Encoding the Etymological Dictionary of the Serbian Language

We came to the Lexical Data Masterclass in Berlin with a clear idea in our minds – to explore the possibilities of retrodigitizing the Etymological Dictionary of the Serbian Language, which is a project both of us are working on, together with other colleagues at the Etymological department of the Institute for the Serbian Language of SASA. Little did we know it would turn out to be such a challenging task, yet extremely exciting at the same time.

We came to Berlin hoping to begin our work on the retrodigizitation of the Etymological Dictionary of the Serbian Language by encoding several entries as test models. In this blog post, we will present one of them – безбатан. Even though it is a shorter entry, it has all the important structural elements – grammatical information, several attestations, a semantic definition, examples, complex etymological information etc.

We tagged the entire entry, but our main focus was the etymology section. As you may notice, the entry consists of three parts, and here we will present the second one, which contains the etymological explanation.

Let’s start by deciphering all the etymological data we have in this entry. All the different parts are marked with different colors.

First of all, there is a Turkish etymon with two phonetic variants – bezbaht and bedbaht. Secondly, there is the implicit etymology of the Turkish word – “персијског порекла” (”of Persian origin”). In the final part of the sentence, there is additional etymological information which contains two cross-references to other entries in the Dictionary. So, we are dealing with three different etymologies standing on two parallel etymological levels.

The question that arose was how to deal with such a complex information and how to tag it in the most precise way.

Before getting into that, we needed to add some changes to our schema in order to adjust it to the structure of the Dictionary and in order to make it as etymologically accurate as possible. Some of them were already accepted by the TEI, just not implemented yet.

The changes we introduced to the schema were following –

  1. Adding the attribute @type to the <etym> element,
  2. Allowing the nesting of one <etym> element within another one,
  3. Allowing the adding of the <gloss> element inside the <cit> element,
  4. Allowing the adding of the <lang> element inside the <cit> element,
  5. Allowing the adding of the <dictScrap> element inside the <etymFree> element.

Before continuing into the detailed representation of the etymological part, here is one of the two XML versions of the entire entry –

<entryFree><form type="lemma"><orth>безбатан</orth></form>
<re><form type="attestation"><orth>безба̏тан</orth></form>, <form type="ending"
><orth>-тна</orth></form>, <form type="ending"><orth>-тно</orth></form>
<gramGrp><pos>adj.</pos></gramGrp>
<sense><def>"худе среће, несрећан"</def>
<cit>
<bibl><placeName>Косово</placeName> (Елезовић I)</bibl>
</cit></sense></re>, <re><form type="attestation"><orth>безба̏ тник</orth></form>
<gramGrp><pos>m.</pos></gramGrp>
<cit>
<bibl><placeName>Косово</placeName> (Елезовић I)</bibl>
</cit></re>, <re><form type="attestation"><orth>безба̏ тњик</orth></form>
<cit>
<bibl><placeName>Васојевићи</placeName> (Боричић)</bibl>
</cit></re>, <re><form type="attestation"><orth>безба̏ тница</orth></form>
<gramGrp><pos>f.</pos></gramGrp></re>, <re><form type="attestation"><orth>безба̏
тнӣче</orth></form>
<gramGrp><pos>n.</pos></gramGrp>
<sense><def>"који, -а, -е нема среће"</def>
<cit>
<bibl><placeName>Косово</placeName> (Елезовић I)</bibl>
</cit>.</sense></re>
<etym type="borrowing">
<metamark>o</metamark>
<lbl>Од</lbl>
<cit type="etymon">
<lang value="tr">тур.</lang>
<lbl>сложенице</lbl>
<form xml:lang="tr" type="variant"><orth>bezbaht</orth></form>
<lbl>поред</lbl>
<form xml:lang="tr" type="variant"><orth>bedbaht</orth></form>
<gloss>"id."</gloss>
</cit>
<etym type="borrowing"><lang value="per">,перс. порекла</lang>
</etym></etym>
<etym type="folk_etymology"><desc>уз наслањање првог дела на</desc>
<xr><ref>без2</ref></xr>, <desc>за други</desc>
<xr><lbl>в.</lbl>
<ref>бахт</ref></xr> "срећа".</etym>
<dictScrap>Тур. bed- / bez- у вези са бет, в. Tietze 1:301-302.</dictScrap></entryFree>

Also, as the XML file is showing, within the <etym> element, and under the <cit type=“etymon“>, the etymon is tagged with both <form> and <orth> tags. The reason to do so was based on the fact that there are often cases in the Dictionary where we have grammatical information, such as gender, within the etymological information, which requires additional tagging with <gramGrp> and <pos>.

And now, let’s go back to the main subject – the etymology and the <etym> tag. During the Masterclass, ideas and solutions were flowing and changes were made daily, thanks to the help of other participants and lecturers, which we are extremely grateful for. Finally, we reached a solution. A two-part solution, to be more precise, where one is preferred to the other. Now we will present both of them and explain why we like one more than the other.

Let’s start with the “less loved child”, sort of speak. Just to remind ourselves once again, we’re dealing with complex etymological information with several different etymologies, existing on two parallel levels.

Solution number 1 was to have sequential <etym> elements, as it is shown in the picture below.

By doing so, we have three separate <etym> tags, one for the Turkish etymon, one for the Persian etymon of the Turkish word (implicit) and one for the additional etymological explanation. Each <etym> tag has its type, two @borrowing attributes, and one @folk_etymology.

When going through solution number 2, we will explain this is a “less loved child”.

Solution number 2 was to nest one <etym> element within the other one, as it is shown in the picture below.

So, why weren’t we happy with sequential <etym> elements? The problem with this one lies within the etymological accuracy. More precisely, these three etymologies aren’t just sequential, they are standing in two parallel levels.

What does this mean? Again, we have the Turkish etymon with two different variants – bezbaht and bedbaht, encoded as the first <etym type= “borrowing”>. The next part is changed – the implicit etymology of the Turkish word – “персијског порекла” (”of Persian origin”) is tagged as <etym type= “borrowing”> and nested within the first <etym>, because the Turkish word was borrowed from Persian. The third part of the etymological information isn’t changed and it remained as an independent <etym type= “folk_etymology”> element.

We believe this is a better solution because it represents the etymological processes explained in the Dictionary in a more precise way.

In conclusion, we are extremely happy with the progress we made during the Masterclass and we feel much more enthusiastic now when it comes to the digital future of the Etymological Dictionary of the Serbian Language.

What do we have to look forward to? First of all, fully encoding the printed volumes (three of them) and, simultaneously, working on defining the guidelines for our future work in the digital environment, thus eventually having a digitally-born dictionary edition.

We’d love to share our future progress with all of you!

Snežana Petrović and Ana Tešić


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.