find all third declension nouns/adjectives

[A. Smith]

I can't figure out how to construct a search to find third declension nouns/adjectives. I think something like *@ [ADJECTIVE] <OR> [NOUN] <AND> but then i get stuck. How do I look for words ending in a consonant? I think I've seen something like this before but can't remember it. Any help would be appreciated. Thanks

[Rod Decker]

I can't figure out how to construct a search to find third declension nouns/adjectives. I think something like *@ [ADJECTIVE] <OR> [NOUN] <AND> but then i get stuck. How do I look for words ending in a consonant? I think I've seen something like this before but can't remember it. Any help would be appreciated. Thanks

 

Third declension would be challenge due to the wide diversity of endings. In principle, here's one way to do it for 2d declen nouns:

 

[noun]@*ος

 

Then use the Analysis window to see a list of the words. (See below for results from this search.)

 

If you checked a morphology text (e.g., Mounce's MBG) and compiled a list of all possible 3d declen endings, you could eventually get the same info. (But then if you check MBG, you'd already have such a list and not need to do the search!)

 

For 2d declen, here's what you'd get, truncated to show only those occuring 100 x or more in the NT:

 

Total number of clauses = 7300

(total number of verses displayed = 6263)

 

[noun]@*ος (10077 total words)

 

Number of different forms = 585:

 

θεός god, God = 1317

κύριος_2 (κύριος) (n) lord, master, the Lord = 717

ἄνθρωπος man, human = 550

Χριστός (χρίω) Christ = 529

υἱός son = 377

ἀδελφός brother; (adj) brotherly, twin = 343

λόγος (λέγω) word, speech, message; book, volume = 330

οὐρανός heaven = 273

νόμος law, principle = 194

κόσμος world, mankind, earth; adornment, decoration = 186

ἄγγελος messenger, angel = 175

ὄχλος crowd; army; people = 175

ἔθνος nation, people, Gentile = 162

Παῦλος Paul; Paulus = 158

Πέτρος (πέτρα) Peter = 156

λαός people = 142

δοῦλος slave; slavish, servile = 124

θάνατος (θνῄσκω) death = 120

οἶκος house = 114

ὁδός road, way = 101

ὀφθαλμός (ὁράω) eye = 100

...

[A. Smith]

Thanks Dr Decker. This is about what I expected. I know MBG would gave the answer, but I was really looking to see if such a search was possible. Thanks for your input.

[Ruben Gomez]

Another approach would be using root searching. Thus, [noun]@+*ξ, [noun]@+*ρ, etc. would return nouns with roots ending in ξ, ρ, and so on. And you can use the <OR> command to join as many different root endings as you like (e.g., [noun]@+*ξ <OR> [noun]@+*ρ).

 

Hope this helps.

[A. Smith]

Another approach would be using root searching. Thus, [noun]@+*ξ, [noun]@+*ρ, etc. would return nouns with roots ending in ξ, ρ, and so on. And you can use the <OR> command to join as many different root endings as you like (e.g., [noun]@+*ξ <OR> [noun]@+*ρ).

 

Hope this helps.

 

Ruben,

Thank you so much. This is exactly what I was looking for. I did not know that +*κ etc would find roots ending in the given letter. I knew about the ? For a character wild card and that * was a general wild card, but did not think that attaching a letter to the end would find such roots. Thanks again. OH! Looking again, this is exactly what Rod said too, (re * plus a letter) but I missed it! Much thanks to both of you!

Edited December 3, 2010 by A. Smith

[DarinL]

It would be really great if there were a direct way to limit searches by declension. This would be very helpful for Greek teachers who are looking for specific examples. The closest I've come so far is to use a Greek Construct search, adding the following to the first column:

COUNT 50–10000 (to get only common words, since my purpose is to find examples for beginning students),

NOUN -proper (to eliminate common words that often don't inflect fully)

LEX =*ος, *ός, *ον, *α, *η, *ης, κύριος_2

Then I drag the NOT over the LEX element.

 

The *ος, *ός, and *ον are intended to eliminate 2nd declension

The *α is intended to eliminate many 1st declension, but also eliminates all the -μα 3rd declensions

The *η and *ης eliminate most of the rest of the 1st declension

κύριος_2 eliminates this one anomaly.

 

I'm sure I've missed a lot, but perhaps this is the beginning of a reasonable work-around??

[Ken Simpson]

Just a couple of off-the-cuff suggestions...

[COUNT 50+] rather than the more clumsy and arbitrary [COUNT 50-10000] though of course that does work

And if you removed the *a one and replaced it with *?(zstxeir)a should return your ma nouns but remove most of the 1st declension pure and doubled stem nouns.

Will give it a little more thought.

Thanks for the challenge though!

 

When I did this, I turned off exact lexeme (that dealt with the accenting issues) and included my wildcard string.

 

This then gave me this list

 

αἷμα blood; bloodshed = 97
  αἰών age, eternity = 122
  ἀνήρ man, husband = 216
  ἀρχιερεύς (ἄρχω, ἱερός) high priest; pontifex = 122
  βασιλεύς king = 115
  γραμματεύς (γράφω) scribe = 63
  δύναμις (δύναμαι) power, strength, capability; authority; (military) force = 119
  ἐλπίς hope = 53
  θέλημα (θέλω) will, desire = 62
  μήτηρ mother = 83
  νύξ night = 61
  ὄνομα name = 230
  πατήρ father, Father = 413
  πίστις (πείθω) faith, belief, trust; value; proof = 243
  πνεῦμα (πνέω) wind; breath; spirit, Spirit = 379
  πόλις city, town; city-state = 162
  πούς foot = 93
  πῦρ fire = 71
  ῥῆμα word, thing = 68
  σάρξ flesh, meat, body, sinful nature = 147
  στόμα mouth = 78
  σῶμα body; slave = 142
  ὕδωρ water = 76
  φῶς (φαίνω) light = 73
  χάρις grace, favor; gratitude; gift = 155
  χείρ hand = 177

 

 

 

Seems to be ok to me :-)

 

 

ADDIT:

 

One more note on this, when I did the "less than 50" nouns, I saw a number of really significant nouns you may wish to include. They would all be caught by using a 20+ count search.

 

the 20-50 nouns are

 

  κρίσις (κρίνω) judgment, decision, legal case; meaning = 47
  θλῖψις (θλίβω) trouble, tribulation, oppression = 45
  σπέρμα (σπείρω) seed, offspring = 43
  ἀνάστασις (ἀνά, ἵστημι) standing up, rising, insurrection, resurrection = 42
  ἄρχων (ἄρχω) ruler = 37
  οὖς ear = 36
  μάρτυς witness; martyr = 35
  ἱερεύς (ἱερός) priest = 31
  συνείδησις (σύν, εἶδος) conscience = 30
  γνῶσις (γινώσκω) knowledge = 29
  παράκλησις (παρά, καλέω) encouragement = 29
  πάσχα (πάσχω) passover; passover lamb = 29
  θυγάτηρ daughter = 28
  κρίμα (κρίνω) judgment, decree, decision = 27
  ἀστήρ star = 24
  νοῦς mind, thought = 24
  παῖς child; slave = 24
  σωτήρ (σῴζω) savior, Savior = 24
  ἀμπελών (ἄμπελος) vineyard = 23
  εἰκών image = 23
  χιλιάς (χίλιοι) group of a thousand = 23
  γονεύς (γίνομαι) parent = 20
  ἐπίγνωσις (ἐπί, γινώσκω) knowledge = 20
  ἡγεμών (ἄγω) governor, leader, chief = 20
  ἰχθύς fish = 20

 

That also gives you a bigger list to play with. The previous list felt a bit small to be terribly useful (to me).

 

This has been fun thinking through. Thanks for the good question!

Edited July 11, 2016 by Ken Simpson

[c. stirling bartholomew]

The obvious solution is tagging the texts for declension. I am still using the Gramcord GNT and LXX so I don't know what tags are currently available.

[Λύχνις Δαν]

I had a similar thought the other day.

 

That not being present I wondered how far one could get without it ...

 

There are two sources of declensional information for Greek in Accordance that I know of - LEH and MBG. Neither is easy to join to the text in a bulk query. I believe that you could extract a list of nouns of the right declension from LEH and then you could set up a text tab selecting those words. It would require extracting the LEH declension data and processing it with a script to produce a query but it could be done, though the list of words would be long. You could then [LINK ..] or xxxx@[HITS ...] to combine that data into other queries.

 

MBG is much harder to use like this because it would require a lot more processing to get it to the point where you could create the list of words. But it would fit the probably GNT better. Perhaps a quick check of words unique to GNT vocabulary not present in LEH would give you an idea as to how worthwhile that exercise would be.

 

This is speculative and perhaps some issues would arise which would render it unworkable, like having thousands of words in a query list.

 

For whatever it might be worth....

 

Thx

D