Bengedian

The Bengedian language (Bengedian: thu nelanem "the speech") is a constructed language I've been working on for some time now and I've decided to be a nice person and share it with you guys.

It's my first serious, linguistically informed attempt at conlanging, and as such I didn't try anything radical with it. A lot of the features in Bengedian will seem very Indo-European-ish compared to some other conlangs. I tried hard to make it complete and usable without turning it into a kitchen sink.

Hope you like it :D

At a glance
Bengedian is a moderately inflecting, SVO-order language with nominative-accusative alignment.

Nouns inflect for case (nominative, accusative, dative, genitive) and number (singular, plural). There is no grammatical gender, but the endings have partially fused to the noun, creating "declensions" of sorts. Adjectives agree in case and number with their head nouns and inflect like nouns.

Verbs inflect for tense (past, present, future) with four personal endings in each tense (1st singular, 2nd singular, 3rd singular, plural). There are also two non-finite forms: an infinitive/gerund/present participle, and a past participle. Other verb forms (e.g. "person who does" forms, what some grammars call "active participles") are part of the derivational system, not inflection.

Consonants
Bengedian possesses the following consonant phonemes:

Vowels
Bengedian's vowels are:

Allophony
The following allophonic realizations (in no particular order :P) are seen in standard Bengedian:


 * Intervocalic voicing of fricatives: sisec "(that which is) stopped" /ˈsi.sek/ → [ˈsiˑ.zɛk]
 * Fricatives are also voiced when they precede voiced consonants: thu cos da "the number two" /θu ˌkos ˈda/ → [θu ˈkoˑz ˈdaˑ]
 * A glottal stop /ʔ/ is inserted between vowels on word boundaries: sa eda "I was" /ˌsa ˈe.da/ → [saˑ ˈʔeˑdə]
 * The retroflex approximant /ɻ/ becomes a retroflex flap [ɽ] between vowels: sre ira "I exist again" /sɻe ˈi.ɻa/ → [sɻe ˈʔiˑ.ɽə]
 * Unstressed vowels tend to relax. That is, /a e i o u/ > [ə ɛ ɪ ɔ ʊ]. Except for the rule /a/ > [ə], this does not apply to final syllables.
 * Vowels are lengthened in open stressed syllables, and lengthened (slightly less) before voiced consonants.

Stress and intonation
Bengedian, like English, is a stress-timed language; that is, stressed syllables occur at a roughly steady rate.

Bengedian has a movable stress accent. Normally, in words of two or more syllables, the stress is on the second-to-last syllable; when this is not the case, the stress is indicated orthographically, which I transliterate as an ácúté áccént.

Derivational and inflectional endings can change a word's stress placement. Again, when this stress is not on the second-to-last syllable, an accent mark indicates the irregular stress.

Bengedian has two layers of stress: lexical stress (primary) and prosodic stress (secondary). Only primary stress is marked on a word in writing.

Phonotactics
Bengedian's general syllable structure is (C)(C)V(r,l,w,j)(C)(C).

These conventions are used here:
 * (a, b, c) means either a, b, or c.
 * [ Brackets ] indicate an optional thingy.
 * + means concatenation.

Cluster-level rules
The following sequences of consonants are prohibited anywhere:
 * Two sibilants
 * Alveolar stop + /l/
 * Geminated (doubled) consonants

Syllable-level rules
The following patterns are permitted in the syllable onset:
 * Any single consonant
 * Zero
 * Sibilant + voiceless stop
 * Sibilant + voiceless fricative
 * Obstruent + approximant
 * Sibilant + nasal

The following patterns are allowed in the syllable nucleus:
 * Any vowel
 * (a,e,o) + semivowel
 * Vowel + liquid

The following clusters are allowed in the syllable coda:
 * Any single consonant except /h/
 * Zero
 * Voiceless stop + sibilant or /θ/
 * Voiceless fricative + voiceless stop
 * Any voiceless obstruent except /h/ or /t/ + /t/
 * /b/ or /g/ + /d/

Thus berm, fašt, blečt are valid, but bdačl and scra are not.

Word-level rules
Medial consonant clusters are limited to the following patterns:
 * Liquid + any two consonants
 * Liquid + nasal + [sibilant] + obstruent + approximant
 * Liquid + nasal + sibilant + nasal

Orthography
Bengedian is written using its own alphabet, which though I could put it up, I couldn't show you texts with it cuz there's no font for it (and all the good font-making programs 'COST MONEY...' blech. Or they're for Unix, and Cygwin didn't work for me.) Anyways...

Or you could just use our trusty Roman alphabet like I'm doing here. The following letters are used:

A B C Č D E F G H I K L M N O P R S Š T U W Y a b c č d e f g h i k l m n o p r s š t u w y

While this transliteration is for the most part one-to-one, there are some complications:
 * The Bengedian alphabet does not distinguish c and k. I do so here to make it easier for Anglophones--that would be you if you're reading this :D-- to make it clear that the letter represents /k/ not /s/ in all cases. I must stress, however that I'm inconsistent with this, and to remember that c is always hard.
 * I use the digraph th to represent /θ/ where the Bengedian alphabet uses a single letter.
 * I use the letters č and š to represent the phonemes /t͡ʃ/ and /ʃ/ respectively.

Declension of nouns and adjectives
Bengedian nouns, pronouns, and adjectives decline for case and number. There is no grammatical gender. Adjectives agree in case and number with their head nouns and take the same endings.

There are six "declensions" in Bengedian. Which class a noun or adjective falls into is determined by form alone, depending on the word's final sound in the nominative singular: a consonant, /a/, /e/, /i/, /o/, or /u/. The noun and adjective declension is given below, one for each class:

Notice how:
 * The nominative and accusative are mostly identical, but nouns and adjectives ending in a consonant take a final -e in the accusative singular.
 * The dative singular -e on words in -i.

Irregularities
Certain adjectives do not take the plural endings in -n. These are usually quantifying adjectives that are inherently plural; these include:
 * The numbers themselves: guim bermen "five persons"
 * Quantifying adjectives: unco bermen "many people", tom bermen "some people"

A note on spacial relations
This is under "Nominal morphology" for a good reason, people.

Bengedian does not have prepositions in the sense that English does. Bengedian prepositions usually convey information about the noun's relationship to other parts of the sentence, not for spacial relations like in English. There are a few prepositions (di "from", im "in", am "on"), but there are only a few.

Bengedian words for spacial relationships are nouns (e.g. ut "top", uš "bottom", indo "inside"). These nouns take the dative case when used for relationships; their objects take the genitive:


 * Sas dume e amuti thus seutos.
 * My house is on top of that hill.

They can also be used for directions by putting the head of the phrase in the dative:


 * Guet cu uti thui seutoi o wicišem sas dumei.
 * Go up to that hill to reach my house.

Prepositions and dire

Verbs
Verbs are conjugated for three tenses (past, present, future) with four distinct endings in each tense: one for each person in the singular, and one in the plural. The plural verb takes the same ending for all persons; person is not distinguished on the verb in the plural. :-)

In addition, a verb consists of two non-finite forms (infinitive, past participle). The infinitive ending is usually -em, but verbs with stems in -m take the ending -an instead. e.g. pernuman "forget". The rest of the conjugation is not affected.

The different classes of verb are given below: Notice:
 * The past tense seems to have a certain element that reminds one of, oh say, Germanic weak verbs for instance. I'll confess to borrowing here :-D
 * The participle ending is normally -ec. For verbs with a stem in -c already (the infinitive -em doesn't count as part of the stem), use -eth instead--they mean the same thing.
 * For the past tense of verbs whose stems end in a consonant other than t or d, the ed of the past tense becomes a t for voiceless consonants, d for voiced.

Personal pronouns
The indefinite pronouns in the 2nd and 3rd persons are used when the subject's identity is unknown. For example,
 * Wa e fašem thu, siset cu thici!
 * Whoever is doing that, stop now!


 * Tom, siset de thici!
 * Tom, stop now!


 * To pilired sas donsen! Toneth, astinet cu ke wica the!
 * Somebody took my stuff! Someone else, help me catch him!

Demonstrative pronouns
The demonstrative thu is also used as a definite article.

Relative and interrogative pronouns
These pronouns correspond roughly to English that/which and who?/what? respectively. There are several contracted forms such as:
 * kie, from ke e, "which is", and its other forms kea, ket, kemon, keda, kedát, ked, kedón, kum, kunt, kume, kumon
 * carke, wac, meaning "when"
 * šoke, wašo, meaning "where"
 * šuke, wašu, meaning "how"

The relative pronoun ke is also used as a general subordinator, used to mark sentential arguments:


 * Loseda ke de e nec.
 * I said that you're bad.

Syntax
Phew, we're all done with the tables, finally. Seriously, do you know how much effort it takes to type one of those things?

Eh, but I digress.

Here is where I cover Bengedian syntax: the types of sentences allowed. I will also cover the semantics of them, that is, what those constructions actually mean.

Basic word order and headedness
Bengedian declarative sentences use SVO word order, just like English.
 * Sa sordeda thu dule.
 * I closed the door.

As for headedness, Bengedian is largely head-final, aside from having SVO word order.

Questions
Simple yes/no questions are just indicated with a rising intonation in speech, or by a question mark in writing. There is no syntactic difference.


 * Fašedat des fašaren?
 * Did you do your work?

"WH" questions (to use the English term) are formed by fronting the phrase headed by the queried item. Unlike English, auxiliaries are not inverted:


 * Was lengeš de cosedat?
 * Whose length did you measure?


 * Wa et?
 * Who are you?


 * O wa de et fašem thi?
 * Why are you duing this?

Tag questions are formed using the set phrase e čel?, meaning "is that right?"


 * Et Tim, e čel?
 * You're Tim, right?

A negated tag question uses the tag nire? (lit. "it doesn't exist"):


 * Net Tim, nire?
 * You aren't Tim, are you?

Imperatives
Imperatives are formed by placing the subject pronoun after the conjugated verb.


 * Siset de!
 * Stop!


 * Gúon san!
 * Let's go!


 * Ringua sa!
 * Let me leave!

The use of pronouns
Demonstratives and quantifiers always precede their heads; relative clauses come after.


 * Sa sordeda thu dule kie yarec yo albeges.
 * I closed the door which is made of wood.

And yes, it is indeed possible to say "I closed the wooden door", but that's in the next subheading.

Bengedian is pro-drop; the subject pronoun can be omitted and the sentence remains grammatical:


 * Sa sordeda thu dule.
 * Sordeda thu dule.
 * I closed the door.

The copula
The primary copula is em, meaning "to be". The conjugation of this verb is irregular and is given above.

Em is also used to express existence:


 * E ušwad.
 * It is raining.


 * Ide bermo andoi thus dumes.
 * There was a man in that house.

Case
Bengedian's four cases are used as follows:

The preposition yo "of" may be used in place of the genitive. In this case, the preposition's object is in the accusative:


 * Thu Englišes los yo "wadu" e "water".
 * The English word for "water" is "water".


 * Loson san yo thin edrin!
 * Let's talk about this food!


 * Dalet de sai da glarcan yo wadu.
 * Give me two bottles of water. (i.e. an amount of water equal to two bottles)

Predicatives (objects of the copula) take the nominative case:


 * Thu fašec nie rešlač.
 * That action is unlawful.

Noun phrase order
A noun phrase in the genitive case can modify another noun phrase, similar to English possessive 's.


 * Thus bermos con e ašo und.
 * That man's dog is very big.

The Bengedian genitive ending is a true inflectional affix like in German and Old English, not a clitic like in Modern English. Because of this, certain constructions involving the genitive are possible which are awkward or even ungrammatical in English (the question mark means "awkward" here):


 * E predo dus sans.
 * ? He's all of our brother. (Though this is syntactically valid (maybe), it is extremely awkward and most prescriptivists would give it a thumbs-down.)


 * Thum edrin thus bermos.
 * ? The food of the man. (A native English speaker would be more likely to say "the man's food".)

Demonstratives and quantifiers always come before their nouns. Quantifiers are not inflected:


 * Thu con "The dog"
 * Da conen "Two dogs"
 * Dus thum conen "All of those dogs"

Adjectives
Adjectives phrases (APs) can either precede or follow their head nouns.
 * Sa sordeda thu laica dule.
 * Sa sordeda thu dule laica.
 * I closed the little door.

This placement affects the semantics of adjective coordination. Compare English "the big and colorful ball" versus "the big ball and the colorful ball" (though it could be argued that the latter is actually a coordination of NPs). Bengedian makes the same distinction, albeit using different construction:


 * Thum alben unden at álbecoren
 * The big and green trees (the trees that are both big and green)


 * Thum unden at álbecoren alben
 * The big and green trees (the big trees and the green trees)

Verb phrases
The Bengedian verb phrase shares many similarities with the English verb phrase.

Direction words that modify verb objects always come before the verb, e.g. am loem "to look at".

Auxiliaries
Bengedian makes use of auxiliary verbs to refer to complex nuances and relationships. An example:


 * Thu šritec sera fašem ec am loec carke sre elada.
 * That paper has to have been looked at when I come back.


 * Ke thu bermo sere fašem ec rin sindem caned.
 * That man could have to have been running away.

Auxiliaries appear in the following order:
 * Mood auxiliary + infinitive
 * Perfect fašem + participle
 * Progressive em + infinitive
 * Passive em + participle

Mood auxiliaries include:
 * canem "to be able", equivalent to our "can, could"
 * norem "to have to", equivalent to our "must, should"
 * lufem "to want, desire", equivalent to our "will, would"

Negation
Verbs are negated using the particle ne. The copula em may optionally contract with ne:


 * Na sisec.
 * Ne ma sisec.
 * I'm not done.


 * Nide thu bermo.
 * Ne ide thu bermo.
 * It wasn't that man.


 * Thi con nie sas.
 * Thi con ne e sas.
 * This dog isn't mine.

A note on the passive participle
The passive participle in -ec, though I call it a "past participle" by analogy with English, actually carries little, if any, implication of past tense at all. Thus tusec dons could mean either "thing that is used (by something)" or "thing that has been used (in the past)". Its primary meaning is that of the passive voice:


 * Im tu funec thus blatičes kindes, thu funtu cacted.
 * In being played with by the aggressive child, the toy broke.

Generally, the intended meaning can be deduced from context.

The passive
The Bengedian passive voice is formed by:
 * 1) Putting the subject in the genitive case
 * 2) Inverting subject and object
 * 3) Using em+participle in the verb phrase.

Example: Thu lumi  ide        edec     thus        bermos. that apple be-3s.PRET eat-PART that-GEN.Sg man-GEN.Sg "That apple was eaten of that man" The apple was eaten by the man.

Subordination
The relative pronoun ke is used as a general subordinator. Ke is used to mark sentential clauses.

A sentential argument is syntactically a noun, so it is possible to attach a demonstrative or quantifier to a sentential argument just as with a regular noun.


 * Thu ke sa edéda thu bunec lumi e nec, carke sreloa the.
 * That I ate the rotten apple is bad, when I look back on it.

Derivational morphology
There are several ways in Bengedian to derive one word from another "base" word:

Affixation
This is the most common form of derivation; Bengedian has a large array of derivational "prefixes" and "suffixes" (which actually behave like infixes to some extent).

A derivational affix is simply appended onto the base at or near the end. The following rules apply:


 * If a "suffix" begins with a vowel, then it's inserted after the last consonant: plešo "market" > plešomo "something that's found at a market"
 * If a "suffix" begins with a consonant, then it goes after the last vowel: con "dog" > cotun "leash"
 * If a "suffix" begins and ends with consonants, then it goes after the last vowel.
 * If the word ends in a consonant, then said vowel is reduplicated after the derivational morpheme: los "word" > lonusos "etymology"
 * If a "prefix" ends with a vowel, then it goes before the first consonant: clam "round" > šuclam "around something"; uwa "sheep" > udiwa "wool"
 * If a "prefix" ends with a consonant, then it goes before the first vowel: rog "heavy" > relog "very heavy"

Some more examples:
 * šrit "write" : šritom "writer" : šritec "text" : šritor "dictionary"
 * ples "trade" : plešo "market" : plestu "money" : plésadir "economy"
 * piclo "star" : piclito "white dwarf"

Derivations on verbs are affixed to the stem, (that is, the bare verb without any endings. The infinitive ending -em, while considered part of the verb's citation form, is not considered part of the stem.) So:


 * irem "to exist" : inerem "to die" : ineryem "to kill".

The copula em "to be" has the irregular stem es:


 * esom "someone who is"

Irregularities
But, as with all systems, nothing is perfect. You'll notice that 2 out of the 3 examples for ples "trade" didn't follow the rules I laid out above. It is true that those words predate the rules I came up with, but all naturalistic languages, constructed or not, should have some quirks, no?

Snarky comments aside, there are some irregular derivations; for example:
 * By regular rules, the nominalizing "suffix" šo, meaning "place associated with X", should go before the s in ples "trade" giving plešos, but this is not the case. The "suffix" went on the end instead, giving plešo "market". The two sounds /s/ and /ʃ/ have assimilated into /ʃ/.

Zero derivation
A lot of words can be converted between two parts of speech by zero derivation:


 * donad "group, category" (noun) : donad "to classify" (verb)
 * dols "to sit" (verb) : dols "chair" (noun)
 * wadu "water" (noun) : wadu "to douse" (verb) : wadu "wet" (adjective)
 * ut "up" (noun) : ut "to rise" (verb) : ut "high, tall" (adjective)
 * rog "load" (noun) : rog "heavy" (adjective)

Vowel change
This is a very rare method (and not even a method per se, but a phonetic by-product of other methods) and the rules governing it are very, very complicated. It shares some aspects with Indo-European ablaut; for example, certain syllables tend to reduce to a sort of "zero grade" if the resulting cluster is phonotactically valid:


 * soc "rock" : scund "boulder"

Compounds
Bengedian, much like English and all Germanic languages for that matter (I personally get kicks out of those super-long German compounds :-D), allows compounding. Compounding is the process by which multiple free morphemes (or compounds) join together into a single unit. A compound is grammatically a single word. Examples of compounds in English include "blackboard", "jumprope", "textbook", "Spongebob".

Bengedian compounds are usually head-final, but can be head initial sometimes, and can be in any of the following patterns:
 * direction+verb: rinserem "to take away, steal, confiscate" from rin "away from", serem "to have"
 * direction+noun: ampucem "to breathe" from am "on, in(to), at", puc "air"
 * verb+verb: nuspirem "to learn" from nus "think", pir "to take"
 * noun+noun: álbecor "green" from alb "tree", cora "color"
 * verb+noun: wicwaduec "beverage" from wic "take", wadu "water", -ec verbal noun/past participle
 * noun+verb: loydal "sun" from loy "light", dal "to give"
 * adjective+noun: yalome "gold" from yal "yellow", ome "metal"

Long words are often phonologically reduced when compounded.

A few words have variant combining forms:
 * nus "think" has the variant stem nud
 * yar "make" has the variant forms yan and y