Compromising My Way Toward A Standard Language

[Yaali Annar]

I remember someone here tried to make a "Standard Language" by taking the most common feature in WALS and come up with a language for them.

I'm attempting the same, and in my progress I found something interesting: Sometimes the most common features clashes against each other. When this happens, I have to take a Compromise. There are several ways I'll compromise:

1. I can use the combination capability of WALS (I want to say combination feature, but it might cause confusion with the term "features" used in WALS)
2. I can modify the feature values as I see fit, or
3. I can choose the second most common feature.

So, I start with phonology. The restriction is pretty simple:

1A. Consonant Inventories: 19 to 25 Consonant. I'll pick 19. I don't like too much consonant
2A. Vowel Quality Inventories: 5 to 6 Vowels. I'll pick 5
3A. Consonant-Vowel Ratio: 2,75 to 4,5. 19/5 = 3,8

4A. Voicing in Plosives and Fricatives: Plosive Only
5A. Voicing and Gaps in Plosive Systems: /p t k b d g/. The full BeGeD KePeT
18A Absence of Common Consonants: All present (the language have bilabials, fricative, nasal)

6A. Uvular Consonants: None
7A. Glottalized Consonants: None
8A. Lateral Consonants: Approximant <l> Only
9A. The Velar Nasal: No velar nasal
19A. Presence of Uncommon Consonants: None, (no clicks, no labial-velars, no pharyngeal, no interdental)

10A. Vowel Nasalization: None
11A. Front Rounded Vowels: None
12A. Syllable Structure: Moderate. I'll go with CVVC
13A. Tone: None

The vowels are pretty much /a e i o u/ and as for diphtongs, there are /ai ei au ou/

But which consonants should I pick? First I'll pick 19 most common consonant in Phoible and then rearrange the results into a neat-ish rectangle. Anyway the 19 most common consonant in the phoible are: /m k j p w n s t b l h ɡ ŋ d ɲ f t̠ʃ ʔ ʃ/

But like I said, I want a symmetric system. So this is what I came up with.

Code: Select all

m n ɲ
p t tʃ k
b d dʒ ɡ
f s ʃ  h
  r j  w
  l  

The language is more restricted on what it can have on the end of the syllables.

Code: Select all

m n ɲ
p t tʃ k
f s ʃ  h
  r
  l  

Point 14A and 15A is when I start having to compromise.

14A. Fixed Stress Locations: No fixed stress (mostly weight-sensitive stress)
15A. Weight-Sensitive Stress: Fixed stress (no weight-sensitivity)

Now, using the combination capability of WALS:
- Fixed stress (no weight-sensitivity) / Penultimate (110 Languages)
- Fixed stress (no weight-sensitivity) / Initial (92 Languages)
- Right-edge: Ultimate or penultimate / No fixed stress (65 Languages)

The Penultimate is the second most common value for feature 14A

Now the result becomes:

14A. Fixed Stress Locations: Penultimate: stress is on the penultimate.
15A. Weight-Sensitive Stress: Fixed stress (no weight-sensitivity).

Incidentally the next feature agrees with the compromise:

16A. Weight Factors in Weight-Sensitive Stress Systems: No weight, or weight factor unknown.

I have no idea what feature 17A means, I'm always bad at rhythm and stress.

17A. Rhythm Types: Trochaic, left-hand syllable in the foot is strong.

That's it for phonology, next post will be about Morphology.

[Yaali Annar]

[MORPHOLOGY]

20A. Fusion of Selected Inflectional Formatives: Exclusively concatenative (not isolating, not tonal, not ablauting)
26A. Prefixing vs. Suffixing in Inflectional Morphology: Predominantly suffixing
27A. Reduplication: Productive full and partial reduplication

Beside the reduplication, this sounds quite European.

[ADPOSITIONS]

48A. Person Marking on Adpositions: Adpositions without person marking.
52A. Comitatives and Instrumentals: Differentiation.

I recall that adpositions with person marking is a Thing in Irish. Also, the language will have two different kind of "with"

[ADNOMINALS]

The section on indefinite and definite article clashes here:

37A. Definite Articles: Definite word distinct from demonstrative.
38A. Indefinite Articles: Neither indefinite nor definite.

Let's combine them then:

- No definite or indefinite article / No definite or indefinite article (198 Languages)
- Definite word distinct from demonstrative / No indefinite, but definite article (59 Languages)

So there won't be any definite or indefinite articles

37A. Definite Articles: Neither indefinite nor definite.
38A. Indefinite Articles: Neither indefinite nor definite.
41A. Distance Contrasts in Demonstratives: Two way. So, the standard this and that.
42A. Pronominal and Adnominal Demonstratives: Same forms.

[PRONOUN]

35A. Plurality in Independent Personal Pronouns: Person-number stem
39A. Inclusive/Exclusive Distinction in Independent Pronouns: No inclusive/exclusive opposition
40A. Inclusive/Exclusive Distinction in Verbal Inflection: No inclusive/exclusive opposition

This means that there are unanlyzable stems for every person/number combinations. In plural, there won't be any clusivity distinction.

As for the third person.

43A. Third Person Pronouns and Demonstratives: Third person pronouns and demonstratives are unrelated to demonstratives
44A. Gender Distinctions in Independent Personal Pronouns: No gender distinctions
45A. Politeness Distinctions in Pronouns: Second person pronouns encode no politeness distinction

No gender, no politeness, very egalitarian.

46A. Indefinite Pronouns: Interrogative-based indefinites
47A. Intensifiers and Reflexive Pronouns: Intensifiers and reflexive pronouns are formally identical

So somebody and something is derived from the word for "who" and "what". Feature 47A implies that there's a reflexive pronoun.

[NOUN]

21A. Exponence of Selected Inflectional Formatives: No Case.
49A. Number of Cases: No morphological case-marking.
50A. Asymmetrical Case-Marking: No morphological case-marking.
51A. Position of Case Affixes: Case suffixes

Ah another clash, let's combine 49A and 51A:

- No case affixes or adpositional clitics / No morphological case-marking (74 Languages)
- Case suffixes / 6-7 cases (18 Languages)

No case it is.

51A. Position of Case Affixes: No case affixes or adpositional clitics
24A: Locus of Marking in Possessive Noun Phrases: Possessor is dependent-marked.
25B. Zero Marking of A and P Arguments: Non-zero marking
28A. Case syncreticsm: Inflectional case marking is absent or minimal

The possessor is dependent marked, which means the marking is on the well... possessor (by using genitive or conjunction like english "of" or romance "de"). But I wonder... what's the most prevalent strategy in a language with no case? So I combine it with 21A:

- Head marking / No case (35 Languages)
- Dependent marking / Monoexponential case (34 Languages)
- Dependent marking / No case (21 Languages)

It seems that head marking is more common in language without case. Since I personally like the idea of having possessive affixes I'm gonna change the value for 24A.

24A. Locus of Marking in Possessive Noun Phrases: Possessor is head-marked.

As for genders and plurals

30A. Number of Genders: None
31A. Sex-based and Non-sex-based Gender Systems: No gender system
33A. Coding of Nominal Plurality: Plural suffix
34A. Occurrence of Nominal Plurality: Plural in all nouns, always obligatory
36A. The Associative Plural: Associative plural marker also used for additive plurals

The plural strategy is basically European. Except for the associative plural. IIMN Associative plural is basically like... "x and friends"

[VERB]

21B. Exponence of Tense-Aspect-Mood Inflection: Monoexponential TAM
22A. Inflectional Synthesis of the Verb: 4-5 categories per word
23A. Locus of Marking in the Clause: P is head-marked

I have a feeling that I'll violate point 22A. The verb will definitely have a lot of affixes attached to it. The head marking part is connected to the point 21A. The language has no nominal case, but the verbs are marked for person.

25A. Locus of Marking, Whole-language Typology: Inconsistent marking or other type.
29A. Syncretism in Verbal Person/Number Marking: Subject person/number marking is never syncretic

This means there's a unique suffix for each person and number marking.

[Ser]

It was vecfaranti and he called it an "average language".

viewtopic.php?f=4&t=36730

[Ser]

24A. Locus of Marking in Possessive Noun Phrases: Possessor is head-marked.

23A. Locus of Marking in the Clause: P is head-marked

25A. Locus of Marking, Whole-language Typology: Inconsistent marking or other type.

I found a clash in your choices. Your whole-language typology is actually "head-marking", not "inconsistent marking or other type", since both 23A and 24A have the head-marking value.

[cromulant]

[VERB]

21B. Exponence of Tense-Aspect-Mood Inflection: Monoexponential TAM
22A. Inflectional Synthesis of the Verb: 4-5 categories per word
23A. Locus of Marking in the Clause: P is head-marked

I have a feeling that I'll violate point 22A. The verb will definitely have a lot of affixes attached to it. The head marking part is connected to the point 21A. The language has no nominal case, but the verbs are marked for person.

Why? You can easily keep the verb in the 4-5 category range. You'll need nom-acc subject/agent and patient agreement (2 categories), tense (past/nonpast, 1 category) and mood (including imperative and epistemic possibility, 1 category). That's 4 categories, all your rules require. You probably won't have aspect: see features 65 and 68. You probably won't have voice: see 107, 108, 109. You could add another category if you wanted.

25A. Locus of Marking, Whole-language Typology: Inconsistent marking or other type.
29A. Syncretism in Verbal Person/Number Marking: Subject person/number marking is never syncretic

This means there's a unique suffix for each person and number marking.

That's still one category, though.

[Yaali Annar]

I found a clash in your choices. Your whole-language typology is actually "head-marking", not "inconsistent marking or other type", since both 23A and 24A have the head-marking value.

Yeah, I think I'll make the language consistently head-marking (it's the second most common after inconsistent marking).

Why? You can easily keep the verb in the 4-5 category range. You'll need nom-acc subject/agent and patient agreement (2 categories), tense (past/nonpast, 1 category) and mood (including imperative and epistemic possibility, 1 category). That's 4 categories, all your rules require. You probably won't have aspect: see features 65 and 68. You probably won't have voice: see 107, 108, 109. You could add another category if you wanted.

Well... it's just a feeling. But I would probably ignore this restriction because it's more... nebulous than simply restriction like "no passive construction" (which I also violate here).

So anyway, after I downloaded the data set, import it into my local DB I found that there are largely two trend-set for languages: Language with VO (Verb-Object) word order have different trend-set than language with OV word order. For this language I sometimes pick the trend-set that is more common with OV languages, than trend-set that is most common in general.

Why OV and VO of all things? Well, one of the reason is because the trend-set is explicitly mentioned it in point 95A, 96A and 97A. This is the most common values for the three of them:

- OV and Postpositions
- VO and NRel
- VO and NAdj

Since I'm sticking with OV language, I have to compromise with point 96A and 97A:

- OV and Postpositions
- OV and Rel-N
- OV and N-Adj

Additionaly, I will sometimes also pick the second most common term not because of compromize but because it jives better with my aesthetics.

So... verb conjugation. There's a clash here between 102A (the language will mark both A and P argument on verb) and 104A (A and P argument do not occur in the verb at the same time). As usual, Let's combine them:

Code: Select all

 
A precedes P      / Double marking      : 96 Languages
No double marking / No person marking   : 82 Languages
No double marking / Only the A argument : 73 Languages
P precedes A      / Double marking      : 57 Languages

So A preceds P. Since there's no rule on whether or not the marking are suffix or prefix I decided to come up with A-verb-P (this arrangement is probaly uncommon, but let's spice up the language a bit)

Code: Select all

  sg          pl
1 ku(h)- -ku  ta(h)- -ta
2 su(h)- -su  mu(h)- -mu
3 ńa(h)- -ńa  ja(h)- -ja

As per 103A, the third person is non zero marking. Verb with no subject prefix will turn into participle/adjectival verb.

Code: Select all

Pul ńamordo : The wolf sleeps
Pul mordo   : The sleeping wolf

This ties in with point 118A where predicative adjectives has verbal encoding. In this way, adjectives behave as stative verbs.

Code: Select all

Pul ńakanlaba : The wolf is white
Pul kanlaba   : The white wolf

Does this mean we can use this strategy for relativization? Kinda... the difference being that relative clause is placed before the noun:

Code: Select all

Pul nińeku ńarukńa : The wolf chases my husband
Nińeku rukńa pul   : The wolf who chases my husband

In order to allow for relativization on subject I choose to allow passive into the language even though it's not the most common trend. Active voice is null, while pasive is a suffix before the object suffix. Making the verb paradigm:

Subject-Verb-Voice-Object

Code: Select all

Nińeku pul ńarukbińa : My husband is being chased by a wolf.
Pul rukbińa nińeku   : My husband whom the wolf chase.

I just realized that the sample sentence is a bit tongue twisty with the whole ń vs n.

I'm not sure what kind of voice would be for relativization on oblique (such as location). I would probably be forced to use pronoun retention.

Code: Select all

Raćap  mana         eć jol   heyem
father it.INANIMATE at dwell house
The house where my father lives.

PS: You can probably tell that I don't come up with the words a posteriori.

[Yaali Annar]

Negation has a lot of compromise between 143, 144, and the language being OV. According to the 143 set (A,E,F) the negative morpheme is a preverbal word similar to spanish "no". But then we go to 144A and it said that the most common form of negative morpheme is morphological. Additionaly 144L says that the position of negative morpheme is a suffix.

Thus the verb gained another affix: -dek

Subject-[Verb]-Voice-Negation-Object

The imperative is considered as a special form of subject prefix. So you might say that there's a fusion between mood and subject agreement. Additionaly, per 72A the hortative is lumped together with the imperative in a system that is "neither maximal nor minimal". This give rise to the following prefixes.

Code: Select all

  sg     pl
2 solu(h)- molu(h)-
3 ńola(h)- jola(h)-

Code: Select all

Solumordo : Sleep!
Ńolamordo : Let him sleep.

As per 71A, the negation of the two mood, known as prohibitive, uses a non-indicative negation: -jan

Code: Select all

Sumordo      : You're sleeping.
Sumordodek   : You're not sleeping.
Solumordo    : Sleep!
Solumordojan : Don't sleep

---

As per 74A a verbal construction is used for situational possibility.

Code: Select all

Tarafku ribil ńagelná         : My brother is reading a book.
Tarafku ribil ńavop gelńa     : My brother can read a book.
Tarafku ribil ńavopdek  gelńa : My brother can't read a book.

While epistemic possibility is conveyed through other means, basically adverbs.

Code: Select all

Tarafku ribil ńavop gelńa   : My brother may read a book.
Tarafku ribil tesop ńagelńa : My brother may (probably) read a book.

---

This time it's about posession, the locus of marking is onthe head, that means there are possessive suffix. The same suffix used for verbs' object is used here.

Code: Select all

Dunuhku : My dog
Dunuhta : Our dog

The 3rd person is used for non pronominal posessions. Additionaly, genitive follows noun:

Code: Select all

Dunuhńa        : His dog.
Saćtis dunuhńa : The hunter's dog.

As per 117A, predicative posession is conveyed by the verb "have" just like in English.

Code: Select all

Dunuh ńavah : I have a dog

[Yaali Annar]

I like the direction this language is going. It's not "boring" at all IMO.

With SOV and no case, when there's an [noun] + [verb] construction is it an SV or OV?

There are two posibility that leads to [noun] + [verb] the first one is intrasitive verb. Which is obvious.

Code: Select all

Dunuhar jadormo : The dogs are sleeping. 

With transitive verb, the [noun] + [verb] construction is caused by pronoun dropping. From the verb itself it should be obvious which pronoun is dropped.

Code: Select all

Yaku dunuhar kutaśja : I like dogs 
Dunuhar yaku jataśku : Dogs like me

Dunuhar kutaśja : I like dogs 
Dunuhar jataśku : Dogs like me

Of course, both pronouns can be dropped:

Code: Select all

Kuhamamu : I love you.

Aspect is not marked in the verb. As for the tenses, I decide to pick the tense through combination of 67A and 66A

Code: Select all

Future       Past

No future  / No past tense                       : 53 Languages
Present    / Present, no remoteness distinctions : 48 Languages
No future  / Present, no remoteness distinctions : 46 Languages
Present    / No past tense                       : 35 Languages

Whoops no tenses as well. It's entirely my Indonesian bias, but I settled with no tense inflection.

So, anyway Reflexives and Reciprocal

As per 47A, the language have reflexive construct. The reflecive construction is redundant. It's achieved by partially reduplicating the verb.

As per chapter 47 the word used for reflexive construct is the word for "body" (porok). This is where it gets a bit interesting, the reflexive pronoun is the word for body (porok) with possesive suffix added to it. Note how the object slot in the verb is not filled.

Code: Select all

Porokku kudidiyu : I see myself
Porokńa ńadidiyu : You see yourself

Lugepse eć porokku kudidiyu : I see myself in the mirror

The reflexive verb can be used as adjective.

Code: Select all

Lireyu riritis : Self driving vehicle.
Dorop kukuba   : Automatic door (self-opening door)

When the verb is not reflexive, the porok- pronoun becomes an intensifier. Replacing a pronoun, basically.

Code: Select all

Porokta meyeh tadańumńa : We build the house ourselves. 

Reciprocal verb is created by full reduplication of the base verb. There are two strategies of reciprocality. The first one involves plural person and doesn't require the object slot to be filled:

Code: Select all

Tadiyudiyu : We see each other

The second form involves two person.

Code: Select all

Mudiyudiyuńa : You and him see each other.
Ńadiyudiyumu : He and you see each other.