Page 1 of 3

The Most Average Conlang Ever, NP: Qmachah!

Posted: 22 Nov 2014 16:31
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Big Update'.

So I'm taking a short break from my Oscan project to present a bit of a speedlang. It is inspired by the ongoing WALS poll conlang running on Google+ by Pete Bleackley. This language is designed to score with the largest proportion of languages for every feature on the World Atlas of Language Features, although there are a few where it doesn't quite conform. So far I'm up to chapter 76, and I thought it would be a good spot to give a little taste.


/m n ɲ/ <m n nj>
/p b t d k g/ <p b t d k g>
/ts dz tɕ dʑ/ <ts dz tj dj>
/f s ɕ x h/ <f s sj ch h>
/w l r/ <w l r>

/a e i o ʉ/ <a e i o u>
/ai̯ ei̯ aʉ̯ ou̯/ <ai ei au ou>

The phoneme /x/ becomes [ç] before /e i/ and /t d/ don't appear before /i/, instead merging with /ts dz/.


The syllable structure is C(R)V(C) where C is any consonant and R is an approximant or trill. /h/ doesn't appear in clusters, except across morpheme boundaries in compounds, and similarly affricates can only occur in clusters of the type /r l t f x/ + affricate + /w/. /w h/ don't appear in the coda, although /x/ is realised as [h] at the end of a word.

Stress is fixed on the penultimate syllable, and follows a trochaic rhythm.

So far this will score average for phonology section (chapters 1-19) except chapter 14, as it's impossible to be in the largest portions of both chapters 14 and 15, and as languages with fixed stress outweigh languages without I sided with chapter 15.

Re: The Most Average Conlang Ever

Posted: 23 Nov 2014 15:24
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Big Update'.


The morphology is heavily agglutinative, so contains no ablaut or isolating morphemes or fused morphemes. It is also predominantly suffixing, with a maximum of 4-5 inflectional categories per word. It is both head and dependent marking, with nominal case, but subject and object being distinguished only by object marking on the verb. There's no case syncretism in nominals nor subject+number marking syncretism in verbs. There is no gender, but there are definite and indefinite articles (although the indefinite article is just the number one). The language features both full and partial productive reduplication. Nouns inflect for 6 cases (yet to be decided exactly) and 2 numbers, while verbs inflect for subject, object, tense (past/future/present) and probably something else.

I'm going to try to make it quite grammatically dense, as I saw on WALS the Mesa Grande Diegueño example ʔ-a-x 1.SUBJ-go-FUT 'I will go' and the morpheme-phoneme ratio of 1:1 in that word appealed to me. So here are a few words:
tjwa - n. man
cho - n. woman
four - n. donkey
sjot- - v. love
a- - v. see


So as I said before, nouns are the same for subject and object, and so I will leave this 'direct' case as just the bare stem, so tjwa and cho are complete words on their own. There will be a dedicated genitive case, and it seems natural to have a dedicated dative case for the indirect object. So from there we need another three cases. An idea would be an instrumentive case (which couldn't be used comitatively). There is no pressing need for certain cases as the language has adpositions which can encode information.

These are all expressed as suffixes, coming after the plural suffix. The plural marker is obligatory for all nouns, except after a quantifier that expressly shows number (this could be subject to change).

Nouns can also take a possessive suffix, available for all three persons and both numbers, although it can only be used without a express possessor, so 'his wife' would be glossed as woman-3SG, while 'the man's wife' would have woman without the possessive suffix and man in the genitive. (I can't build sentences at all until I hit the syntax section; I don't even know whether it's SVO or SOV at this stage.)

I don't know about agreement yet, so I don't know how much of this is true for adjectives (or whether the language has adjectives - although it probably does). I will develop nouns more in a future post.


So I know that verbs obligatorily mark subject and object if there is one, as well as marking for either past, present or future. The tense markers will probably come before the person markers, although I'm open to changing that. Verbs can also show epistemic possibility (which can't be used deontically). So far the only markers I know are:

-∅- - 3rd person singular subject and object marker
-(w)i- - past tense
-(a)ch(a)- - 1st person subject marker

So some valid words are: a - '(he/she/it) sees [he/she/it]', ai - '(he/she/it) saw', ach - '(I) see', achi - '(I) saw', sjotchi - '(I) loved'. More verbal morphology later.


So for the WALS morphology chapters 20-80, this will conflict with chapters 28, 38, 67 and 75.

The following post will look at basic syntax and then I'll revisit specific parts of the morphology, as this time I'll be able to build example sentences.

Re: The Most Average Conlang Ever

Posted: 24 Nov 2014 11:58
by cedh
I don't want to spoil the fun, but this has been done before. (It might be interesting though to see if (and where) your ideas are different from vec's.)

Re: The Most Average Conlang Ever

Posted: 24 Nov 2014 15:28
by kanejam
cedh wrote:I don't want to spoil the fun, but this has been done before. (It might be interesting though to see if (and where) your ideas are different from vec's.)
Goddamn. Ah well, the conlanging principle of ACADEB is just as true as ANADEW... I might keep going after looking at Vec's lang; just agreeing with WALS leaves an awful lot of questions unanswered and it should be possible to get a very different looking lang out of it. Especially if our interpretations differ, which it looks like they do.

As an aside, looking through WALS is worthwhile for conlangers who have the basics down. They tend to explain all the grammatical features really well and describe plenty of alternatives constructions to things you might not think about.
Edit: So Vec's lang led me to CALS where the exact same idea is actually shown on the front page under CALS challenges...

Re: The Most Average Conlang Ever

Posted: 24 Nov 2014 17:18
by DesEsseintes
It doesn't matter. Please continue regardless. I for one am interested in seeing what you come up with. [:)]

Re: The Most Average Conlang Ever

Posted: 24 Nov 2014 17:36
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Big Update'.

Thanks Des :)


Some additions to phonotactics: I've decided /oi/ is also a valid diphthong, and that /h/ is allowed in the single cluster /hw/. Also, clusters of the same morpheme form geminates and the only other clusters not allowed are /tk tp/ which metathesise if they occur. Also /w/ doesn't appear before /u w l r/. Edit: also, /r/ between vowels is phonetically a tap.

Syllables other than initial syllables must start with a consonant. The language makes use of epenthesis of w and a to remedy this. The past tense affix in the previous post is phonemically /i/, but after /i u/ it appears as -wi.

Similarly, the first person subject marker is phonemically /x/, but after a single consonant other than an affricate, it appears as -cha and after affricates and clusters it appears as -ach. When more than one consonantal affix appears, the language applies epenthesis from left to right to maximise clusters. For example a theoretical word /t-r-x-p-p-s/ surfaces as trachpapsa, but as trappas if the /x/ morpheme is removed.


So the 3SG subject marker can never be null (which is a crying shame) and I've discovered that a null 3SG object marker leads to some ambiguities that I'd rather avoid, so for now, disregard the forms in the morphology post. I'm not precluded from using a zero morpheme for first or second person though, and in the interest of conciseness (and also avoiding the somewhat boring repetition that agglutinative languages are prone to) it may well appear.

I'm also not completely sure about the indefinite article; 37 conflicts with 38, but the second option for 38 doesn't really agree with this map. So the indefinite article may disappear.


Basic word order is SOV, with oblique or indirect arguments preceding the object but following the subject. It has postpositions, genitives precede the noun they modify, although demonstratives, numerals, adjectives (and possibly relative clauses)* follow the noun. Degree words such as very precede the adjective.

The language is nominative-accusative, although this distinction is only shown through verbal morphology (and possibly treatment of action nominals)*. The language is pro-drop in terms of subjects and objects as these distinctions are maintained on the verb, at least in all non-subordinate clauses.

There is a separate question particle for polar questions, which occurs at the end of the clause. Negation is also expressed through a separate particle which occurs just before the verb. Negation sometimes causes a collapse of one of the grammatical categories; I'm not sure about this but at the moment, I'm thinking that either some tense or some aspectual (or possibly the subjunctive if it exists) collapses or isn't marked in certain situations. The WALS features for this are very open: 113 and 114, where it is necessarily SymAsy/Cat. If anyone has suggestions for that I would be very grateful.

The language has both a morphological causative as well as a periphrastic causative, although the syntax for this is a bit interesting. This will be covered when I get to the verbs again.


So this will conflict with chapters 84, 92, 95, 96, 97, 104, 144 but otherwise the language will agree with all other features. So the language is not in the largest proportion of languages for only 12 features of 144 (3 of which are inapplicable to spoken language) which I think is pretty good, although it's probably possible to optimise it down to 10 or 11.

I have one problem with relative clauses: they are supposed to follow the noun, but have no marking (the language employs the gap strategy) and so might end up confusing sentences. A simple remedy might be to make relative clauses precede the noun, which is much more common for SOV languages anyway. Another fix might be around optimising features 81-84 and 95-97 which might eliminate that ambiguity. I will investigate that.

So the next posts will look more like conventional conlang posts. You'll notice that the information above is pretty sparse to cover all of WALS, but the trouble is that either the features are very broad or they cover specific constructions, so I won't explicitly mention the broad features but try to keep them in mind, and I will cover the specific constructions as I get to them. I will probably move back to nouns first.

Re: The Most Average Conlang Ever

Posted: 24 Nov 2014 23:29
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Big Update'.


So the indefinite article is still kicking, mainly because I don't get more WALS points for getting rid of it and would mean I take the fourth commonest feature for chapter 34 rather than the second commonest. I'm also getting rid of the future tense and epistemic marking on verbs which does give me more WALS points, and I've also discovered I wrongly counted 95 as inconsistent, so the number of inconsistencies has reduced to 9.

I'm also allowing obliques and indirect objects to appear in between the verb and object as well as before the object. Of the inconsistencies, only 96 and 144 don't score in the second commonest category.

Also, I realised I had confused the relative clauses and I probably don't need to worry about it. It may be pretty ambiguous but should be okay.

For the phonology, I hadn't ruled out R+R clusters in the same syllable, which I'm doing now. Also I realise now that I haven't ruled out different voicing clusters, which I'll roll with for now but reserve the right to remove later on (at least for stops). The exception is for stops at the same POA (or /t d/ followed by an affricate) which do assimilate to the voicing of the second element, e.g. /ab-p/ becomes appa.

I'm also thinking there will be nasal assimilation, and I will probably update the clusters that affricates are allowed in.


Sample nouns: tjwa - man, cho - woman, four - donkey, dju - house, ban - field. Also we're going with a me definite article for now.


Nouns mark for plural completely regularly by attaching the suffix -l directly to the stem and before any case marking.


Nouns inflect for six cases: first is the direct case. This is used for both the subject and direct object of a sentence. It takes no marker, and so just the bare stem of the noun is used: me tjwa me cho sjotsan 'the man loves the woman'. It will probably also be used as a locative case before certain locative postpositions and maybe even locative verbs.

The dative is used for the indirect object, and for expressing the goal of an action. It is marked by the suffix -p: me tjap 'to the man', me fourlap 'to the donkeys'.

The genitive is used to modify nouns, often indicating possession, and as an oblique for certain verbs. It is marked by the suffix -dj: me tjwadj four 'the man's donkeys', me choldja tjwal 'the women's men'.

The ablative is used to signify the source of an action. It is marked by the suffix -tta: me djutta 'from the house', me banlatta - 'from the fields'.

The locative is used to signify being in or on something. It is expressed with the suffix -enj: me djuwenj 'in the house', me banenj 'in the field'.

The instrumental is used to signify the instrument or path of an action, and not comitatively. It takes the suffix -char: me djulchar 'with the houses', me fourchar 'with the donkey'.


The genitive forms of the personal pronouns can only be used adverbially. The language makes use of special suffixes to express a concept such as 'my house'. These suffixes are the same as the subject markers for verbs except for the first person singular:

Code: Select all

-w 'my'           -ar 'our'
-k 'your'         -pe 'your
-s 'his/her/its'  -so 'their'
For example: foursa 'his/her donkey', chou 'my wife'. These markers come after the case and number markers: me djulattso 'from their houses'.


... are welcome! If you think I'm going overboard on the reductive suffixes or if anything else in the language strikes you as odd or ugly, please let me know.

Re: The Most Average Conlang Ever

Posted: 24 Nov 2014 23:44
by kanejam
Here are some example paradigms (too lazy to format just now):

Four 'donkey'

Code: Select all

       sg        pl
dir.  four      fourla
dat.  fourpa    fourlap
gen.  fourdja   fourladj
abl.  fouratta  fourlatta
loc.  fourenj   fourlenj
ins.  fourchar  fourlachar
Note that morphemes in suffixes stick together, so /four-tt/ doesn't become fourtat.

Dju 'house'

Code: Select all

       sg       pl
dir.  dju      djul
dat.  djup     djulpa
gen.  djudj    djuldja
abl.  djutta   djulatta
loc.  djuwenj  djulenj
ins.  djuchar  djulchar
The ablative plural of a monosyllables ending in an affricate will currently be four syllables, which I don't really like but I can solve this at a later date. Otherwise no form is longer than three syllables.

Re: The Most Average Conlang Ever

Posted: 25 Nov 2014 11:09
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Big Update'.


Personal Pronouns

Their are three persons and two numbers, pretty standard: u 'I, me', ner 'we, us', kas 'you (SG)', krai 'you (PL)', ol 'he/she/him/her/it', sok 'they, them'. They are usually dropped in subject position unless being emphasised, but are usually obligatory in object position. This prevents ambiguities: me tjwa sjotsan 'he/she/it loves the man' vs me tswa ol sjotsan 'the man loves him/her/it' vs sjots 'he/she/it loves (in general)'. This language doesn't have a problem leaving transitive verbs as objectless, as long as it's marked on the verb.

Here is a cheeky image showing the whole paradigm. Not very exciting as so far they take the same case markings as nouns. Note they can't take the plural nouns or the possessive suffixes of nouns:


Reflexive and Reciprocal

Time for a new noun: ab - body. The reflexive pronouns are simply the word for body plus the appropriate possessive suffix, e.g. abar 'ourselves', abka 'yourself', as in abwa aun - I see myself. The reflexives are used for object pronouns if the subject and the object are the same. Note that when used as a pronoun, ab doesn't take plural marking.

The reflexive pronouns can be used as intensive pronouns, similar to the English sentence 'the president himself signed these papers': me tjwal abso four aisoron 'the men themselves saw the donkey'.

The reflexive pronouns can't be used as reciprocal pronouns, which instead takes the form helma and is invariant: helma awarartsa 'we see each other'. Edit: The reflexive pronouns, as in English and unlike French, can't be used as a middle voice marker.

Re: The Most Average Conlang Ever

Posted: 26 Nov 2014 13:33
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Big Update'.


Types of Verbs

So there are transitive verbs and intransitive verbs. The language isn't too strict about transitive verbs being used intransitively. Worth mentioning are that adjectives sometimes behave like stative verbs; we'll see more about them later.


There are only two tenses, the present which is unmarked (and would be better called a non-past) and the past which is marked by the suffix -i: ai 'I saw' from a 'I see'. It appears before the person markers.

Person Marking

So as you should know by now, the verbs agree in person and number for both the subject and object. Here are the endings:

Code: Select all

  Subject      Object
   SG  PL       SG   PL
1 -0  -ar      -wn  -rts
2 -k  -pe      -kn  -pen
3 -s  -so      -n   -ron
The subject marker comes before the object marker, and they both come after the tense marker: sjoktan - 'I love you (SG)', aisortsa 'they saw us'.


So there is a periphrastic causative in the language, especially for the more complicated causatives, but there is also a causative marker for verbs. This takes the form of a suffix -dzwi and comes before the tense marker.
It can only be applied to a verb used intransitively, as the causer becomes the subject and the causee or agent of the verb becomes the object, so an original patient has nowhere to go. It can be implied but not expressed overtly, although it is possible with the periphrastic construction which will be covered later.

- This is what I had written originally but then I thought it might still be possible. Any ideas on how to do it? I was thinking maybe marking the original patient as genitive?
Examples: as 'he saw', adzwin 'I made him see' - notice how, sort of like in English, the person doing the seeing is marked as the object of the causative.

Anything else

This is some pretty bare verbal morphology, mainly because there can't be an imperfective/perfective distinction, a perfect or a passive (or an antipassive or applicative). If anyone wants to suggest anything else I'm all ears; I'm allowed one more inflectional category for the verbs but don't need it, and I can also fuse some aspect with tense so that little past tense morpheme doesn't feel so lonely. On the walk home just now I entertained the idea of a telicity contrast.

Edit: I probably also need an infinitive or subjunctive form. Will have to think about this.

Re: The Most Average Conlang Ever

Posted: 29 Nov 2014 12:17
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Big Update'.


So here are a couple of adjectives: rweg 'big', mar 'small', uts 'kind'.

To modify an adjective, simply place it after the noun: me tjwo mar 'the small man'. To stack adjectives, the verbal conjunctive dza 'also' is required: me fourwa rweg uts dza 'my big, kind donkey'. The adjectives don't agree with their head nouns at all.

Adjectives can be used substantively without any overt derivation, and taking the markers of nouns: me marlatta 'from the small ones'.

Adjectives can also happily behave as verbs, simply taking verbal marking to make an adjectival predicate: me chol rwegso 'the women are big', me dju marsa 'the house is small'.

Adjectives can often take an ablative or locative suffix and be used adverbially, distinguished from an oblique by coming after the verb. No examples of that just yet.


So we've discussed adjectival predicates above, but we'll discuss to others, after some a new word: kritjan 'Christian (noun)' (let's assume this is an isolate colonised by the English, just for simplicity's sake).

Firstly, nominal predicates. To either equate two nouns or to specify that one noun is a subgroup of another, the word we is used. It marks for the subject but not the object, as it is essentially transitive: me cho kritjan wes, me tjwal kritjanla weso. Note that the second noun can't take articles in this position. In order to specify a definite meaning, demonstratives must be used.

Next, locational predicates. These have a different word than nominal predicates, gro. The location can be a noun in any case except the instrumental and the direct. The ablative simply expresses direction from, the locative expresses location in or on depending on the noun, and the dative expresses direction to.

When the genitive case is used, it means 'at, by', which is separate to locative case for many nouns: me djudj gro 'I'm at/by the house' vs. me djuwenj gro 'I'm in the house'. This distinction is lost for some nouns, depending on what the default meaning of the locative is: me bandja/banenj gro 'I'm in the field'.

Re: The Most Average Conlang Ever

Posted: 12 Dec 2014 11:41
by ol bofosh
I've been wondering how to merge English, French and Spanish. I sort of had an idea about the phonology, but I didn't know how I should decided things about the grammar. This thread has given me the idea to compare the three trough CALS and use the most common elements. So now I have a place to start with an average Franco-Anglo-Hispano-language.

Edit: Franglispan Just a start.

Re: The Most Average Conlang Ever

Posted: 19 Jan 2015 11:38
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Makeover Edition'.

Big Update

So I'm not super happy with this lang and the direction it's going. One major problem is that it's very boring and, well, average. This is definitely to be expected, but I think I should be able to create a properly interesting language while staying within the bounds of a WALS avrelang (term stolen from vec). So I'm going to give it a total makeover in the next day or two while I'm waiting for the relay torch from Linguifex. I'm hoping the short deadline will improve my otherwise terrible work ethic.

So, let's go from the top:


/m n/ <m n>
/p b t d tɕ dʑ tʂ dʐ k g ʔ/ <p b t d tx dx ch zh k g '>
/f s ɕ ʃ x/ <f s x sh h>
/l ʋ j/ <l v j>

/i ɨ ʉ/ <i y u>
/æ o/ <a o>

Slightly Mandarin inspired here, and now I'm feeling a lot better about the aesthetic. The vowels are there as a bit of a fuck you to the usual five vowel systems.

Phonotactics and Allophones

The syllable structure is C(R)V(Q), where C is any consonant, R is an approximant (/l ʋ j/), and Q is either /ʋ j/ or a nasal (homorganic with the following consonant) or the first part of a following geminate stop. The glottal stop is not written before a vowel word-initially where it doesn't contrast as null onsets are disallowed.

The only exception to the above syllable structure is that word-initially, syllables of the type NPV(Q) are allowed, where N is a homorganic nasal and P a plosive or affricate - this feature is totally stolen from Albanian.

The sequence /ji/ as well as any approximant+approximant sequence is disallowed. /t d k g x/ become /ts dz c ɟ ç/ before /j i/. Of the approximants, only /ʋ/ can begin a word or appear after an affricate. /l j/ can't appear after any of the sibilants and /l/ also can't appear after a nasal in the same syllable.

/ʋ j/ in the coda form diphthongs with the preceding vowels, altering some of the vowel qualities: /iʋ ɨʋ ʉʋ oʋ æʋ/ [ɪu̯ ɪ̈u̯ ʊu̯ oʊ̯ ɑʊ̯] and /ij ɨj ʉj oj æj/ [ɪi̯ ɪ̈i̯ ʊ̈i̯ ɵɪ̯ ɛɪ̯]. The coda /ʋ j/ are written <u i>, with /æj/ additionally being written <ei>.

Sample words:

Ndan, kjei'va, yttxum, hlambjuu (none of particular meaning). If I can think of more allophony (or if anyone wants to suggest any?) I will add it in.

Re: The Most Average Conlang Ever, NP: Revamp

Posted: 20 Jan 2015 01:05
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Makeover Edition'.


The language is still suffixing and strictly agglutinating, with no syncretism, no gender, double person marking on verbs and case marking on nouns. It has both definite and indefinite articles, although it is unsure whether they will inflect.

I'm slightly mixing up the morphonology, instead of adding an epenthetic a, the language will delete vowels of suffixes of the form CV if it can get away with it. It also has a tiny amount of vowel harmony in affixes, where i/y alternate in a suffix depending on the frontness of a preceding vowel, or coda approximant. This can be seen in the possessive below.


So nouns decline for six cases. The reference form is the direct case, which is just the bare stem and is used for subjects and objects of verbs, as well as in a vocative capacity. Examples are txva 'a man' and hon 'a woman'.

There is also an indirect case used for recipients or benefactors, especially in ditransitive clauses. It takes the form of -ni/-ny/-n such as txvan 'for a man' and honny 'for a woman'. It has a durative sense when used with time expressions.

The language features two genitive cases, one representing direct ownership and control or an especially inalienable relationship, and the other representing association. The first will be called the possessive for the purposes of the grammar and takes the form -ti/-ty. The second will be called the genitive proper, and takes the form -ba. So we have txvati, txvaba 'of a man' and honty, homba 'of a woman'.

The difference between the two can be shown with a phrase 'the person's city'. If the person were in the possessive case, it would imply that he owned or ran the city as mayor or baron or similar, rather than just living there. o honty o txva 'the woman's husband', o homba gy txva 'a male friend of the woman'.

Note that the language keeps the articles in genitive constructions (where the genitive precedes the noun), and so can distinguish between 'the man's friend' and 'a friend of the man's' without resorting to circumlocutions. The genitive case can be used with time expressions to mean 'from' or 'ago'.

The adessive signifies at or on something, either spatially or temporally, and is formed with the suffix -appi such as o xvilo homappi 'the clothes on the woman'. It also has a partitive sense when used with some quantifiers and numerals, e.g. o gy ou txvalappi 'one of the men'. It is also used in comparative statements.

The essive signifies as or like something. The ending is -zhi/zhy.

Plurals: Nouns have both a singular and a plural form, the plural forms being marked with the suffix -li/-ly gy txvali 'men' and gy honly 'woman'. It comes after any case marking, often causing vowel deletion (although the form of the vowel is still conditioned by the deleted vowel): gy hombli 'of (some) women', gy txvanli 'for (some) men'.

Re: The Most Average Conlang Ever, NP: Revamp

Posted: 20 Jan 2015 11:39
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Makeover Edition'.


The independent personal pronouns are as follows:
1SG ala
1PL hanym
2SG ntuva
2PL ntanyva
3SG hou
3PL nta

They decline for case in the same way as the nouns, although the two genitives are very rarely used as there are also dedicated possessive suffixes:

Code: Select all

      POSS    GEN
1SG  -'inna  -'anna
1PL  -hin    -han
2SG  -vinno  -vanno
2PL  -nvin   -nvan
3SG  -hiva   -hava
3PL  -tin    -tan
There is are reflexive pronouns which is formed out of putting the genitive possessive suffixes (not the possessive possessive suffixes... I might need some new terminology) on the noun ba 'body', e.g. ba'anna 'myself'.

The reflexives are used for object pronouns if the subject and the object are the same. Note that when used as a pronoun, ba doesn't take plural marking. The reflexive pronouns can be used as intensive pronouns, similar to the English sentence 'the president himself signed these papers'.

The reflexive pronouns, as in English and unlike French, can't be used as a middle voice marker or as reciprocal pronouns, which instead takes the form heilva and is invariantly 'each other'.

Other pronouns will appear as I need them, I think.

Re: The Most Average Conlang Ever, NP: Revamp

Posted: 22 Jan 2015 07:09
by kanejam
So my cool idea to have the indefinite pronoun as null in the singular is a failure as I need it to be the same as the word for one, so gy is now the indefinite article, which doesn't change for number, as well as the word for one. The definite article is o in the singular and ou in the plural, using the pronominal plural marker -va which also appears on the demonstratives. (I think that's allowed).

So far verbs are looking like they'll be very similar to the last iteration. There will be an added infinitive or subjunctive, although I'm not sure which, and there will be a particle where the passivity is based on the transitivity of the verb itself (which I think is the same as Albanian).

Re: The Most Average Conlang Ever, NP: Revamp

Posted: 25 Jan 2015 07:08
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Makeover Edition'.


These haven't changed much. They still follow the form ROOT-TENSE-MOOD-SUBJ-OBJ, where tense is either past (-ju in affirmative statements, -ja in negative statements) or non-past (null), and is straightforward to use.

Mood is either indicative (null), causative (-oshy), imperative (-ki/-ky, also null subject marker for 2SG) or subjunctive (-ga). Indicative is used in normal declarative statements. The causative is used with intransitive verbs (or objectless transitive verbs) only, otherwise the periphrastic causative construction must be used. The imperative is used for giving commands but is only used for second person. The subjunctive is used variously, but alone it expresses a wish. It is used in complement clauses of many verbs, including the periphrastic causative. In a complement cause, if the subject is the same as the main clause then it isn't marked.

The subject and object markers are as follows:

Code: Select all

  Subject      Object
   SG  PL       SG   PL
1 -la  -zhei      -'i  -'va
2 -du  -do      -va  -nva
3 -hu  -hvi      -'a  -tva
The i's in the above paradigm as well as those of the possessive suffixes agree in vowel harmony.

The participle affix is txu-, one of the only prefixes in the language. This is a derivation and so removes any other inflection on the verb. The participle acts as an adjective that can't be used as a stative verb, and takes an active sense for transitive verbs and a passive sense for intransitive verbs.

Examples to come.

The negative marker is zhi, although there is a separate prohibitive particle for imperatives azhai. They both precede the verb.

Re: The Most Average Conlang Ever, NP: Revamp

Posted: 04 Feb 2015 22:21
by kanejam
Edit: This information is out of date, please scroll down to the post titled 'Makeover Edition'.

I'm looking at revamping this again slightly, although I'm getting happier and happier with how it's going. The main things I'm considering are a Croatian based orthography, sibilant harmony and more concise affixes so that the language doesn't feel as clunky.

The Morphological Causative

So we have two causatives. The first is the morphological causative. This fits into the mood slot and as (for the moment at least) it can't be used as an imperative or in (most?) subordinate clauses. For these, the periphrastic causative is available.

It surfaces as the suffix -oshy. The subject of the verbs marks the causer and the object the causee. It can be used with intransitive verbs, verbal adjectives or objectless transitive verbs:

- Dxunkajudu 'you sat' -> dxunkaju'oshylado 'I seated you'.
- 'Ilanjuzhei 'we are dark' -> 'ilanju'oshydu'va 'you made us dark'.
- O bjozha tyvjuhvy'a 'they saw the girl' -> tyvju'oshydutva 'you made them see'
But not: **o bjozha tyvju'oshydutva **'you made them see the girl'. This construction can be handled by the periphrastic causative.

There are some verbs commonly translated into a causative form:

- Haxaju'oshydu'a 'you killed it' from haxa 'to die'. Note the word haxa is not a passive form; the word itself has a patientive subject. There are no passive forms in this lang.

The Periphrastic Causative, and the Subjunctive

This makes use of the verb do'la, which in normal statements means 'make, do'. However, it is also used in the periphrastic causative, which is used where the morphological causative can't be.

The subordinate clause isn't marked by a particle, but the verb takes the subjunctive mood, -ga. There isn't a whole lot to say about the subjunctive, it can be used alone to express a wish: haxagahu 'may he/she die'.

The periphrastic causative follows the form 'I do [that] you do this', as in the following:

Do'lajudu o bjozha tyvjugahvy'a
do-PAST-2SG.S DEF girl see-PAST-SUBJ-3PL.S-3SG.O
'you made them see the girl', which is the causative form of the sentence above.

The Verb Want

The verb ncha is translated as 'want', but is only used when the want is a subordinate clause. In this case, the subordinate clause also takes the subjunctive, but it is special in that it can't take a subject prefix. If the subject is the same as as the subject of ncha, then it is unmarked; otherwise a full pronoun must be used.

O hon nchahu ou txvali gy xvilo klunga'a
DEF woman want-3SG.S DEF-PL man-PL INDEF clothing decorate-SUBJ-3SG.O
'the woman wants the men to decorate the clothes'

Pretty simple, but slightly interesting in how it works. I'm quite happy with myself because these are relatively complex constructions, which I got with the help of WALS. Also, yeah, I'm lazy with the glosses, sorry about that.
Here are the examples with the new orthography I'm considering:

doqlajudu o bjoďa tyvjugahvyqa
o hon nčahu ov ćvali gy śvilo klungaqa

I'm still undecided about q/' for the glottal stop and ď/ž/dž/ǧ for zh. I'm also thinking of changing y, but I'm not sure what to. Thoughts?

Re: The Most Average Conlang Ever, NP: Revamp

Posted: 17 May 2015 14:25
by J_from_Holland
/i ɨ ʉ/ <i y u>
/æ o/ <a o>

No <e>? No /a/ sound? Man, your language isn't average!

Re: The Most Average Conlang Ever, NP: Revamp

Posted: 17 May 2015 19:21
by idov
His language is very average if you ask WALS. The language would be just as average if it was /a i u e o/ according to their chapters, but kanejam wanted to be as interesting as possible within the parameters. That's partly why the great revamp happened in the first place.