Saturday, February 4, 2017

Ah, Grammar

When I was a child, I spake as a child... and I spake naturally without thinking about rules of grammar. I didn't think about rules of grammar until I began to learn other tongues like French  and Latin. The teaching approach was appalling. It gets a No.

Now that I am old, I will not put away this childish thing, even though my eyes and ears are weakened and my memory is leaving its post. And my aural immersion is limited.

One cannot learn a language by rules. But today I am 10 years old as a student of Hebrew, so its time for me to do a little analysis of the words. I have learned when writing Hebrew to use just about any tense that seems to fit the story or the poem. Here is my Yes to a lost art, my lost art of learning a language.

I am looking at the amo, amas, amat of my Hebrew paradigms. It's an old copy of Fred Putnam's Book. I see it is revised and easier to see online here. (Page 285). The problem with paradigms is that they use the Masoretic vowel marks. Or in this example online, they use plene spelling.

Another problem in general is that some Hebrew textbooks present the persons in the order 3 - 2 - 1 instead of the usual 1 2 3. Putnam has abandoned the 3 2 1 in this revision of his text.

But my question today is how do these verbs look in their base form without vowels (but perhaps with accents since the music may imply the form in a few cases).

I need to reduce the volume and pick one example: Putnam used to have the traditional amo,  but in Hebrew tradition is קתל, to kill! Not the Latin amo, to love.

My Latin teacher was a killer - and that's no joke. My pliable brain was ill-used in the 1950's by such a man. How fearful we are of the incredible gift of our lives. How powerful we are in our weakness.

Part of the problem with verbs is how we express our experience of time. Time is present and past and future, but much more subtle, since the presence of the past determines our future. Time in a story is also relative to the story, and may reflect story time or narrator time. It also will reflect mood, like conditional or subjective, and the aspect of completeness, continuing, or indeterminate. Nice lecture here on Tense-aspect-mood- and a fourth, evidentiality. The more you use another language, the more you realize how flexible your own is beyond the simplistic rules you might have been taught.

It appears to me that the binyan markers in Biblical Hebrew have been inferred from context, and are not absolute. Nothing about verbs seems to be absolute.

Putnam now uses משׁל, to govern, or perhaps to tell a story. This is also the word for parable. Here is tiny bit of the 7 column tables.

If you stare down this typical massive paper example of a paradigm, you see that the three columns for both prefixed and suffixed forms of the verb for Qal, Piel and Pual are indistinguishable if you don't have the vowels. The original manuscripts have no vowels, so the context of other words must indicate a clue to the verb form, simple, intensive, or passive intensive. Three of the seven columns just merged.

Similarly Hifil and Hofal merge. Somehow, the native Biblical Hebrew writer knew to use an A for Qal, an I for Piel (and double the consonant following) and a U for Pual. Similarly for the causative and its passive form, an I for Hifil and a U for Hofal.

What remains is the twofold division of prefixed and suffixed forms: for the form that distinguishes person and number by suffix (like amo in Latin), a נ prefix begins the passive Nifal, a ה prefix begins the causative, and an הת prefix begins the reflexive form in the perfect. Four columns only.

For the form of the verb that uses prefixes to distinguish person and number, the situation is worse: the prefix now requiring yod, taf, aleph, or nun, qal, piel, pual, niphal and hofal all lose their distinction - that's five columns distinquished only by context. Hifil adds a distinct yod to the stem and the reflexive form, having had a double prefix already, drops the first prefix and substitutes the single letter prefix yod, taf, aleph, or nun. Only 3 columns required here.

Every other nuance of meaning is determined or inferred by the child's knowledge. The distinguishing pronunciation would still be evident to ears: the o for qal, e for niphal, doubled consonant for piel, and the u for the passive of the intensive or the causative (or sometimes an a).

As for the number and person, they are largely unambiguous: This is my exercise in typing.
1 cs תי taf-yod, 2 mfs ת taf (distinguished by vowels and accent), 3 ms root only, 3 fs ה heh
משׁלתי I governed, משׁלת you(s) governed, משׁל he it governed, משׁלה she governed.
1 cp נו nun-vav, 2 mp תם taf-mem, 2 fp, תן taf-nun, and 3 cp ו vav
משׁלנו we governed, משׁלתם you(mp) governed, משׁלתן you(fp) governed, משׁלו, they governed

1 cs א aleph, 2 ms and 3 fs ת - taf no distinction, 2 fs ת and לי lamed-yod suffix, 3 ms י yod
אמשׁל I govern, תמשׁל you(ms) she governs, תמשׁלי you(fs) govern, ימשׁל he it governs
1 cp נ nun, 2 mp ת with vav suffix, 2.3 fp, ת with נה suffix, 3 mp, י yod with ו suffix.
נמשׁל, we govern, תמשׁלו you(mp) govern, תמשׁלנה you(fp) they govern, ימשׁלו they(mp) govern.

whew! There are more paradigms for weak, hollow, geminate, and guttural verbs, but the patterns are not significantly different. It's just that letters appear and disappear in these forms. And there are more forms, imperative, infinitives, participles. Some other day.

I have been muttering a little for this exercise. I'll let you know if it has improved my reading.

PS: my thought is that the 7 column format is an implausible structure in the brain. An algorithm is more likely. I have summarized the possible rules in this table, together with some ambiguities...
Spread sheets are difficult to form and and buggy to copy with both ltr and rtl, but maybe this will work.

Suffixed Simple   Passive
Intensive Reflexive
vowel  A   I short A   Act I / Pass U I + dag / U + dag
I short A
 Qal   Nifal   Hifil / Hofal Piel / Pual = Qal
1cs משׁלתי נמשׁלתי המשׁלתי exc vowels+dag התמשׁלתי
2ms (A), fs (:) משׁלת נמשׁלת המשׁלת התמשׁלת
3ms משׁל נמשׁל המשׁיל התמשׁל
3fs משׁלה נמשׁלה המשׁילה

1cp משׁלנו נמשׁלנו המשׁלנו
2mp משׁלתם נמשׁלתם המשׁלתם התמשׁלתם
2fp משׁלתן נמשׁלתן המשׁלתן התמשׁלתן
3cp משׁלו נמשׁלו המשׁילו התמשׁלו
Prefixed    E / I       A      short A / A        I short A / : U  I short A
1cs אמשׁל
אמשׁיל אתמשׁל
2ms / 3fs תמשׁל
תמשׁיל תתמשׁל
2fs תמשׁלי
תמשׁילי תתמשׁלי
3ms ימשׁל
ימשׁיל יתמשׁל
1cp נמשׁל
נמשׁיל תנמשׁל
2mp תמשׁלו
תמשׁ׳לו תתמשׁלו
2/3 fp תמשׁלנה
תמשׁלנה תתמשׁלנה
3mp ימשׁלו
ימשׁילו יתמשׁלו

So what are the summary rules: leading נ passive or 1cp prefixed
leading הת reflexive suffixed, or second letter ת prefixed
leading ה causative suffixed 
additional yod, causative prefixed
leading א prefixed 1 cs
leading yod, prefixed 3 mp or ms depending on suffix
leading taf, prefixed 2 m or s depending on suffix
no leading letter, suffixed - q.v.
It only gets worse with pronomial suffixes added - later, later...