Fixed, permanent surnames as we know them came about in England and France in the aftermath of the Norman Invasion. By the late 1200s or early 1300s, most English people were using them, though it took until the mid-1400s before they were widespread in the northern counties of England, and it really wasn’t until the late 1500s, when the Anglican church began requiring that baptisms be recorded in local parish record books with the baby’s “Christian name”, that fixed surnames essentially became mandatory.
From England and France, the fixed surname system gradually spread throughout continental Europe, with most of Catholic Europe adopting them when they, too, began requiring parish baptism records in the late 1500s and early 1600s.
Hence, non-Catholic and non-Anglican areas of Europe were most of the later adopters of fixed surnames: the Netherlands, Scandinavia, Russia, etc., leaving Iceland as the only part of Europe that does not use fixed surnames today.
Around the same time as fixed surnames began to appear, a competing system of formal hereditary patronymic names developed: John’s son William would be William Johnson, whose son’s name would be Williamson, and so on. These took different forms in different ethnic groups–the Dutch used -sen (“Petersen”), the Italians used -o (“Pietro”), the Russians used -vich (“Petrovich”), the Poles used -ski (“Piotrowski”), the Spanish used -ez (“Perez”), etc.
There was often more than one form of patronymic (for instance, some Dutch used -se instead of -sen) and there were often different forms for males and females (for instance, the daughter of Peter in Poland would be Piotrowska), but these were some of the more widespread forms.
Before either fixed surnames or hereditary patronymics, the older system was called bynames or epithets. These often look exactly like a fixed surname, often being an occupation, or a father’s name, or a local land feature, or a description of what someone looked like. But the difference was, they were not hereditary, at least not strictly.
So John the Smith might have a son named William on the Hill, and another named David the White, and another named George John’s son. They could be quasi-inherited in that John the Smith was likely to teach his trade to his son’s, and any son that stayed on the same manner as their father would often be taxed under the same byname.
Thus, John the Smith’s son George who lived nearby would be known as George the Smith, but the other eight sons who wound up moving to other manors where there were other Smith families had to give up the byname for something else. Likewise, a who son didn’t share the occupation yet still lived on the manor might end up adopting something else.
In any case, under the older byname system, it wasn’t strictly hereditary, so John the Smith’s sons could be named anything, even the sons who lived nearby engaged in the blacksmith trade.
It was also customary to name your older sons after their grandfathers and their father in some cultures, so in those cultures where patronymics were adopted, it limited the pool of both first and last names. For example, in Holland, there was a lot of Pieter Jansens and Jan Pietersens and Willem Andersens and Anders Willemsens and Pieter Pietersens and Jan Jansens, and so on.
This necessitated that bynames were kept around as a third name (technically speaking this is called a cognomen, which is exactly like a byname except that it’s the third name). So anytime two Jan Jansens had to be differentiated from each other, they’d still be known as Jan Jansen de Witt (“the White”) and Jan Jansen de Smid (“the Smith”) where the cognomen worked more like a middle name than a true last name, in that it wasn’t strictly hereditary and it was usually only mentioned in settings where there the two Jan Jansens might be confused (such as on tax assessment lists).
This may be a contributing factor to the demise of patronymics; at least in New York, shortly after the English took over from the Dutch, it became mandatory in the 1680s for the Dutch to drop their “Dutch name” in favor of a fixed surname style “English name” except for those “grandfathered in” (because they were grandfathers and weren’t having new kids and passing down the custom anymore, so they were allowed to keep their old form of the name).
But this took a long time: the Netherlands itself didn’t make permanent surnames mandatory until they were under occupation by Napoleon in 1811 and he ordered a civil registration; the change was even later in Sweden, with some older Swedes still using patronymics into the first decades of the 20th Century; and Iceland has famously never given them up, continuing to use a patronymic last name custom.
Everywhere else, the fixed surname custom eventually took over. As to why it took over, as David Hey writes in Family Names and Family History, it was mostly to benefit record-keeping, such as taxes and inheritance:
“Fashion no doubt played a large part in the spreading use of surnames but the process may also have been connected with the change from an oral culture to a written one in the manorial courts. From the 1230s onwards, and particularly after 1260, manorial courts throughout the land began to record property transactions on rolls.
Local deeds became much more numerous about the same time. The systematic recording was obviously necessary to a well-run estate, but it was also to the advantage of tenants to have surnames which would help to prove a right of inheritance.”
However, Hey tempers with some evidence that the introduction of written records didn’t kill off the byname system entirely, and it continued to varying degrees in various parts of Europe until church baptisms essentially made them mandatory.
If you wanted your child to be baptized, then they needed a baptismal name, which would include a family name that would be recorded in the parish record books generationally. So as Hey says, the signs point to surnames replacing bynames and patronymics because it made record-keeping easier, but it “is a puzzle for which we do not have a satisfactory answer”.
To get back to the question:
By this, we mean that surnames in English backgrounds are mostly related to professions (Miller, Archer, Cooper, Smith, etc.) but there are also other big groups such as colors (Brown, Green, White, etc.) and parental relationships (Johnson, Anderson, Thompson, etc.). Did all of these surname groups start happening at the same time?
Essentially, yes, because these were all originally different versions of bynames and while patronymic cultures gravitated toward the aforementioned system of parental relationships, even in non-patronymic systems, paternal surnames appeared right from the beginning of fixed surnames.
Again, this is thought to be due to record-keeping: a lord might have three Johns living on his manor, so one would be John the Smith, another would be John on the Hill, and another would be John Peter’s son, in order to differentiate between the three. The color names of Brown, White, Gray/Grey, Black, Read/Reid/Rudd were descriptions of hair colors, though the latter two could also be a description of skin complexion.
“White” was used instead of “Blond” because the word “blond” didn’t enter the English language until a bit later than fixed surnames did (though it does essentially appear in non-English/Germanic surnames, e.g., le Blanc). “Green” was a bit different because “John on the Green” lived near the village green, i.e. town square.
You may notice that English surnames dropped the prepositions and articles (of, the, at, etc.) while other languages did not (le, de, y, van, von). While there isn’t documentary evidence confirming why this happened, we do know when: they disappeared mostly between 1300 and 1450, during the Hundred Years War, and when English kings first started speaking English natively. So it’s thought it was intentional, as they were perceived as foreign since a lot of nobility at the time were using French forms such as “le” in their surnames.
2. The Oxford Dictionary of Family Names in Britain and Ireland ed. by Patrick Hanks, Richard Coates, et. al, 2016, especially the section of the introduction entitled “Origins and Development of Family Names in England”
3. Family Names and Family History by David Hey, 2006
4. Christian Names in Local and Family History by George Redmonds, 2004
5. A History of British Surnames by Richard McKinley, 1990
6. A Dictionary of English Surnames by P.H. Reaney and R.M. Wilson, 1970
7. “What’s in a Name?: Family Surnames and Social Upheaval in Medieval England” by William F. Lanahan, 2015
8. Homes of Family Names in Great Britain by H.B. Guppy, 1890
9. English Surnames: Essays on Family Nomenclature by Mark Anthony Lower, 1842
10. The History of the Norman Conquest of England by Edward Augustus Freeman, 1876