Telling Students Vocabulary Coverage Statistics

[Emma Ehrhardt]

I'd love to hear perspectives from both teachers and from students:
Is it helpful to students for instructors to share statistics about how much of the New Testament they can read after they've learned X vocabulary items?

This post inspired by reflecting on James Tauber's post from a couple months ago: http://jktauber.com/2015/10/26/updated- ... tatistics/

Mounce's Basics of Biblical Greek emphasizes these types of stats (learn 80% of NT words in 1st year!), and I remember being frustrated when I realized it was giving me false hope. False expectations about what I could functionally DO with Greek after a single year of study.

On the other hand, looking at Tauber's presentation of stats (viewed through the lense of how many verses/sentences are learnable after X words), can come across as depressing for the student. After so much work and I will barely be able to read a few sentences?!?

So what type of measure would be of better service to the beginning student? I think 2 things: 1) A combined measure somewhere between lemma & form, and 2) organize the rows by the number of items comfortably learned after 1 year, 2 years, etc. through 5 or 10.

Even if we can never confidently derive an accurate measure of learnability, some approximation would still be functional in terms of giving more encouragement than the 'forms' numbers, and being more realistic than the 'lemma' numbers. And setting the number of items learned within a longer length of time context allows teachers to emphasize that learning Greek is a life-long pursuit, not something that you cram for a year or two and then never use again. If we want today's Greek students to maintain their language skills, it's important to equip them with the skills to continue self-directed learning once they're done with formal education, and with excitement about reading Greek rather than duty and dread. In my Greek studies I got the message loud and clear that Greek was important and significant and valuable, but at that time I didn't develop a love for it because my functional skills were simply not there. I couldn't read enough Greek to appreciate it as a language that told its own stories.

[jtauber]

The context of those vocabulary stats (both why I calculated them in the first place and how I talk about them in presentations) is largely to point out (to teachers, not students) how much can be improved by not *just* considering frequency of lexeme. A large part of the "New Kind of Graded Reader" work I did around 2008 was to improve ordering with the goal of reading full verses sooner. For example [1] demonstrates one algorithm that allows 12 verses after 16 forms (rather than 0 after 100 like a pure frequency-of-forms approach). [2] introduced another approach of ignoring the least frequent form when assessing verse suitability.

Of course in the case of most introductory text books, the problem is made "worse" by ordering based on inflectional class (in the case of nominals) and a combination of inflection class / mood / tense/aspect forms in the case of verbs. As I point out in [3], aorist 3rd person singulars of athematic verbs (e.g. ἔδωκεν) don't need to be be delayed until present paradigm of δίδωμι is taught. (also: things like ἀπεκρίθη Ἰησοῦς καὶ εἶπεν αὐτῷ can be taught as a construction before getting into "passive-in-the-aorist" (if ever!) or 2nd aorists; μὴ γένοιτο could be taught as a construction without waiting to learn all about the optative and fully conjugate it.)

I think you're spot on about the need for something between forms and lemmas and, in fact, that's one of the big motivations for me doing my big stem and theme analysis the last couple of years. [4] has some of my thoughts on that topic and I think I'm finally getting close to being able to revisit much of the New Kind of Graded Reader stuff with those ideas incorporated. Watch my blog for more soon!

I haven't answered your core question, though: what do we tell the students? I largely have to defer to those people who have vastly more experience actually teaching language than I do, but my suspicion is showing them real text with inflected words in context sooner (rather than gating based on frequency and when they learn the paradigm) is better encouragement than either "10 words make up 37% of the text" or my "100 forms gives you 0 verses".

James

[1] http://jktauber.com/2008/03/26/just-how ... -improved/
[2] http://jktauber.com/2008/03/26/if-only- ... rare-word/
[3] http://jktauber.com/2015/05/06/my-bibletech-2015-talk/
[4] http://jktauber.com/2015/11/13/initial- ... ning-form/

[Tim Evans]

As a student, the stats in Mounce always seemed nonsense right from the start for two reasons.

1. As you said, being able to read all the occurrences of και is not actually impressive, no matter how many times it occurs.
2. As a learner of several languages, trying to actually use it in real life, being able to read και is the least useful word in a real sentence. Consider the sentence "Do you want to go to the Zoo?", zoo will be the least commonly occurring word in that sentence, yet knowing Zoo, and hearing the intonation in English, is the only thing you need to know to understand the speaker. Conversely, knowing all the other words, i.e. "Do you want to go to ???????" Does not help you much at all. Unless the speaker knows you have three year old English and can modify to suit you, i.e. "Do you want to go see lots of animals?"

My measure of success, and what has always been the most motivating is "number of verses I can read without a dictionary" but to even consider this type of metric, Mounce would need to let go of this idea that learning more 300 words in the first year of Greek is too difficult. (Well that is the impression he gave me in his videos)

[jtauber]

If I had a reverse interlinear in machine-actionable form, it would be a fun exercise to show a passage from the NT in *English* with only the common words (by underlying Greek frequency) shown: "And the _____ day, Jesus ____ to _____ and _____ the ______. And they said to him ______. And he replied and said ______"

That said, "Do you want to go to ???????" is a great example in English of what construction grammarians would call a pivot schema. It's a construction you can learn with the single slot open and then apply it with lots of different fillers.

It would be a worthwhile exercise (I think) to come up with a list of pivot schemas to learn for Greek.

[jtauber]

Here's a little prototype I wrote up a couple of years ago focused just on the pivot schema: the ____ of God.

[Jonathan Robie]

Here's a little prototype I wrote up a couple of years ago focused just on the pivot schema: the ____ of God.

Looks like you are limiting the things that can appear in that slot to a known vocabulary? How can I best learn about using pivot schemas in language teaching?

[jtauber]

Looks like you are limiting the things that can appear in that slot to a known vocabulary?

Just a search in the GNT where DET N τοῦ θεοῦ occurs. Relevant, I think, to Emma's original question is that verse references are given. I can imagine that might encourage early students "oh wow, I can read part of all these verses now" in a much more constructive (no pun intended) way than individual lexemes.

How can I best learn about using pivot schemas in language teaching?

Well, I don't know about language teaching. I've only come across them in construction grammar approaches to language acquisition and 1LA in particular.

Tomasello is *the* guy in this area. His "Constructing a Language: A Usage-Based Theory of Language Acquisition" is a book-sized treatment. See http://www.eva.mpg.de/psycho/pdf/Public ... ids_06.pdf for a short version.

[Emma Ehrhardt]

The context of those vocabulary stats (both why I calculated them in the first place and how I talk about them in presentations) is largely to point out (to teachers, not students) how much can be improved by not *just* considering frequency of lexeme.

Yes, I did rather rip your stats out of context. And while I agree the current measures are useful to point out to teachers, they just got me thinking about what would be helpful for students as well.

[Stephen Hughes]

Looking at such claims in another way, it could be said that the more frequent words are the ones that it might be possible for a student to pick up through repeated exposure during reading, watching some form of media or carrying on a conversation. The uncommon ones need to be "learnt" through some more direct application of the faculties such as glossing.

Grouping them by topic or location (work) would then make sense. To read the book of Romans, you will need the following special words, might be a more realistic statement. Comprehension of 80% of the words in a passage is REALLY different to comprehension of 80% of the meaning of the passage. Why would anyone want to spend 3 years of learning to understand less than they could in translation. The aim has to be total (aka 97+%) mastery of the vocabulary and preferably mastery of extra (120-500%) vocabulary than is needed to read a passage.

Actual numbers might be better than statistical percentages. If a student is told that they need to learn 35 words more than the basic 50x or more words, to read a passage fluently - bareback and without aids - then that could make learning vocabulary a series of tasks with direction and tangible outcomes. Clustering passages by topic might be good. To read passages about daily life, farming, fornication, fasting, law, or marriage for example, then such and so vocabulary needs to be mastered. Real outcomes could then be achieved.

Rather than thinking of effort scattered all over the place and getting nowhere. Perhaps, piece by piece, passage by passage is better. Comprehending the macarisms to 97% in Greek, might be better than the usual situation.

I find it satisfying to read through a book that I have previously worked through and be able to comprehend what is written. That euphoria is short-lived when in the case that as reading continues more questions arise, which can not be answered by a superficial glance at the Greek.

Of course I did learn by the 50 times or more, then 10 times or more and and so on working down method. Actually, I find learning vocabulary a great way to relax, reduce stress and decrease brain function. The offshoot is that I usually get a only a handful of unknown words per chapter now (some chapters, like those in 2 Corinthians and the Peters have big, strong hands that can grasp a lot).

[Jonathan Robie]

I'm a beginning teacher, thinking these things through for stuff that I'm doing. Suppose we use statistics for the most common words in the New Testament as a guide, following Mounce's example.

Code: Select all

<word count="19793" class="det">ὁ</word>
<word count="8196" class="conj">καί</word>
<word count="5552" class="pron">αὐτός</word>
<word count="2894" class="pron">σύ</word>
<word count="2788" class="conj">δέ</word>
<word count="2743" class="prep">ἐν</word>
<word count="2569" class="pron">ἐγώ</word>
<word count="2458" class="verb">εἰμί</word>
<word count="2255" class="verb">λέγω</word>
<word count="1766" class="prep">εἰς</word>
<word count="1558" class="adv">οὐ</word>
<word count="1407" class="pron">ὅς</word>
<word count="1388" class="pron">οὗτος</word>
<word count="1312" class="noun">θεός</word>
<word count="1242" class="adj">πᾶς</word>
<word count="1215" class="conj">ὅτι</word>
<word count="1091" class="pron">τίς</word>
<word count="1038" class="conj">γάρ</word>

ὁ is the most common word, you've learned that word, congratulations. But you haven't learned it yet, it doesn't mean the same thing as the English definite article, and declining it is going to take a bunch more work. The next word is καί, which isn't declined, but also means something a little different from the English equivalent. δέ has similar issues, but is even more subtle.

Looking at verbs, εἰμί and λέγω are extremely frequent, but look which forms are the frequent ones. And knowing the word λέγω does not mean that you know what λέγοντες means.

Code: Select all

<word rank="1" count="899" lemma="εἰμί">ἐστιν</word>
<word rank="2" count="612" lemma="λέγω">εἶπεν</word>
<word rank="3" count="338" lemma="λέγω">λέγει</word>
<word rank="4" count="315" lemma="εἰμί">ἦν</word>
<word rank="5" count="211" lemma="λέγω">λέγω</word>
<word rank="6" count="202" lemma="γίνομαι">ἐγένετο</word>
<word rank="7" count="200" lemma="ὁράω">ἰδού</word>
<word rank="8" count="177" lemma="λέγω">λέγων</word>
<word rank="9" count="157" lemma="εἰμί">εἰσίν</word>
<word rank="10" count="149" lemma="λέγω">λέγοντες</word>
<word rank="11" count="138" lemma="εἰμί">εἰμί</word>
<word rank="12" count="124" lemma="εἰμί">εἶναι</word>
<word rank="13" count="117" lemma="εἰμί">ἔσται</word>
<word rank="14" count="104" lemma="ἔχω">ἔχει</word>
<word rank="15" count="103" lemma="λέγω">εἶπαν</word>
<word rank="16" count="94" lemma="ἀποκρίνομαι">ἀποκριθείς</word>
<word rank="17" count="94" lemma="εἰμί">ἦσαν</word>
<word rank="18" count="92" lemma="εἰμί">εἶ</word>
<word rank="19" count="92" lemma="εἰμί">ἐστε</word>
<word rank="20" count="89" lemma="ἔρχομαι">ἔρχεται</word>
<word rank="21" count="89" lemma="ἔρχομαι">ἦλθεν</word>
<word rank="22" count="86" lemma="ἔχω">ἔχων</word>
<word rank="23" count="82" lemma="ἀποκρίνομαι">ἀπεκρίθη</word>
<word rank="24" count="78" lemma="λέγω">ἔλεγον</word>

I'm inclined to think that teachers need at least three lists:

• Most frequent words (teach pronouns early, really teach the semantics of little words and how to distinguish lookalikes)
• Verbs
• Nouns

Nouns are the easy ones. If people have to look up nouns, that's not as hard as trying to figure out how the little words work or how the verbs work.