Trivia: Which chapters have the most obscure words?

Forum rules
Please quote the Greek text you are discussing directly in your post if it is reasonably short - do not ask people to look it up. This is not a beginner's forum, competence in Greek is assumed.
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Trivia: Which chapters have the most obscure words?

Post by Stephen Hughes »

It's been over a year since the last trivia, but I have another question now.

Which NT chapter has either; the numerically highest number of words occurring say 5 times or less (or 3 times or less), OR the highest concentration of rarely occurring words (or words occurring in obscure senses)?

(An answer to either question is okay, I'm wondering where the thickest mud for a reader to wade through would be).
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Emma Ehrhardt
Posts: 58
Joined: December 23rd, 2014, 10:28 am
Location: IN
Contact:

Re: Trivia: Which chapters have the most obscure words?

Post by Emma Ehrhardt »

Numerically Highest — Acts 27 (by either measuring standard)
  • 137/752 words occur <=5x, or 18.2%
    106/752 words occur <=3x, or 14.1%
Highest Concentration — 2 Peter 2 (by either measuring standard)
  • 79/372 words occur <=5x, or 21.2%
    63/372 words occur <=3x, or 16.9%

Top 10 chapters, by concentration of lemmas occurring 5 or fewer times in the GNT:

Code: Select all

1.  2Pe 2	79/372	21.2%
2.  Mt  1	83/426	19.0%
3.  2Ti 3	43/235	18.3%
4.  Lk  3	98/587	16.7%
5.  Ac 27  137/752	18.2%
6.  1Ti 6	46/343	13.4%
7.  Tit 3	29/219	13.2%
8.  1Pe 5	25/208	12.0%
9.  1Ti 2	26/186	14.0%
10. Ro 16	44/383	11.5%
Data calculations were based on MorphGNT.
Emma Ehrhardt
Computational Linguist
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: Trivia: Which chapters have the most obscure words?

Post by Stephen Hughes »

Emma Ehrhardt wrote:Numerically Highest — Acts 27 (by either measuring standard)
  • 137/752 words occur <=5x, or 18.2%
    106/752 words occur <=3x, or 14.1%
Thank you Emma, I am so happy to hear that!

It was actually that I finished reading Acts 27, put down the book wondering why I was still reading at a level of comprehension that I did in around 1990, and then wrote this question.

While those numerical figures are high in themselves, what they don't make clear is the number of times that bread and butter words are used here in unfamiliar senses, further complicating matters.

How can I generate a list of those 106 words to try to rote learn them at least?
Emma Ehrhardt wrote:

Code: Select all

2.  Mt  1   83/426   19.0%
That is definitely the most difficult chapter to pronounce!

Can the MorphGNT produce a list of chapters showing the chapters with the highest numbers of infrequent three (or four) letter strings in them? E.g. frequent στρ (or στρε) vs. infrequent χαβ (or αχαβ)?
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Emma Ehrhardt
Posts: 58
Joined: December 23rd, 2014, 10:28 am
Location: IN
Contact:

Re: Trivia: Which chapters have the most obscure words?

Post by Emma Ehrhardt »

You're welcome, Stephen. Matthew 1 was the first chapter that jumped to mind when I first read your question. :)

As for a list of the rare words in Acts 27 the script I wrote to count frequencies was easily adjusted to also report the actual rare words themselves as well. How is a good way to share that file?
Emma Ehrhardt
Computational Linguist
Emma Ehrhardt
Posts: 58
Joined: December 23rd, 2014, 10:28 am
Location: IN
Contact:

Re: Trivia: Which chapters have the most obscure words?

Post by Emma Ehrhardt »

Stephen Hughes wrote:Can the MorphGNT produce a list of chapters showing the chapters with the highest numbers of infrequent three (or four) letter strings in them? E.g. frequent στρ (or στρε) vs. infrequent χαβ (or αχαβ)?
MorphGNT is just the SBLGNT text with extra data attached to it:

Code: Select all

050527 V- -AAPNPM- Ἀγαγόντες Ἀγαγόντες ἀγαγόντες ἄγω
050527 C- -------- δὲ δὲ δέ δέ
050527 RP ----APM- αὐτοὺς αὐτοὺς αὐτούς αὐτός
050527 V- 3AAI-P-- ἔστησαν ἔστησαν ἔστησαν ἵστημι
Someone with basic programming ability could query many things with that dataset, especially related to frequency. (The sections are book/chapter/verse, part of speech, parsing code, text (including punctuation), word (with punctuation stripped), normalized word, & lemma.)

As for your specific question, can you clarify what you mean by 3-letter strings? Are these strings within a word only, or can they cross word boundaries? Are you counting the strings within the inflected word forms that actually occur in the text, or whether their lemma contains that string? Do you ignore accents and breathing marks or treat them as marking the same character? The reason why you want such a list will influence these answers.
Emma Ehrhardt
Computational Linguist
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: Trivia: Which chapters have the most obscure words?

Post by Stephen Hughes »

Emma Ehrhardt wrote:As for a list of the rare words in Acts 27 the script I wrote to count frequencies was easily adjusted to also report the actual rare words themselves as well. How is a good way to share that file?
With gratitude, I would say that a PM is an okay way for something in that range of characters 500 - 600. The character limit for a PM is about 66,000 characters - which I only exceeded the once.

The results that I got from a Perseus vocabulary tool search of Acts 27 are at best difficult to deal with.
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Emma Ehrhardt
Posts: 58
Joined: December 23rd, 2014, 10:28 am
Location: IN
Contact:

Re: Trivia: Which chapters have the most obscure words?

Post by Emma Ehrhardt »

Done. I hadn't used that feature of the forum yet...
Emma Ehrhardt
Computational Linguist
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: Trivia: Which chapters have the most obscure words?

Post by Stephen Hughes »

Emma Ehrhardt wrote:As for your specific question, can you clarify what you mean by 3-letter strings? Are these strings within a word only, or can they cross word boundaries? Are you counting the strings within the inflected word forms that actually occur in the text, or whether their lemma contains that string? Do you ignore accents and breathing marks or treat them as marking the same character? The reason why you want such a list will influence these answers.
Well, the thinking behind it is the pronunciation of the Greek. I have list of all possible vocal combinations in all words that occur in the New Testament, which is interesting but not adequate. To get the list, I took the whole text of the New Testament and reduced the vowels to their phonetic 7-vowel restored Imperial Koine equivalents as an extra piece of data attached to the words themselves, and retained the accent - in effect a 14 vowel vocalic system (seven vowels *2 for stressed and unstressed - vowel length is affected by the intonational pattern of words in Modern Greek in ways that would not be noticeable to English "ears", so if that were in incorporated into learning, effectively an extra dimension of flexibility / control needs to be taught / mastered).

I then reduced the list by eliminating duplicated - I would have preferred to have acquired data rather than just eliminated duplicates. Then I created a third data element with each word - the vowel only pattern, and sorted all words according to the number of vowels - unfortunately, by now the data was corrupt because enclitics were separated and lost in the list. That produced a list ordered by single vowel, followed by that the single vowel in combination of other vowels - Christmas tree of stacked data triangles if you like. Sorting the consonantal patterns within the equivalent vowel combination strings of the third data element became difficult, because the data became to large to be successively cut and pasted between Excel for sorting and Word for the manual string substitutions and textmechanic.com/. I then took to manually extracting data to try to create the big triangle of data arranged by consonants, but didn't continue - but wanted to - because the data was getting to be too simplistic to be useful. In the end, I produced only 1 and 2-syllable single difference pairs, and a few 3-syllable ones. Using them has been very helpful for me to practice my own pronunciation, but far short of producing phonetic drilling material that I wouldn't be ashamed to offer others. The total time spent on that was about 75 hours, with little in the way of results, but I did come away with a much greater appreciation of what is involved in producing more than a subjectively suitable list of reading-pronunciation exercises.

Word breaks etc. are good, but what I'm really interested in in producing a scientifically accurate set of the phonetic structure of the Greek text when read in the 7-vowel Imperial Koine pronunciation system, so that I can easily - relatively easily at least - produce single difference training pairs for "all" combinations - by all I mean stressed and unstressed combinations - and from my manual handling that the examples starts to get more and more laconic as you move from 4 to 7 vowels. In that way then what I'm looking for is more syllabic than it is actual length. What is missing here is the secondary stress patterns when the word has two accents.

The basic structure I want to produce is to get 1, 2, 3 or 4 vowel patterns, that can be pronounced without consonants - like a toddler or some deaf-mute speakers, and then successively build up words - add orals, then nasals, unvoiced stops, voices stops then fricatives etc. according to the order that the literature on Modern Greek child language acquisition suggests that children acquire the consonants - very similar to the Koine consonantal system. Training for vowels also starting from the basic vowel and moving out - initial reference differentiation from one point in the mouth, not relative to each other - I expected to be able to more clearly see the significant differences that would need to be pointed out clearly.

Asking you for 3 or 4 letters together is really just a single calculation rough estimate of which chapter would be the most difficult to read. I personally haven't found a more difficult passage to read, and honestly with the foreign names in reading English aloud, I process them as single unit memory triggers - indeclinable in Greek terms - rather than try to process them into morphological units, consonant patterns or the like - as French words have come into English indeclinable.
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Stephen Hughes
Posts: 3323
Joined: February 26th, 2013, 7:12 am

Re: Trivia: Which chapters have the most obscure words?

Post by Stephen Hughes »

Emma Ehrhardt wrote:Done. I hadn't used that feature of the forum yet...
Thanks.

As ever, the world is easy when you know how.

Moreover, that's my rote learning task for the next month set out for me - three a day will square it away.
Γελᾷ δ' ὁ μωρός, κἄν τι μὴ γέλοιον ᾖ
(Menander, Γνῶμαι μονόστιχοι 108)
Emma Ehrhardt
Posts: 58
Joined: December 23rd, 2014, 10:28 am
Location: IN
Contact:

Re: Trivia: Which chapters have the most obscure words?

Post by Emma Ehrhardt »

It sounds like you've done a considerable amount of work on this already. I think I get the gist of what you're aiming for, but I wonder, if you want to practice lots of sound combinations, is it necessary to use attested GNT forms? Especially for 1-2 syllable words, why not just create a list of possible phone combinations, regardless of whether they're attested? (e.g. "aba,ada,aga,apa,ata,aka...ebe,ede,ege", etc.) And if this is also intended for use by students who aren't yet very proficient in the language, they won't recognize words from non-words yet anyway.

I personally doubt that finding a chapter with a higher percentage of rare phoneme combinations will be of much use in pronunciation training. When we talk about lexical items, some are rare enough that it's worth seeking them out for special practice if we want to learn them. But the phoneme inventory of a language is quite small. Some combinations of phones may be less common, but they will still be covered in general usage. That said, you're probably right to emphasize that it is desirable to practice phone combinations in context, so maybe short sentences with a tongue-twister flavor could be useful, perhaps.

If I understood correctly that you're primarily looking for shorter words, why the drive to combine enclitics with the word they lean on? Just for more data?
Emma Ehrhardt
Computational Linguist
Post Reply

Return to “New Testament”