Doing a search for grammatical forms
-
- Textkit Neophyte
- Posts: 16
- Joined: Wed Jul 25, 2018 9:55 pm
Doing a search for grammatical forms
On Perseus it is easy to search for words and find passages, but I want to find examples for grammatical forms. For example I would like to see usages of future participles, but I couldn't find a way to collect examples other than searching every verb one by one. Does anyone know an easier way to do it?
-
- Textkit Enthusiast
- Posts: 593
- Joined: Fri Nov 09, 2007 1:35 am
Re: Doing a search for grammatical forms
I don't imagine how there could plausibly be one, but computers will keep finding ways of surprising us.
-
- Textkit Member
- Posts: 156
- Joined: Fri Apr 19, 2013 12:01 pm
Re: Doing a search for grammatical forms
You can do this in Diogenes by entering regular expressions and searching on the PHI corpus. There's no separate search mode for regex. Just enter the expression as you would a normal word, e.g. for future participles:
tur[aeiou] (with no spaces either side)
https://community.dur.ac.uk/p.j.heslin/ ... /index.php
tur[aeiou] (with no spaces either side)
https://community.dur.ac.uk/p.j.heslin/ ... /index.php
- opoudjis
- Textkit Member
- Posts: 116
- Joined: Tue Oct 03, 2017 2:54 am
Re: Doing a search for grammatical forms
The TLG (i.e. me) has implemented grammatical search, and it's not actually that big a deal, once you have lemmatised your corpus. Of course, disambiguation is going to be a big problem, and I was not comfortable with letting computers do massive disambiguation automatically. Can't show you a sample, since I no longer have access to it, but you can indeed do things like search for future participles (and be prepared for trouble if you're searching for a dative plural, which will be ambiguous with an indicative 3pl.)
- jeidsath
- Textkit Zealot
- Posts: 5332
- Joined: Mon Dec 30, 2013 2:42 pm
- Location: Γαλεήπολις, Οὐισκόνσιν
Re: Doing a search for grammatical forms
I was going to suggest that you could do this through the TLG's wildcard search, if you were willing to type in the form-endings.
Given that all of the words seem to be tagged in TLG (that's not real-time, I expect) there should be a way to search based on grammatical form, but I'm not aware of it being open to the enduser through an interface.
@opoudjis I often think that there are a lot of interesting things that could be done with database access to the TLG. Isn't there any sort of developer program for that?
Given that all of the words seem to be tagged in TLG (that's not real-time, I expect) there should be a way to search based on grammatical form, but I'm not aware of it being open to the enduser through an interface.
@opoudjis I often think that there are a lot of interesting things that could be done with database access to the TLG. Isn't there any sort of developer program for that?
“One might get one’s Greek from the very lips of Homer and Plato." "In which case they would certainly plough you for the Little-go. The German scholars have improved Greek so much.”
Joel Eidsath -- jeidsath@gmail.com
Joel Eidsath -- jeidsath@gmail.com
- opoudjis
- Textkit Member
- Posts: 116
- Joined: Tue Oct 03, 2017 2:54 am
Re: Doing a search for grammatical forms
I have indeed implemented search based on grammatical features (several years ago), based on offline lemmatisation, and also benefitting from some contextual disambiguation. I'm not in any position to confirm whether it has gone live (once I stopped working there, I lost even demo access), but if it has, it is an option alongside text search and lemma search.jeidsath wrote: ↑Fri Dec 07, 2018 1:10 am I was going to suggest that you could do this through the TLG's wildcard search, if you were willing to type in the form-endings.
Given that all of the words seem to be tagged in TLG (that's not real-time, I expect) there should be a way to search based on grammatical form, but I'm not aware of it being open to the enduser through an interface.
@opoudjis I often think that there are a lot of interesting things that could be done with database access to the TLG. Isn't there any sort of developer program for that?
The TLG is as closed source as closed source can be, and quite paranoid about any external access. An API is simply not going to happen under present management; ditto bringing external devs in.
- jeidsath
- Textkit Zealot
- Posts: 5332
- Joined: Mon Dec 30, 2013 2:42 pm
- Location: Γαλεήπολις, Οὐισκόνσιν
Re: Doing a search for grammatical forms
That's really unfortunate. Aren't they part of UC Irvine? Or are they private somehow?
Regardless, I suspect that they have a very fragile monopoly, given how much better OCR is going to get over the next few years.
“One might get one’s Greek from the very lips of Homer and Plato." "In which case they would certainly plough you for the Little-go. The German scholars have improved Greek so much.”
Joel Eidsath -- jeidsath@gmail.com
Joel Eidsath -- jeidsath@gmail.com
- opoudjis
- Textkit Member
- Posts: 116
- Joined: Tue Oct 03, 2017 2:54 am
Re: Doing a search for grammatical forms
Oh, Greg Crane is actively working to undermine that monopoly, and he's probably going to get there. But that's classical texts; the Byzantine and Early Modern texts are not in his scope.