Biblical Hebrew (and Semitic) questions

[zompist]

I'm looking at Biblical Hebrew for my book, and I have questions... I hope some of you can help!

To start with, I'm looking at the pronominal object suffixes for verbs— e.g. כָּתַבַני kāṯaḇanī "he wrote to me." A correspondent tells me that often these constructions aren't used, and points to examples in Joshua 1, e.g.

אֲשֶׁר נָתַן לָכֶם מֹשֶׁה (that Moses gave you)

If I'm following, this puts the affix on the preposition lə rather than the verb.

Anyway, my correspondent couldn't explain the rules, and the grammars I've looked at don't either. So, what are they? When do you apply the object suffixes to the verb, and when it is optional, and why...?

[Kuchigakatai]

נָתַן nāthan takes an indirect object in that sentence from Joshua, so it uses a preposition as Biblical Hebrew normally does. The real question is where it was that you read that כְּתָבַ֫נִי kəṯāḇa´-nī (notice the different vowels compared to כָּתַב kāṯaḇ without a pronoun) means 'he wrote to me', because that might not be a very reliable source?

According to the Brown-Driver-Briggs, כָּתַב kāṯaḇ is used with a direct object as:
- to write [words, commandments]
- to write [a book, a letter], inscribe [a tablet]
- (more often as a passive participle than an active finite verb) to record/register [a number, a man, peoples] (in a list or some sort of record)

Then it also mentions that to express the indirect object of the person for whom something is written, the prepositions אֶל־ ʾel, עַל־ ʿal and sometimes לְ lə are used.

Some examples (MS = masculine singular, MP = masculine plural):

(Exodus 24:4, direct object: words)
וַיִּכתֹּב מֹשֶׁה אֵת כָּל־דִּבְרֵי יהוה
way-yiḵtōḇ mōše ʾēṯ kol diḇrê YHWH
WAW_CONSEC-write.3MS Moses ACC all words.CONSTR God
'Moses wrote all the words of the LORD.'

(Jeremiah 22:30, direct object: person)
כִּתְבוּ אֶת־הָאִישׁ הַזֶּה עֲרִירִי
kiṯḇû ʾeṯ hāʾîš hazze ʿărîrî
register.IMPER.2MP ACC man this.MS childless.MS
(God is throwing a curse at King Jeconiah) 'Register (pl.) this man as childless.'

(2 Samuel 11:14, direct object: a thing, indirect object: person)
וַיִּכְתֹּב דָּוִד סֵפֶר אֶל־יוֹאָב
way-yiḵtōḇ dāwiḏ sē´p̄er ʾel yôʾāḇ
WAW_CONSEC-write.3MS David scroll to Joab
'David wrote a letter scroll to Joab.'

[zompist]

Thanks for the response! OK, it looks like KTB is a bad verb to use. I really just want a transitive (strong) verb. Maybe ZKR 'remember'?

Focusing on one bit--

כְּתָבַ֫נִי kəṯāḇa´-nī (notice the different vowels compared to כָּתַב kāṯaḇ without a pronoun)

This is tricky for me because a bunch of grammars list the suffixes without giving paradigms. So they don't even say that the vowels change.

If you do have a direct object, can you freely use the suffixes, or are explicit pronouns preferred sometimes?

[Richard W]

Well, in the paradigms I couldn't find any combinations of 1st person subject or object or of second person subject and object; it may just be that these combinations are two rare to occur. If subject and object have the same reference, so far as I am aware, a reflexive would be used.

[Richard W]

This is tricky for me because a bunch of grammars list the suffixes without giving paradigms. So they don't even say that the vowels change.

If you use the transcriptions with macrons and circumflexes, then the rule is that shifting the accent affects vowels with macrons and short vowels in open syllables.

[Kuchigakatai]

Thanks for the response! OK, it looks like KTB is a bad verb to use. I really just want a transitive (strong) verb. Maybe ZKR 'remember'?

Biblical Hebrew doesn't have object differential marking based on animacy. Your correspondent just meant to say כָּתַב kāṯaḇ, specifically, is not used that way. Verbs can take inanimate or animate direct objects, or inanimate or animate arguments that need a preposition, like English (grab a book, grab an engineer, listen to a song, listen to a representative). (This is also true of Standard Arabic.) Biblical Hebrew just has a further pseudo-preposition to mark direct objects: אֵת ʾēṯ (which may also appear as אֶת ʾeṯ), which can take suffix pronouns.

Incidentally, זָכַר zāḵar can be used either way. According to the DBD, animates can be either a direct object or appear after the preposition לְ lə-. Inanimates are almost always direct objects with this verb, but there is at least one example of לְ lə- plus an inanimate (Jeremiah 31:34: וּלְהַטָּאתָם לֹא אֶזְכָּר־עוֹד û=lə=haṭṭāṯ=ām lōʾ ʾezkār ʿôḏ "and=to=sin=3MP not remember.1S more" 'I won't remember their sin anymore').

(Job 14:13: זָכַר zāḵar and other verbs with an animate direct object)
מִי יִתֵּן בִּשְׁאוֹל תַּצְפִּנֵנִי תַּסְתִּירֵינִי עַד־שׁוּב אַפֶּךָ
mî yittēn bi=šʾôl taṣpinē´=nî tastîrê´=nî ʿaḏ šûḇ ʾappe´=ḵā
who/if_only give.3MS in=underworld hide.2MS=1S, hide.2MS=1S until return.INF.CONSTR nose/anger-2MS
(Job praying to God) 'If only in the underworld you hid me, concealed me until the pass of your wrath,
תָּשִׁית לִי חֹק וְתִזְכְּרֵנִי
tāšîṯ l-î ḥōq wə-ṯizkərē´-nî
set.2MS to-1S due/allotment, and-remember.2MS-1S
set me a due time, and remembered me.'

(Exodus 32:13: זָכַר zāḵar with objects marked with the preposition לְ lə-)
זְכֹר לְאַבְרָהָם לְיִצְחָק וּלְיִשְׂרָאֵל עֲבָדֶיךָ
zəḵōr lə=ʾaḇrāhām lə=yiṣḥāq û=lə=yiśrāʾēl ʿăḇāḏệ´=ḵā
remember.IMPER.2MS to=Abraham to=Isaac and=to=Israel servants=2MS
(Moses speaking with God at Mt. Sinai) 'Remember Abraham, Isaac and Israel, your servants...'

Focusing on one bit--

כְּתָבַ֫נִי kəṯāḇa´-nī (notice the different vowels compared to כָּתַב kāṯaḇ without a pronoun)

This is tricky for me because a bunch of grammars list the suffixes without giving paradigms. So they don't even say that the vowels change.

Oh, believe me, I have my great share of complaints about existing Biblical Hebrew grammars and their usual absence of paradigms... I think they tend to assume you're familiar enough with the language after using textbooks to apply analogy to what you need, and that you can look up whatever you need otherwise, or something? For example, Joüon and Muraoka's grammar (1996) has paradigms for non-suffixed and suffixed verbal forms for qal verbs in the first appendix at the end, but nevertheless only includes non-suffixed forms for the other binyanim, and doesn't include any pausal forms for anything. It appears that in the field of Hebrew, you can even publish books entitled Phonology and Morphology of Biblical Hebrew (by Joshua Blau, 2010) and not include suffixed forms or pausal forms in your paradigms (Blau does not include them in that book).

So yeah, when it comes to Biblical Hebrew, you're really expected to have done your homework with textbooks and readings. If you want to use reference resources, I'd suggest doing exactly that. (Same goes for Standard Arabic, Latin and Ancient Greek, where scholars tend to not produce materials that are friendly to general linguists. It is an unfortunate convention among people who study classical languages...)

If you do have a direct object, can you freely use the suffixes, or are explicit pronouns preferred sometimes?

You have to use the suffixed pronouns for pronominal direct objects, and also for pronominal objects of prepositions. Because sentences can be VSO or SVO in some sense, the independent pronouns are always the subject (if there is no pronominal object, see below). So הוּא יָבֹא hûʾ yāḇōʾ "3MS come.3MS" and יָבֹא הוּא yāḇōʾ hûʾ "come.3MS 3MS" both mean 'he is coming'. (Same goes for Standard Arabic.) Joüon & Muraoka remark: "The pronoun can precede or follow the verb, apparently without any difference in meaning; in most cases it precedes" (§146, p. 540 in 1st edition).

On the other hand, any suffix pronoun can be emphasized with its independent pronoun version right after, no matter if the suffix pronoun is attached to a verb, a noun or a preposition. (Same goes for Standard Arabic.) For example:

(2 Samuel 18:33)
מִי־יִתֵּן מוּתִי אֲנִי תַחְתֶּיךָ אַבְשָׁלוֹם בְּנִי בְּנִי
mî yittēn mûṯ=î ʾănî ṯaḥtệ´=ḵā ʾaḇšālôm bən-î bən-î
who/if_only give.3MS die.INF.CONSTR=1S 1S under/instead_of=2MS Absalom son=1S son=1S
'If only I had died instead of you, Absalom!, my son!, my son!'
(more literally: "who will give the dying of me, me, under you, Absalom, my son, my son?")

Well, in the paradigms I couldn't find any combinations of 1st person subject or object or of second person subject and object; it may just be that these combinations are two rare to occur. If subject and object have the same reference, so far as I am aware, a reflexive would be used.

I believe he's asking about how direct objects are handled w.r.t. animacy... Otherwise yeah, Joüon and Muraoka's grammar says semantic reflexives are handled either with a particular binyan (usually hitpa'el, sometimes others, especially nif'al) or with נֶ֫פֶשׁ ne´p̄eš 'a lifeform's spirit, soul' + pronoun suffix.

They mention that 3rd-person suffix pronouns can be used reflexively if they're attached to a noun or preposition, but not a verb. However, they mention that there are some very rare cases of a reflexive 3rd-person suff. pron. attached to the accusative marker אֵת: Exodus 5:19 וַיִּרְאוּ ... אֹתָם בְּרָע way-yirʾû [subject NP omitted] ʾōṯ-ām bə-rāʿ 'and [the officials of the children of Israel] saw themselves in a bad situation'.

[Richard W]

Well, in the paradigms I couldn't find any combinations of 1st person subject or object or of second person subject and object; it may just be that these combinations are two rare to occur. If subject and object have the same reference, so far as I am aware, a reflexive would be used.

I believe he's asking about how direct objects are handled w.r.t. animacy... Otherwise yeah, Joüon and Muraoka's grammar says semantic reflexives are handled either with a particular binyan (usually hitpa'el, sometimes others, especially nif'al) or with נֶ֫פֶשׁ ne´p̄eš 'a lifeform's spirit, soul' + pronoun suffix.

I was thinking along the lines of something like *hiɡ̄daltînû 'I made us great'.

[Kuchigakatai]

I was thinking along the lines of something like *hiɡ̄daltînû 'I made us great'.

I meant to imply that Joüon and Muraoka say that that doesn't happen ("can be used reflexively...but not a verb").

[Richard W]

I was thinking along the lines of something like *hiɡ̄daltînû 'I made us great'.

I meant to imply that Joüon and Muraoka say that that doesn't happen ("can be used reflexively...but not a verb").

I wouldn't call this reflexive.

[zompist]

Thanks, all this is very helpful.

I found a book with paradigms! It even explains changes to the root. And I was lucky, because I got it from the library literally hours before the library shut down due to coronavirus.

They use פקד. However, they have a gap for the 2p, because 3s>2p forms are apparently not attested in the Tanakh.

I'm guessing the forms are

פְּקָדכֶם
פְּקָדכֶן

Or maybe it's this?

פְּקָדְכֶם
פְּקָדכֶן

Or something else?

[Richard W]

Something else according to the paradigms I copied 40 years ago: sheva, pathah, sheva, seghol, (nothing), e.g. pᵉqaḏḵem. The same form is shown in Teach Yourself Bibilical Hebrew. The vocalisation and lenition for these 2pl suffixes is weird. They argue to me that lenition had already phonemicised.

[Kuchigakatai]

So, basically zompist's first guess, with the mid-word sheva he missed. I'd agree those would be the expected forms as well, seeing the shape כְּתָבָם kəṯāḇām (which *kəṯāḇ=hām underlies) and less relevantly כְּתָבךָ kəṯāḇ=ḵā.

What is interesting about the vocalization?

Regarding the spirantization of k, Tiberian Hebrew is not what ancient Hebrew was actually like anyway... I'd hypothesize, cautiously, that it could be an influence of early medieval Aramaic on a very common morpheme. Also, I've noticed that in Aramaic parts of Ezra and Daniel, the plural of me´leḵ appears with a spirantized ḵ (malḵîn), e.g. thrice in Daniel 7, unlike the lk of מַלְכָּה malkā 'queen' in the Hebrew tanach, and I wonder if that's relevant here...

Most of my interest in Biblical Hebrew is about its syntax anyway, what do we know about spirantization among late native speakers? Do we have evidence that in the 1st c. AD, or for that matter the 3rd c. BC, there was spirantization already? Maybe there are some relevant personal name or place name transcriptions in the Septuagint, which was written in the 3rd century BC?

[Richard W]

So, basically zompist's first guess, with the mid-word sheva he missed. I'd agree those would be the expected forms as well, seeing the shape כְּתָבָם kəṯāḇām (which *kəṯāḇ=hām underlies) and less relevantly כְּתָבךָ kəṯāḇ=ḵā.

Er, no. In each of the four cases, Zompist has qoph qamets, whereas my sources all give qoph pathah.

What is interesting about the vocalization?

Given the other suffixed forms for the 3s, I would have expected *pᵉqāḏᵉḵém. *pᵉqaḏᵉḵém would have been unusual in a fully stressed word, whereas we seem to have *pᵉqaḏḵém

Regarding the spirantization of k, Tiberian Hebrew is not what ancient Hebrew was actually like anyway... I'd hypothesize, cautiously, that it could be an influence of early medieval Aramaic on a very common morpheme. Also, I've noticed that in Aramaic parts of Ezra and Daniel, the plural of me´leḵ appears with a spirantized ḵ (malḵîn), e.g. thrice in Daniel 7, unlike the lk of מַלְכָּה malkā 'queen' in the Hebrew tanach, and I wonder if that's relevant here...

Internal levelling is also possible. There are various hints that tav and thav had come to be contrasted, e.g. preposition
תַּחַת táḥaṯ 'below' v. שָׁלַחַתְּ šāláḥat ‎ 'you have sent'. Do we know when the plosives and spirantised forms split in Aramaic?

As for Biblical Aramaic malḵîn, it might be relevant that the Hebrew construct plural of méleḵ is malḵê - there seems to be no sound evidence of it being malᵉḵê.

Most of my interest in Biblical Hebrew is about its syntax anyway, what do we know about spirantization among late native speakers? Do we have evidence that in the 1st c. AD, or for that matter the 3rd c. BC, there was spirantization already? Maybe there are some relevant personal name or place name transcriptions in the Septuagint, which was written in the 3rd century BC?

Unfortunately, the Greeks seems to have heard the plain v. emphatic distinction as aspirated v. unaspirated, so there was no room to record a distinction of plain v. spirant. We know that the Romans heard spirantisation in Punic, as shown in the English word suffete.

[Whimemsz]

The Septuagint does contain evidence in the transcription of proper names that Hebrew at the time still distinguished PS *x (written with khi) from *ħ (unwritten) and *γ (written with gamma) from *ʕ (unwritten). So at least in the third century BC, presumably /k/ and /g/ could not have been subject to spirantization, though it's possible one set was uvular and the other velar or something.

[zompist]

Given the other suffixed forms for the 3s, I would have expected *pᵉqāḏᵉḵém. *pᵉqaḏᵉḵém would have been unusual in a fully stressed word, whereas we seem to have *pᵉqaḏḵém

Just to be sure, you mean פְּקַדְכֶם?

My textbook definitely has קָ for the other 3sm> forms, e.g. פְּקָדַנִי for 3sm>1s.

Side note, did the Masoretes have really good eyes? I have to zoom in to make sure I'm typing the right ones.

[Richard W]

The Septuagint does contain evidence in the transcription of proper names that Hebrew at the time still distinguished PS *x (written with khi) from *ħ (unwritten) and *γ (written with gamma) from *ʕ (unwritten). So at least in the third century BC, presumably /k/ and /g/ could not have been subject to spirantization, though it's possible one set was uvular and the other velar or something.

The deep question is whether spirantisation is common Canaanite or independent influence from Aramaic (or, I suppose, a development of a common pre-disposition).

[Richard W]

Given the other suffixed forms for the 3s, I would have expected *pᵉqāḏᵉḵém. *pᵉqaḏᵉḵém would have been unusual in a fully stressed word, whereas we seem to have *pᵉqaḏḵém

Just to be sure, you mean פְּקַדְכֶם?

My textbook definitely has קָ for the other 3sm> forms, e.g. פְּקָדַנִי for 3sm>1s.

Side note, did the Masoretes have really good eyes? I have to zoom in to make sure I'm typing the right ones.

Yes, and my three sources (Davidson, Tregelles and Harrison) all also have qoph qameṣ for the non-2pl objects with 3sm subject.

The Masoretes probably had better fonts (as in, if you want a job done properly, do it yourself). Perhaps the secret is to use a larger font size for pointed text. The BHS uses a much smaller font for the marginal notes (masora parva) and a smaller font for the footnotes, though the vowels are still legible there.

I can only read the pointing in preview - Firefox renders the point a character or two to the right in normal view! To be sure of what you'd written, I ended up doing a hex dump of the text and checking the codepoints against the Unicode Standard.

[Kuchigakatai]

Er, no. In each of the four cases, Zompist has qoph qamets, whereas my sources all give qoph pathah.

Ah, I misread what you wrote, and assumed the expected *פְּקָדְכֶם *pəqāḏə-ḵem.

As for Biblical Aramaic malḵîn, it might be relevant that the Hebrew construct plural of méleḵ is malḵê - there seems to be no sound evidence of it being malᵉḵê.

I don't understand what you mean here. Did you mean to write māl^eḵê with qamaṣ at the end of that sentence? How would mal^eḵê be distinguished from malḵê?

Unfortunately, the Greeks seems to have heard the plain v. emphatic distinction as aspirated v. unaspirated, so there was no room to record a distinction of plain v. spirant. We know that the Romans heard spirantisation in Punic, as shown in the English word suffete.

Well, I was thinking along the lines of, like, if the Hebrew of the Septuagint translators transcribed /x/ with chi and /ħ/ as a null, then spirantized /k/ [x] may have been left untranscribed like /ħ/. It wouldn't be conclusive evidence for or against spirantization regardless though...

Side note, did the Masoretes have really good eyes? I have to zoom in to make sure I'm typing the right ones.

Default font sizes are optimized for the Latin alphabet without diacritics (i.e. English), not pointed Hebrew or Arabic. Even tildes on letters (ãõñ) often don't display well on the default font that Firefox/Chrome/Chromium use on Linux (Liberation Sans), looking about the same as macrons. (They look just fine on Windows at the same point size, because Arial is a more reasonable font than Liberation Sans.) I remember having trouble distinguishing ă from ā in Latin while on Linux some time ago, which was particularly painful (I think that was an older version of Liberation Sans?). Look at products that have a text translated in both Latin text and Chinese: you'll notice that often the Chinese font is in a slightly bigger size to make characters more readable.

Fonts themselves may not be all that good either. Google's Noto Naskh Arabic (the default font for Arabic text on Android) is a much better serif font than Arial / Times New Roman, even at 12pt and 14pt, largely because the dots are easier to see. (Arial doesn't have sans-serif Arabic shapes, but simply grabs the serif Arabic from Times New Roman, so for Arabic they're effectively the same font.) This is also why it used to be that Arabs on online forums would either specify Tahoma with a font tag, as Tahoma is a lot more readable at default font sizes than Arial, or otherwise would just pump up the font size, often while bolding the whole thing too. (Amusingly, they'd also center text if there was no right-to-left option available, because ending a paragraph with centered text was still more acceptable than ending it on the left side.)

Your problem is not the Masoretes' oddly good eyes, it's today's anglocentrism in computing. For some tragic cases, you may refer to Tibetan and Thai.

[Richard W]

I've found one of the text books I used online - Davidson. The object suffix inflections are given on p108, p210 and p229.

[Whimemsz]

The Septuagint does contain evidence in the transcription of proper names that Hebrew at the time still distinguished PS *x (written with khi) from *ħ (unwritten) and *γ (written with gamma) from *ʕ (unwritten). So at least in the third century BC, presumably /k/ and /g/ could not have been subject to spirantization, though it's possible one set was uvular and the other velar or something.

The deep question is whether spirantisation is common Canaanite or independent influence from Aramaic (or, I suppose, a development of a common pre-disposition).

The limited evidence to me looks like it wasn't common Canaanite. You mentioned Roman transcriptions of Punic, but these at first just used <p>, <t>, <c> everywhere for both plain and emphatic stops (this was before the practice of using <ph>, <th>, <ch> to render Greek aspirates arose), as still attested in some of the Punic dialog in Poenulus. Later, with the introduction of <ph, th, ch> digraphs in Latin, writers followed the Greek thing with Phoenician/Punic and Hebrew and wrote plain stops as <ph, th, ch>, as in the revised Poenulus dialog, and emphatic ones as plain. And then in the Latino-Punic inscriptions in North Africa itself, the orthography used did indicate that these had become fricatives through spellings with <f> for Phoenician /p/ (plus the loan into Latin of suf(f)et- as you note), very occasionally <h> for Phoenician /k/ (which...I'm not sure if this is actually meant to indicate a fricative or what, since <h> should have been silent in spoken Latin by the time of the Latino-Punic inscriptions?), and spelling historical /b/ with <v> or <f> (and rendering foreign [w] or [u] or [β] with <b> in Punic, as in the city Thugga as TBGG or the Roman name Flavius as things like <flabivs> [earlier spelled as <flavi> and so on]), etc. So it's fair to conclude that Punic spirantized its non-emphatic stops some time in this period. But the fact that /p/ was originally spelled with <p> and then <ph>, not with <f>, would suggest that it was originally not a fricative. Also, the distribution of these spellings is different from those of begadkefat spirantization in Hebrew and Aramaic: they indicate that all Punic stops were spirantized, in all positions (note for instance ubiquitous Latino-Punic <fel> for older */paʕal/ "he made").

Of course, this can still be squared with a scenario in which the change *was* shared by Hebrew and Phoenician-Punic, but just indicated differently in Latin script over time (maybe whatever precise sound spirantized /p/ was was different enough from Latin /f/ that it was spelled differently at first? idk), and that Punic later extended the change to cover all non-emphatic stops, not just postvocalic ones. I think that scenario seems less likely than they were independent changes, though. Especially given how intense Aramaic influence on Hebrew was and that it was the Jewish spoken language in Syria-Palestine by this point, and long before.