Wiktionary talk:About Hebrew/archives/2007

Romanization

Where do we want to provide Latinizations of Hebrew words, and what scheme do we want to use for doing so?

I think we should provide a Latinization whenever we link to a Hebrew entry, such that we'd never put just [[אשר]] but always [[אשר]] (''asher'') or the like. (I'm not sure if that's the best way to format it — parentheses and italics — but you get the idea.)

As for the Latinization scheme to use, I think it makes sense to use one based on a sort of average Modern/Standard/Israeli Hebrew pronunciation, with a bias toward distinguishing more sounds rather than fewer, but I know that a lot of people feel strongly about their own Hebrew pronunciations, so some sort of compromise might be necessary here.

Ruakh 19:55, 19 January 2007 (UTC)

I agree that Romanizations should be included in the translation sections of English words, but not in the article. Some sort of standard procedure for this should indeed be written up, however, I do not envy the person who has that task. I believe the standard format is parentheses for Romanization and quotation marks for translations (when present). Cerealkiller13 20:56, 19 January 2007 (UTC)

Oops, you're right, romanization is the right word. *changes section heading*

Proposed romanization scheme:

• All romanizations should be wrapped in the {{romanization of Hebrew}} template, which will simply let through its contents. (This enables the template's what-links-here to be used to find all romanizations of Hebrew, so we can go through and make changes to all of them if necessary.) (proposal continues below)

  I've been adding this template around all the new translations I've been adding, but it doesn't seem to be in very widespread use... Is this what I should be doing, or is this just a suggestion for now? AggyLlama 07:27, 12 February 2007 (UTC)

  I only proposed it three weeks ago, which is why it's not in very widespread use, but no one has objected to it, so I think we should probably stick with it (unless you have an alternative suggestion, in which case I'm all ears). —Ruakh_TALK 07:50, 12 February 2007 (UTC)

  Wouldn't it be simpler to type, say... {{he-roman}} instead of {{romanization of Hebrew}}? Just a thought.--EncycloPetey

  Should {{he-translation}} be used on the About Hebrew#Romanizations page in place of {{romanization of Hebrew}}? I think it should at least be mentioned there...

• (proposal continued from above) The letters are romanized as follows:

letter	romanization	notes
א	(')	omit word-final א, except optionally1	ל	l
ב	b/v	b when with dagesh, v when without	מ,ם	m
ג	g	optionally1 gh when without dagesh	נ,ן	n
ד	d	optionally1 dh when without dagesh	ס	s
ה	(h)	omit word-final ה, except when with mapik	ע	(')	optionally1 `; omit word-final '
ו	w	see below for ו as vowel marker	פ,ף	p/f	p when with dagesh, f when without
ז	z		צ,ץ	ts	optionally1 TS
ח	kh	optionally1 ḥ	ק	k	optionally1 Q
ט	t	optionally1 T	ר	r
י	y/i	i when the latter part of a diphthong, y elsewhere;    see below for י as vowel marker	שׁ	s(h)	sh when shin, s when sin
כ,ך	k/kh	k when with dagesh, kh when without	ת	t	optionally1 th when without dagesh

• ¹ used when it is particularly important to represent a form of Hebrew in which the sound is distinguished, or in the case of צ, a form in which emphatic consonants are emphasized (whether glottally or pharyngeally).
• Dagesh khazak is optionally (as described above) indicated by doubling the letter. In the case of שּׁ and צּ, this produces shsh and TSTS.
• Vowels are romanized as follows:

vowel	romanization	notes
בְ	(')	an apostrophe when na`, omitted when nakh or when adjacent to א or ע
בֱ, בֶ, בֵ	e
בֲ, בַ, בָי	a
בָ	a/o	a when gadol, o when katan
בֳ, בֹ, בוֹ	o
בִ, בִי	i
בֻ, בוּ	u

• The position of the stress may be indicated using an acute accent on the main vowel of the stressed syllable (á, é, í, ó, ú).

—Ruakh_TALK 20:49, 23 January 2007 (UTC)

Someone over at Wiktionary:Christmas Competition 2008 just read this section and came away with the conclusion that "[t]he About:Hebrew talk page does not show any objections to the posted system". Let me, then, file an objection for the record. Well, not an objection precisely, but a preference for some differences from the current proposal. I prefer (chet and chaf transliterated as ch not kh (think German), yod as y even if it's the latter half of a diphthong, and, most especially,) vav transliterated as v not w.—msh210℠ 19:46, 3 December 2008 (UTC)

Since my Hebrew is so limited, I think that my opinion should be considered to hold half the weight of anyone else's, but here it goes. First of all, I agree with msh210 concerning vav. While it certainly used to be w, my impression is that it is v in a modern context. I strongly disagree with using y for yod when the latter half of a diphthong, as that simply makes no sense in an English context (which is really the context in which this scheme needs to be meaningful.) As for kh versus ch, that is tougher. Admittedly, ch is probably the most common method of representing that sound in an English context (a very rare instance, note), however, my fear is that it will get confused with the much more common (in English) /tʃ/, such as in chalk. — This unsigned comment was added by Atelaes (talk • contribs) at 20:20, 3 December 2008 (UTC).

Re: ch vs. kh: Yeah, maybe we should. Ch does seem to be more common in the world. I prefer kh, but my attempts at world domination have yet to bear fruit. :-(

Re: y vs. i: So for example, you'd write "aláy" for עלי, "ókey" for אוקיי, and "gilúy" for גילוי? Weird. I don't know how I feel about that.

Re: v vs. w: I totally agree, have been using v, didn't realize I proposed w, and have no idea why I did. (Maybe I meant to write that w was optional, like gemination and emphatic consonants, and just messed up? I do think we should allow w optionally for Ancient Hebrew.)

I also think alef and ayin should be optional in more cases, such as word-initially. I think the above proposal is a result of my not wanting to figure out how to express all those cases.

—Ruakh_TALK 20:26, 3 December 2008 (UTC)

In answer to your question re okey et al., yes, I would. And if we allow as an option different spellings for ancient Hebrew based on the supposed pronunciation, then, sure, allow w as such an option. And, yes, alef and ayin can be optional often, especially alef. Or, otoh, we can have no options at all, just one rigorously defined and enforced transliteration system. That's much better, imo, but, of course, we're not likely to agree on one.  :-) —msh210℠ 17:15, 4 December 2008 (UTC)

Well, personally I prefer a more Biblical style for roots (e.g. q-w-_ instead of k-v-_ for the root of tikvah), for citations from the Bible, and for etymologies of English words that come from Biblical Hebrew. But it might be good to have two rigorously defined and enforced transliteration systems, with other variations (such as mixing-and-matching) being forbidden, and not-so-rigorous guidelines for choosing the appropriate transliteration for a given word in a given context. —Ruakh_TALK 19:17, 4 December 2008 (UTC)

Ktiv khaser and ktiv maleh

I think we can all agree that where emot kria (matres lectiones) appear in Biblical Hebrew, we should certainly include them. My preference is that wherever they're used in Modern/Standard/Israeli Hebrew, the main entry should be at the article title including them, but there should also be an entry at the article title excluding them with a note directing people to the main entry. Does anyone object to this? Ruakh 20:00, 19 January 2007 (UTC)

I agree. The main entry should use the spelling which is most likely to be searched for; and when using unvocalised Hebrew the searcher will probably use standard (full) spelling. Shai 21:59, 19 May 2007 (UTC)

I disagree. The reason the emot k'ria are used is not because a linguist who only knows Modern Hebrew thinks the word is spelled that way; rather, they are used only to aid reading (hence the name em k'ria). Proof of this is that on the (admittedly rare) occasions Modern Hebrew is written with vowels, the emot are omitted. (That is, they are usually omitted, and, I suspect, always omitted in well-edited works.) So the "real" spelling is then without the em k'ria, and that's where the entry belongs. There should, however, also be an entry with the em k'ria, directing users to the entry without. —msh210 18:59, 3 July 2007 (UTC)

I don't follow. You're saying that the "real" spelling is the one we'd use if we wrote with vowel signs, except without the vowel signs? —Ruakh_TALK 19:36, 3 July 2007 (UTC)

No. (1) I maintain that the "real" spelling is independent of whether we use vowel signs (henceforth "vowels"). (2) I have not voiced an opinion on whether vowels should be included, or on whether the "real" spelling includes vowels. (3) But those (1 and 2) are merely to answer your question, and do not reflect my main point, which was: The "real" spelling is without the emot k'ria.

I backed up this main point with a two arguments: (a) Linguists will agree with me. (Note, though, that I have no proof of this.) (b) Books with vowels omit the emot k'ria.

And, finally, I stated my opinion/advice that the entries with the emot k'ria should exist also, but direct users to the other entries. —msh210 14:41, 4 July 2007 (UTC)

Linguists are scientists who study language as it is used. I can't imagine a linguist deciding that the "real" spelling is a rarely-used one. —Ruakh_TALK 17:35, 4 July 2007 (UTC)

You're right, of course, that linguists (most of them) are antiprescriptionists (is that a word?), who study language the way it's used. But my point was (in part) that the language is used -- spelled -- without the emot k'ria (as evidenced by books with vowels); the only reason people write them is to (to an extent) replace vowels.

Cf. the failed system, years back, of using slashes and other symbols between letters to replace n'kudos (n'kudot). The system was designed to make typing (on a typewriter) with vowels easier by having all symbols appear on one line instead of above and below. Now let's suppose hypothetically that this system had become widespread, but only for typing; for writing by hand, people still used ordinary n'kudos (when they used anything). Would a linguist, in our hypothetical world, say "the 'real' spelling is the one with the slashes"? Certainly not: the slashes would be a device used to maintain the information inherent in the n'kudos while keeping typing easy. End of hypothetical case. Our case is parallel to it:

	hypothetical	real
what are used	slashes and things	vavs and things
why they're used	to keep vowel information	to keep vowel information
what they simplify	typing on a typewriter	typing or writing
where they're not used	handwriting	children's books (e.g.)
why they are not used there	they don't simplify the activity (but still would carry enough information)	they don't carry enough information (but still would simplify the activity)

So the hypothetical is slightly different from our case, in the last line. But that just makes my argument stronger : {not carrying enough information about the pronunciation of a word} is more reason to say "this is not the correct spelling" than {not being easy to write} is! —msh210 19:08, 4 July 2007 (UTC)

Your analogy is flawed. Emot kria are much older than nikud; nikud were devised as a more precise way of indicating vowels than the existing system. Now, nikud, when used, didn't replace emot kria; rather, nikud were added in addition to emot kria — except that over time, more and more emot kria were used in nikud-less writing, but weren't added to nikud-ated writing. (There are minor exceptions in both directions: re-transcriptions of ancient texts don't add emot kria, so while there isn't new nikud-less writing omitting emot kria, there are new printings of old nikud-less writing omitting some; and conversely, there are some texts, especially bilingual dictionaries, that include both nikud and modern emot kria, using emot kria because they're part of the normal spelling, and using nikud to indicate pronunciation.)

Also, in general your argument strikes me as specious. What does it mean to say that a letter isn't part of a word's spelling, it's just added to make it easier to read? Isn't the point of spelling to make a word easier to read? Would you say that our entry for "viz" should actually be at "vi", because the "z" is just there to indicate it's an abbreviation?

—Ruakh_TALK 22:37, 4 July 2007 (UTC)

To respond to your first paragraph: Certainly some emot k'ria are as old as n'kudos (older, you claim; I don't know). Those are the ones you mentioned at the top of this discussion: I think we can all agree that where emot kria (matres lectiones) appear in Biblical Hebrew, we should certainly include them. We're not discussing those: we're discussing the new ones.

And to respond to your second paragraph: You ask What does it mean to say...? I think that the analogue I cited gives a good explanation of what I mean when I claim that the emot k'ria are there only to make reading easier, but are not part of the spelling. But you knew that: your point was more your second question: Isn't the point of spelling to make a word easier to read? Well, yes, but not conversely. That is, not everything that makes word easier to read is part of the spelling of the word. A good example is my analogue, above. Another might be serifs. (Serifs are just an example of things that make a word easier to read (or harder, but we can agree for argument's sake that they make it easier, and, moreover, that everyone agrees that they make it easier), so I'm using them as an answer to your question Isn't the point of spelling to make a word easier to read?. I am not claiming that they are a good analogue for emot k'ria: the differences between the two cases are many.)

As far as viz goes, two things. The word historically was an abbreviation, but currently, in English at least, is not. It's a word, and it includes the letter z (zee or zed). So it should be listed under its current spelling. Should we include vi, list its language as Latin, and have the sense be, as an abbreviation, whatever it stands for? Yes: that belongs properly s.v. vi (which I haven't checked) and not s.v. viz (which I also haven't checked). But the English word belongs under viz. (Evidence that the Latin word was spelled vi and the English is spelled viz lies in the use of the letter z in the English word, but a symbol indicating truncation in the Latin. But I grant that that's not extremely strong evidence.)

You will argue, then, as follows: "Aha! You're [i.e. msh210 is] saying that the historical spelling doesn't matter: we only care about the current spelling. But in Hebrew the current spelling is with emot k'ria!" To which I respond, no, the current spelling is without emot k'ria, and repeat all my arguments above. —msh210 20:55, 5 July 2007 (UTC)

By the way, included in this discussion, of course, are words like תכנית / תוכנית, which is spelled, when it's vowelized, with a kamatz katan. (I say "of course" but mention it anyway, in case someone was unaware.) —msh210 15:26, 4 July 2007 (UTC)

Vowel sign inclusion

I don't have much opinion on where vowel signs should be included, except that I agree with the current seeming consensus against including them in article titles, and I think it's strongly preferable to provide a version with vowel signs for each definition (especially since a given string of letters often produces different words with different vocalizations, e.g. dever vs. davar vs. ktiv-khaser-diber). Ruakh 20:46, 19 January 2007 (UTC)

I concur. Cerealkiller13 20:56, 19 January 2007 (UTC)

You can see two ways this is handled for Latin by looking at pilus and liber. --EncycloPetey 17:27, 24 January 2007 (UTC)

Cool, thanks. :-) —Ruakh_TALK 18:22, 24 January 2007 (UTC)

Regarding vowel signs in general (both in articles on Hebrew words and in Hebrew translations in English articles), one of the problems with including the vowel signs is that it is often difficult (at least for me) to decide when to put a patakh (ַ) vs a kamatz(ָ), or a tsere (ֵ) vs a segol (ֶ), and when to not include a vav with a holam or a yod with a hirik, or when to include a regular vowel or a hataf (ֱֲֳ), etc. AggyLlama 16:56, 10 February 2007 (UTC)

That's an argument for giving unvocalized (without-vowels) versions everywhere, but it's not an argument against giving vocalized (with-vowels) versions anywhere, as it's fully possible to give both vocalized and unvocalized versions, and if an editor doesn't know the vocalized version, (s)he can simply leave it out and wait for someone else to provide it. (We can also have a Category:Articles needing vocalizations of Hebrew for this purpose.) As noted above, I think we should give only unvocalized translations in articles on English words, but that we should give both vocalized and unvocalized versions in articles on Hebrew words. Every respectable Hebrew dictionary or Hebrew-English dictionary gives vocalized versions of words. —Ruakh_TALK 18:48, 10 February 2007 (UTC

I agree that article titles with vocalization are problematic, especially because most post-biblical words are - and were - never written with vocalization (except in dictionaries). However, for very long entries, especially when different roots are concerned, I wonder if some form of a disambiguation page won't be better, with links to vocalized entries. I know this is probably unprecedented (and unnecessary) in Indo-European words, but in Semitic words it might make things more "user friendly" for those who search for a word. Just an idea. Shai 22:22, 19 May 2007 (UTC)

Having vowel signs in the article would seem necessary if it were that there are three to seven possible words with simply the letters and no nikudot. Would then the version without nikudot lead to a disambiguation page with the possibilities? Neskaya 06:26, 10 July 2007 (UTC)

Root format

This issue has come up with Arabic, and I imagine that we may as well try and preempt it in Hebrew, because someone will certainly bring it up at some point. It is my opinion that Hebrew (and other Semitic languages) should be allowed the usage of a non-standard header "root" which would have the triletter root and all of the various nouns, verbs, etc, that come from it, their meaning, and a link to their full articles elsewhere. The title of the full article for each case would be the lemma form. This would mean that some full articles would appear on the full page, especially QAL conjugated verbs. Cerealkiller13 20:56, 19 January 2007 (UTC)

Hmm. I agree that there should be a "root" section (in addition to any "verb" sections, "noun" sections, etc.), but I think I disagree about what it should have. I think it suffices to give a one-sentence summary of what the words using the root tend to relate to (so for the "root" section of שטח, for example, we'd have a definition along the lines of "Words formed from this root typically pertain to area, to spreading out, or to flatness"), plus an etymology (if known), plus links to cognate roots in other Afro-Asiatic languages (if any — and especially when the etymology is not known). I don't think we need to list all of the words that come from the root, because that would duplicate the "related words" section, and nor to define them, because that's what their own articles are for. Ruakh 21:08, 19 January 2007 (UTC)

The way we do it for Arabic is simply to add an ==Etymology== section, as in Template:ARchar or Template:ARchar. Each different deriviative that shares an etymology gets the same etymology statement, but that is the only connection between them. —Stephen 21:25, 19 January 2007 (UTC)

So what happens if a root itself has an etymology (if it's a borrowing from another language, or is formed from an acronym or proper noun, or the like)? In such a case, does each individual word's etymology give the root's etymology? —Ruakh_TALK 21:35, 19 January 2007 (UTC)

A word borrowed from another language would merely have the usual ===Etymology=== From English poly, etc. I cannot think anything like an acronym in Arabic. If Arabic borrows an acronym from English, such as NATO, it just becomes a (borrowed) word. Also, Arabic doesn’t really have proper nouns. Proper nouns are an invention of the users of the Roman alphabet, with its capitalization (recently copied by Greek and Cyrillic). About the closest thing Arabic has to proper nouns is definite nouns. In any case, if a word is taken from another language, it gets the expected etymology. If it’s a native word, it usually has a triliteral root (although there are also some mono-, bi-, and quadriliteral roots). —Stephen 22:36, 19 January 2007 (UTC)

There are some more complex cases at Template:ARchar and Template:ARchar. This is because the entries are primitive verb roots, and Arabic verbs have eleven different derivative classes. —Stephen 22:46, 19 January 2007 (UTC)

Sorry, but that doesn't really answer my question. Maybe this doesn't happen in Arabic? In Hebrew there are actual roots borrowed from other languages, e.g. טלפן from English telephone (and its counterparts in other languages); so there's not necessarily a dichotomy between word-from-another-language and word-from-an-n-literal-root. And even native Hebrew roots sometimes have recognizable etymologies, e.g. תרם from a reanalysis of תרומה (which comes from the root ר?ם). So would these etymologies be given in the entries for each word that uses these roots? —Ruakh_TALK 23:10, 19 January 2007 (UTC)

Yes, Arabic borrows some words such as telephone as well (Template:ARchar from English, along with the native Template:ARchar). The etymology of Template:ARchar will say it’s from English; the etymology of Template:ARchar will say that it’s from the primitive root Template:ARchar, meaning to shout, to call out. The biggest difficulty, as far as I can see, is that some words are a little tricky to figure out, especially where one of the radicals is a, w, y, or glottal stop, because they often morph into one another and fade in and out of existence. Double consonants can also be confusing, since they are not written except in forms that have an intervening vowel. —Stephen 21:41, 23 January 2007 (UTC)

Maybe that approach makes sense for Arabic for a reason that I don't understand (not surprising: I don't know Arabic), but that just doesn't make sense to me for Hebrew. The etymology of the verb טלפן (tilpen) is that it comes from the root טלפן (T-l-p-n), which in turn comes from the noun טלפון (telefon), which in turn comes from the English noun telephone (and other languages' counterparts). I think it would be misleading to say that the verb טלפן comes from the English verb to telephone, because that's really not quite how it happened. (Hebrew doesn't borrow verbs very easily; it has a strong tendency to borrow nouns, then infer roots from those nouns, and finally form verbs.) —Ruakh_TALK 22:04, 23 January 2007 (UTC)

Yes, it’s the same with Arabic, which also does not adopt foreign words easily. Arabic will typically borrow a noun, then make a verb of it, and other forms from that. I still don’t really understand what the problem is. In the case of native Template:ARchar (hatif), it’s from the primitive Arabic root Template:ARchar (hatafa); in the case of טלפן, it’s from טלפון, which is from telephone. Sometimes it’s hard to know whether the noun came first or the verb, but in Arabic dictionaries, it is standard convention to move from the verb to the other parts of speech. Therefore, the provenance of the noun Template:ARchar (tilifún) is (even if only conventionally) from the verb Template:ARchar (talfána), and that is from the English noun telephone. I wonder if we’re talking about different things, because I can’t see a difficulty. —Stephen 22:50, 23 January 2007 (UTC)

Ah, I see. The difficulty is just that that's not the Hebrew convention; even though Hebrew and Arabic are the same in this regard, Hebrew articles can't use the convention for Arabic articles, because it's just not the Hebrew convention. (I happen to think the Hebrew convention makes more sense, but I suppose I'd feel differently if I had grown up with the Arabic convention.) Thanks for your explanations. :-) —Ruakh_TALK 17:20, 24 January 2007 (UTC)

On second thought, I agree that the root section shouldn't include the definitions, a simple list will suffice. Here's how I would format a fictitious entry:

Hebrew

Root

xxx Generally pertaining to pancakes, and the eating, making, and pouring on of syrup of them.

Derived terms

• xx1 - noun
• xx2 - noun
• xx3 - QAL verb
• xx4 - Peal verb
• xx5 - Niphal verb
• xx6 - Hitpalel verb

Noun

xx1 (with vowel markings)

1. pancake

Verb

xx3 (with vowel markings)

1. to make pancakes

Verb

xx4 (with vowel markings)

1. to eat pancakes

What we could do is in the root section have the vowel markings listed on the page, but not in the link. So it would be *[[xx1 (with markings)|xx1 (without)]] . I realize this is somewhat redundant, for example with xx1 being listed under the root, and then being again listed with its own section, but I think this is necessary in that it makes the format much more flexible, while still allowing for a great deal more standardization. I apologize that this example is so sloppy and ridiculous, but for some reason I have the damndest time trying to get a bunch of Hebrew words to line up properly, they always seem to interfere with each other, and I lost patience. Cerealkiller13 21:55, 19 January 2007 (UTC)

I don't think your example was sloppy or ridiculous at all; if nothing else, made-up examples have the advantage of allowing us to discuss general principles without worrying about the details of specific roots.

I'm mostly O.K. with your set-up, but you seem to be using "derived terms" in the sense of "words using the root in question", which I guess makes some sense, but which is a bit misleading, because that can include some words that probably aren't really derived (in the non-Semitic-specific sense) from a given root (e.g. midbar and dever from dalet-bet-reish; or sapar and histaper from the same samekh-pei-reish), and can exclude terms that really are derived (in the non-Semitic-specific sense) from that root (e.g. taram from reish-vav/yud/?-mem, and din v'kheshbon from dalet-vav/yud/?-nun and khet-shin-bet). Is there maybe a different word we can use here instead of "derived"? —Ruakh_TALK 23:22, 19 January 2007 (UTC)

Yes, I think you're right. Perhaps Hebrew will, yet again, be forced to use a non-standard header. In which case, we'll have to think of a word which will clearly convey what these things are. Perhaps we could substitute a "forms" header? Cerealkiller13 00:19, 20 January 2007 (UTC)

Again, I agree on almost all points discussed here. There is one issue I'm afraid I have to disagree on. Unlike Arabic, roots in Hebrew have been traditionally set apart from regular words by typographic means: either by quotations marks (בו"ש) or by hyphens (ב-ו-ש). This tradition has been reaffirmed by the Academy of the Hebrew Language as being the official guideline, and it is widely accepted (also by the Hebrew wiktionary).
For wiktionary it has the advantage of (1) not confusing roots with verbs/nouns while linking; (2) preventing the formation of super long pages containig all roots, all root forms, and all real forms of both Hebrew and Aramaic on one page; (3) emphasizing that this is not a word, but a hypothetical, theoretical, unattested form; (4) uniformity with the Hebrew wiktionary.
I recommend that wiktionary will adopt this widely used practice in handling Hebrew roots. You can see an exapmple here: ב-ו-ש. Shai 17:06, 24 May 2007 (UTC)

You know, that's how I always write it when I'm actually writing (the makav way, I mean, not the gershayim way), but for some reason I felt weird doing that here. I don't know why, though; that's obviously the sensible way to do it. (The funny thing is that even here, I've written some romanizations that way, e.g. at ראש and איחוד. I really don't know what I've been thinking.) So yes, agreed, thanks. :-) —Ruakh_TALK 17:25, 24 May 2007 (UTC)

Except: all entries in the en.wikt are words spelled the way they are actually written in use. If the he.wikt wants to have a different class of entries we can't help that (they seem to do everything differently ;-). Typographical conventions of dictionaries don't enter into it. (For example, Random House has headwords like dic·tion·ar·y as do others, we don't do that.) The "headword repeater"/"inflection line" sometimes shows a different form (e.g. Latin, which is very unhelpful because the only thing in entries that shows how it is actually written is the entry title ;-) ב-ו-ש really has to be at בוש in the en.wikt. Robert Ullmann 12:28, 25 May 2007 (UTC)

So does this mean that -ed is actually written with a hyphen? ;-) —Ruakh_TALK 22:04, 25 May 2007 (UTC)

Inflected forms (1)

Wiktionary does generally include all inflected forms of an English word, but I think in Hebrew we probably need to limit that a bit, seeing as a Hebrew noun can have up to 28 forms (what with singular/dual/plural and indefinite/definite/construct/construct-with-pronoun), and a Hebrew verb can have many more than that (if we count all the permutations of subject and direct object, since in Biblical Hebrew the direct object could be reflected as a suffix on the verb). For nouns, I suggest we give only the singular, dual, and plural indefinite and construct forms, where they exist, since the definite forms are always obvious from their indefinite counterparts, and we can provide a Wiktionary Appendix on noun declension that covers how to get a construct-with-pronoun form from a construct form. For adjectives, likewise, I suggest we give only the singular and plural masculine and feminine indefinite forms. For verbs, I don't know what to suggest, but we obviously can't give full conjugations for all verbs. And for prepositions, I think maybe we should give all the forms, at least for the irregular ones, since there aren't that many prepositions anyway. —Ruakh_TALK 22:10, 19 January 2007 (UTC)

I believe it is best to have every form. Ancient Greek has a boatload of forms, and we're trying to include them all. As an example, see φιλέω (make sure to hit the "show" button under the inflection header). I don't think its necessary to have separate entries for every form, but I think it's good practice to at least show them on the lemma entry. Perhaps, in time, we'll have bots to write entries for all the forms for us (like the Daverossbot is doing for Spanish). Cerealkiller13 22:07, 23 January 2007 (UTC)

Oh, I forgot about the togglable things. Yeah, you're right, we should just give full conjugations, behind one of those. (Well, I'm not sure we should include direct-object-containing verb forms, because that seems over-the-top even if it's going behind one of those, but all the "normal" conjugation/declension/etc. should definitely be included.) —Ruakh_TALK 22:44, 23 January 2007 (UTC)

I just added a @#%$load of translations to with. As with any preposition, this means a LOT of translations. Does this seem reasonable? Or would it be better to just point to the base form of the preposition, which would have all the different forms on a separate page? AggyLlama 03:17, 4 March 2007 (UTC)

Wow, that's a tough call. Hebrew is rather unique in having inflected prepositions (even Ancient Greek doesn't do that), and I don't know if we've come across that issue before. Perhaps someone who's been here longer will know. My first inclination is that what's there now is too much. Only the lemma form should be there, and it should link to the entry, which should have all the forms. But I'm open to other opinions. Atelaes 04:35, 4 March 2007 (UTC)

I agree. Firstly, with other parts of speech, we only give lemmata in translation sections; secondly, since néged (against) and negdó (against him/it-MASC) differ, it's reasonable to view néged itself as translating the actual preposition, and -ó as translating the pronoun. (Contrast, say halákh, which reflects the person, gender, and number of the subject whether or not an explicit subject is included.) —Ruakh_TALK 06:48, 4 March 2007 (UTC)

So I should create a new entry for [עם] and for [נגד], translate only to that base form, and then expand upon the various conjugations of the preposition in the page for the preposition itself? AggyLlama 18:42, 4 March 2007 (UTC)

That sounds about right to me. What would be really useful would be if you created a full inflection template for prepositions and then created those main pages and used them. That way, the creation of future pages would be much easier. Hebrew will have some unique inflection issues, and I'd be more than happy to help you work on the templates, if you should need any. If you don't feel like going through all that, you're certainly not obligated to do so. Atelaes 20:15, 4 March 2007 (UTC)

How's {{he-prep-inflection}}? —Ruakh_TALK 04:52, 5 March 2007 (UTC)

Hmmmm.....that's not quite what I was thinking. What I think would be best is a template which will do the work for you. That template still requires the user to know and enter in all the forms by hand. What I was thinking of is a template in which you enter the three letters, and then it inflects them and creates all the forms for you, by magic. This could be tough, because I don't know if there's a way that it can "attach" the vowel markings to the letters, or if we'd have to program them all in by hand. In addition, I find it nice to have things ordered with rows of 1st, 2nd, and 3rd person and columns of singular or plural. Unfortunately I don't have time to write a template like that tonight, but I may be able to tomorrow. I'll try and be as prompt as I can. Atelaes 05:17, 5 March 2007 (UTC)

The problem is, it doesn't seem to me like the various forms are perfectly predictable; there are various patterns, but they don't seem terribly hard-and-fast. :-/ —Ruakh_TALK 05:56, 5 March 2007 (UTC)

Huh, I thought they inflected rather normally. Would you be willing to write up an entry for two prepositions with different inflection, so I can take a look at what we'd be dealing with? Templates can handle certain exceptions, but there admittedly are limits to what they can do. Atelaes 06:27, 5 March 2007 (UTC)

Oh, I didn't address your other points. Yes, it's possible to attach vowel markings to template-parameter-provided consonants; to avoid editing complications, I'd recommend using numeric character references (e.g., ָ for the kamatz). And a table would be quite nice; I'm just really bad at formatting those nicely, so thought I'd leave that to someone else. ;-)

At any rate, I don't know how much it'll help you, but below are the inflectional paradigms of two common Hebrew prepositions (both meaning to, but in slightly different senses). In a sense, they represent the two broad categories of prepositions: those that inflect rather like singular nouns in the construct state, and those that inflect rather like plural nouns in the construct state. (Not all prepositions are so simply classified, though; for example, Template:he-onym is quite strange, and Template:he-onym is actually suppletive.)

    Template:he-onym

Non-personal-pronoun-including form				Template:he-onym
Personal-pronoun- including forms	Singular		Plural
	Masculine	Feminine	Masculine	Feminine
First person	Template:he-onym		Template:he-onym
Second person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym
Third person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym

    Template:he-onym

Non-personal-pronoun-including form				Template:he-onym
Personal-pronoun- including forms	Singular		Plural
	Masculine	Feminine	Masculine	Feminine
First person	Template:he-onym		Template:he-onym
Second person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym
Third person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym

—Ruakh_TALK 08:02, 5 March 2007 (UTC) and later fixed mis-inflection of l'-

I just have a comment about the specific forms you used for אל... The 1st person singular would never be spelled with two yod's, it would always be written as אלי. I think the same goes for 2fs, although I find it more plausible that that one would be spelled with a double yod. Also, I would say say that colloquially (and almost ubiquitously in spoken Hebrew), 2xp and 3xp are pronounced with an 'e' sound in the beginning rather than an 'a' sound. AggyLlama 08:24, 5 March 2007 (UTC)

I'm not sure where you've gotten the idea that elái is never spelled with two yud's … google:allintext:אלי pulls up 2.55 MGhits, of which (looking through the first few pages of hits) about 90% seem to be the given name Éli, whereas google:allintext:אליי pulls up 656 kGhits, of which all seem to be elái. A bit of math, and we have that on the Web at least, it seems that elái is spelled with two yud's more than 70% of the time — probably much more, but I'm not sure. (Of course, there's the complication that nikud-less spelling differs from spelling with nikud. Maybe it would make sense to use ds and dsvw, but not wv, since when vowels are used, so is defective spelling?) (comment continues below)

I don't particularly care to get in an argument about it, especially since I seem to converse far less frequently than the two of you, and that would make for a rather frustrating discussion at least on your end. As for the one vs. two yods thing - a one yod search on Google turns up about 8 or 9 times more hits than a search with two, so comparing the relative number of relevant hits on the first page is not much good - plus the one with two yods is not ambiguous. Anyway, if you check for a phrase such as בא אלי vs. בא אליי, which seems to be a much more balanced comparison, the version with one yod shows up about twice as much as the one with two. Regardless, I stand corrected about the version with two yods never appearing. I, personally, don't recall ever having seen it, and it struck me as quite odd when I first saw it written here. However, the facts do speak for themselves, and given that בא אליי does return at least 14K hits, obviously someone is using it, so I hereby rescind my previous objection. — This unsigned comment was added by AggyLlama (talk • contribs) at 05:21, 7 March 2007 (UTC).

It looks like a lot of those are from Bible sites and whatnot (so not really reflective of Modern Hebrew), but even discounting those, it looks like the single-yud version is more common than I'd thought. I'm no longer so sure which is more common in Modern Hebrew. —Ruakh_TALK 06:01, 7 March 2007 (UTC)

(comment continued from above) As for a vs. e — I agree, but I wasn't sure what to do. I felt like I had to give the romanization that corresponded to the nikud, even though I'm not sure I've ever heard anyone use an a sound there (if only because for most speakers that would make it homophonous with forms of Template:he-onym). What do you think we should do? Give the "grammatically correct" nikud, but a romanization that reflects how normal people say it? Or the "grammatically correct" nikud, but both romanizations? Or something else entirely?

This is actually a discussion that will affect a lot of things — there are a lot of words that people don't usually pronounce "grammatically correctly".

—Ruakh_TALK 18:09, 5 March 2007 (UTC)

The Romanization should simply be done from the spelling, regardless of pronunciation, the pronunciation (done in IPA) can take care of the pronunciation. As for the preposition table, I suppose it is too irregular to make a template for it. Sigh. Well, I'll try and reformat the table you made into rows and columns sometime today. But, I'm sure we'll want to make some templates for nominal and verbal inflection at some point. Atelaes 18:43, 5 March 2007 (UTC)

Ok, I've formatted the template in a way I think is a little easier to take at a glance. Someone should probably change the colour scheme, as it's atrocious, and I'm no good with colours. I attempted to implement it into אל, but could not get it to go properly. For some reason, whenever I try and use Hebrew characters, I get characters disappearing and rearranging themselves at will. I don't know exactly why, but I think my computer does not quite have a handle on right to left, mixed with left to right text. It's incredibly frustrating. In any case, someone who owns a computer which is not inept should implement it on a real page, and then we can see how well it works. Atelaes 21:55, 5 March 2007 (UTC)

Well, I've done it now, but really, you could have just copy-and-pasted the above. :-)

One confusing thing with mixed directionality in a wiki is that the wiki markup mostly consists of characters with neutral directionality (square brackets, curly brackets, equals signs, pipes), so if they're preceded or followed by English characters, they display left-to-right in editing, but if they're surrounded by Hebrew characters, they display right-to-left in editing. (Especially confusing is that right-to-left square and curly brackets are inverted — for example, right-to-left open-square-bracket looks like ] — though not all browsers handle it correctly.) The way I address this is by using lots of named parameters in templates I create, so you can order things such that there no special characters with Hebrew on both sides. This way everything is strictly left-to-right, except for the Hebrew words themselves. As far as I can tell, all modern browsers handle things fairly intuitively in this case (though it takes a bit of practice to develop the right intuition).

I've posted a few thoughts/questions on the template's talk-page; I'd appreciate it if you could take a look when you get a chance.

—Ruakh_TALK 05:12, 6 March 2007 (UTC)

Actually, the direction handling is a function of the font, not the computer or operating system. Arabic and Hebrew Unicode fonts have left-to-right built right into each alphabetic letter. Punctuation and symbols are neutral is this regard and they get their direction from the surrounding alphabetic letters. Until you get used to the logic, the easiest way to do something like this is to lay down all the English and punctuation and symbols first, putting [[]] as a place holder for each Hebrew word. Afterwards, type the Hebrew out on a separate line, then copy and paste each Hebrew word inth the empty [[]]'s. It will appear to become jumbled, but it will work correctly.

You asked for my input on this, and I tried to write a template at Template:he-prep, but it only works for certain words (likr ל).

    Template:he-onym

Non-personal-pronoun-including form				Template:he-onym
Personal-pronoun- including forms	Singular		Plural
	Masculine	Feminine	Masculine	Feminine
First person	Template:he-onym		Template:he-onym
Second person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym
Third person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym

You should ask User:Robert Ullmann, who is especially good with writing templates. —Stephen 14:51, 6 March 2007 (UTC)

Thanks! I've used your work to made some improvements, but unfortunately, as it stands, it's only correct for three prepositions: Template:he-onym, Template:he-onym, and Template:he-onym. Even for these it's not absolutely perfect: I added stress-indicating accent marks, but some forms of Template:he-onym and Template:he-onym are monosyllabic, making it a bit silly to indicate stress. (Ah, well).

To support the other singular-construct-form-like prepositions, I think what we need to do is have a toggling parameter that tells the template whether the preposition is like the above three (forming second- and third-person plural forms in -akhé[mn] and -ahé[mn]), or like Template:he-onym and Template:he-onym and all other singular-construct-form-like prepositions (forming second- and third-person plural forms in -'khé[mn] and -á[mn]).

—Ruakh_TALK 16:21, 6 March 2007 (UTC)

O.K. I think I've mostly managed. Here are Template:he-onym and Template:he-onym:

    Template:he-onym

Non-personal-pronoun-including form				Template:he-onym
Personal-pronoun- including forms	Singular		Plural
	Masculine	Feminine	Masculine	Feminine
First person	Template:he-onym		Template:he-onym
Second person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym
Third person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym

    Template:he-onym

Non-personal-pronoun-including form				Template:he-onym
Personal-pronoun- including forms	Singular		Plural
	Masculine	Feminine	Masculine	Feminine
First person	Template:he-onym		Template:he-onym
Second person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym
Third person	Template:he-onym	Template:he-onym	Template:he-onym	Template:he-onym

The only things that seem to need improvement still are: (1) shouldn't stress the ending if there's only one syllable overall; (2) the 2nd-person endings (except feminine singular) sometimes start with a sh'va na', sometimes with a sh'va nakh, depending on the stem. The romanizations need to be able reflect that (by including an apostrophe or not).Done. Is there anything else, that I'm missing?

—Ruakh_TALK 17:30, 6 March 2007 (UTC) and 17:42, 6 March 2007 (UTC)

The first person plural form of אצל is coming out wrong - the kamatz in the lamed should actually be a tsere. Similarly, the second person singular feminine should also have a tsere under the lamed rather than a kamatz. AggyLlama 16:07, 7 March 2007 (UTC)

Oh, darn, you're right. When I designed the template, I was actually thinking of Template:he-onym, and then when I plugged in Template:he-onym, I didn't pay enough attention to notice that problem (I guess I got sidetracked by the problem with the sh'va na'). So, what to do? The most general solution is to have {{he-prep-sing}} support all the same parameters that {{he-prep-sing}} does, but have them all be optional; that is, so you could specify (in this case) values for 1p and 2fs to override the values that {{he-prep-sing}} would otherwise compute. —Ruakh_TALK 16:40, 7 March 2007 (UTC)

Maybe this is a naïve question, but why do we need a template to do the work for us? It's not like prepositions are an open class - wouldn't it be less work to just enter the data manually? If the push to have a template is that strong, perhaps you can include the ability to override each form individually in case the default template is wrong: Have the optional parameters 1ps, 1pp, 2psm, 2psf, 2ppm, 2ppf, etc. AggyLlama 16:55, 7 March 2007 (UTC)

Well the new template looks quite good. I personally disagree with the need to include Romanizations of every form, but as I've stated before, I don't know all that much about Hebrew, and so will abstain from taking a strong stance on it. The question of whether we really need a template for prepositions has merit. Prepositions are a relatively small group, and the need for automation isn't urgent. However, there will need to be templates for nouns and verbs, and I figured that prepositions would be a good place to start, since they're a smaller group, they're a bit easier to start with. I still wish the template was a bit more automated (I still would like to see a template which takes the three root letters as the only input), but sadly I don't know enough Hebrew to really contribute to this project. I'm really sorry that I keep offering criticism but can do little to help accomplish anything. As for the idea of having every form have an override for irregular forms, I think that's an excellent idea, and it is what I've started working towards with the Ancient Greek templates. I'm going to put a technical question concerning Hebrew on the Grease Pit, if anyone reading this thread has anything to add to it (either in the form of answers or further questions) please do so. Again, excellent work on the template Ruakh. Atelaes 21:49, 7 March 2007 (UTC)

You keep talking about "three root letters", but few prepositions have real roots; indeed, Template:he-onym against is the only preposition I can think of off-hand that is related to other words, not counting prepositions that are actually composed of other words like Template:he-onym before (literally "to the face of") and Template:he-onym behind (literally "from behind of"). —Ruakh_TALK 04:06, 8 March 2007 (UTC)

Oh. Well, I feel quite silly. Atelaes 04:20, 8 March 2007 (UTC)

Inflected forms (2)

When the existence of homographs doesn't prevent this, I think inflected forms should simply redirect to the corresponding citation form. —Ruakh_TALK 00:12, 20 January 2007 (UTC)

I think you'll run into problems if you try and do this. Redirects are generally frowned upon in Wiktionary. What I would suggest is that you simply leave inflected forms alone for the time being. If you must, put a soft redirect page up: Hebrew- noun- xx1- suchandsuch form of [[xxx]]. Such a page is short and sweet, gives all the necessary information for the particular form, and then directs the user to the lemma page, which has all the information for the word in general. Also, this works even in the presence of homographs. Cerealkiller13 22:51, 23 January 2007 (UTC)

k, thanks. :-) —Ruakh_TALK 03:37, 24 January 2007 (UTC)

O.K., so it occurs to me that we need to name the various inflected forms. My proposal is that we describe inflected forms in the following fashion (omitting any descriptors that don't apply):

1. <person> <gender> <number> <modern-ish name(s)> (<ancient-ish name>) of <lemma>

for example:

1. First-person plural future tense (prefix conjugation) of Template:he-link.
2. Masculine singular present tense and present participle of Template:he-link.
3. Plural construct form of Template:he-link.

I propose the following "<modern-ish name(s)> (<ancient-ish name>)" descriptions:

• for verb forms:
  • "past tense (suffix conjugation)"
  • "present tense and present participle" (many of these will warrant separate adjective entries as well)
  • "future tense (prefix conjugation)"
  • "affirmative imperative" ← Template:he-link etc. (pu'al and huf'al verbs don't have this)
  • "jussive subjunctive" ← Template:he-link, Template:he-link, etc. (most verbs lost this in pre-historic times)
  • "to-infinitive" ← Template:he-link etc. (pu'al and huf'al verbs don't have this)
  • "bare infinitive or gerund (infinitive construct)" ← Template:he-link etc. (pu'al and huf'al verbs don't have this)
  • "infinitive absolute" ← Template:he-link
  • "action noun or gerund" ← Template:he-link etc. (pu'al and huf'al verbs don't have this) (many or most of these will probably warrant separate noun entries as well)
  • I do not propose any name for passive/past participles (Template:he-link etc.), because it's not clear to me that these have any consistency in sense between verbs (what do Template:he-link and Template:he-link have in common, meaning-wise?), which is really the most important consideration for a dictionary. Further, they're not used in actual verb forms (unlike past/passive participles in English), so it makes sense to treat them as derived adjectives, noting their source verbs and roots in their etymologies.
• for adjective forms:
  • "definite form"
  • I propose that we leave indefinite forms unnamed, as they're very much the basic form; "Masculine plural of Template:he-link" is perfectly clear.
• for noun forms:
  • "definite form"
  • "construct form"
  • "construct form of <lemma> containing the <person> <gender> <number> person pronoun"
  • I propose that we leave indefinite forms unnamed, as they're very much the basic form; "Plural of Template:he-link" is perfectly clear.
• for prepositions:
  • "personal-pronoun-containing form"

Some further points:

• For pu'al and huf'al verbs:
  • The entry for the lemma should have the definition "Passive of ___." (It won't be just a soft redirect, because it will give the conjugation and may have a list of derived terms, but it won't have a real translation.)
  • Entries for inflected forms should link to both lemmata, using the formula "___, the passive of ___"; for example, "Third-person masculine singular future tense (prefix conjugation) of Template:he-link, the passive of Template:he-link."
• Even outside of pu'al and huf'al, many verbs can simply be defined as "Passive of ___" or "Reflexive of ___" or "Causative of ___", where ___ is another binyan of the same root; however, since there are no consistent rules for these relationships, I think it makes sense to consider these to be separate but related verbs, and propose that we give a real translation as well as the simple definition.
• For feminine nouns derived regularly from masculine counterparts with no change in sense besides the natural gender of the referent:
  • The entry for the lemma should have the definition "Feminine counterpart to ___." (It won't be just a soft redirect, because it will give the declension and may have a list of derived terms, but it won't have a real translation.)
  • Entries for inflected forms should link to both lemmata, using the formula "___, the feminine counterpart to ___."

Do those make sense as ideas, both overall and in the details? Am I being stupid about anything? Please comment/criticize/accuse/attack/assault! :-)

—Ruakh_TALK 04:28, 27 March 2007 (UTC)

שטח

שטח is a first iteration of matching the above discussions, except that it doesn't include full conjugations (yet).

What do y'all think? Does it look reasonable?

—Ruakh_TALK 18:33, 25 January 2007 (UTC)

I think it looks excellent. One minor thing, if I had done it, I would have included the types of verbs at some point (Niphal, QAL, hitpalel (sp?)). I think you know rather more about Hebrew than I, so I'll leave it at your discretion. Cerealkiller13 18:51, 25 January 2007 (UTC)

That seems mostly unnecessary to me, because the names of the binyanim are just the lemma forms of verbs formed from the root pei-'ayin-lamed; so, the verb diber is of binyan pi'el (replace the dalet with pei, the bet with 'ayin, and the reish with lamed to the get the name of the binyan) — and when there is ambiguity, as in the verb nigash (which in the present and past tenses acts like nif'al and in the infinitive, imperative mood, and future tense acts like pa'al), it seems like we wouldn't want to have to deal with that in the listing of forms. It also has the complication that nouns and adjectives also usually belong to mishkalim (a mishkal being the noun/adjective equivalent of a binyan), so I'd feel that if we start doing that, we'd want to identify nouns and adjectives with their mishkalim, only that's much blurrier, because there's not as clearly fixed a number of mishkalim and not all nouns and adjectives belong to one. It seems like in all the cases where it's complicated enough for it to be useful to include the binyan/mishkal, it's complicated enough for it to be non-trivial to decide what binyan/mishkal to include.

That said, as a non-native speaker it seems like you're in a better position to judge what would be useful for an English speaker using Wiktionary's Hebrew entries, so if you think it would be useful, then I guess we should include it.

—Ruakh_TALK 01:05, 26 January 2007 (UTC)

I guess I must admit that I do think it would be helpful to an English speaker. If memory serves correctly, the different binyanim have the same (or at least similar) conjugation patterns, and so a person could remember that שטח means to spread out in the פָּעַל formation and would thus be able to write or say the verb in its various forms, only having to memorize five or six (or however many binyamin there are) conjugation tables instead of a new one for each word. As for the mishkalim, I think those are less necessary. And if you don't include them, and someone comes along and wants them to be there, you can always reply with, "Go for it, and good luck to you" in true Wiktionary spirit. One other thing that could be done with this is link it to a help page, in a similar way as "first declension" is linked to on Σατανᾶς (it just happened to be the last entry I edited). Again, if you don't feel like doing it, don't bother. If I want them to be there, I can always get up off my lazy ass and do it myself. Cerealkiller13 01:58, 26 January 2007 (UTC)

Oh! I see what you're saying. I misunderstood. For some reason I thought you were talking about the list of forms in the "Root" section; but you're actually saying that within the "Verb" section it should identify the binyan, aren't you? Yes, in that case I definitely agree. —Ruakh_TALK 04:01, 26 January 2007 (UTC)

Well, to be honest, I was originally thinking in the root section, so you did not misunderstand me. But, as I think about it, the verb entry is in fact a better place for it. So, well, at long last I suppose we're on the same page.  :) Cerealkiller13 04:19, 26 January 2007 (UTC)

prefixes

I just edited the entry for of and had need to mark two of the translations into Hebrew as prefixes. I wasn't able to find any template that marks a prefix, and I'm loath to just write it as text in the entry. Is there some sort of prefix template that I missed? If not, would anybody mind terribly if I added one? AggyLlama 16:45, 13 February 2007 (UTC)

I'm a bit confused; there don't seem to be any {{pronoun}} or {{preposition}} or any other such; are you sure there's supposed to be a {{prefix}}? —Ruakh_TALK 17:24, 13 February 2007 (UTC)

I don't know what you mean by supposed to be... When the translation is just a particle and not an entire word, It'd be nice to be able to mark it as such...no? AggyLlama 20:21, 13 February 2007 (UTC)

You may want to discuss this with some people first. If everyone put parts of speech on all words in the translation tables, it would be quite a mess. Perhaps prefixes could be an exception, but I'd get opinions on it first. Personally, I think it's best to just have the word and its romanization, and nothing else (it is linked, after all). But that's just me. Atelaes 20:43, 13 February 2007 (UTC)

I'm not talking about marking the POS, just the fact that this is not a complete word, but rather attaches to the word following it. AggyLlama 04:34, 14 February 2007 (UTC)

Oh, I see. The usual way to indicate a combining form is to mark it with a hyphen indicating where it connects to the other word: מ- (mi-), -ות (-ut), etc. Unfortunately, you need to precede the link with <span dir="rtl"> and follow it with </span> to make sure that the hyphen ends up on the correct side. (Maybe we should make a template for this?) —Ruakh_TALK 20:58, 13 February 2007 (UTC)

I think that's not a bad idea. I could create a template called {{rtl}}, which takes in as a parameter text in Hebrew. Or, I could follow the lead of Arabic and create a {{HEchar}} template, which takes in as a parameter text in Hebrew. The Arabic template doesn't actually contain a dir='rtl' bit, but the hebrew one could. What do you think? AggyLlama 04:34, 14 February 2007 (UTC)

For that matter, we could just combine the Hebrew text and the romanization into one template, which wraps the first parameter in a span='rtl' and wikifying brackets, and puts the second in parens italicized. Call it something like {{he-translation}}. AggyLlama 04:39, 14 February 2007 (UTC)

We can't have a {{rtl}}, or if we do we can't call it that, because 2- and 3-letter template names are reserved for ISO 639-1 (alpha-2) and 639-2 (alpha-3) language codes. (rtl isn't currently an alpha-3 code for any language, but that could change.)

The {{he-translation}} idea sounds great — it will really minimize the coding people need to do. Will you make it, or shall I?

—Ruakh_TALK 05:36, 14 February 2007 (UTC)

I've never created a template, so if you don't mind, I'd like to try. AggyLlama 06:14, 14 February 2007 (UTC)

On second thought, I'm not feeling to great, and I don't really understand the format of what you did in {{romanization of Hebrew}}, so why don't you go ahead and do it, and I'll just start using it... AggyLlama 06:30, 14 February 2007 (UTC)

Done. Let me know if there are any problems with it. —Ruakh_TALK 08:06, 14 February 2007 (UTC)

I've been using it for a while now and it seems great. There's a problem however for people who want to display the voweled form but link to the unvoweled form - there is no way to do that currently afaics. AggyLlama 16:43, 15 February 2007 (UTC)

I wasn't sure quite what you had in mind; can you take a look ({{he-translation}}) and tell me if that's what you meant? (By the way, if you have any questions about how the template works, feel free to ask them at User talk:Ruakh and I'll do my best to explain.) —Ruakh_TALK 21:23, 15 February 2007 (UTC)

It looks cool. I didn't really have anything specific in mind, as I don't like giving vocalized translations. But I was editing a page which had a vocalized spelling in it, and I wasn't quite sure how I could convert it to using this template, which is why I asked. The named parameter workaround seems like a great solution :). AggyLlama 04:24, 16 February 2007 (UTC)

"Biblical Hebrew"

This is used as the language header in some 324 pages (see User:Robert Ullmann/L2). In ISO/SIL, the language code is hbo and it is called Ancient Hebrew. I don't think we really want to call this Biblical Hebrew? It is the language of the Torah. And the people of the time, and still used in some dialects by some modern groups.

In any case we want to code Template:hbo and use a standard name. Robert Ullmann 11:26, 3 March 2007 (UTC)

Incidentally, I do believe all of those entries are by Dubaduba, and need to be cleaned up. If they're anything like his Ancient Greek entries, they need to be cleaned up a lot. I'm about halfway through his Greek entries, and I had considered going through his Hebrew entries afterwards. However, my knowledge of Hebrew is scant compared even to my minimal knowledge of Ancient Greek, and I think they would greatly benefit from the graceful touches of someone who knows Hebrew a bit more than I, if anyone's up for that challenge. Atelaes 22:25, 3 March 2007 (UTC)

One question that these entries raise is whether Hebrew should be divided between Ancient and Modern. Now, I know that modern Hebrew is remarkably similar to ancient Hebrew, considering the span of time which has passed between them (much like ancient and modern Greek), but I guess I really don't know exactly how different the two are. I think a large part of the reason why it makes sense to distinguish between Ancient and Modern Greek is because of the different orthography. With Ancient using breathing marks and polytonic accentuation, and modern using monotonic accentuation and no breathing marks, many entries would get split simply because of that. If it wasn't for that, I think the decision might be different. How does modern Hebrew compare with that of the Bible? Is it spelled and inflected similarly? Are the meanings relatively similar? Atelaes 06:15, 7 March 2007 (UTC)

There are a lot of differences, but in most cases the Ancient Hebrew forms still can be used, they just usually aren't. The major differences, as far as would affect Wiktionary, seem to me to be:

1. Modern Hebrew uses yud and vav to indicate vowels in many places where Ancient Hebrew did not; but then, writing with vowels (poetry, writing directed at young children, and so on) still almost always uses the traditional spellings, and even in writing without vowels, some writers use more such emot kria than others. So even if we separate Ancient Hebrew from Modern Hebrew, we'll still have this problem with many Modern Hebrew words; we might as well find a decent solution, and apply it in all cases.
2. Ancient Hebrew had two conjugation paradigms for finite verbs, usually called the prefix and suffix conjugations, whereas in Modern Hebrew has three (past/present/future). The forms are identical (the ancient prefix and suffix conjugations becoming the modern future and past tenses, respectively, and the modern present tense being a duplicate of the present participles), but the terminology is different. This is probably a minor wrinkle; we can just use both terminologies, saying things like "Third-person masculine singular future tense (prefix conjugation) of Template:he-link."
3. Ancient Hebrew had a much stronger tendency to incorporate object pronouns into whatever they were the object of: a verb, a preposition, or another noun (for genitive-like constructions). In Modern Hebrew, object pronouns are usually only incorporated into prepositions, with the prepositions Template:he-link and Template:he-link hosting pronouns that are objects of verbs or of other nouns (respectively). Now, they still can be hosted by nouns, so that's not a huge issue (?Template:he-link, "my telephone", is absolutely ridiculous, but arguably grammatical), but they really cannot be hosted by verbs (*Template:he-link, "you called me", is not grammatical, and *Template:he-link, "I loved you", was grammatical in Ancient Hebrew but really no longer is). But obviously English articles have dealt with this problem somehow (the same complication appears with forms like *"uploadeth", as in "He uploadeth files to the Internet for all to download"), so I don't see why we can't as well.

All told, I think it's probably best to treat the two as one language, using markers like "obsolete except in liturgy" and so on, but I think it would also be quite workable to try to distinguish them.

—Ruakh_TALK 07:56, 7 March 2007 (UTC)

Ok. That was sort of my original inclination, but I really didn't know and thought it best to ask. And I think some of the grammatical issues really wouldn't affect Wiktionary, as we largely deal with words in isolation. And we can certainly always use alternate spellings, with perhaps a note running along the lines of "more common in Ancient writings" or something of that flavour. Thanks. Atelaes 08:20, 7 March 2007 (UTC)

Vocalisation of Ayin

Are we treating ayin like it has it's own sound with no vowel, or not? I know some people speak Hebrew and give ayin the unique, slightly gutteral sound, but a lot do. I've personally learnt that there is a sound to it, and that it's like a gutteral "a"ish vowel. If so, how are we transliterating it? Thanks. Neskaya 06:07, 11 July 2007 (UTC)

Relatively few Israelis pronounce the ayin — even fewer than distinguish khet from khaf. Myself, I pronounce the ayin in what you might call "careful speech" (especially prayer), but not usually in ordinary conversation, and even this is going further than most Israelis do. (Many Israelis do pronounce both alef and ayin as glottal stops in careful speech, though.) So, I've been a bit inconsistent:

• In romanizing Biblical verses, Mishnaic texts, etc., I've been using a backtick ` (which is how the corresponding sound in Arabic is romanized).
• In romanizing roots, I've been using a backtick `.
• In romanizing individual words, and passages from modern texts, I've been using an apostrophe ', except at the start or end of a word, where I've been omitting it completely.

I'm not particularly advocating this inconsistency, though. It might just be best to always romanize it with a backtick `.

—Ruakh_TALK 16:44, 12 July 2007 (UTC)

Thanks. I was fairly unclear as to how to romanize it. I almost always pronounce it to the extent of that it's one of those unconscious things for me. And hrm --- I never really had too much contact with Israelis, the Hebrew-speaking community I know is an extension of one of the Orthodox communities. Thank you for clearing that up. :) Neskaya 02:07, 19 July 2007 (UTC)

Lexical form of verbs

There are numerous entries in Wiktionary (e.g. go) where a verb is translated by a form with the prefix ל-. All the Hebrew dictionaries I've seen use the 3sm form or the root as the lexical form (הלך, in this case) rather than the prefixed infinitive. Should we change these entries? -PierreAbbat 23:50, 14 September 2007 (UTC)

Yes, please. :-)   —Ruakh_TALK 02:44, 15 September 2007 (UTC)

Template:he-noun

I've created a {{he-noun}} template; please take a look. It's kind of crappy, in that it requires an editor to supply each form, because I don't think there's any way to come up with sensible default values; but still, it should help standardize presentation. (And if a needed form isn't supplied, the template adds the article to a corresponding request category, so it's O.K. if an editor doesn't know all the needed forms.) Please let me know what you think. :-) —Ruakh_TALK 03:47, 16 October 2007 (UTC)

It allows for the plural; what about word for which the dual is more common (like yad, hand)?—msh210℠ 05:49, 22 October 2007 (UTC)

Good point. What do you suggest? —Ruakh_TALK 17:45, 22 October 2007 (UTC)

Well, I took a stab at it, as documented at Template talk:he-noun. Does this seem good? —Ruakh_TALK 06:22, 14 November 2007 (UTC)

I haven't tried it out, but based solely on the documentation it looks good. (The code itself is too hard for me to read.) But we also need dualwv then, I suppose, if possible.—msh210℠ 17:18, 14 November 2007 (UTC)

Yeah, we have that (second bullet point in the list of optional parameters). —Ruakh_TALK 18:54, 14 November 2007 (UTC)

Oh, sorry: I'd missed it.—msh210℠ 19:01, 14 November 2007 (UTC)

See יצר הרע and the (nonexistent) sg= parameter I added to its he-noun template. Is there a way to add that functionality, or is that asking too much?—msh210℠ 19:09, 14 November 2007 (UTC)

Done. It's actually a synonym for wv=, heh. Don't tell anyone. ;-) —Ruakh_TALK 20:18, 14 November 2007 (UTC)

Also, see סבתא and the (nonexistent) m= parameter I added to its he-noun template, counterpart to the f= parameter. Any way to add that?—msh210℠ 19:24, 14 November 2007 (UTC)

Yeah, I don't know why I didn't add that to begin with. I think I was imagining we'd define feminine counterparts in terms of their masculine counterparts, but that's probably not a good idea (and even if we want to do it in some cases, we need to support the cases where we don't). So, I'll do that in a few minutes. Thanks for all your input, BTW! :-) —Ruakh_TALK 20:18, 14 November 2007 (UTC)

And now you've done it, I see. Thanks much, and thanks for sg.—msh210℠ 21:08, 14 November 2007 (UTC)

Vocalisation in links (e.g. [[רִאשׁוֹן]])?

Now most link words in wiktionary contain only consonants, not vowels, because one would have to write consonants twice to make it work, e.g. [[ראשון|רִאשׁוֹן]]. Would it make sense to modify MediaWiki software so it strips all (Hebrew) vowels from links? After that one could just write [[רִאשׁוֹן]] and that word would point to ראשון anyway. I'd like to discuss about this here before I make a bug report without understanding whole issue.--Thv 08:43, 21 October 2007 (UTC)

Also all those templates would not need addition wv (with vowels) parameters, it would happen automatically at least in most cases.--Thv 08:48, 21 October 2007 (UTC)

That's a smart thought; unfortunately, I don't think we can do that, because the Hebrew script is also used for Yiddish, which apparently does need vowels in certain article titles? (I don't speak Yiddish, but that's how it seems from Yiddish articles here.) Also, you couldn't just write [[רִאשׁוֹן]] in any case, because we use templates to control the fonts for Hebrew-script text (so it's more readable for people with certain font sets). What would work better, I think, is if we could get a vowel-removing parser function; then Hebrew-specific templates would only need the vowelly version, and could generate the vowelless version internally. —Ruakh_TALK 18:36, 21 October 2007 (UTC)

How about letting [[רִאשׁוֹן|]] be saved as [[ראשון|רִאשׁוֹן]]?—msh210℠ 05:49, 22 October 2007 (UTC)

That's a really good idea. I wonder how we go about getting that implemented? —Ruakh_TALK 15:56, 22 October 2007 (UTC)