Clauses and Particles
Posted 13 December 2011 - 01:37 AM
2. Since the clauses I am interested in are always dependent clauses, I thought it would be reasonable to place two dependent clauses within an independent clause in the construct window. Nope. No results. There are definitely cases where an independent clause contains two dependent clauses, so it makes no sense to me that it would return no results.
3. I am also trying to find instances where a particular word occurs as the first word in a clause. I see no way to specify this in the construct window. I also want to find all the places where it is not the first word in a clause.
4. Finally, I am having trouble searching for particles that are not part of phrases in the trees because they begin a clause. If I wanted to construct a search that specifies which particle begins the clause, how would I do so? Whenever I simply use LEX and place it at the beginning of the clause I get no results. This even happens with waw, which obviously occurs often at the beginning of a clause.
Posted 13 December 2011 - 03:43 AM
The particle כי may function in several different ways. Unless I am mistaken, the first step in determining how a particular כי is functioning is to determine whether the כי clause precedes or follows the main clause.
Is Accordance capable of running a search to determine where a כי clause precedes the main clause and where it follows the main clause?
Are there any other searches that would help in identifying the various ways כי may function?
Obviously, I could rely on the grammars, but I am looking at its uses in Qumran Hebrew, and the secondary sources aren't much help there.
Posted 13 December 2011 - 11:17 AM
Tracking information within consecutive clauses is a feature that it still being programmed.
Finding multi-use function words requires knowing the three basic functions/positions: conjunction (e.g., "and, or but"), clausal subordinator ("that, because"), and exclamative (e.g. "indeed!"). Not all function words are used with all three functions, of course. But the function word כי/KY certainly is. Here is how you would go about sorting them out.
1. Exclamative: the easiest — insert "EXCLAMATION" and "LEX = כי" into the same column and hit search.
2. Subordinator: easy — insert CLAUSE=DEPENDENT and then LEX = כי into the first column. The result in the current corpus is 1074 hits. Now, these may include a few cases of the Exclamative כי, so you could exclude these by inserting EXCLAMATIVE in the column and negating it with NOT. You could also exclude any (admittedly rare) cases of כי that appear in the middle of subordinate/dependent clause by using the PLACE = 1-2 feature.
3. Conjunction: easy — insert CLAUSE=INDEPENDENT and then LEX = כי into the first column. The hits will include conjunctive (non-subordinating) cases of כי *AND* Exclamative cases. To further refine, insert EXCLAMATIVE and negate it with NOT. The result will be only cases of כי that were not judged to be subordinators or exclamatives.
Finally, for your question about finding כי subordinate clauses that precede or follow the main clause, one way would be to search like this:
A. Insert PHRASE=PREDICATE. In the first column insert CLAUSE=DEPENDENT. Then in the second column insert PREDICATE. This forces a search for clauses in which a dependent clause (adjunct or complement, although you could specify which) precedes the predicate of the main clause. You could go further and insert LEX=כי in the first column of the dependent clause.
B. Insert PHRASE=PREDICATE. In the first column insert PREDICATE. Then in the second column insert CLAUSE=DEPENDENT. This forces a search for clauses in which a dependent clause (adjunct or complement, although you could specify which) follows the predicate of the main clause. Again, you could go further and insert LEX=כי in the first column of the dependent clause.
Note that I have uncovered simple tagging errors in a couple (just a couple!) of the hits returned in the searches I've described in this response. I have corrected them for the next release.
Thank you for asking syntax searching questions. Each one helps us to clarify how to use the modules and to find tiny tagging errors.
Ok, so this response wasn't so "brief", but I hope it's helpful.
Edited by Robert Holmstedt, 13 December 2011 - 12:56 PM.
Dept. of Near and Middle Eastern Civilizations
The University of Toronto
2 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users