Semantic Domains for the Old Testament available in English Bibles
In the New Kings James version Lowe-Nida Semantic Domains are available so I can create a visual filter for all the People and Places in the New Testament. Anderson-Forbes also has Semantic Domains for the Old Testament although these are not the same as Lowe-Nida nevertheless it would be nice if the NKJV Old Testament was tagged with Anderson-Forbes Semantic Domains for the Old Testament. As far as I know the Only Bible that has these is the Hebrew Test for Anderson-Forbes.
Another point is the Dictionary of Biblical Languages has Lowe-Nida domains in each Hebrew defination. Maybe this could be used so you could have Lowe-Nida domains for both the Old and the New Testament. Anyone else interested in this?
Comments
-
John Brumett said:
Anyone else interested in this?
As a matter of fact, yes! However, this is not as easily done as it looks like. The United Bible Societies are working on a Hebrew version of Louw-Nida called the Semantic Dictionary of Biblical Hebrew (see here: http://www.sdbh.org/). For an explanation of the project and the complexity of it see the following article: Towards a New Dictionary of Biblical Hebrew based on Semantic Domains by Reinier de Blois, available at http://www.academia.edu/2963593/Towards_a_New_Dictionary_of_Biblical_Hebrew_Based_on_Semantic_Domains. Unfortunately, this is a work still in progress. But I'd think Logos is aware of it and would want to include it as soon as it is released.
0 -
Our Bible Sense Lexicon is essentially this kind of database, and it was designed to work with both the OT and NT.
0 -
As a stop-gap, the afore-mentioned BSL includes the approximate LN# in the 'See also' section.
But it's very touch-and-go. I was looking at the NT's use of Isa 6:10 (Acts 28:27 Mat 13:15 alluded Mark 4:12). NRSV goes with 'mind' in Isaiah (even though LXX went with heart), but moves to heart in the NT quotes. The BSL points to LN#26 in Isaiah, but tightens it up to LN#26.3 in the NT.
Interestingly the semantic relationships are also assigned to God. So presumably, whatever meanings are assigned in the BSL/LN are also assignable to the divine.
"If myth is ideology in narrative form, then scholarship is myth with footnotes." B. Lincolm 1999.
0 -
Bob Pritchett said:
Our Bible Sense Lexicon is essentially this kind of database, and it was designed to work with both the OT and NT.
This is true. And I like to have one database that interacts with both, the OT and the NT at the same time. Yes, the BSL is similar to LN but without the semantic domains attached. Semantic domains come into play for those who do Bible translation, transferring meaning from one language to another. Now WordNet has a separated dataset with domain labels. These domains don't line up with LN, just like the BSL at times doesn't line up with LN. And there is a reason for that. WordNet uses English as a starting point. It's domains would reflect English thought, the way the English speaking world categorizes things etc. LN on the other hand looks at the Greek of the NT and its domains reflect Greek thought. And that's why in the article by Reinier de Blois above he makes mention, that the LN domains don't work for OT Hebrew. So he started all over again to come up with a Hebrew lexicon according to semantic domains.
So yes, for most of us the BLS works great. And I'm very grateful for that tool! But for those of us who work cross-culturally, we would still appreciate to include SDBH all the way to the reverse interlinear Bibles. And that might be a lot of work for a small but growing percentage of Logos users who work in different languages. Something to look into!
0 -
One more thing, dictionaries with semantic domains not only cover nouns (objects), verbs (events), and adjectives and adverbs (attributes). They also cover connectors and other markers that signal relationships. BSL does not cover those words just like WordNet doesn't. But that is one way I use LN a lot, to get a clue on how two clauses (propositions as they are called in semantics) are semantically related. There is a difference between grammatical relations which one can look up in the clause analysis tools and semantic relations which are not available in Logos, despite many efforts to include discourse tools into Logos. Semantic relations (also called inter-propositional relations or communication relations) signal how the flow of thought or an argument develops. They signal for example whether two propositions are sequential, are in a reason-result, means-purpose, generic-specific, grounds-exhortation etc. relationship. Now, a dictionary with semantic domains covering words that signal such relations do not cover every instance. Because a lot of times there is no overt signal, just juxtaposition. But it is a starting point.
0 -
Schumitinu said:
without the semantic domains attached
If you move up the hierarchy, you should have semantic domains if the BSL is coded correctly. That is one of the advances of the WordNet approach to a dictionary. By
are you referring to http://wndomains.fbk.eu/labels.html for natural language processing purposes?Schumitinu said:WordNet has a separated dataset with domain labels.
BSL although based on WordNet, Logos has attempted adapting it to Greek and Hebrew thought although they don't explicitly adjust for time. I agree that the SDBH should be included in Logos, but I also believe that where the BSL needs tweaking, we need to tell Logos so they can correct the problem.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
Schumitinu said:
BSL does not cover those words just like WordNet doesn't.
Actually one of the developers of the BSL has indicated he would enjoy doing this expansion. If we can build the demand, Logos may follow.[:)]
My particular bias: I'd like to see Logos tag for semantic relationships in addition to the grammatical relationships which are comme ci comme ça.
Orthodox Bishop Alfeyev: "To be a theologian means to have experience of a personal encounter with God through prayer and worship."; Orthodox proverb: "We know where the Church is, we do not know where it is not."
0 -
On semantic domains: we're hoping a future release of the Bible Sense Lexicon will include more on this. MJ is correct that some of that can be found in the hierarchy: "to understand ⇔ see" is a more specific kind of "to understand". But the broader notion of semantic/contextual domains provides a way to group things that are related "functionally" or "culturally" but aren't related hierarchically. The WordNet book describes this as "the tennis problem": there's no place in the WordNet hierarchy that really combines all the aspects of "net", "court", "racquet", "to serve", "to volley", "love", etc.
We're working on this by defining specific cultural practices and then using them as the "frame" for multiple senses in the BSL. So "cooking" as a cultural practice will include senses like "pot", "oven", "to boil", "to bake", etc.
0 -
On semantic relationships: we haven't included functional connectives in the BSL (prepositions, etc.) so far, because
- it's harder to clearly define their semantics
- we wanted to provide both a lexicon and an annotation of the text (note LN only does the first of these). But deciding which semantic relationship is appropriate in each context is both an enormous amount of work, and more subjective than deciding the lexical semantics of content terms.
We'd like to do it eventually, of course, but I'm not sure when (or if) that will happen.
0 -
Schumitinu said:
Yes, the BSL is similar to LN but without the semantic domains attached.
Just to reiterate what MJ has said and try to relate what we have done with the BSL to SDBH, we have semantic domains if you move up the hierarchy that are akin to what de Blois refers to as "lexical domains." These are the green links in SDBH entries as opposed to the red ones, which are referred to as "contextual domains." Part of de Blois's proposed improvements to Louw and Nida, as I understand him, is that Louw and Nida didn't really keep these two types of domains as distinct as they could have whereas he tries to.
As an aside, in SDBH there is still some overlap between these two kinds of domains, whether intended or not. For example, if you look at the entry for 'av - meaning "father," both the lexical and contextual domains are"kinship."
All that to say, there is a degree of overlap between what we've done with the BSL and what de Blois has done. Only we have not dealt with relationals, as you have noted, and have dealt primarily with what he calls lexical domains.
0 -
First, thank you for your interaction here. And thank you Logos team for your hard work. Things like the BSL are indeed amazing tools! It is good to learn more about it. And it is good to hear Logos' thoughts on "improving" it.
MJ. Smith said:are you referring to http://wndomains.fbk.eu/labels.html for natural language processing purposes?
Yes, I am. And no, they are not of much help to the translator.
Sean Boisen said:We're working on this by defining specific cultural practices and then using them as the "frame" for multiple senses in the BSL. So "cooking" as a cultural practice will include senses like "pot", "oven", "to boil", "to bake", etc.
You might want to consider the domains suggested here: http://www.semdom.org/v4/1 But maybe you are already aware of it. That list is used universally by linguists and would serve translators, if that's what you have in mind. But then again, I don't know whether you want to use a numbering system or just labels. The hierarchy suggested would give you a place to put all your examples above under one "frame". You had 5.2.1. Food preparation (cooking), then 5.2.1.1. Cooking methods (to boil, to bake) and 5.2.1.3. Cooking utensils (pot, oven). This goes further than the lexical hierarchy already in the BSL which would only differentiate between more general and more specific verbs/events (cooking - to boil, to bake)
Jeremy Thompson said:in SDBH there is still some overlap between these two kinds of domains, whether intended or not
There is nothing wrong with overlap. Of course they overlap some. Like in our cooking example, they overlap on a verb (event) side of things but don't on a noun (object) side of things. And that is what in the WordNet book is described as "the tennis problem". The lexical domains are limited. Despite the overlap I think de Boil wants to keep the two realms separate. And it looks like the two need to be kept separate in the BSL as well. At least that's what it sounds like to me if Sean talks about assigning "frames" to the "meanings". And that's why I have looked at the BSL and LN as two separate things, even though related.
0 -
Sean Boisen said:
On semantic relationships: we haven't included functional connectives in the BSL (prepositions, etc.) so far, because
- it's harder to clearly define their semantics
- we wanted to provide both a lexicon and an annotation of the text (note LN only does the first of these). But deciding which semantic relationship is appropriate in each context is both an enormous amount of work, and more subjective than deciding the lexical semantics of content terms.
For the NT there are some helps out like LN and the Exegetical Summary Series (consider also the Semantic Structure Analysis Series by SIL). But does anybody know of any resources that give help and insight on these relationships for the OT? This can be both, commentary on a case to case or verse to verse basis or a summary of devices/words and there discourse function (like chapter 2 'Connective Propositions' in Steven Runge's book "Discourse Grammar of the Greek New Testament").
Also, don't be afraid of annotations being subjective. They always are. I don't agree with some "meanings" of words that I look up in the reverse interlinear BSL section. And I sometimes don't agree with the "referent" annotations neither. But the same happens when I read a commentary. But I'm still happy to have these datasets and commentaries. And I take them as what they are, as interpretations of one or more person. We don't all have the same knowledge and understanding (and I'm certainly not a scholar) and sometimes we look at things differently. But you have some in-house scholars that are very capable of doing this, and even analyzing propositional relations beyond connector words. People like Steven Runge have spent uncountable hours interacting with the text on a discourse level and have already shared some of their insights. It's "just" a matter of taking it to the next level and apply it to the text. Yes, it is a lot of work, like many other things Logos has done and did not shy away from. Having somebody do it that has already an initial knowledge of the subject would certainly speed it up. And if it happens five years down the road, we will still take it!
0 -
Schumitinu said:
and even analyzing propositional relations beyond connector words. People like Steven Runge have spent uncountable hours interacting with the text on a discourse level and have already shared some of their insights. It's "just" a matter of taking it to the next level and apply it to the text. Yes, it is a lot of work, like many other things Logos has done and did not shy away from. Having somebody do it that has already an initial knowledge of the subject would certainly speed it up. And if it happens five years down the road, we will still take it!
Thank you Logos! The Propositional Bible Outlines Dataset is a big step forward in this direction! I hope we will see the same for the OT in the near future as well.
0 -
Schumitinu said:Schumitinu said:
and even analyzing propositional relations beyond connector words. People like Steven Runge have spent uncountable hours interacting with the text on a discourse level and have already shared some of their insights. It's "just" a matter of taking it to the next level and apply it to the text. Yes, it is a lot of work, like many other things Logos has done and did not shy away from. Having somebody do it that has already an initial knowledge of the subject would certainly speed it up. And if it happens five years down the road, we will still take it!
Thank you Logos! The Propositional Bible Outlines Dataset is a big step forward in this direction! I hope we will see the same for the OT in the near future as well.
You're welcome! I'm really excited about this feature, and appreciate all the hard work the Content Innovation team did for this.
More discussion about our future plans for Propositional Outlines is in this thread: https://community.logos.com/forums/t/92621.aspx .
0
