Combine search keywords and morphological parameters in same search

Jonathan Rhein said:

verb-morph:VA verb-root:λεγω AND subject:Jesus AND addressee:Disciples

And for your specific example, does using agent / patient instead of subject / addressee get you close to what you are looking for?

agent:Jesus patient:Disciples verb-root:λεγω verb-morph:va

Jonathan Rhein said:

It would be great if FL could expand search functionality to allow all available search keywords (e.g. addressee, person, subject, recipient, agent etc. etc.) to be used in the same search if this is technically feasible?

To the best of my knowledge FL already provides for it - the same search terms are available in Bible, Book, and Morphology searches. Clause arguments are not.

The clause arguments have the following field names as per the Help document:

Verbum Help said:

Clause Search Fields

The following fields are available in a Clause Search (some are unique to certain resources):

• adverbial — An entity in an adverbial clause modifier
• adverbial-lemma — A lemma in an adverbial clause modifier
• adverbial-sense — The sense of an adverbial sense modifier
• adverbial-surface — An original-language word in an adverbial clause modifier
• agent — An entity acting as a semantic agent, the instigator of a verb; direct entities
• benefactive — An entity acting as a semantic benefactive
• case-frame — The arguments needed by the verb, not sensitive to null complements
• case-frame-brackets — The arguments needed by the verb, supports null complements (specified in brackets)
• comitative — An entity acting as a semantic comitative
• comment — An entity acting as a semantic comment
• comparative — An entity acting as a semantic comparative
• content — An entity acting as a semantic content
• cost — An entity acting as a semantic cost
• current — An entity acting as a semantic current
• event — An entity acting as a semantic event
• experiencer — An entity acting as a semantic experiencer
• goal — An entity acting as a semantic goal
• indirect-object — An entity acting as an indirect object of the clause
• indirect-object-lemma — A lemma in an indirect object of the clause
• indirect-object-sense — The word sense of an indirect objects of the clause
• indirect-object-surface — An original-language word in the text indicating the clause’s indirect object
• instrument — An entity acting as a semantic instrument
• location (OT only) — An entity indicating the location of a clause
• location-lemma (OT only) — A lemma in a location clause modifier
• location-lemma-sense — The word sense of the location of a clause
• locative — An entity acting as a semantic locative
• manner — An entity acting as a semantic manner
• object — An entity acting as a direct object of the clause
• object-lemma — A lemma in an direct object of the clause
• object-lemma-sense — The word sense of a direct object of the clause
• object-surface — An original-language word in the text indicating the clause’s direct object
• participant — An entity that is participant of a clause
• participant-lemma — A lemma in phrase that indicates a participant of a clause
• participant-sense — The word sense of a participant of a clause
• path — An entity acting as a semantic path
• patient — An entity acting as a semantic patient, the recipient of a verb action
• person — A person mentioned in the clause
• place — A place mentioned in the clause
• polarity — Whether the clause in normal (positive) or negated (negative)
• related — An entity mentioned in the clause but not falling into another semantic role; returns indirect entities
• related-lemma — The lemma of an entity mentioned in the clause but not falling into another semantic role
• related-sense — The word sense of an entity mentioned in the clause but not falling into another semantic role; returns indirect senses
• result — An entity acting as a semantic result
• resultant — An entity acting as a semantic resultant
• sense — The word sense of an entity, see Bible Sense Lexicon; direct sense
• source — An entity acting as a semantic source
• stimulus — An entity acting as a semantic stimulus
• subject — An entity acting as a subject of the clause
• subject-lemma — A lemma in the subject of the clause
• subject-sense — The word sense of the subject of the clause
• subject-surface — An original-language word in the text indicating the clause’s subject
• temporal — An entity acting as a semantic temporal
• theme — An entity acting as a semantic theme
• thing — A thing mentioned in the clause
• topic — An entity acting as a semantic topic
• topic-and-comment — An entity acting as a semantic topic and comment
• verb — The main verb of the clause—synonymous with verb-gloss and gives the same results
• verb-gloss — Translation-language text main verb of the clause
• verb-lemma — The lemma of the main verb of the clause
• verb-ln — The Louw-Nida semantic domain of the main verb of the clause (Greek only)
• verb-morph — The morphology of the main verb of the clause (Logos morphology, as in morph search)
• verb-root — The root form of the original-language text main verb of the clause
• verb-sense — The word sense of the main verb of the clause
• verb-surface — The original-language text main verb of the clause

Verbum Help (Bellingham, WA: Faithlife, 2018).

So yes, the grammatical and semantic roles are available only in the clause search.

Jonathan Rhein said:

But I might be the only one seeing some value in that.

No, you are not. I am very concerned about the limitations put on semantic roles in searches which matter more to me than the grammatical roles. I am also concerned about the misuse of morphology to answer grammatical and semantic questions. But I have gotten used to most forum members not caring much about precision, so the issue of abuse gets little attention.

Jonathan Rhein said:

Also maybe that is technically not feasible.

It is not simple as they draw data from different sources with different units of text. However, many queries can be done by running a morphology query and a clause query, saving to passage lists, then using the appropriate logical operators to "merge" the two passage lists. There will be an occasional false positive especially if a clause ends mid-verse, but generally it suffices.

Vote for change at Integrate Clause search terms | Faithlife

Because the clause search is temporarily disabled, I can't build an example. But from a prior post:

MJ. Smith said:

You should not expect there to be a set relationship between subject and agent. For example:

• Mary hit Tom with a ball.
• Tom was hit by a ball thrown by Mary.
• A ball, thrown by Mary, hit Tom.

The subjects are Mary, Tom, ball respectively ... but in all cases it was Mary who was the agent and poor Tom who got the bruise. Yes, there are discernible constraints within a given language as to possible grammatical roles for a semantic role. Unfortunately, I can't find a straightforward article on mapping semantic roles to grammatical functions.

This illustrates why finding the "subject" does not always tell you "who did it?" This is compounded by the assumption that the nominative always identifies the subject. But there are at least 9 uses of the nominative in Greek only one of which is "subject'. Therefore, using the nominative to find all the times Jesus healed someone is wrong for multiple reasons:

• the nominative may not be the subject of the sentence
• the subject may not be the agent of the sentence

Jonathan Rhein said:

But unlike you I am somewhat cautious when it comes to semantic keywords knowing that they are not determined by the original text itself but by someone interpreting the text and ascribing a certain meaning to it (which of course can be erroneous).

I consider morphological data, grammatical data, and semantic data to be equally defined by the text i.e. the rules for determining them are equally accurate/inaccurate.  All of them are rules induced from the text projected back onto the text.

excellent points Dave