pedigree, many of them will be unfamiliar as personal names:

Grufudd ap Cynan ab Iago ab Idwal ap Meurig ab Idwal Foel ab Anarawd ap Rhodri Mawr ap Merfyn Frych ap Gwriad ab Elidir ap Sandde ab Alcwn ap Tegid ap Gwair ap Dwg ap Llywarch Hen ab Elidir Lydanwyn.

Eighteen generations working back from Gruffudd ap Cynan who died in 1137 to Elidir Lydanwyn who would have been around in about the year 500AD, not long after the Romans went home. You can take a piece out of any part of a pedigree as above and make a person's name so that Gruffudd ap Cynan ab Iago ab Idwal might be his formal name but the short patronymic for a man as famous as he was would be Gruffudd ap Cynan or just plain Gruff to his friends.

That has shown you one end of the story, back in the dim and distant past. What happened between then and now to leave us with the types of family names we have ended up with is what I want to show you next. Skipping 500 years or so of Welsh history after Gruffudd ap Cynan we come to a major upheaval which was to affect the customs of the Welsh people.

Tudor Tree Harri Tudur

As you can see from this short pedigree of the House of Tudor, had the family adopted English ways and settled on a fixed family name at a different time, England might have had the Goronwy or Meredith dynasty rather than the Tudors in the sixteenth century.

The conquest of Wales by Edward in the thirteenth century had had little effect on Welsh customs but, ironically, the rot set in when a Welshman became King of England. The nobility started to drift to London and, even if they did not actually remain there, they started to adopt English ways. Not least among these was taking a fixed surname. Imagine the ridicule in the Royal Court if your name was just a recitation of your pedigree. How could you possibly be anyone if you did not have a family tag? The changes did not happen overnight and the families reacted to English influence in different ways. Some took the family seat as the name, e.g. Mostyn, in north-east Wales. Others froze the patronymic and took the current second name, as the Tudors had done. We are talking here about the upper echelons of society and these families are well documented and of limited interest to most of us mere mortals. But, of course, what the nobility do today the common folk do tomorrow and gradually – very reluctantly in some areas – the patronymic system was dropped. It did not happen overnight and it faded away in a manner to cause the most confusion to family historians. Well, why should they make it easy? They didn't want to change their customs anyway! The usual question at this stage, and the one I cannot possibly answer, is when the change took place in a particular area or in a certain family. All I can say is that the higher up the heap and the nearer to English influence you were, the quicker you made yourself look less Welsh. On the other hand, if you had nothing to gain by change and were tucked away in the West, then there was no reason to take on foreign ways.

Even before the time of the Tudors, however, there had been a gradual change in the names that children were given. Most of the names seen in the old pedigrees were no longer used and names such as John, David, William and Edward were becoming the most popular, with profound effects later.

I now want to go on to show how these changes resulted in the sort of surnames that are predominantly Welsh. Let us look first of all at the formation of surnames in the male line. Take the patronymic Edward ap Griffith ap Thomas as an example of a name you could well come across in a parish register or legal document as late as the eighteenth century. The use of 'ap' between the names in a patronymic was often dispensed with, so that our Edward ap Griffith ap Thomas could just as well be known as Edward ap Griffith Thomas, Edward Griffith ap Thomas or even Edward Griffith Thomas. This last form, which is quite common, has been the downfall of many an uninitiated genealogist. To us it looks just like a present day name, that is, two Christian names and a surname. You, the initiated, know that it is nothing of the sort and still means Edward son of Griffith son of Thomas.

The next stage was for the patronymics to get shorter so that our Edward ap Griffith ap Thomas would become Edward ap Griffith. No problem for the genealogist there, but would you still be so sure that this was a patronymic if he dropped the 'ap' and became plain Edward Griffith? Do not forget that 'ap' means 'son of' so that, when it was dropped, there was still a strong urge to imply 'of' by putting the English possessive 's' on the end so that we could end up with our man calling himself Edward Griffiths. Now I defy you to recognise that as a patronymic!

Did I hear a sigh of relief? At last Griffiths is something that is recognisable as a surname. No such luck I am afraid. Our Edward's son might just as easily call himself Edwards as Griffiths! Are you perhaps beginning to see the reason for the second definition of paranoia being so apt? I think, though, that you should make the tendency to suspect and mistrust normal rather abnormal. Sooner or later one of these names would settle as a surname, with or without the final 's', giving most of the common surnames such as Roberts, Williams, Griffiths etc.. Thomas already had the 's' in place. Jones is a little harder to understand but, if you consider that John would be spoken as Siôn, the 'j' sound not existing in Welsh at the time, then the development to Shone, Shones and Jones is not too difficult to see. Davies probably came by a similar route with the colloquial form Dafydd giving rise to Davis and Davies.

Back to the patronymics. Strictly speaking 'ap' was used before a consonant. Before a vowel it should be 'ab' but, as the two sounds are quite similar, the rule was not strictly adhered to. The Welsh vowels are as the English but with 'w' and 'y' in addition. This short grammar lesson is by way of introduction to another method of forming a surname. For this example we can look at the patronymic John ab Evan. If we say this quickly it begins to sound like John Bevan and hey presto, another surname. There are many, many examples of this type of formation so that any surname beginning with a 'b' or 'p' is suspect, especially if by dropping the initial letter you end up with something that sounds like a Christian name.

Examples of b and p names are legion but some the most common are Badam, Bithel, Boliver, Bowen, Prandle, Prees, Price, Prichard and Probert. 'Ap' before H is a special case so that ap Harry becomes Parry, ap Henry becomes Penry and ap Hugh becomes Pugh, and so on.

You may have noticed from the examples given that the same patronymic can give rise to different surnames quite apart from minor spelling variations. Thus Henry can end up as the surname Henry, Penry, Harry, Parry, Harris, Harries and so on. Hugh can become Pugh or Hughes. Richard becomes Richards or Prichard and many, many more.

For the last main group of names we go back to a patronymic such as Griffith ap Griffith. In order to distinguish between the two Griffiths the son might well be called Griffith Fychan ap Griffith, that is, Griffith junior, son of Griffith. If this situation coincided with the freezing of the patronymic the children of Griffith Fychan could well take the surname Fychan, anglicised to Vaughan.

In a similar way, some attribute of the person could be used to differentiate between two people with similar patronymics. The most frequently used were the colours of llwyd (grey), gwyn (white) and coch (red) and it does not take much imagination to see how these developed into surnames Lloyd, Gwynne or Wynne and Gough. Llwyd and gwyn could also be used in the sense of holy or blessed when attached to a name.

The ladies, who have been more or less ignored up to now, come into their own to illustrate one more aspect of the patronymic system that may not have occurred to you. What is the wife to be called after marriage? Yes, of course, she keeps her maiden name. She does not cease to be the daughter of her father by getting married and no other arrangement of names makes sense. Therefore, when you come across a baptismal entry such as John son of William Jones and Mary Edwards it does not necessarily mean that John was illegitimate. If he was, it will generally say so. Very often the entry will be John son of William Jones and Mary Edwards his wife, which can cause confusion for those not up on their patronymics.

It is time now to start working the other way round because that is what you will be trying to do as you delve into your Welsh ancestry. You will be approaching the problem from the base of a fixed surname. For the first hundred years or so as you work back there will be no difference but, from the middle of the nineteenth century, you will have to tread carefully. Yes, you may have great difficulty even getting to the mid nineteenth century because of the common surnames but, with a combination of Civil Registration and Censuses and a good smattering of non-conformist records, you should make it back to 1841. Earlier than that, especially in West Wales, you must start becoming paranoid in the second sense of the word. Doubt anything that may be a patronymic in disguise, which is any surname that looks as though it may have been a Christian name at some time. If it is obviously an English word, then you can forget this problem and face the other one of finding where it came from to Wales!

Just a few figures, not so much to blind you with science as to serve as a word of warning. Ignore patronymics in your Welsh ancestry at your peril. I have used the computer to search through the post 1837 marriages in the North Wales Marriage Index to see who was using patronymics when they got married. Only the grooms have been studied to see where the father had what appeared to be a different surname from the groom. It is not so easy with the brides because their surnames would be different from their fathers' if they were widows, which would complicate the issue. This method does not include marriages where the father's name was of the form John Jones or William Williams etc., of which there are many, as it is impossible in that case to tell whether the son was using a patronymic or not, so that the numbers quoted are an underestimate. The Marriage Index covers the counties of North Wales only.

As you might expect from what I have already said, in the eastern counties, Flintshire, Denbighshire and Montgomeryshire, the number of patronymic marriages after 1837 is quite small, less than 1% between 1837 and 1900, although they are to be found as late as the 1880s even in those counties. However, when we look at the western counties that now make up Gwynedd and Anglesey, the story is very different. For all marriages between 1837 and 1900, Anglesey had 7%, Merioneth 8.5% and Caernarfonshire 13% of grooms using patronymics. If we look at the period up to 1850, the proportion is about double, so that in Caernarfonshire, about one in four grooms used the patronymic. I suspect, when we take account of the marriages which I could not be sure about and those where the bride was using a patronymic, somewhere between one third and one half of all marriages in that period would have had one or both partners using patronymics. Do not forget, I am talking about marriages in the time of the great-great grandparents of most of us.

Remember the open mind I mentioned at the beginning. For anyone coming new into family history, it may take a while to accept that the spelling of your surname may not be more than a couple of generations old. Now I am telling you that, in Wales, the surname itself may not be more than a couple of generations older than that. Do not let that put you off because, once you get back to the patronymics, it opens up a whole new vista.

To see how suspicious you have become see if you can identify the patronymics in the following extracts from the Llannefydd parish records:-


  1. Edward Jones of this parish son of John Roberts and Jane his wife aged 24 and Anne Vaughan spinster of this parish daughter of David Vaughan by Catherine his wife aged 22 were Married by Licence the 19th day of February.
  2. Edward Pirs of the Parish of St Asaph bachelor son of John Pirs by Mary his wife formerly Mary Peters aged 26 and Mary Jones spinster of this parish and daughter of Richd Jones by Dorothy his wife formerly Dorothy Jones aged 18 were Married the 9th day of July by Licence.
  3. Robert Williams of the parish of St Asaph bachelor son of John Williams by Jane his wife aged 43 formerly Mary Davies [that's what it says!] and Jane Jones spinster of this parish and daughtr of Richard Jones by Dorothy his wife formerly Dorothy Jones aged 20 were Married the 9th day of July by Licence.
  4. Richard Williams of this parish widower son of William Patrick by Jane his wife formerly Jane Williams aged 76 and Elizabeth Williams spinster of this parish and daughter of John Williams by Mary his wife formerly Mary .... aged 44 were Married by Banns the 10th day of Oct.
  5. Thomas Parry widower of this parish son of Henry Salsbury by Mary his wife aged 55 and Elizabeth Price widow of this parish were Married by Banns the 20th day of December aged 55 being the daughter of Rice Roberts by Jane his wife formerly Jane Williams
  6. Isaac Roberts bachelor of the parish of Llansannan son of Robert Roberts by Mary his wife formerly Mary .... aged 26 and Elizabeth Edwards of this parish daughter of John Edwards by Mary his wife formerly Mary Jones his wife aged 25 were Married by Licence 31st of December.

The date is 1803 and it shows a complete mixture of patronymic and family names being used at the same time, often differing between the bride and groom in the same marriage. What a pity all marriage registers don't give such a wealth of family history details as this one does.

The following hints may help those new to the idea of patronymics, when working back through the parish registers, to identify the all important change from fixed surname to patronymic but there is no substitute for experience and developing a 'nose' for it!


  1. In the early censuses look for siblings using different surnames. Some may have taken the father's first name as a patronymic.
  2. Similarly look for cases where all the children apparently have a surname other than their father's. If it is a patronymic the surname should correspond with his first name.
  3. In the parish registers look for the occurrence of 'ab', 'ap', 'ach', 'uch', 'vch' and 'vz' the last four being common forms of 'ferch'. This means you are in patronymic country but your family may not necessarily be using patronymics at that time.
  4. Look for instances of non-gentry having apparently more than one Christian name.
  5. Look for examples of the wife retaining her maiden name.
  6. Be on the lookout for aliases.
  7. When in patronymic company in the register then, when looking for a baptism, search for the name as a surname and as a patronymic. This may make elimination more difficult but can pay dividends if it identifies the change of system.
  8. When looking for a female burial, check maiden name as well as husband's name.
  9. Generally, when working back, once on patronymics then on patronymics you shall stay. The main exception to this is if there is an English ancestor lurking further back in which case you may well be back on a fixed surname. These are quite easily spotted!


This page was last revised 8th September 2012
Copyright © 2012 Clwyd Family History Society