Nederlandse Familienamenbank


Grinding one’s teeth. Linkage of surnames in the Database of Surnames in The Netherlands

by Leendert Brouwer
21st International Congress of Onomastic Sciences
Uppsala, August 19-24, 2002

Ladies and gentlemen,

Our database is available online on the Internet. It has been installed on the website of the Meertens Institute in Amsterdam, where the implementation takes place. You’ll notice the web-address on the cover of the handout. I will begin with a few words about the history of the Database, its purpose and progress, and I’ll dwell on some advantages of an electronic reference-work compared to words on paper. Especially the advantages of a hypertext structure will be brought to attention. We have used hyperlinks to produce relations between surnames, for instance. This makes it possible to hop from one name to another, so to say. It is just an onomastic issue, by the way. So, you don’t have to be afraid that you have to follow a computer-technical explanation. The title of this paper reveals that the extension of dimensions doesn’t make work easier. But hopefully the visitors of the website take advantage.

Once upon a time the Database of Surnames in the Netherlands was nothing but a nameless card-index box. After it had been digitalized around 1995 it grew into an information retrieval system, waiting to supply large-scale onomastic research. But in the year 2000 the database was submitted to the internet public as a reference-work in progress. Its purpose is to offer onomastic and genealogical information, the history of names and information about semantic and morphological features. Because people have to live with a name, we think they have a right to know more about it. And they want to know! We always have many visitors: between 1,000 and 1,500 every day.

Let’s have a look at the print-outs in your file. The first page is a print-out of the initial page of the site. It shows a text entry field and by entering a surname followed by a mouse-click a search will be generated. The result is information about that name we have collected and processed in a record of the database. This name-page - print-out number two - is divided in several sections. In the first part the meaning of the surname has been explicated. If your name is X, for example, you may learn that the name indicates that the first one with this name probably was an X, lived in the X, came from X, looked like an X, or had a father whose first name was X. The section below this is meant for historical name-records, quotations, summaries and biblio-graphical references. The name explication will be based on this information, although different viewpoints can be mentioned here as well. The next section provides links to pages and paragraphs of specific interest concerning this particular name as a name in a category of names. The name is qualified or classified here chiefly by its semantic characteristics and morphological components. Each term or marker will direct you to information about patronymic surnames, for example, or to information about parts of names such as the Frisian suffix -ma. Other features outlined here could be archaic spelling or adaptation of immigrant names. In a way this mechanism, this hyperlink structure, can be seen as the application of immediate linkage from an entry in a name-dictionary to a paragraph of the introduction. Also comparable to the concept of a name-dictionary is the application of a cross-reference system. The next lines of our name-page provides references to other names, but I’ll come to that later.

The section at the foot of the page shows the geographical distribution of the name in the Netherlands according to the census taken by the government in 1947. It’s presented here as a list of the Dutch provinces and the quantity of persons with that name who lived there at the time. On the next page you’ll notice a map presentation. In the future we’ll also process the quantity of names in the provincial towns, so we’ll be able to generate local and regional specifications on the map.

An advantage of electronic publication on the internet is that you can take the risk to publish work in progress, as we do. The database contains at the moment 27.000 entries of names. We certainly have included the most frequent names already, but I believe we are only halfway. We try to help impatient visitors of the database, who don’t find what they are looking for yet, by giving them the opportunity to e-mail a questionnaire to us. This way we are also able to collect information from our website visitors.

The Database of Surnames in the Netherlands actually serves as an electronic dictionary or encyclopedia, making use of several advantages of the medium. Spaceis a remarkable benefit, for instance. Printed dictionaries are limited by volume. In digital space we only have to deal with the willingness of the reader to scroll. We are able to give extensive quotations and a large amount of bibliographical references. Though, as editors we have to check the relevance, of course. Another advantage, as I have mentioned before, is the opportunity to link marked terms or names to information elsewhere. The available information is laid out in the constellation of the main database and its satellites, such as a bibliographical file. Also links to other online databases, such as the Database of First Names, were realized. It is obvious that we also have to take advantage of a linking system for cross-references between names. We’ve figured out a system with see and compare references borrowed from the thesaurus concept.

I’ll try to show you that by using this structure we accomplish higher realistic values compared to the limited possibilities of a printed dictionary. First we have to acknowledge that surnames are clustered in groups with mutual hierarchical relations. (Figure number three gives an impression of how these groups of names stick together, overlap and belong to larger groups.) Because of semantic and morphological similarities surnames are related to each other and we could think of several ways to define hierarchical relationships between names. But for the cross-reference system in the database we distinguish them mainly by the most practical criterion. And that is quantity. In a group of names that can be considered as variations, simply one name will be borne by the most persons. This is the main form, which we call the top-name. The other names fit as minor names in the hierarchical relationship. This may change due to demographic developments, and nowadays also due to surname-choices by parents, but the name corpus used for the database is static - at the moment. The hierarchical relationship between names expressed in the database has been defined by the situation of the year 1947, as I said before. In the future we’d like to compare this profile to recent figures.

The cross-reference system has been set up as follows. All minor names will have a see-reference to a top-name. As you’ll understand, through linkage you’ll be able to jump directly from a minor name to the top-name, which, by the way, may lead to another top-name and so on. On the top-name’s page relevant information has been given that could be shared by the whole group of names. But specific information about a secondary name may be relevant too for the knowledge of this group of names, so it will be useful to link the top-name by a compare-reference to several minor names. A list of variations has been given in a section of a top-name’s page. Now, there may be a dozen or more variations, each with their own detailed information. However, it doesn’t seem to be adequate to give so many linkage options here. So, instead we have developed a multi-staged linkage system with a limitation of three compare-references on each name-page. From top-name A you may find links to the names AB, AC and AD. These names are selected from the list of variations because they are next in line concerning quantity, or because they represent, as a secondary top-name so to say, a distinguishing collection of names with specific features, in its entirety belonging to the group of names represented by top-name A. A group of names can be split three ways from its top-name into its constituant parts. On the next stage, for example on the level of name AB, again three compare-references may lead to ABA, ABB and to ABC. And so on: the relationship constitution of a top-name and its variations should look like a beautiful christmas tree this way. But ..., in figure number four of the handout I tried to set up such a tree diagram, based on an actual constellation of names, and despite its decoration it looks more like a christmas tree that is alive and kicking. (And the whole figure looks like a tea-pot, or like a dinosaur ...)

When surnames do not just belong to one group of names the peaceful image of a tree will be disturbed. If we try to explicate surnames properly, actual linkage will tie many surnames also to names representing other groups with certain differences. Therefor the linkage system we’ve developed makes it possible to link a name to superior names by two see-references. The three optional compare-references do not only lead to minor names of the same group or category, but possibly also to names that have only partial similarities. In that case the criterion of quantity has been neglected. From name A you may go down to AB and AC, and sidewards to name XA, for instance. In fact through the name XA name A has been linked to name X. Name XA functions as a hinge joint between A and X. This connection from A to X may be irrelevant, but it makes it also possible to focus on aspects these names may have in common. Sometimes creative solutions have to be found when the origin of a name can only be guessed at. Sometimes we can only suggest a relationship with a big question mark. We’ll need more information about the history of this name to analyze it properly. Even then we may have to give several possibilities in an explication, which will tie them to many optional variations. Sometimes the interpreter of Dutch surnames has to be agitated. To lay out linkage in these situations is like unravelling a tangle.

To illustrate what surname reality looks like, I have figured out a kind of surname labyrinth which represents connections between names. (I use it as a place-mat, you’ll find it as the inside of the handout.) Some names are made up, some connections result from folk etymological suggestions, but still this rather holistic picture answers the description of the surname world closely.

You’ll find the name Kies there, spelled KIES, the name that figured also in the print-out and in the disturbed tree diagram. It has been integrated in this picture and I’ve chosen this name as an example of a complicated name. If we simply translate the name Kies according to the dictionary into English, we’ll have the name Molar, which name probably doesn’t exist. Probably the name Kies has nothing to do with the noun ‘kies’ either, but as long as we cannot cross out this connection to the Dutch lexicon, we’ll have to acknowledge it as a serious possibility. The noun ‘kies’ or molar may be reproduced as a nickname for a dentist, or for someone complaining about his tooth-ache all-day, which nickname subsequently may have been adopted as a surname by his descendants. This possibility connects the name Kies to the name Tant or Tooth in English, and subsequently to the surname Silvertand (Silvertooth). In the picture you have before you, I also made a connection to the name Knarsetand, which is an imaginary name, an imperative or a “Satzname” to use the German term here, for someone who grinds his teeth. Now, if the interpreter doesn’t have a hollow tooth to hide possibilities in, he has to continue to list the possibilities he can think of as an explication of the name Kies. They have to make sense of course, so he has to decide if certain possibilities are worth mentioning. Nevertheless, to the name Kies we can attribute several relevant connections. I won’t go into details concerning the meaning of Kies, but a persisting interpreter will be able to match this name, not only to dentists, but also to daintiness, the verb to choose, hallmarks, the Food and Drug Administration, gravel banks, fast-flowing rivers, villages all over Europe, the Hungarian Kiss, to several patronymics derived from Germanic or Christian names of course, and so on. I’d only like to illustrate the uncredible outgrowth of meaning here. (The christmas tree diagram, by the way, does not suggest that Kies is another derivation of the Christianname Christian. It only shows that possible variations of Kies can be possible derivations of Christian as well.) We do have some genealogical information about Kies that clears up a little, but in the meantime also increases the possibilities. It seems that we have at least two different families in the Netherlands with the name Kies. One of them is known from 15th to 17th century records in Holland and might have died out. The other family is of German origins, immigrated with the name Gies, with a G instead of a K. We have a link in the database to a website with the address where family ties have been sorted out.

The reason why I point at these complicated names and the way they are imbedded in the whole surname constellation, is that I think we’ve been given a generous medium to present this complexity fairly by means of applying a hyperlink structure in an electronic reference work. I deeply admire the makers of surname dictionaries published as books, especially the onomasticians gathered here today. Restricted as they are by space and paper they have to find practical solutions to adjust and arrange their onomastic information into an alphabetical A to Z structure, applying references and cross-references from names to names. Of course in an electronic dictionary this concept simply has been copied, taking advantage mainly of the increase of space. But I believe that taking advantage of a hyperlink reference-structure with relevant links from names to names answers better to the complicated onymic reality. This is an important surplus value. We’d like to express that according to several onomastic features a certain name fits into certain categories of names and that this particular name may be related to name A and also to name B, as well as name C, D and maybe even more names, as long as the highly interested reader will be able to follow and understand traces. We certainly won’t forget that we serve the reader. Although the surname reality is complex and unclear so often, transparency should be an onomastic achievement.

Background information