Oro Mpaa


2003 by Christian "Catharsis" Thalmann
cinga (at) gmx (dot) net


Table of Contents


1.  Foreword

2.  Phonology
2.1.  Phoneme Inventory
2.2.  Word Structure
2.3.  Pronunciation of Vowels and Glides
2.4.  Pronunciation of Kernels
2.5.  Stress
2.6.  Phrasal Sandhi
3.  Nomina
3.1.  Nouns
3.2.  Pronouns
4.  Verbs
4.1.  Verb Forms
4.2.  Conjugation Paradigms
4.3.  The Prefix me(m)–
4.4.  Tense and Aspect

5.  Syntax
5.1.  Phrases
5.2.  Simple Clauses
5.3.  Serial Clauses
5.4.  Sentences
5.5.  Relative Clauses
5.6.  Questions


1.    Foreword



Oro Mpaa is a mostly isolating, analytic a priori language with only few (but often irregular) inflected forms.  The phonology is not very trivial, considering that the sound of vowels changes under the influence of nearby glides, and the sound of consonant phonemes is dependant on their environment within the phrase.  

The design goals of this language project are to make experiences with a grammatical structure unlike the case-driven IE-esque ones of the languages I've created or worked with so far, as well as to achieve a good measure of both acoustical and typographical aesthetics.

As for notations, I will use "apostrophes" for English glosses, boldface for the Latin alphabet transliteration of Oro Mpaa, /slashes/ for phonemic notation and [brackets] for phonetic notation.  The latter two will make use of the X-SAMPA ASCII encoding of the International Phonetic Alphabet.  For the discussion of syntax, I'll also use {braces} for dummy translation, where each word is translated into English by itself, leaving the original word order unchanged. 




2.    Phonology


2.1.   Phoneme Inventory


The good news is that both the native script and the Latin alphabet transcription of Oro Mpaa are completely phonemic, i.e. one letter corresponds exactly to one phoneme.  The bad news is that the phonetic realisations of these phonemes are rather non-trivial at times.

Here follows a table of the consonant phonemes.  Notice that there is no phonemic distinction between voiced and unvoiced consonants, but many consonants have voiced and unvoiced allophones.  Remember that boldface indicates transliteration of the native script, while the slashes indicate the correspondent phonemes.

The correspondence between consonant phonemes and their phonetic realisations will be discussed in chapter 2.4.


Consonants
Alveodentals
Labials
Velars
Stops
t /t/
p /p/
c /k/
Fricatives
s /s/
v /f/
h /x/
Nasals
n /n/
m /m/
g /N/
Liquid
l /l/
 
 
Tap/Trill
r /r/
 
 


Oro Mpaa has two glides, which are treated as a category of phonemes separate from consonants and vowels.  Glides can influence the pronunciation of both consonants and vowels, and sometimes become silent themselves in the process.


Glides
Palatal
Labiovelar

i /j/
u /w/


The vowel inventory of Oro Mpaa appears very spartan at first.  There are only three vowel phonemes, as given in the following table.  However, interactions with glides will split these phonemes into a surprising range of phonetic realisations.

The correspondence between vowel phonemes and their phonetic realisations will be discussed in chapter 2.4.


Vowels
High Front
Low Center
High Back
 
e  /i/
a /a/
o /u/


The Latin transcription might sometimes appear a bit far-fetched.  For example, it would have been intuitive to represent the vowels /a i u/ by **a i u and the glides /j w/ by **j w.  The main reason for the deviations is aesthetics.  A word like */jawtwijsj/ [jQdeS] is arguably more pleasantly rendered as *iautueisi than as **jawtwijsj.


2.2.   Word Structure


In this paragraph, I will use the following conventions, which might not coincide with the official linguistic ones...  but I'm an amateur after all.


There is only a limited number of allowed initial, medial and final consonant clusters, as detailed in chapter 2.3.  


2.3.   Pronunciation of Vowels and Glides


At first, the vocalic inventory of Oro Mpaa may appear spartan, comprising only three vowel phonemes /a i u/ which are represented by the letters a e o. futhermore, there are two glides /j w/ which are represented by the letters i u.

The two letters e o /i u/ dissimulate into [e o] when preceded or followed by the corresponding glide i u /j w/.

If a glide precedes a vowel within its syllable (GV), it is pronounced as a glide; however, if a glide follows a vowel within its syllable (VG), it merges with the vowel into a new sound according to the following table.  Note that a combination VGV is syllabized as V-GV, never VG-V, thus the glide is pronounced.  Phrasal Sandhi (see Chapter 2.5.) can cause a -VG combination to medialize into -VGV-, thus making the silent glide audible again.


GV-Combinations
VG-Combinations
ia /ja/ [ja]
ua /wa/ [wa]
ai /aj/ [E]
au /aw/ [Q]
ie /ji/ [je]
ue /wi/ [wi]
ei /ij/ [e]
eu /iw/ [y] *
io /ju/ [ju]
uo /wu/ [wo]
oi /uj/ [y] *
ou /uw/ [o]


It is also possible that a syllable contain a glide both before and after the vowel (GVG).  In those cases, the initial glide is simply appended before the sound of the following VG-combination, e.g. iau /jaw/ [jQ].  However, such GVG combinations are only legal if the included VG combination doesn't form the sound [y] by itself (these are marked with an asterisk in the table above).

As the following chapter will explain, the consonants t s can become palatized under certain conditions.  If a glide i u follows such a consonant, it toggles its state of palatization.  In such cases, the glide is "consumed" by the consonant, and omitted from speech.  Example:  se [Si], sa [sa], sue [si], sia [Sa].


2.4.   Pronunciation of Kernels


Not all combinations of consonants yield valid kernels.   All valid kernels are listed in the large table below.   It might look a bit overwhelming at first, but it's really quite regular and predictable.

The pronunciation of a kernel depends on its position within its word.  As mentioned in chapter 2.2., one distinguishes initial, medial and final kernels.  Thus, each kernel in the table is followed by brackets containing three pronunciations: [initial medial final].  A dash indicates that the kernel is illegal in that position.

The alveodentals are pronounced palatized if followed by e or i.  As mentioned above, the glide i is not pronounced after such a consonant, it is "consumed" in the palatization.  Similarly, the glide u, which prevents palatization, is not pronounced after a palatizable consonant. Examples: tata [tada], tete [tSidZi], tiatia [tSadZa], tuetue [tidi].

The Latin transcription of some kernels might be slightly non-intuitive (e.g. nt rather than nd for [d nd —]), but keeps true to the native script of the language (which admirably represents its phoneme inventory).  The only concession made for the sake of the reader is the notation of the combinations gc [g Ng —] and gcr [— Ng4 —] as nc, ncr.


Alveodentals (Regular)
Alveodentals (Palatized)
Labials
Velars
t [t d t]
t [tS dZ S]
p [p b p]
c [k g k]
tt [t tt —]
tt [tS ttS —]
pp [p pp —]
cc [k kk —]
s [s z s]
s [S Z S]
v [f v f]
h [h h x]
ss [s ss —]
ss [S SS —]
vv [f ff —]
hh [x xx —]
nt [d nd —]
nt [dZ JdZ —]
mp [b mb —]
nc [g Ng —]
tr [t4 d4 —]
 
pr [p4 b4 —]
cr [k4 g4 —]
ntr [— nd4 —]
 
mpr [— mb4 —]
ncr [— Ng4 —]
ts [ts dz —]
ts [tS dZ —]
 
 
nts [— ndz —]
nts [— JdZ —]
 
 
n [n n n]
n [J J J]
m [m m m]
g [N N N]
nn [— nn —]
nn [— JJ —] mm [— mm —]
gg [— NN —]
r [4 4 4]
 
 
 
rr [— r —]
 
 
 
l [l l l]
l [L L L]
 
 
ll [— ll —]
ll [— LL —]  
 


As can be seen, the alterations of voiced and unvoiced sounds is the same in all four columns.  The one exception is h, where the initial sound is [h] rather than the expected [x].  The sound [x] has a palatized allophone [C] in the vicinity of e and i.


2.5.   Stress and Accent


Most words, especially nouns and verb forms, receive a weak (and often neglected) stress accent on the first syllable of their stem.  This means that productive affixes such as the passive voice prefix me(m)-don't change the location of the stress.  If one wishes to emphasize a given word, one can exaggerate this stress accent, though the pitch of the syllable shouldn't be raised significantly — in fact, lowering the pitch is more appropriate.  The stress accent is notated as [,] in the phonetic transcriptions in this document.

More importantly, all verb forms receive a pitch accent on their first syllable, i.e. it's pronounced at a markedly higher musical pitch than the surrounding syllables.  This shall be represented as ['] in this phonetic notation.


2.6.   Phrasal Sandhi


When several words form a phrase together (e.g. a predicate phrase or a prepositional phrase, see chapter 5), some "fusing" can occur between the word borders.  In analogy to Indian languages, these phenomena are called "Sandhi".  Here follows a list of the different Sandhi processes that can occur.


Medialisation.  This phenomenon happens when
or
In these cases, the two words are effectively fused together.  The kernel that was initial or final before becomes medial, and is pronounced like a medial kernel.

Examples: oro ['u4u], mpaa ['baa], oro mpaa [u4um'baa];  ot [ut], aren ['a4in], ot aren [u'da4in].


Discrimination.  This phenomenon happens when
In these cases, the vowel-glide combination of the first word breaks up, and the glide is pronounced medially between the two vowels (-VGV-).

Examples:  neu [Jy], aren ['a4in], neu aren [Ji'wa4in];  helai ['hilE], aren ['a4in], helai aren [,hila'jarin].


Assimilation.  This phenomenon happens when
In these cases, the nasal of the former word is assimilated to the point of articulation of the latter.  

Examples: an [an], coa ['kua], an coa [aN'kua];  mpa [ba], an mpa [am'ba].


Condensation.  This phenomenon happens when
In these cases, a voiced epenthetic stop condenses between the nasal and the r or s.  The point of articulation of the stop depends on the nasal in the case of r, but is always alveodental for s.  

Examples: seev ['Siif], an seev [aJ'dZiif];  com [kum] "some", rae [4ai] "pretty", com rae [kum'b4ai] "some pretty (ones)".




3.    Nomina


3.1.    Nouns


There is not much to say about nouns in Oro Mpaa, as they don't inflect in any way.  There are no compulsory articles and no inflections for number, so the noun ttou [to] can mean "man, a man, the man, men, the men".  This level of detail is often sufficient in speech.  Should the definiteness or number be important, it can be expressed with a pronoun or quantifier (see following chapter).

Unlike English, Oro Mpaa builds its noun phrases head-first; that is, adjectives and appositions follow the main nouns.   Example: ttou  [to] "man",  cele ['kiLi] "old", mpolen [,buLin] "beard", aaia ['aaja] "white", ttou cele mpolen aaia [,to'giLim,buLi'naaja] "the/an old man with a white beard".

There is a gender distinction in Oro Mpaa, but it strictly follows the biological gender, much as in English.  All living things with a gender are represented with the animate pronouns, while everything else is mapped onto the inanimate ones.


3.2.    Pronouns


Considering the very isolating nature of Oro Mpaa in general, the personal pronouns show a somewhat surprising number of forms that distinguish numbers and genders.

The numbers are singular, plural and categorical.  Plural is used for a specific given group or amount of things, even if it's not easily measured or counted, while categorical refers to the totalty of the things, the things in general.  For example, consider the sentence "Alligators came near the settlement yesterday.  I killed them."  In Oro Mpaa, you would use the plural pronoun om for "them", because you are talking about a specific group of alligators.  If however, you said "I hate them", meaning alligators in general rather than some particular specimens, you'd use the categorical pronoun neu.


Inflected Pronouns
Singular
Plural
Categorical
1st Inclusive
sia [Sa]
siaum [SQm]
neusia  [,nyZa]
1st Exclusive

sseum [Sym]
neusse [,nySSi]
2nd Informal
le [li]
lom  [lum]
neule [,nyli]
2nd Honorific
ahan [a,han]
ahon [a,hun]
neuhan [ny,han]
3rd Animate
an [an]
om [um]
neu [ny]
3rd Inanimate
ei  [e]
el [il]
nel [nil]
Indefinite Animate
cuan [kwan]
com [kum]

Indefinite Inanimate
cuei [kwe]
cuel [kwil]



The personal pronouns can be used by themselves, or as the head of a noun phrase, in which case the function can become similar to a strong definite article: ttou [to] "man", an ttou [an'to] "he, the man (litterally: he man)".  This also works well persons other than the third: Ttaro tuer, neusse ttou tsan neule vel. [,tarudir nySSi,to'dzan nyli,vil] "We men never understand you women."

When placed after a noun, the personal pronoun can act as a possessive pronoun:  sia ttou [Sat,to] "I as a man ..."; ttou sia [to,Za] "my man".  This also works well with names: corampa ttou [,ku4ambat,to] "the man Kuramba", ttou laare [to,laa4i] "Laari's man (husband)".

The indefinite pronouns mean "a, one, someone" in the singular and "some, several" in the plural.




4.    Verbs


4.1.    Verb Forms


The verbs of Oro Mpaa are even easier to use than those of English.  Each verb has only two forms in which in can appear.  A professional grammarian would probably call them finite and non-finite, but I shall refer to them as strong and weak according to the usage of the native speakers.  

The two forms are often related to each other by one of a handful of regular paradigms, but there are many irregular pairs as well.  For example, the verb "talk to" is mpa ['ba] in its strong form and mpaa ['baa] in its weak form; these are clearly related.  However, one might not expect that oua ['owa] is the weak form of mo ['mu].

Strong verb forms...
Weak verb forms...

4.2.    Conjugation Paradigms


It might be a bit pretentious to speak of two form per verb as conjugation, but for lack of a better word, I'll call it that way.

It might also be a bit modest to speak of Oro Mpaa verbs as regular; most of them are paradigmatic, so it's easy to remember a given conjugation pair, but it's not possible to reliably predict a conjugation when only one the base form of the verb is known.

What is the base form, then?  Strong or weak?  The answer isn't quite so simple.  Oro Mpaa can basically describe two kinds of concepts: Actions and states.  Actions are words that have been "born as verbs", their strong (finite) form is the base form, and their weak form is usually an obvious derivative of the strong one.  States, on the other hands, correspond to the adjectives of English: They are mostly used as appositions in noun phrases, thus the weak (non-finite) form is the base, and the strong form is usually an obvious derivative of the weak one.  The strong form of a state is translated as a predicate in English, e.g. "is old", "is pretty" etc.

To respect this distinction, there is the lexical convention of writing the forms of an action as strong/weak and the forms of a state as weak\strong.  This makes sure that the base form, which is usually more recognisable, is mentioned first in word lists and dictionaries.  The direction of the slash still indicates unambiguously which of the forms is the strong.

Here follow a few common derivation processes for the non-base forms.  As mentioned above, they are not predictable with certainty, so one should always memorize both forms when learning a verb.

Action paradigms:
State paradigms:

4.3.    The Prefix me(m)–


Adding the prefix me(m)- to a verb causes it to swap its subject and object, that is, the syntax of that particular phrase becomes OVS rather than SVO.

One might be tempted to view this as a passive voice, but its scope of application is wider than in English, and even intransitive verbs can make use of it.  The prefix doesn't convey a connotation of passive involuntariness like the English passive voice, thus its use in speech is less restricted.

Examples:

4.4.    Tense and Aspect


Marking past or future tense is not a necessity in Oro Mpaa, and is indeed neglected in most cases, since context usually establishes the time frame well enough.  For the sake of precision, one can always start out with (tuer) uaneh, ... [(ti4) 'waJiC] "(time) downstream" for the future and (tuer) heleh, ... [(ti4) 'hiLiC] "(time) upstream" for past tense.  As ever so often, Oro Mpaa makes use of an analogy from its jungle environment here, equating life with the journey of a leaf on a river, never knowing what surprises the next river bend might bring, but certain to be swept up and put to rest on the riverside sooner or later.  

Another mechanism for marking tense is to apply the state verbs uane\uaneh "downstream" and hele\heleh "upstream" to the subject instead of dedicating a clause to them.  

Examples:
As for aspect, Oro Mpaa provides idiomatic uses of the action verbs ntevo/ntevoa "come from" and golua/goluaa "face, look towards, stand before":

Examples:
Obviously, tense and aspects can be marked or neglected quite independently of each other.




5.    Syntax


5.1.    Phrases


There are essentially two sorts of phrases:  Noun phrases and predicate phrases.  In both types, sandhi happens between all constituent words wherever possible.

A noun phrase is an uninterrupted string of nomina (nouns and pronouns).  The first of those nomina is the head, which all subsequent nomina modify.  For example, the aforementioned ttou cele mpolen aaia [to'gilimbuli'naaja]  {man old beard white}  "the/an old man with a white beard" is a noun phrase with the head noun ttou "man".  All other nouns tell us what kind of man it is.

A predicate phrase is a noun phrase plus a directly following verb.  The noun phrase acts as the verb's subject.  Any following object is excluded from the predicate phrase.  Example:  Ttou cele mpa ttou tame [to'gilim'ba to'dami]  {man old talk_to man young} "the old man talks to the young man".  There is phrasal sandhi within the verb phrase ttou cele mpa and the noun phrase ttou tame, but not between mpa and ttou, since these words are not in a phrase together.


5.2.    Simple Clauses


A clause is a very flexible definition in Oro Mpaa:  It's an alternating sequence of noun phrases and verbs which are parsed serially.

A simple clause contains at most two noun phrases and a verb in between.  They are parsed as subject, verb, and object (SVO).  Every one of those components is optional and can be left away.  Both strong and weak forms of verbs count as verbs.  Remember that even some words rendered as an adjective in English, e.g. pale "soft", are actually the weak forms of a verb (pale\palema "to be soft").

Even a single verb or noun phrase can be a valid clause.  Such vestigial clauses usually only make sense in the context of their whole sentence.

5.3.    Serial Clauses


Things get more complicated but also more flexible if we introduce more than one verb in a clause.  Its components are parsed serially from left to right.
Sometimes, the promotion of an object to subject might not be desirable.  It can be hindered with the conjunction hau "and" or a comma to break the phrase flow (see next chapter for punctuation):
In order to express relations beyond subject and direct object, the English language uses prepositions.  In Oro Mpaa, one makes use of idiomatic serial phrases instead.  Some verbs like ee/eia "to go to..." or cala/calaa "serve, benefit" have taken the role of prepositions like "to" and "for".

5.4.    Sentences


A sentence is a sequence of clauses separated by punctuation marks (comma, semicolon) and conjunctions.  In speech, the global tone rises in the last one or two syllables before a comma, while it dips before a semicolon.

The comma is a rather weak separator, it allows the following clause to inherit pretty much all of the context of the previous clause: The most recent subject, object and adverbial circumstances.  If the following clause has a subject and/or object of its own, they override the ones inherited from the previous clause.  

The semicolon is a stronger separator.  While it still keeps the context alive, it puts a greater distance between clauses and brings older subjects and objects back to the surface.

Examples:

5.5.    Relative Clauses


Most English relative clauses can be translated into Oro Mpaa without the use of any special structures for this purpose.  The serial verb syntax, the weak verb forms and the order-swapping prefix me(m)- can already go a long way by themselves.

Examples (subjects and objects):
Examples (adverbials):

5.6.    Questions


Absolute questions are formed by adding the adverb miati [mjaS] to the end of a sentence, separated by a comma.  It's a compressed form of meiat sia [me'jat Sa] {SWAP:ask I}, meaning "that I ask".

Example:
Wh-questions are formed with the state verb oe\oie [ui uje].  Its weak form oe can be used as a noun phrase of its own, meaning "what?", or as an adjective "which? what kind of?".  Its strong form is a predicate meaning "... is what?".

Example:



2003 by Christian "Catharsis" Thalmann
cinga (at) gmx (dot) net