Sangi

=Orthography=

Basics
The basic letters, i.e. those without diacritics are pronounced as in English, except “j” which is pronounces at a “y” would be and “y” in Sangi is pronounced as it is in Finnish. There is no “k”, “q” or “x” but there are two additional letters “þ” and “ð” which pronounced as the “th” is in “bath” and “bathe” respectively. These 2 extra letters are more commmonly written as ŧ and đ in more informal writing.

Diacritics
An acute accent over a vowel makes it long. An umlaut indicates I-Affection and a circumflex indicates A-Affection. There pronunciations are given below. The letters “ę” and “ų” represent the schwa and the “u” in “but” in RP. An acute accent over a consonant indicates that it is palatal, so “c”, “g”, “s”, “z” will be pronounced “ch”, “j”, “sh” and “zh” respectively. The Cedilla under the consonants “t”, “d”, “n”, “l”, “r”, “s” and “z” create the “retroflex consonants”. In spite of the name, most of these consonants are true palatal consonants, unlike the “palatal” consonants, which are alveolar-post-alveolar affricates. “t”, “d”, “n” and “l” with the cedilla are pronounced “c”, “ɟ”, “ɲ” and “ʎ” respectively, while an “r” with a cedilla is a retroflex “ɽ”. The pronunciation of “s” and “z” with a cedilla is not generally set, and they can appear as retroflex (ʂ/ʐ) alveo-palatal (ɕ/ʑ) or even velar (x/ɣ). Usually they are pronounced as alveo-palatal consonants before “i” and “e”, velar before “a”, “o” and at the end of a word, and retroflex in other environments. The sound they take when next to a consonant is the one most similar to that consonant, but is more prone to differences between speaker to speaker. They do also appear as “ç” and “ʝ” when in a cluster with true palatals. If a retroflex consonant is created by a following "r" then this "r" is lost completely. If the consonant is created by a preceding "r" then this following "r" is retained as a retroflex one instead.

=Phonology=

Introduction
The most basic of all words in Sangi are derived from Modern (British) English words, mostly of Germanic origin, through a series of phonological changes.

There are also a large number of consonant and vowel mutations which occur at morpheme boundaries and within the words themselves.

When creating a new word from an English base, the consonants are changed first according to their environments. Next the vowels are changed. This linear process is followed with only two exceptions. They involve the change of medial and final consonants. If the vowel before the medial or final consonant becomes a vowel followed by a consonant, i.e. r, then this change occurs before the consonant change.

consonants
ɕ and ʑ also exist as allophones of ç/ʂ/x and ʝ/ʐ/ɣ (possible underlying ç and ʝ)
 * ɮ is an allophone of ɬ which appears before voiced consonants.

vowels
Allowed combinations of vowels are very varied and depend on local traditions on epenthesis.

Phonological Changes
The changes which a sound is subject to are dependent (usually) on its position in the word and the surrounding sounds.

Initial Consonant Changes
'w' also becomes 'Ø' before o unless the w becomes v in which case the w disappears and the o becomes an a.

Medial Consonant Changes
'w' also becomes 'Ø' before o unless the w becomes v in which case the w disappears and the o becomes an a.
 * See the table of Other Consonant Changes

Final Consonant Changes
¹ The forms of this new vowel complex can be found in th table for Vowel Changes.
 * Retroflex after VVC#>VrC#

Consonant Cluster Changes

 * These also occur in the spoken and written language at the beginning of words when the final sound of the previous word was a vowel, e.g. “an e ssil – and yet still”. It is a kind a sandhi. This process also happens after a prefix which ends in a vowel.

¹ These sounds, or there alveolar parts, become retroflex under the normal conditions, i.e. preceded and/or followed by an "r"

Other Consonant Changes
¹ This applies only to plosives, sibilants and nasals in there basic form.

² The j affects the fricative consonants in the following way: f > p; v > b; th > t; dh > d; ch > c; gh > g; l > jj; r > s.

³ This change means that the only Cl combinations are cl, gl and hl. After all other consonants l becomes i but becomes j before i to avoid the creation of a long vowel and causes palatalisation of the preceding consonant. If l is preceded by an original initial “p”, now absent, it is still either “i” or “j”.

Palatalisation
Palatalisation is a change which occurs after the new word has been made from the original English. It occurs when a word ends in an -i but is then followed by a vowel ending, e.g. “-a” for female nouns in the nominative case. It also occurs when and l becomes a j. It only affects a small number of consonants and should not be confused with I-affection, which is a morpho-phonological process.

t > ć

d > ǵ

s > ś

z > ź

When the stem undergoes any form of consonant mutation, the original combination of “Cj” or “Ci” re-emerges. It is this combination which goes through the mutation, not the palatalised form. The “j” is, however, ignored under these processes if the new form can be palatalised, if it can not then the “j” is taken into account and the stem will not mutate, for example “-tj-” becomes “-ć-” under palatalisation. When subjected to plural mutation “t” becomes “ð”. “ð” cannot be palatalised so the plural form of “-tj-” remains “-tj-” and therefore becomes “-ć-”, even though “-ć-” can undergo plural mutation. The “t” of the same “-tj-” combination, when subjected to consonant gradation, becomes “d”, which can be palatalised. The new sequence will be “-dj-” which becomes “-ǵ-”. This process does not occur for -Ci- sequences. Here the -i- re-emerges and the consonant is changed regardless. The -i- only disappears again when the new consonant can be palatalised otherwise it remains intact. So the sequence -ti- will change to “-ć-” when palatalised and the “-ti-” sequence will become “-ði-” in the plural and remain as such. If the original "-ti-" sequence were weakened, however, it would become “-di-” which would palatalise to “-ǵ-”. If the sequence were “-Cji-” then it is seen as “-Cj-” followed by a “-i-”. This means the "C" will palatalise but will not undergo any further form of mutation as the "-i-" is seen as the stem ending.

Vowel Changes

 * Syllabic “m”, “n”, “l”, and “r”. The inserted “e” is lost when the consonant is followed by a vowel though. It acts more as an epenthetic vowel that still causes change in consonants when forming new words from English roots. When a compound is formed and the first word ends in “e” (original “r”) and the next begins with a vowel, the original “r” will be restored and the “e” lost. A special circumstance for this change is when forming surnames in which the second begins when “l”. The “r” returns causing the “l” to become retroflex and the “r” to disappear again.

NOTE: if the vowel becomes -Vr- and the following consonant can be retroflex, it becomes retroflex and the -Vr- sequence is reduced to -V-. This process does not affect word final “t” which is not counted as a sound as it disappears earlier than in-word “t”s meaning that any long vowel before final “t” which would become -Vr- do not do so. In other words VVt# becomes V#, e.g. cart>sa which is in fact a homophone of the word for cat (sa).

Phonological Processes at Morpheme Boundaries
As part of the nature of the phonology of the language only voiceless consonants (includding clusters) and the “r”, “l”, “n” and “m” can occur at the ends of words. If a voiced consonant, a voiced-final consonant cluster or a geminate cluster occurs, then the letter “-i” is placed after them.

No sound is immune from these processes so some ambiguity may be caused during declension and conjugation, but context usually clears this up. There are exceptions but these are rare and usually carry some meaning.

These changes can occur at word boundaries, also, if a final epenthetic vowel is removed or does not occur. In the cases where only one consonant is formed the punctuation mark ` is used in place of ’ to represent that 2 final sounds are missing. This process only occurs in certain styles of writing, but more regularly in speech. The only ones which are never used at word boundaries are the ones such as “mp>mm” which are dependent on the second sound over the first (go across, not down on the table).

In consonant combinations of -r+dental- the dental consonant is made retroflex while the -r- disappears. If an “r” occurs before a retroflex consonant it also becomes retroflex while remaining intact. This feature is also expanded to suffixes of this structure and is quite common, almost obligatory for most people. There are also a set of sound changes which occur with retroflex consonants and plain alveolars. When a retroflex precedes and alveolar, the alvelor becomes retroflex and the same pattern of changes occurs as with plain alveolar combinations, e.g. t.+n = t.+n. = n.n.. The same patterns also occur in retroflex-retroflex combinations, e.g. t.+z. = t.s.. In these changes, a retroflex "r" is preserved, as no process of retroflex creation can occur. The processes are more complex, however, when alveolar consonants precede retroflex ones. The retroflex consonants tend to become the "palatal" equivalent or the alveolar one plus the sound "i".

Epenthesis: When a vowel follows another vowel and does not form an allowable diphthong then a consonant is inserted between them. An “h” [x/ç] is placed before “e” which becomes “ę” at the end of a word, and “r” before “a” and “i” and a “z” before other vowels. When a consonant follows another and cannot be put together in the above to ways, then a vowel is inserted between them. A “ę” is placed between like consonants, and “i” is placed after labials, an “a” is placed after dentals and an “e” is placed after velars.

Examples of Epethesis:

i+a > ira “I hit”

An exception to these rules occurs when r is followed by another consonant. Instead of placing an epenthetic vowel between the r and the consonant an ę is placed after the combination.

Morpho-Phonology
As said above, there are a large number of consonant and vowel mutations which occur at morpheme boundaries and within the words themselves. There are, in total, four consonant mutations and four vowel mutations.

Consonant Mutations
The four different consonant mutations are I-affection, A-affection, plural mutation and stem gradation. I-affection and A-affection do not affect the meaning of the word and are simply phonological processes which affect the last consonant[s] of the word. Plural mutation and stem gradation, on the other hand, help in changing the meaning of the word.

I-Affection
I-affection occurs when the last consonant of the stem is immediately followed by the vowels i and e (long, short and diphthong forms) and also the consonant j.

I-Affection does not affect geminate l, though.

The "pf" form occurs with a singular "p" mutating, "ff" with "pp".

A-Affection
Like I-affection, A-affection occurs when the last consonant of the stem is immediately followed by a vowel, in this case a.

A slight irregularity in this system occurs as “w”, “h” and the velar stops are affected in the above way by a following “o”. A-Affection doesn’t affect consonant clusters, unlike I-Affection, except when the final consonant is an h, in which case the h, unexpectedly, becomes an the a-mutated form of the preceding consonant, or the final consonant of this if it is a cluster itself, e.g. nh>nd, rh>rź, etc.

Plural Mutation
This mutation occurs when forming the plural of the four noun stems (see below).

When the plural stem is weakened or strengthened then the retroflex form of the plural is used as the stem consonant.

Stem Gradation
Stem gradation is a weakening of the final consonant of the stem to form different noun and verb stems which are used to built up the meaning of the noun or verb as well as to derive words from different classes of words.

pp > p > b > v > f > w > h > hi > he

tt > t > d > j > i > e*

cc > c > g > j > i > e

m > mb > mm > mp > mi > me

n > nd > nn > ns > nt > ni > ne*

ň > ŋ > ng > nc > nct > ŋi > ŋe

l > ld > ll > ls > lt > li > le*

ł > łd > łł > łs > łt > łi > łe

r > rd > rr > rs > rt > ri > re*

ř > řd > řř > řs > řt > ři > ře*

s > t > d > j > i > e

sp > sw > sl > śi > śe

st > ss > ś > si > se

sc > ś > si > se

h > p > b > w > hi > he

Ø > c > g > j > i > e (if the Ø was originally a c)

Ø > t > d > j > i > e (if the Ø was originally a c)*


 * These gradation series can also be applied to retroflex consonants. So if the sequence was originally VVt# then the t is reinstated as retroflex, removing the ambiguity caused by the process VVt#>V#. The “j” however is pronounced as “hj” and written “ɧ”, while the “i” and the “e” are pronounced as normal.

It should be noted that “s”, “tt”, “cc” and “Ø” all eventually weaken to the same pattern, causing a certain amount of ambiguity if the gradation is carried through enough times. Also, grades of the pattern “Ci” and “Ce” can never be the general stem, being analysed as “C-i/e” where C is graded, but “i” and “e” can be strengthened.

Vowel Mutations
There are two sets of vowel mutation, both with two further subsets. One set is pure mutation which causes a semantic change is the word. The other is affection, which is caused by pure vowel mutation.

Pure Mutation
Pure mutation has two subsets; A-mutation and I-mutation. I-mutation involves the raising of the final vowel of the stem, while a-mutation involves the lowering of the final vowel.

¹ long “a” becomes “œ” in a-mutation, but is written as a two circumflexed “a”s.

² long “i” becomes “ia” in i-mutation and long “u” becomes “wá”, but are both written as a two umlauted letters.

In the orthography the base vowel is written as it is with a diacritic to mark that it is mutated. I-mutation is shown by the use of the umlaut and the letter “y” for the sound “ü”. A-mutation is shown by the use of the circumflex. If the word changes semantic class, e.g. a verb becomes a noun, then the vowel with the diacritic is written with the vowel that marks its pronunciation, e.g. “ü” > “au”. This is not the, however, when a plural noun is used as a stative verb, the vowel marked with the diacritic is kept to mark the plurality and a special tense infix is added.

Vowel Affection
Affection comes in two subsets; I-affection and A-affection. Under I-affection the vowel is raised and under A-affection the vowel is lowered, just like in pure vowel mutation and they are represented in the orthography by a grave accent and a breve respectively. The difference is that vowel affection has no semantic effect on the word and occurs under different circumstances. I-affection occurs when the last vowel of the stem was raised to either an “i”, “y” or “e”. A-affection occurs when the last vowel was lowered to either an “o” or an “a”. This means that vowel affection can only occur in disyllabic or polysyllabic words and only on the penultimate syllable of the stem. “i” and “u”, though, are not affected by vowel i-affection and “a” is not affected by a-affection. Affection also occurs when a derivational suffix is added. This changed the root vowel, but it is marked by pronunciation and not by diacritics as with affection caused by pure mutation, e.g. lap (laugh) > läp (laughs) but lap (laugh) > lepfi (chuckle). Vowel Affection, also, does not affect truly mutated vowels. Unlike normal affection, that which occurs when a new stem is formed, acts with each subsequent suffix added.

=Morphology=

Verbs
Verbs are possibly the most important class of words in Sangi. Because their conjugation includes different endings for the subject, object and secondary objects, it is possible to drop all pronouns used in a sentence and it will be understood. This means that the verb is the only part of the sentence necessary for the sentence to be complete.

Nouns
Nouns are nowhere near as complicated, in terms of structure, verbs are, but the number of suffixes available to choose from for certain slots is much higher. They also have a complex set of phonological rules through which number and case are defined. Full use of all slots leads to a number of forms for each noun exceeding fifty-four thousand, more specifically the full number of forms is 54,181.

Adjectives
Sangi Adjectives decline in a simialr way to nouns, but for fewer cases and no number. They have, however, a complex comparitive system and the ability to carry demonstrative and indefinite suffixes in place of the pronouns themselves. With adjectives not declining for case and possessive and the like, full adjective declension leads to 155 forms.

Derivation
Sangi derivation usaually affects verbs in there form. Nouns and adjectives can be used interchangeabley with only a change in how the can be declined.

Pronouns
There are seven classes of pronoun in Sangi, each one declined as if it were a noun. These classes are personal, demonstrative, interrogative, temporal, spatial relative and indefinite.

=Syntax=

Unlike in English, the main sentence structure is VSO (verb-subject-object). The sentence is built around the principle of what is viewed as the most important feature first. This means that verbs always come first, nouns afterwards, adjectives come after nouns and adverbs come after verbs. Pronouns are rarely used in written form, but long winded verb constructions mean that in the spoken language they are used for all cases but the nominative. The word order, however, is not fixed and entirely up to the individual, VSO is simply the most common, giving more specific information about what happened as the sentence develops.

A note on "I love you"
The number of intensifiers and pronouns means that the single English phrase “I love you” can be said many different ways in Sangi using a single word “Lawel”. If it is conjugated with the intensive suffix it can mean “adore” if this is then used with the desirative pronoun it can mean something even stronger. If used with the honorific pronoun it can mean “respect” and “love” at the same time. The combinations vary in meaning from person to person as verbs of emotion in any language never convey the actual feeling intended to be described by the speaker. This idea is representative of a large scale phenomenon in which suffix and pronoun choice can affect the overall meaning of a sentence, which occurs most often with verbs of emotion. Overall, this can make describing the exact feeling easier than it is in English.

=Context and Culture=

Real-World Development
Sangi was developed using my knowledge of phonetics, morphology and linguistic change. I started with a base vocabulary, i.e. English words of Germanic origin and applied a set of sound change rules to these. Next I developed a series of grammatical suffixes influences by Finnish and Estonian. I wanted to add a bit of a natural feel to the language so I developed a set of sound changes to create a degree of spoken ambiguity. For example, plurality was originally marked by the suffix -i or -t depending on the final sound of the word. This later became -[i]t and then the -t was dropped all together. I looked at languages like Welsh and Norse, dropped the -i and developed an i-mutation rule. I added other morpho-phonological rules to cause further changes of the stem based on context. So the language is almost entirely unnatural but the ambiguity in the spoken language gives it a kind of natural feel.

Fictional History
I had the idea that I should develop a society for this language to actually have some speakers, so I developed the society around what I saw as the perfect one for me. This was a kind of anarchist society (not to be confused with anarchy, but based upon anarchism as well. Trust me, there is a difference, but people are not taught this by those in charge of society). There exist no rules, only personal moral codes, no hierarchy, but direct democracy among the communitites that are affected by an issue. In this type of democracy, no action is taken unless 100% of the community agrees with the idea, so there is a lot of negotiation and debate before anything is done, but so much focus is put on these debates that they can last, at the most, 1 week, or there abouts. War is non-existent so military decisions or not necessary (just in case anyone thought war would be slow or their armed forces would easily be defeated). For this society to work it would have to be the only kind of society in place, so I placed it on another world. The geography of this new world was never considered when colonised by it's settlers from Earth after they knew it could support life long-term. If you want to go somewhere, you know only the climate will change. The people will hold the same general beliefs as you but, as they do on Earth, their personalities will differ.

Art is different in this society. Writing is highly practiced and many novels, poems, etc. of all genres exist. The static visual arts, i.e. painting and scultpure, are rarely produced and those that are produced are rarely viewed. The realm of visual art is dominated by calligraphy and theatre. Theatre is divided into two classes, the opera, which combines the auditory arts and the visual, and the play, which combines the written arts with the visual. The auditory arts are divided in three; instrumental, vocal and mixed. All forms of art focus on the ideals of society. War is only mentioned in terms of a moral war or the war against the oppressive regime of the old home-world, which is never called Earth, but is instead called "the World of Rulers" or "œlso rujambil"

Numbers
The numbers can easily be seen to have derived from English as many of them have the same general sounds.

1 - an

2 - su

3 - ţi

4 - wo

5 - wáw

6 - dikra

7 - deben

8 - é

9 - nán

10 - ćen

Numbers 11 to 19 are built up on the system "10+number", e.g. ćen-an (11), ćen-wo (14) and so on. Numbers 20 to 99 are built on the system "number-multiplicative-10-number", e.g. suntćen (20), éntćen-nán (89) and so on. Higher numbers are built up in this manner, attaching the multiplicative suffix to the number and placing the multiple of ten after it.

The words for higher multiples of ten are:

100 - aņțes

1000 - túðen

10000 - ćenden

100000 - aņțesten

1000000 - mijjen

An example of a higher number is dikrantmijjen-suntaņțes-wontćen-su (142)