Names in the Low Countries before 1150

Early-medieval personal names in The Netherlands and Flanders

Analysis 5: Composition rules

In his study on given and family names in Holland, Van der Schaar mentioned several principles and rules that were applied in the Middle Ages when composing Germanic names. He illustrated every rule with a few examples and het stated that the rules became less strict over time, but his study lacked a solid quantitative foundation. In an article by Ebeling the same rules appeared, plus a few extra, but again without a basis. (1)
On these pages it has been examined to what extent the theory was put into practice: first the rules regarding the second elements of Germanic names (the so called deuterothema), next the rules regarding first elements (protothema) and finally the principles for the composition of names. (2)

Second elements: gender difference

The second part of a man’s name is always a masculine word, and of a women’s name it’s a feminine word . (3)

The sharp gender difference in second elements weakened over time, according to Van der Schaar, but the only examples he gives to substantiate this were from the 13th to 15th century. In our compilation it would not be noticeable.
Still, the distinction between masculine and feminine deutherothema appeared not to be so strict as the rules required. There were 97 different roots that were used as second element. Of these 97, 58 were exclusively masculine and 15 exclusively feminine. The remainder, a quarter of all roots, occurred both among men and women. Part of those could be considered ‘predominantly’ masculine or feminine (based on a s score over 95%), but in 15 roots the distinction was unclear.
The least ‘gender sensitive’ roots were -laifa (survivor, son or daughter) and -kin (no meaning) which occurred in roughly the same proportion as would be expected on the basis of chance (that is 85:15). Other frequently occurring deuterothema that were not linked exclusively to one or the other gender, were -bera/berin (bear), -radi (council), -ka (no meaning), -viga (to battle, to defeat) and -var (guard, protection / truth).
Thus, there are a few things to be said against the statement that there was an absolute distinction between male and female name endings in the early Middle Ages. It has not even been considered that one in every seven Germanic names was singular, so a deutherothema was totally absent. Apparently, such a second, gender defining element was not really required.

Second elements: no weapons in women´s names

Names of weapons do not appear in women´s names in the second part . (4)

Only four real weapon names were found as deuterothema: -branda and -gairu and (rarely) -agil and -brord. These four were all exclusively masculine.
Three armour names were also exclusively masculine: -bord, -helma and -grima. In contrast, the elements -brunja and -lindi, both of which could mean ‘shield’, were exclusively feminine. It should be stressed that these two roots could also have other meanings. Elements that had to do with battle and war occurred both as male and as female deuterothema (-badu, -gundi, -harja, -hathu, -hildi, -sintha, -valha, -viga).
Thus, the rule as formulated by Van der Schaar was practiced indeed. It is uncertain whether this rule also applied to armour names; it certainly did not apply to roots referring to battle.

Second elements: not starting with a vowel

No word starting with a vowel in the second position. (5)

In many names the first consonant of the second part was omitted, such as in -(w)ald and -(w)olf. When such names are excluded, and only the official spellings of the roots are considered, it appears that deuterothema with a vowel were rare, but not completely absent: in 1% of the bipartite names (52 of 4996) the second element begun with a vowel. The vowel involved was usually an i, and otherwise an a. The prevalent vowel-endings were -ing and -in. The elements -agil and -ansi were rare as deutherothema; here, it must be added that the identification of these elements was sometimes questionable.
Deuterothema beginning with a vowel occurred in all centuries of the study period. It was not a phenomenon that developed over time.

First elements: all roots can be used

With regard to the second part, people were much more limited in their choice than for the first: a number of words was used as first, but never as second part . (6)

Several restrictions applied to second element of Germanic names but not to the first element. Although Van der Schaar does not say it literally, this implicates that every root could be used as a protothema.
Most roots by far (236 of 266) were indeed used as protothema, and for the 30 roots that were not, there was usually a good explanation: these were either short endings without a meaning (-ko, -za, etc.), or roots that occurred so infrequently that it could not be concluded that they were exclusive second elements.
Three roots that dit occur frequently, but never as first element, were -branda (sword), -ing (descendant) and -laifa (son or daughter). Förstemann mentioned some examples of Branda and Laifa as protothema, but these were all from other regions. In the Low Countries they were used exclusively as (male) second element. In Förstemann’s survey the element -ing was also an exclusive deuterothema. (7)
Van der Schaar’s suggestion, that in principle all roots can be used as protothema, seems to be largely corroborated.

The lacking of restrictions for first elements would also imply that each root can be used in both male and female names as protothema.
On the basis of the number of individuals registered, it would be expected that for each root the male:female ratio would be about 85:15. For 9 of the 10 most popular protothema, this was indeed the case (that is 80-90% male and 10-20% female). Hrothi was an exception: this root appeared predominantly among men and rarely among women (96% versus 4%).
Among the protothema that were somewhat less popular (57 roots that did occur at least 20 time each), nine were exclusively masculine (Aba, Arin, Asca, Franc, Hugu, Huni, Sahs, Un, Vaz), and five predominantly masculine, with a score over 95% (Ag, Agil, Balda, Ebur, Haimi). Exclusively feminine protothema did not occur, but there were five roots that could be considered predominantly feminine on the basis of a score higher than 50%, where 15% would be expected (Avi, Ermin, Haila, Id, Im).
The remaining 169 protothema (from a total of 236) occurred less than 20 times each, a frequency that was considered too low to pronounce upon the question whether they were neutral, masculine or feminine.
In conclusion, it can be said that the ‘gender neutrality’ was relative: 19 of the 67 most popular protothema (that is 28%) were not distributed equally between men and women.

First and second element: no rhyme

Alliteration within a name was avoided, and so was rhyming between the two elements. (8)

Alliteration within a name, or repetition of the first consonant of a root, occurred in 18 names with a total of 45 bearers (m/f) (nearly 1% of the persons with a bipartite Germanic name). (9) The definition of alliteration has been taken rather broadly here: in many cases, one could be in doubt.
The most widespread alliterating name was Rothardus, which was counted as alliteration because the first element was officially spelled Hrothi. From the 11th century onwards, the initial H was no longer used, so actually there was no more alliteration. The same applies to Hunruco (Huni-hroc) and Rotherius (Hrothi-harja). The opposite also occurred: Referic (Hraban-ricja) from the 12th century contained an alliteration but this would not have been so in the earliest centuries. Names like Helenardus (Helan-hardu), Illehere (Hildi-harja), Uuerinhold (Varin-vald), Wichaldo (Viga-vald) did formally contain an alliteration, but this had disappeared because the intial consonant had been dropped from one of the elements. Names beyond doubt were Likelini (Li*-lin), Liudalog (Leudi-lac/lec), Thiaddag (Theuda-daga), Verwinus (Var-vini) and Vualauuayn (=Walewin, Valha-vini).
Based in the considerations above, the number of 45 alliteration bearers could be reduced by about a half. This indicates that alliteration did occur, but that it was very rare, with a frequency of only 0,5%.
Shortened namest with obvious alliteration have been excluded her. There were quite a few such names, for instance Poppo, Dodo and Tetta.

Just like alliteration, end rhyme between the first and second element was rare (23 names and 58 persons, 1,2%). The most popular rhyme name was Werner, followed by Berenger and Berner. Again, the definition of rhyme has been taken broadly here: not everybody might consider a combination like Ger-har to be a rhyme. However, the number of dubious cases was less than in alliteration, so it can be concluded that end rhyme within Germanic names was rare, but not as rare as alliteration.

The frequency of alliteration showed an increasing trend over the centuries, although the scores never got very high (in the 12th century: 1,3%). End rhyme fluctuated around 1% throughout the study period.

Meaningful combinations

Originally, all combinations must certainly have been meaningful. (10)

According to Van der Schaar, the Germanic personal names consisted originally always of meaningful combinations of elements. However, the consciousness of the meaning of the names waned in the course of the centuries. Already in Carolingian times meaningless combinations appeared. The only example given by Van der Schaar is the name Wilhelm: 'the striving helmet’ . (11) However, 'helm' may also be a metaphor of the king or the leader. (12) In that sense, 'helm' in combination with 'Wil' does have a meaning, and it is even very appropriate that through the ages many monarchs bore the name Wilhelm.
A better example of a strange combination is the name Fredegonde (= peace-war), spouse of the Frankish king Chilperic I. (13)
Since it is impossible to draw a sharp distinction between meaningful and meaningless, no attempt has been made here to examine whether or not, over time, people paid less attention to the meaning of the name elements they employed.

1 - Van der Schaar (1953); Ebeling (1993).
2 - See the Table of roots for the numbers upon which the analyses have been based.
3 - Van der Schaar (1953, p. 41-44 and 192).
4 - Ibid., p. 41.
5 - Ebeling (1993, p. 18).
6 - Van der Schaar, 1953, p. 44.
7 -Förstemann (1900, p. 333, 956, 996).
8 - Van der Schaar, 1953, p. 44; see also Ebeling, 1993, p. 18).
9 - The 18 names were: Berberto (Bera/Berin-berhta), Goltgart (Goltha-gardi, the only women’s name in this list), Hegihardus (Hag/Hagan-hardu), Helenardus (Helan-hardu), Hunruco (Huni-hroc), Illehere (Hildi-harja), Likelini (Li*-lin), Liudalog (Leudi-lac/lec), Referic (Hraban-ricja), Rothardus (Hrothi-hardu, 19 persons), Rotherius (Hrothi-harja), Ruthericus (=Roric, Hrothi-ricja, 5 persons), Rutradus (Hrothi-radi), Thiaddag (Theuda-daga), Uuerinhold (Varin-vald), Verwinus (Var-vini), Vualauuayn (=Walewin, Valha-vini), Wichaldo (Viga-vald).
10 - Van der Schaar (1953, p. 44).
11 - Ibid. (p. 93).
12 - Heaney (1999, p. xxix and 16).
13 - Ebeling (1993, p. 17).


Material & Methods

Male names
Male names extended
Female names
Female names extended
Table of roots
Non-Germanic names
Single rooted names
Composition rules Namesakes
© Dr. Kees C. Nieuwenhuijsen
Last update: September 2012