Saturday, October 20, 2012

Snake Phonotactics

In my version of Parseltongue, here are the phonotactics:


(The tap is the alveolar flap and the click is dental.  ugh.)  You can start anywhere.  There is no boundary between words.  Epenthetic 'a's will probably abound.  I need to wrap my head around 's' being the resting sound, like 'uh' in English.

Lexical Intentionality in Fluid-S Alignment

So, I'm imagining a language which uses Nominative (N), Accusative (A), Ergative (E), and Absolutive (B) cases.  The verbs do not inflect for voice, but have a lexically expected case.  For active verbs, this is either N or E.  For stative/descriptive verbs, this is either A or B.   I imagine a good N-verb might be "to eat", an A-verb "to be red", an E-verb "I punch", and a B-verb "to be reddened".  Here goes:

Case Expected
Case Given N A E B
N Normal Intentionality Average Causality Unintentionality Unintentional Causality
A Passive Voice Normal Descriptive Unintentional Passivity Pseudo-Passive Object
E Screaming Intentionality Screaming Causality Strong Intentionality Intentional Causality
B Intentional Passivity Pseudo-Passive Voice Anti-passive Voice Normal Stative
OK, that kinda worked. Let's try it with the verbs I mentioned, imagining an Ergative-Absolutive alignment when needed:
Case Expected
Case Given N A E B
N I ate I made (it) red I punched (him) on accident I was reddened on accident
A I was eaten I am red I was punched (by him) on accident I got reddened
E I ate on purpose I made (it) red on purpose I punched (him) I reddened (it)
B I was eaten in purpose I got (it) red I was punched (by him)* I was reddened
I have no idea if I'm murdering Fluid-S or not. I better ask some people.

Friday, July 13, 2012

English Tense-Aspect vs. Hebrew Aspect

English Tense-Aspect:

Name Example Past Future Imperf. Perf.
Simple past I walked. +
Past Progressive I was walking. + +
Pluperfect I had walked. + +
Past perfect progressive I had been walking. + + +
Simple present I walk.
Present progressive I am walking.
+
Present perfect I have walked.
+
Present perfect progressive I have been walking.
+ +
Future I will walk.
+
Future progressive I will be walking.
+ +
Future perfect I will have walked.
+ +
Future perfect progressive I will have been walking
+ + +

English Aspects:
  • Perfective - focuses on the end
  • Imperfective - focuses on the middle
  • Both - from the beginning to the end
Hebrew Imperfect Aspect:
  • Conative - before the beginning
  • Inceptive - at and just after the beginning
  • Progressive - in the middle
  • Egressive - just before the end
  • Resultative - before and after the end
  • Gnomic - none
Hebrew Perfective Aspect:
  • Ingressive - at the beginning
  • Constantive - from the beginning to the end
  • Constantive! - from the beginning to near the end
  • Perfect - at the end
  • Gnomic - none
Definition of Hebrew aspects
The imperfective aspect is a close-up o a small section of the event where the progressive action is made visible.  The perfective aspect is a view, as if from some distance, of a great part, or of the whole of the event, where the details of the progressive action are not made visible.  The imperfective aspect may include either the beginning or the end, the perfective aspect includes either the beginning and not the end, or both beginning and end.

-The Verbal System of Classical Hebrew.  An Attempt to Distinguish between Semantic and Pragmatic Factors. - Rolf Furuli, University of Oslo

Tuesday, June 19, 2012

Lenition and Fortition

Again, collecting the napkins and typing them up so I can throw them away:

Lenition
Remember, consonants in Perelandran are divided into the "lemon ring" (/ɫ/, /m/, /n/, /r/, and /ŋ/), "the azuře views" (/ð/, /ž/, /ɣ/, /v/, and /z/), the stops (/b/, /d/, /g/, /ʔ/, and the affricative /dž/), and the "breathing" (/j/, /w/, /ʁ/, and ◌ aka nada).

Lenition happens when you add to the end of a verb (nearly everything is a verb in Perelandran).  It's as if the words are all trying to "shorten up" in a rush to get to the end.  Lenition happens differently in the onset than it does in the coda.
Initial

  • The stops all reduce to the glottal stop, except for the glottal stop itself which disappears:
    • b → '
    • d → '
    • g → '
    • j → '
    • ' → ◌
  • In the "lemon ring", nasal become stops and sonorants become glides:
    • m → b
    • n → d
    • ŋ →  g (this can happen at the beginning of a word, only across syllable boundaries)
    • l → y
    • r → w
  • "The azuře views" also become stops (well, one affricative)
    • v → b
    • z → d  (I'm considering adding d̪ ... not phonemicly)
    • ž → dž
    • ð → d
    • ɣ → g
Final
  • The stops cannot occur in the coda
  • The fricatives/"the azuře views" universally become the glottal stop
    • ð → '
    • z → '
    • ž → '
    • v → '
    • ɣ → '
  • The "lemon ring" consonants all turn into fricatives:
    • m → v
    • n → ž
    • l  → ð
    • r → z
    • ŋ → ɣ
Fortition
Well, about now is when I googled lenition and fortition and discovered I conceived of these backwards from everyone else.  :-(  I picture fricatives and nasal as being able to be held for a long time, versus there is no way to make a stop last longer.  For now, I'll stick with my original schema.
Again, there is a difference between onsets and coda.
Initial
  • The stops become fricatives or a trill
    • b → v
    • d → ð
    • g → ɣ
    • j → ž
    • ' → R
  • The fricatives become sonorants
    • v → m
    • ð → l
    • z → r
    • ž → n
    • ɣ → n (well, at the beginning of words.  Internally, it becomes ŋ.)
  • The "lemon ring" are already at the top of their game.  If there is a consonant between them and their vowel, then they become syllabic.  Otherwise, they get growled into being their own syllable.  (It was either that or follow them with a glottal stop, which does sound nice.)
    • m → Rəm:
    • n → Rən:
    • ŋ → Rəŋ: (this wouldn't be happening at the start of a word
    • ɫ → Rəɫ:
    • r → Rər:
Final
  • Stops can't occur in the coda
  • The fricatives are as listed above, but with out the concern over ŋ.
  • For the "lemon ring", I realized I typed colons instead of ː's.  It's clear to me that Perelandran will require gemination, that is, the doubling of consonants.  I remember thinking to myself a few days ago, "Why isn't it important in English if I syllabize the participle of "run" as 'ruh-ning' vs. 'run-ning' vs. 'run-ing'?"  This will be hard for me to remember to say right, but I think I can get it.
Examples
If I have the tri-vowelic root "a-u-ai" and some imaginary verb conjugation is "m+žð+ŋ+l", then it would be realized as "maž.ðu.ˈŋaiɫ".  Some suffix (let say, "ðri") comes along, and it would become bajdugaiððri.  Or, just suppose, the prefix byo'o snuck up on us.  The root would need to beef up and the whole thing would become byo'oRəmmanɫuŋŋaiɫ.

Monday, June 18, 2012

Revising Phonology

Having listening to Conlangery #29: Sound Systems and Romanizations and #48: Designing a Sound System, I think I will revise my phonology to be more symmetrical:

Labials Coronals Velar Uvular Glottal
Bilabial Lab.dent. Dental Alveolar Post-Alv.
Nasal m/מ n/נ ŋ/כ
Stop b/ב d/ד g/ג ʔ/ע
Affric. dž/צ
Fricative v/פ ð/ת z/ז ž/ש ɣ/ח
Trill r/ר R/ה
Lat. Approx. ɫ/ל
j/י w/ו
Yellow is for the "lemon ring" - למנרכ.  The velar nasal will be more limited than I previously thought.  I think it can still open a syllable, but not if any other consonants are around.  I think it can be syllabic still.
Blue is for "the azuře views" – תשחפז.  The big change is here.  The lone retroflex /ɻ/ has been replaced with /ɣ/.  I think this may get Romanized as γ (wait, that's Hellenizing) or ř.  The Hebrew ח  will help with remembering to rasp.
Grey is for the "breathings".  Like Greek -- ὀ vs. ὁ -- Perelandrans think of "smooth breathing" (i.e. nothing – א), "front breathing" (/w/ – ו), "middle breathing" (/j/ – י), and rough-low breathing (/R/ – ה).  These are spelled א and י and ו and ה respectively.
The stop (signs) are in red - בדצגע.  These are considered "weak", and are typically lenited fricatives.

Vowels don't need major revising.  I think the long vowels will all be rising and the diphthongs all falling.  I need to read a lot more, but I'm considering the much-lauded-on-Conlangery two-tone idea.  Previously, I had considered nasal vowels and I'm thinking about putting it in now.  I need to practicing saying them for a while longer.

Saturday, June 16, 2012

Tense-Aspect-Mood

Real quick, I've decided on a tense scheme:

  1. Near Past - This is the default tense.  English speakers should note that "near" is relative, so cenozoic vs. hadean, 100 years ago vs. a million, last month vs. last decade.
  2. Non-past - This could be called "present plus near future".
  3. Anterior Past - Again, this is a relative.  In the retelling of legends, the beginning is in this tense, but in a longer telling, it doesn't get sustained.  "Long ago there was a boy named Neal ... Neal was walking around one day."
  4. Posterior Future - a.k.a. far future
Aspects:
  1. Perfective - On action viewed without parts
  2. Imperfective - There is no generic imperfective.  There are four subdivisions
    1. Stative - This is un-changing but ongoing
    2. Continuative/Episodic/Continuative - I think 'CONT' is all I'll write on this in the future, but the idea is an action viewed as having parts, durative, and on-going, it can be unchanging or changing.  If in the presence of the other three imperfective aspects, it would mean "still verbing" but that's over-translating.
    3. Progressive - Ongoing, evolving and "increasing" (see next)
    4. Regressive - Ongoing, devolving and "decreasing".  The PROG and REG are opposites of each other.  The notion of "increasing" vs. "decreasing" deserves some explanation.

      If someone said Hitler was a "good murderer" we would typically balk at the use of the word "good" in the same sentence as "Hitler" and make them restate it as "he was good at murdering".  Greek was far more sensible and had two words for good!  In the same way, a verb in the PROG aspect is "good", that is, becoming more of what it is.  A verb in the REG is becoming less and less of what it is/ought to be.  So, "I was walking-PROG to the store" could mean I was walking "faster and faster" or "better and better" or "ever closer to my goal".  "I was walking-REG to the store" could mean I was walking "slower and slower" or "in the wrong direction" or "I got worse and worse at walking  (due to injury?)".
The gnomic would be perfective in the far-past for etiologic truth and perfection far-future for teleologic truth.  Also other aspects (momentane, habitual, terminative, inceptive, inchoative, etc.) will be handled through adverbs or verb serialization.

Mood:

I am too close to Attic Greek on this.  I need more sources/reading:
  1. Realis
    1. Topical - This is normally a case marking in most languages, but because all "subjects" are process expressed through verb in Perelandran, it is a mood.  A sentence initial word, phrase or phrasal grouping may end in the topic mood, thereby indicating the controlling genitive, dative, indirect object or just all-around topic of the following sentences.  Perelandrans consider this mood more realis than the indicative.
    2. Indicative - the statements of facts are in this mood
  2. Irrealis
    1. Subjunctive - this mood signifies the lower amount of unreality.  It may be translated "may", or "should" or "let us" or "possibly".  In the apodosis of conditional verbs it indicates contrary-to-fact suppositions.
    2. Optative - this mood signifies the most amount of unreality.  It may be translated "might" or "ought" or "would (that)".  In the protasis of conditional sentences, it indicates not actually true sentences that are unlikely to occur.
    3. Imperative - the aspect is crucial in determining whether a congoing command is being exressed or not.
  3. None of the above - None-finite.

Friday, June 15, 2012

Objectlessnessish

An astute linguist by the handle of Taipo comments on my post Hiding Nouns in Verbs, that merely switching from encoding pronouns (e.g. you, me, them) to encoding deictics (e.g. this, that, those) isn't really doing away with nouns.  He says a better concept is "vectors", as if the verb is an arrow on map, and you turn it to point to and from whatever you want.  He comments about doing that is his conlang, t.a.f.k.a. Tepa.
I agree with him that a lot of what I'm doing is just hiding the nouns grammatically.  However, I want to do more, to model a way of thinking.  What is noun-less-ness?  Look at an English dictionary.  It's all nouns.  Even verbs are turned into nouns.  e.g. 'run' becomes 'to run' and is defined as 'the act of propelling one's legs quickly'.  Further proof of English's noun-centric-ness can be seen in that all our adjectives are nouns and immediately recognizable as substantive nouns when alone.  e.g. 'the good, the bad and the ugly' immediately becomes three men, not three actions.
Korean has prepared my brain to think slightly differently.  There are stative verbs in Korean.  "To be pretty" is not the copula and 'pretty-the-adjective'.  It is all one verb: 예쁘다.  If you want to say 'the pretty person' you use the verb: 예쁜 사람 - the-being-pretty person.  Korean is a verb-centric language.
When English-speakers try to conceptualize the world, we see it as solid, unchanging things that are annoyingly interrupted by action.  We ask stupid questions like, "What is the essence of running" and expect a noun in reply.  Perelandrans wouldn't be like that.  They live on the ocean all their lives and everything the use is organic.  Their own bodies would be the only thing not constantly in flux.  There is no sun in the sky or moon at night, only fuzzy, continual cloud cover.  The essence of things isn't an idea or a state, it's a process.  Hence my goal is not to eliminate nouns per se, but "solid, unchanging objects" and replace everything with processes.


In a lot of sentences, nouns outnumber verbs.  "I hit the ball" and "Suzy ate cake" each have a 2:1 ratio.  More complex sentences only get worse: "Aunt Sally's ball was given to Uncle Bill by Steve the butcher on a Wednesday night to ease his frustration over Cardinal's baseball." 13:2.  Korean and Japanese often do better by drop the subject as a matter of course.  But I really want to keep the attention on processes.  If there are multiple actors and patients in a sentence, then there must be role marking (typically case or word order) and the language emphasizes nouns, whether present or absent from any given sentence.  The solution -- as I see it -- to limit each phrase to one noun or none.

The key topic becomes voice or valency.  Because of the birthday, the mother made the child use her breath to blow out the candle.  Each of these five nouns could have it's own sentence:

  1. Birthdays causes blowing
  2. The mother made there be blowing
  3. The child was made to blow
  4. The breath blew
  5. The candle was blown out.
These five "voices" of the verb "to blow/to blow out" with their five nouns would have to be five phrases in Perelandran.  Each "noun" would have to be a process.  Notice that this means "nouns" would change their "name" if they did something different.  In a conversation "the speaking one" becomes "the hearing one".  If you are getting wet, you are not the same process as when you are drying off and when you leave you become another process.  Perelandran would clue the listener in that the "vector orientation" had or hadn't changed with a switch-reference marker.

For my own reference, the "voice roles" called and numbered as follows:
  • 5) The Reason, Stimulus, or Etiology ("because of x, verb....")
  • 4) Agent, Cause, Force.  ("x made someone make there be verbing")
  • 3) Actor, Enable ("x makes there be verbing")
  • 2) Means, Instrument ("x actually does the verbing")
  • 1) Patient, Direct Object, Experiencer, Recipient ("x is verbed")
  • 0) Dummy subjects in English "It is raining" is there same as "There is raining" ("There is verbing")
  • -1) Result, Teleology, Effect ("Verbing caused x")
Additionally, processes can have the topical marker.  So, put it all together, if the above situation were being described to a Perelandran, you would have to say

Earth-TOPIC Birthday-Blow5 Parent-Blow4 Child-Blow3 Mouth-Blow2 Candle-Blow1 SWITCH Happy-Blow-1.

"On Earth, because of 'birthdays', parents make kids use their mouths to blow out candles, and then we are happy."

Trial Pronouns

I'm not going to have pronouns in my language, but I got to thinking about them anyway.  Here they are in English:

sing. pl.
1st I we
2nd you *y'all
3rd anim. he/she them
3rd inanim. it them

Māori has singular and dual, for two of something.  Therefore, plural means 'three of more'.  Māori also distinguishes between 'we' that includes the hearer, and 'we' that excludes the hearer.  In English, we'd have to resort to "you and us", "y'all and us", "you and I", and "y'all and I".  Here's Māori:

Sing. Dual Pl.
1st. excl. Ahau, Au Māua Mātou
1st. incl. - Tāua Tātou
2nd Koe Korua Koutou
3rd Ia Rāua Rātou


Na'vi has the singular, dual and trial numbers.  That means 'plural' means 'four or more'!  Here is the system of pronouns:

Sing. Dual Trial Plural
1st excl. oe moe pxoe ayoe
1st incl. - oeng pxoeng ayoeng, awnga
2nd nga menga pxenga aynga
3rd anim. po mefo pxefo ayfo, fo
3rd inanim. tsa'u, tsaw mesa'u pxesa'u aysa'u, sa'u

But one words seems weird: pxoeng.  It's the #3 prefix -- pxe+ -- plus the root 'I' -- oe -- plus most of the word for you -- ng.  Why doesn't that mean "the three of us and you'?  Don't get me wrong, I learned a lot from Na'vi, especially about the trial number.  I hadn't thought clearly about 1st person plural (or dual or trial) personal pronouns before, and the separation into 'exclusive' and 'inclusive' was enlightening.  Translating down the 'dual' column into English would read: "both of us, you and I, both of you, both of them, the two things."  But if we can split the English 'we' into two things in the dual, why can't we split the English 'we' into more things in the trial?  Let me illustrated with another table in *hypothetical Na'vi:

Word Person A Person B Person C English
pxoe Me #1 Me #2 Me #3 "The three of us"
*moeng Me #1 Me #2 You #1 "The two of us and you"
*oemenga Me #1 You #1 You #2 "You two and I"
*moepo Me #1 Me #2 Her #1 "The two of us and her"
*oemepo Me #1 Her #1 Her #2 "Those two and I"
*oengpo Me #1 You #1 Her #1 "You, me and her"
pxenga You #1 You #2 You #3 "The three of you"
*mengapo You #1 You #2 Her #1 "You two and her"
*ngamofo You #1 Her #1 Her #2 "The two of them and you"
pxefo Her #1 Her #2 Her #3 "The three of them"

Now I know this isn't as succinct, but couldn't you imagine that if a language was serious about the trial, it might do this?

Thursday, June 14, 2012

Hiding Nouns in Verbs

The first step in getting rid of nouns -- which I already knew about -- is encoding the subject in the verb.  Speakers of Spanish know that 'soy', 'eres' and 'es' all require two words in English ('I am', 'you are', 'she is').  The same is true in Greek and many other languages.

  • εἰμί - I am
  • εἶ - you are
  • ἐστίν - he/she/it is.
The breakthrough came when I heard William A. on Conlangery mention that many languages families (Athabaskan and Uto-Aztecan for example) encode both the subject AND THE OBJECT in the verb!  So, for example, one word in Navajo:
  • yishạ́ - I am eating it.
This gained me a lot of milage, but it felt a bit like cheating.  Aren't I just hiding the (pro)noun, not eliminating it?  How could I get rid of it more completely?  As I thought about it, I remembered that 'here' and 'there' are adverbs, so if I could replace person in the encoded subject/object with demonstratives, I would be closer to my goal.
Deixis is the over-arching concept behind 'this' vs. 'that'.  My experience with Korean taught me that there are (at least) three levels of deixis:
  1. This.  By me.  Proximal.  Here.  사람은/this person
  2. That.  By you.  Medial.  There.  사람은/that person
  3. Yon.  Not by either of us.  Distal.  Yonder.  사람은/yon person
As you can see, English previously had three levels, but has generally lost this 3-way distinction in favor of a two-way one.  Sad.  R.I.P. 'yon'.

So far, Perelandran would then be able to say, in one word, 'that-hit-this', which could mean 'you hit me' or a host of other things.
Next, I discovered the Seri language, which has nine demonstratives that are mixed up with definiteness ('a' vs. 'the').  I've decided to go with eight in Perelandran:
  1. Definite, Proximal, standing Up (DPU) - "this standing person"
  2. Definite, Proximal, Sitting (DPS) - "this sitting dog"
  3. Definite, Proximal, lying Down (DPD) - "this mat which is lying here"
  4. Definite, Medial, Coming (DMC) - "that approaching storm"
  5. Definite, Medial, Going (DMG) - "that woman who is walking away"
  6. Definite, Distal, Coming (DMC) - "yon attacking (us) army"
  7. Definite, Distal, Going (DMG) - "the horse riding off into the sunset"
  8. INDefinite (IND) - "some people"
Subject marking will be required, that is, not optional, on all verbs with a valency greater than zero.  I also thought of an optional animacy marking, which implies intentionality.  This is a simplification and adaptation of the classificatory system from Navajo:
  1. Persons/Lightning/(dynamic) Forces (PLF) - uncontrollable, (mostly) unpredictable, intentional things.  cp. "acts of God/Nature"
  2. Living/Animate Beings (LAB) - more predicable, but very alive.  Cows, cats, and gravity.
  3. Receptacles/Openings/Controls (ROC) - that which is manipulated.  An element of unpredictability remains, since anything can go in a box.
  4. Inanimate objects -- predicable and controllable -- are all at this level, but subdivided into three sub-categories:
    1. Solid/Discreet Things (SDT) - balls, fruits, non-containers with clear boundaries
    2. Mushy, Porous Areas (MPA) - fuzzy boundaries, fuzzy definitions, e.g. mud or a mirage/reflection on a hot day.
    3. Bendy, Flat Fractals (BFF) - On Perelandra, "land" is actually great clumps of tough "seaweed" growing together into "islands".  This categories sees the interlocking, patterned nature of bushes, carpets, rope and shower curtains together.
  5. Flowing/Undulating Flocks (FUF) - a school of fish is not the sum of individuals.  You can't study individual water molecules and understand a river.
  6. Cloudy, Random Hazes (CRH) - There is continual, complete cloud cover on Perelandra.  It's always moving and changing, but means nothing.
  7. Unchanging, Dead Stuff (UDS) - This is how we see a majority of things in our world, whereas Perelandrans consider this the least populated, least important category there is.
On our world, we 'name' everything.  On Perelandra, only PLF's get names.  If you talk to your dog, you've promoted them to a PLF.  But even people progress down through the levels.  Babies are LAB's.  When a person is a sleep, they become ROC's, contains for dreams (which are very important on Perelandra).  A corpse is an SDT, but quickly decomposes into a MPA, then a BFF (!) and finally, because everything is buried "at sea", a FUF.  Once the deceased memory has begun to fade in the collective subconscious, they are a CRH.  The 'are not now' are distinguished from the 'never have been' as CRH's vs. UDS's.  If someone has no idea what a pencil is for, it's a SDT.  Once they learn, it's an ROC.

Perelandran Phototactics

The phototactics of Perelandran are straightforward enough.  Simply put, a syllable can be (C)(C)V(C).  The only acceptable onset consonant clusters are [/v/, /z/, /ð/, /ʒ/, or /ɻ/] (a.k.a. "the azure views") + [/ɫ/, /m/, /n/, /r/, or /ŋ/] (a.k.a. "lemon ring").  Be sure to read the previous blogpost for the complete phonology.   The complete list of consonant clusters, with the vowel 'a' is: 'vra', 'vla', 'vma', 'vna', 'vnga', 'zra', 'zla', 'zma', 'zna', 'znga', 'ðra', 'ðla', 'ðma', 'ðna', 'ðnga', 'žra', 'žla', 'žma', 'žna', 'žnga', 'rla', 'rma', 'rna', 'rra', and 'rnga'.  The best way to practice is to start buzzing/growling, say the last two letters and then repeat and try to shorten it.

"The Azure Views" and "Lemon Ring" are the "strong letters" or litteræ fortes.  Any littera fortis may be a syllabic consonant.  Any littera fortis can be the coda.  But there are two major rules:

  1. A "lemon ring" (/ɫ/, /m/, /n/, /r/, or /ŋ/) cannot precede "the azure views" (/v/, /z/, /ð/, /ʒ/, or /ɻ/) within a syllable.
  2. A syllabic consonant (littera fortis) cannot take a coda.

Morae
  • Onset consonant(s) do not count.
  • Coda consonants count as one.
  • Regular vowels count as one.
  • Diphthongs, long vowels and syllabic consonants count as two.
The penultimate mora is always stressed, as in Hawai'ian.  For words longer than four morae, the initial mora is also stressed.
  • לַאֻבִיֶאַאֻעָ – /ˈɫau.bi.je.ˈau.ʔo/ - LAU.bee.yeh.OU.'o
  • פרַלדֻ – /ˈvɾaɫ.du/ - VRAL.doo, said too slowly as /vɾa.ˈɫ̩.du/
  • ולְתְ – /wɫ̩.ˈð̩/ - wll.TH  - even a Perelandran would consider this the beginnings of a tongue twister.  Practically, there is often a glottal stop between the syllables, since "lemon-rings cannot precede the azure views"!
You might notice that I changed how I write diphthongs from last time.  I would like to put two vowels under a letter, but the variability from font to font is two high.  Only on a LaTeX PDF do I have enough control to do it my way.  On the web, I will have to keep using extra א's.

Wednesday, June 13, 2012

Not Romanized, באותיות עברית.

So I gave up (for the time being) on creating the ultimate theological language ("Trinitarian") and decided to take on the challenge of creating a language without nouns: Perelandran.  I had stumbled upon Kēlen, created by Syvia Sotomayor, a language supposedly without verbs.

Actually, it has four verbs.  I hope to be even more rigorously anti-noun than she is anti-verb.

Anyway, I noticed how everyone "Romanizes" their languages and IPA is very Latin-based.  What if, just for variety's sake, my language had been discovered by Korean, Chinese or Hebrew speakers?  Since I am studying Hebrew is class this summer, I decided to go with Hebrew.  (Note, if you do not have Courier New font install on your computer, this will look nothing like what I intend.)

Vowels

  • /a/ - אַ - called a 'páṯaḥ'
  • /e/ - אֶ - called a 'sĕḡōl' - allophony with /ɛ/
  • /i/ - אִ - called a 'ḥḯreq' - allophony with /ɪ/
  • /o/ - אָ - called a 'qấmeṣ ḥāṭûf' - allophony with /ɤ/
  • /u/ - אֻ - called a 'qibbûṣ' - allophony with /ɯ/
Some Perelandrans never round their vowels, many round 'o' and 'u'.  I decided to go with all the diphthongs from Hawai'an.  Writing vowels in this way is not standard Hebrew.
  • /ae/ - אֶַ - not in English, "ah-eh" quickly
  • /ai/ - אִַ – like the English "aye/I"
  • /ao/ - אַָ – not in English, "ah-oh" quickly
  • /au/ - אַֻ - very close to English "ow"
  • /ei/ - אִֶ - like the English "A/mate"
  • /eo/ - אֶָ - not in English, "eh-oh" quickly
  • /eu/ - אֶֻ - not in English, "eh-oo" quickly
  • /iu/ - אִֻ - very close to English "eww .. that's gross"
  • /oi/ - אִָ - as in English "boil"
  • /ou/ - אָֻ - We actually say our long 'o's in English as "oh-oo"
  • /aː/ - אֲ - a glyph already used in Hebrew for something else
  • /eː/ - אֱ - a glyph already used in Hebrew for something else
  • /iː/ - אְִ - *could be tricky with fonts*
  • /oː/ - אֳ - a glyph already used in Hebrew for something else
  • /uː/ - אְֻ - *could be tricky with fonts*
You'll notice the aleph (א) is a placeholder for a syllable with no onset.
The nasal-liquid consonants are affectionately called 'lemon-ring' consonants:
  • /ɫ/ - ל - notice this is "darker" than a Hebrew 'lāmeḏ' or and English 'L'.
  • /m/ - מ - Perelandran does not use the 'final mêm' (ם).
  • /n/ - נ – never 'final nûn' (ן).
  • /ɾ/ - ר - there is allophony with /r/ and even /ɻ/.
  • /ŋ/ - כ - Perelandran does not use the 'final kaf' (ך).  Like many Earth-languages, syllable in Perelandran can start with an 'ng'.  This requires a lot of practice for English-speakers.
These letters behave similarly, so it will be convenient to remember them as 'lemon-ring' (לֶמָנרִכ).  Practice saying them as stand-alone sounds.
The next set is called 'the azure views':
  • /ð/ - ת – as in the, not as in thin.
  • /ʒ/ – ש – as in pleasure.  This sound doesn't occur in Hebrew, so 'šîn' has been recruited.
  • /ɻ/ - ר – This letter does double/triple-duty, as the alveolar flap/tap or trill and as the retroflex approximate.  Circumstances dictate some pronunciation rules, but some of it is personal choice.  When it is considered a 'lemon-ring', it's alveolar.  When it's one of 'the azure views', it's /ɻ/.
  • /v/ - פ – There are no voiceless consonants in Perelandran, so this /p/,/f/ letter has been recruited as a 'V'.  Never use 'final peh' (ף).
  • /z/ - ז - buzz like a bee!
"The Azure Views" can also be stand-alone syllables.
The last set are the stops:
  • /b/ - ב - in Ancient Hebrew one had to put a dāḡēš inside the letter to distinguish /v/ from /b/ (i.e. ב vs. בּ).  Not so now.
  • /d/ - ד - never ð, as in the ancient pronunciation.
  • /g/ - ג
  • /ʔ/ - ע - the dash in English 'uh-oh'.  Very hard for English speakers to hear at the beginning and end of words.
  • /ʤ/ - ח - standard English 'j'
There are two semi-vowels/approximants:
  • /j/ - י – like the English 'y'
  • /w/ - ו
There is one, last letter, and I'm having trouble finding in IPA:
  • /x/ - ה - but I don't think the IPA is right.
I thought for sure the German phonology page would show me how I can say "Bach" or "loch" and hold out that last rasp forever.  I think I am making it voiced.  It's basically the 'hocking a loogie sound'!  Please comment if you know a better way to represent that.

3+ Roles

I was considering how "Trinitarian" might play out grammatically and I had some constraints.  I wanted there to be three parts to every sentence: subject, object and verb.  I wanted the grammatical voice of every verb to be either active, middle or passive.  I quickly saw that this would not cover everything, so I decided to write myself a loop-hole: a topic marker.

I had decided words could be either nouns or verbs, depending on their position in the sentence.  (I had picked SOV at the time.)  But as I really considered valency for the first time, I saw that I had a real problem.  Voice could be either active, middle or passive, but what if it was unmarked?  One would still have two nouns marked as SUBJECT and OBJECT and VERB.  I stumbled upon the pseudo-passive.  Here are two example of it:

  1. The Ancient Egyptians had their cats killed.
  2. Vain bald men get their heads waxed.
The subjects stay subjects (Egyptians and men) and the objects stay objects (cats and heads) but the subject need not be the one acting on the subject at all.  I decided that would be my option for "no voice specified but and object and subject still given."
-------
But what about three argument sentences, either with indirect objects or genitive phrases?  Ex.:
  1. I gave the cat the turtle.
  2. You ate food in France.
Given that it is my second-best language, it's no wonder that my thought turned to Korean first.  In Korean, one could use the equivalents of the English "to the cat" /고양이한테 and "in France"/프랑스에서.  But there is another option: the topical.

  1. 고양이는 내가 거북이를 주었다.  In regards to the cat, I gave the turtle.
  2. 프랑스는 내가 음식을 먹었다. In regards to France, I ate food.
These examples are a little strange, but you get the idea.  Actually, you might not get the idea, since English isn't a topic-comment language.  Listen to Conlangery episode 53 for help.  Anyway, it's a solution to eliminate a lot of genitives, datives and double accusatives.

Tuesday, June 12, 2012

Bubble Burst

As I busily try to catch up with the Conlangery Podcast, I blew through episode three without thinking as carefully about them as I should have.  A day or two later it me, "I'm trying to do exactly what that featured conlang did.  I better go read."  The language in question is Ithkuil.  Let me break it down for you:

It's a conlang.  That means its not a naturally occurring language.  If that emotionally bothers you, grow up.  Most conlangers (someone who invents languages) don't believe in God, so they take the symbol of the Tower of Babel tongue-in-cheek.  I do, however.  The Bible tells us the end of the story, that we will understand every tribe, tongue and language.  We will all be linguists naturally in heaven!

It's an a priori language.  That means it's not even supposed to seem like it is related to any other language.  Frankly, unless you're very creative in your mixing-and-matching, this seems like the only way to go.  Why would you invent another Indo-European language?  Now to be honest, I would be interested to see a Sino-Romance language, but you would have got A LOT of research ahead of you before you could even conceive of a project like that.

It is a philosophical language.  That means it has a particular mindset or paradigm it's trying to achieve.    The Whorfian idea behind Ithkuil is that language limits how you think, so the language should avail you of every opportunity at every moment, so you're never limited by the language, only yourself.  Ithkuil has maximum variety at every turn.  I should have seen the parallels between this and my ideas right away.

The nature of my hypothetical creation didn't hit me until I starting thinking about creating a lexicon.  In my hypothetical "Trinitarian" language, every 'thing' to be lex'ed would have to be weighed and considered under this rubric: "What thing is it's dialectical opposite and what 'glue' unites those two?"  These three components correspond to the Father, Son and Holy Spirit.  One is the unbegotten, unsent God, the other is the begotten, sent, subordinating Man and the third is Love, the connective tissue that binds them together and spans the chasm between the Two.

The trouble came when I considered "The cat sat on the mat."  How would I say this in Trinitarian?  Cat's could be the antithesis of dogs, untrainable, not-really-domesticated animals.  Sitting might be the polar opposite of running.  Mats are for getting dirt of one's shoes, so their antinomy would be dirty shoes.  But what about "The wallaby lounged on the ottoman"?  How would I make a Trinitarian sentence for that?  Or "Steve played chess with Sally"?  Philosophically, I had let the One conquer the Many.

Aside from the fact that John Quijada used a "binity" instead of a "trinity" system of thought, Ithkuil had already thought of everything I could think of and much, much more.  If you haven't attempted his grammer yet, it's well worth your time.  You won't make it through it all, but just to see how much thought he's put into his categories is awesome.  I saw (by the tenth page) that this is not what I wanted to do.   I wanted a language sinful, fallen, post-Babel people (like me) could speak.  I decided to just aim for "the opposite of English" and then let "Trinitarian" be the meeting of the two.

Monday, June 11, 2012

There's Nothing New Under the Sun

Two weeks ago, I got started preparing to actually make a conlang.  Given my disposition, I naturally wanted to make a language that would encourage theologically sophisticated, Biblically sound thought. What is the distinctive foundation of Christian thinking?  Obviously, it the Trinity.  There are other monotheistic religions and there are other "multiform" religions, but only Christianity espouses the equal ultimacy of "the One and the Many".  Only we say that God is One and God is Three.

How would that fact manifest itself in language?  I wasn't sure at first.  Some trite examples came first: three "levels" for consonants, as in Korean.  For those who don't know

  • ㅂ means b/p (unaspirated)
  • ㅃ means bb (tensed)
  • ㅍ means p' (strong and aspirated)
Now that I've been exposed to ejective consonants (thanks to Na'vi), I imagined three levels that would be clear to me:
  • ㅂ could come to mean /b/ (voiced)
  • ㅃ could mean /p/ (unvoiced, aspirated)
  • ㅍ could mean /p'/ (ejective)
Interestingly, there are 9 triples like this that English speakers can readily distinguish:
  1. /b/, /p/, and /p'/
  2. /ʒ/, /ʃ/, and /ʃ'/ (zh, sh, and sh')
  3. /ʤ/, /ʧ/, and /ʧ'/ (j, ch, and ch')
  4. /d/, /t/, and /t'/
  5. /g/, /k/, and /k'/
  6. /v/, /f/, and /f'/
  7. /ð/, /θ/, and /θ'/ (they, thin, BATH! - as an angry, two syllable word)
  8. /z/, /s/, and /s'/
  9. /ʣ/, /ʦ/, and /ʦ/
There are three clicks that I can distinguish by hearing alone:
  1. ʘ - like a kissing sound but without the lip pursing
  2. ! - your tongue sucks to the roof of your mouth and then slaps down
  3. | - like the tsk of calling a horse.  I can't tell the difference with ‖.
The vowels can easily be divided up into 9, 3 high, 3 mid and 3 low: beet, bit, boot, bate, soot, boat, bet, bat, bot.  There are also three nasal consonants (and thanks to Na'vi, I can say 'ng' at the start of a word now!) I can hear:
  1. /m/
  2. /n/
  3. /ŋ/
Also thanks to Na'vi, here are three liquid consonants
  1. /l/
  2. /r/
  3. /ɾ/
I had trouble thinking of three glides (just /j/ and /w/) but then again, the lack of one could be the third!

There are three tenses: past, present, future.  There are three persons: first, second and third.  Hebrew has three numbers: singular, dual, plural ... or maybe I should go with the Kantian: singularity, plurality, universality.  Greek has three voices: active, middle, passive.  Aspect could have perfective, imperfective or none-specified.  Korean has three level of deixis (just like English used to have): here, there and yonder.

Obviously, I was seeing three's because I wanted to.  I will write tomorrow about how I set this vision aside for something harder and less idealistic.

Sunday, June 10, 2012

About

What?!  Another blog?  Robert, are you nuts?

Yeah.  Yeah, I am.
However, I am rapidly filling up journals and I think languages have been an abiding interest of mine for a long time.  My first word was in Korea: 먹어/eat!  I went to preschool in Japan (in Japanese) had German neighbors at my British primary (elementary) school, studied Klingon in middle school and went to high school in Korea.  There I learned lots of Chinese characters, studied Latin and French and took etymology classes at a college.  At university, I studied Ancient Greek, computer languages and took several linguistics classes.  I've taught myself Hebrew (I'm in a formal class now) and studied Na'vi pretty rigorously.  I'm a very language-minded person.

The thing that tipped me over the edge has been the Conlangery Podcast.  They hit upon topics that make me read several dozen Wikipedia pages each.  I supposed I am very old-school and too Whorfian in my approach to language creation.  I am presently interesting in what linguistics can tell us about the human mind and the nature of reality.  I really like having my mind blown.  I hope I can elucidate some topics for random, Google-searching passers-by!