|Triconsonantal Root Systems
|Page 1 of 9|
|Author:||Mecislau [ Sat Apr 23, 2005 6:45 pm ]|
|Post subject:||Triconsonantal Root Systems|
Some people've expressed some interest in me explaining how exactly these work. I've decided to write up a description of the TRS in Hebrew, although with how it works together with Hebrew's binyan structure.
I'm working on it now, but if you have any questions right now, feel free to ask and I'll try my best to answer them.
|Author:||Glenn [ Sat Apr 23, 2005 11:01 pm ]|
|Author:||Soap [ Sun Apr 24, 2005 6:08 am ]|
|Author:||Mecislau [ Sun Apr 24, 2005 2:53 pm ]|
Alright, here goes. And yes, Merc, there are hundreds when you count all of the inflections on each word, but I'll list out the major categories...
Triconsonantal Root Systems, such as those used by the Semitic languages, consist of a collection of a few hundred roots (consisting of three consonants, as the name implies). When a certain vowel pattern is placed over these roots, a wide range of meanings can be assumed.
I've heard the system described almost like a multiplication table. On one axis is the root, and on the other is the pattern. Neither of these can stand alone, but when combined, their meanings mix.
To give an example of this, I'll use the root *K-T-V ("write").
If the pattern CoCeC (masc sg pres verb) is applied, it becomes kotev, meaning "I/You/He/She write(s)". The pattern CoCCim (masc pl pres verb) makes it kotvim "We/Y'all/They write". These exact same patterns can be applied to almost any root with the same result:
*CH-SH-V "think" > choshev "I/You/He think(s)"
*?-CH-L "eat" > 'ochel "I/You/He eat(s)"
*R-?-H "see" > ro'eh "I/You/He see(s)"
These roots also can form nouns and adjectives, given the right patterns. For example, maCCeC makes a verb into a tool:
*CH-SH-B "think" > machshev "computer" (lit. "thinking tool")
*P-T-CH "often" > mafteach "key" (lit. "opening tool"; p > f is normal lenition in Hebrew, and the additional /a/ is a normal phonological change before /X/)
The plural for nouns in the structure "CeCeC" is "CCaCim":
sefer "book" (from *S-P-R "tell, count") > sfarim "books"
seret "film" > sratim "films"
Some patterns can even be derived from other nouns. For example, the noun tochnah "software" has led to the development of the pattern "CoCCah", making a root into a piece of software:
*L-M-D "learn" > lomdah "educational software"
*G-N-V "steal" > gonvah "pirated software"
The list goes on and on.
For Mercator, here's all of the forms of K-T-B that I know of (there may be more). I don't list out the derivations of each, but I'll mention that they're there. Keep in mind that /b/ > /v/ allomorphic lenition is common in Hebrew:
kotev "write" (plus 25 other forms of the verb, in other tenses and conjugations)
ktivah "writing, cursive" (plus a dual and plural form, plus 10 singular possessive endings and 10 plural possessive endings; ie, +22 forms)
katuv "written" (plus a masculine plural, feminine singular, and feminine plural form; ie, +3 forms)
nichtav "be written" (plus 25 other forms of the verb...)
hikatvut "the act of being written" (plus 22 other noun forms)
mitkatev "correspond (+25)
hitkatvut "correspondence" (+22)
machtiv "dictate" (+25)
hachtavah "dictation" (+22)
muchtav "be dictated" (+22)
muchtav "dictated" (+3)
machetev "engrave" (+25)
ktav "text, piece of writing" (nowaydays these two are pronounced the same, but they didnt used to be)
ktuvah "marriage contract"
ktovet "address, inscription"
michtav "letter, note"
katav "journalist, reporter"
katavah "article (in a newspaper, magazine, etc)"
kitviyah "a collection of texts"
(Each of the above having 22 other forms)
There's certainly many more as well...
If anyone wants to calculate how many forms that is, be my guest!
Sometimes the adjective forms can be slightly idiomatic as well. For example, the root *G-D-L "to grow" yields the adjective gadol "big".
In foreign borrowings, roots can be extracted by pulling the vowels out of a word. For example, Hebrew borrowed the noun telefon from English. The root *T-L-P-N was extracted, and this can be used as a verb (eg, 'ani tilpanti otach means "I phoned you"). This is a rare four-consonant root. It functions the same as a triconsonantal one, as the middle two consonants are just treated as one.
Each of these patterns are productive. Even if a given word doesn't exist in the dictionary, you can just apply a given pattern to any root and you will be understand (just like how you can add "-ize" to many words in English, and you will be understood). For example, you could say telefoniyah, not something you'd find in a dictionary, but be understood as meaning "a collection of telephones".
EDIT: Actually, telefoniyah is a word. It's a telephony.
Next Up, Binyanim. And how they work in combination with triconsonantal roots.
|Author:||Soap [ Sun Apr 24, 2005 6:49 pm ]|
|Author:||blank stare [ Sun Apr 24, 2005 7:19 pm ]|
|Author:||Mecislau [ Sun Apr 24, 2005 8:19 pm ]|
|Author:||Herra Ratatoskr [ Sun Apr 24, 2005 8:29 pm ]|
|Author:||Glenn [ Sun Apr 24, 2005 11:12 pm ]|
Thanks for the info, Maknas! I knew the basic outlines (as well as the way in which foreign words can be adopted into the root system), but your explanation certainly helps. Like Mercator, I found the noun > noun software examples fascinating.
It would also be helpful if you provided at least one "expanded" example each of the noun and verb forms you mentioned (although I understand that it might be time-consuming).
I will have to be off-board for at least a few days for work reasons, but I will continue to follow this thread when I drop back in again (among other things, my conworld has always contained a place for an Arabic/Hebrew-inspired conlang, and I would like to construct a proper one eventually.) I look forward to seeing the next installment.
|Author:||Mecislau [ Mon Apr 25, 2005 3:46 pm ]|
|Author:||Trebor [ Tue Apr 26, 2005 5:52 pm ]|
|Author:||Mecislau [ Tue Apr 26, 2005 8:39 pm ]|
|Author:||Trebor [ Tue Apr 26, 2005 8:53 pm ]|
|Author:||Mecislau [ Tue Apr 26, 2005 9:04 pm ]|
|Author:||Noriega [ Fri Apr 29, 2005 2:31 pm ]|
Here are some reanalyzed loans into Arabic:
film plural 'aflaam
bank plural bunuuk
yacht plural yuchuut
raasib (recipe) plural ruasiib
Not entirely sure about the one above, then I remember seeing TaZin (dozen), but I've forgotten the plural.
And my teacher claims that he, in a TV debate, heard one man refer to the others as hataalir, the plural of Hitler!
|Author:||brandrinn [ Wed May 04, 2005 7:19 pm ]|
|Author:||Bryan [ Thu May 05, 2005 10:08 am ]|
Please do, Brandrinn. This thread is awesome. I aksed about triconsonantal roots a while back and mostly folk were unco-operative. 'Just study hebrew' kinda crappy answers. But this thread is bop-seleckta. Or summat. I am bookmarking it!
|Author:||brandrinn [ Thu May 05, 2005 10:41 am ]|
well, i dont have a lot of time, but i can start off with some simple things:
k-t-b (writing, to write, etc)
verb, perfect, third person singular kataba
verb, imperfect, thid person singular yaktab
desk/office (i get them mixed up) [/i]maktab[/i]
as you can see, making nouns feminine is at least as productive a derivational process as changing the vowels. the fun thing about Arabic, though is the crazy plurals. here are some common patterns:
CaCC > CuCuuC
CaaCiC > CawaaCiC
CaCiC > 'aCCiCa' (glottal stops on either end!)
also, in Arabic, we use the word faxal (where X is a pharyngeal voiced fricative) instead of pa'al, and we say binaya instead of binyan, although we dont call the forms "buildings," we call them "wazn," which just means form.
more to come after i finish my Russian final!
|Author:||Bryan [ Thu May 05, 2005 11:12 am ]|
hehee. what about biconsonantal roots? Do these exist in some languages?
|Author:||Soap [ Thu May 05, 2005 11:17 am ]|
|Author:||brandrinn [ Thu May 05, 2005 11:31 am ]|
|Author:||Mecislau [ Thu May 05, 2005 5:13 pm ]|
Alright, I'll take a few minutes to continue this...
Within each verbal binyan, there are many sub-patterns. The main one is called shalem (plural shlemim), meaning "intact". This represents a perfectly regular pattern. For example, the shalem forms of the present tense of a pa'al verb are as follows:
Each binyan has, of course, its own shalem pattern. The pi'el shalem, for example, is meCaCeC, meCaCeCet, meCaCCim, and meCaCCot (pi'el and pa'al are closely related).
However, this doesn't always work. See where the shalem patterns force two consonants together? For some consonants, that would become rather tricky to pronounce. And if the last consonant is /h/, then it is at risk of easily dropping altogether. For these situations, and many others, exist alternate, but completely regular and predictable, patterns called gzarot (singular gizrah). Each binyan at least a half-dozen major gzarot, and many more minor ones.
The notation for identifying the gzarot is relative to the root *P-?-L again. The "P" represents the first consonant of any root, the "?" the middle, and the "L" the last. For example, gizrat L"H bepa'al means "the pattern of roots whose last consonant is an H, within binyan pa'al". The L"H means that the last consonant of the root (the "L" of *P?L) is an H in these roots.
I'll list some examples from Binyan Pa'al.
The simplest gizrah deals with 'gutteral' consonants in the middle of the root. "Gutteral" consonants in Hebrew are defined as /? h X R/. Gizrat ?"gronit inserts an epenthetic /a/ to break up the cluster in the MascPl and FemPl forms. With the root *?-H-V "love":
MascPl: 'ohavim (not *'ohvim)
FemPl: 'ohavot (not *'ohvot)
Gizrah L"H (final letter = H) shows the loss of the /h/ in all forms of the present tense, and some vowel changes in the singular forms to help account for that. With the root *R-TZ-H "want":
MascSg: rotzeh (final orthographic h is silent)
FemSg: rotzah (not *rotzehet)
MascPl: rotzim (not *rotzhim)
FemPl: rotzot (not *rotzhot)
Gizrah ?"I and ?"U deal with roots whose medial consonants are the semivowels yod and vav. These consonants completely drop when conjugated, meaning these patterns deal only with the first and last consonants (although the yod and vav may appear again in some other forms of the verb as full vowels). With the verb *R-U-TZ "run" (Infinitive: larutz):
Gizrah L"` (the last letter is a glottal stop as 'ayin), the 'ayin just screws up all of the vowels around it. Gizrah L"CH is very similar, but with a few different patterns. With the verb *Y-D-` "know":
(And of course many others, including other patterns in different tenses)
Now, these are not quite irregular verbs. These adjusted patterns are completely predictable. That is, all roots ending in H and conjugated within binyan pa'al will take the gizrah L"H pattern, never the shalem one.
Nouns also have "gzarot" of sorts, though they're referred to as mishqalim "weights" (same root as "sheqel"). Certain vowel patterns in the singular form of a noun dictate a certain plural form different from the expected one, yet these are still completely predictable. Many of these involve a certain vowel dropping, as Hebrew generally likes to keep inflected forms of nouns the same number of syllables whenever possible.
As I mentioned earlier, one pattern is CeCeC, as in sefer "book" or seret "film", whose plurals are all CCaCim (sfarim and sratim). Nouns of the form CeCVC shorten to CCVC whenever a prefix containing its own vowel is added: rechov "street", barchov "in the street". It remains when the prefix lacks a vowel: brechov "in a street".
Sometimes the consonants may change as well, although this is rarer. Feminine nouns of the form CVCVt have a plural in CVCVyot (possibly including other shortenings, depending on what the vowels are): chanut "store", chanuyot "stores".
That covers most of it. Unless anyone has any more specific requests, my next post here should be about other misc topics dealing with triconsonantal root systems, as well as more examples of derivation of meaning.
|Author:||Mecislau [ Sat Jun 11, 2005 10:54 pm ]|
|Author:||Glenn [ Sun Jun 12, 2005 1:06 am ]|
|Author:||Grath [ Sun Jun 12, 2005 4:13 am ]|
|Page 1 of 9||All times are UTC - 6 hours [ DST ]|
|Powered by phpBB® Forum Software © phpBB Group