This is my go at dewrad's 48 hour challenge---a bit early, but it's time for bed. I'm putting everything in more tags so it's easier to jump around or over. There are definitely inconsistencies and errors, but what can you do? It was fun, and hopefully the results are interesting.
Phonology
More: show
1. Inventory
I'll write ŋ for the place-assimilating nasal coda when it assimilates to a velar, but there's no contrastive velar nasal. [r] occurs intervocalically, but I take it to realise geminate /ɾ/; I will write the rhotics accordingly as r and rr.
All vowels occur long. Semi-officially I take vowel length to be suprasegmental in nature, but I'll still write the long vowels as sequences.
There are two marked tones, high and low, as well as syllables that are unmarked for tone. Contour tones are possible on long vowels, though this is very rare within roots. Only two contours are possible, the obvious rising LH and falling HL.
I'll usually represent a high tone with an acute accent on the hosting vowel, and a low tone with a grave accent; but floating tones will get superscript ᴴ or ᴸ, as appropriate.
2. Phonotactics
Vowel sequences are common, with adjacent vowels always in hiatus. Within roots, adjacent vowels always differ in quality, and morphophonology enforces this constraint more generally.
All consonants may occur as onsets.
r, l, and w can occur in coda before a plosive, an affricate, or h, and h can occur before a plosive or affricate. Moreover, both nasals can occur before h, and a homorganic nasal before any plosive or affricate.
The voiceless plosives, ts, and r can occur geminate between vowels. (Geminate ts will be written tts. And remember that r represents /ɾ/, and rr is pronounced as [r].)
Note that the syllable constraints ensure that digraphs ts and dz will never be ambiguous.
3. Stress
So far I haven't found any reason to talk about stress.
4. Processes
There are phonological processes that are sensitive to phonological words, clitic groups, and tone groups. A phonological word includes a root and any affixes; a clitic group includes in addition any clitics; and a tone group might include multiple words in a phrase (I can't tell you the details because I don't know them yet).
5. Vowel merger
Whenever two short vowels of the same quality become adjacent within a phonological word, they will merge to form a single long vowel. If the original vowels differ in tone, the resulting long vowel will have a contour tone.
6. h-epenthesis
When two vowels of the same quality become adjacent within a clitic group, they will be separated by an epenthetic h; this process can be bled by vowel merger, just discussed.
7. Nasal assimilation
A coda nasal will assimilate in place to a following consonant other than h.
8. i-epenthesis
i will be inserted to break up any illegal consonant clusters produced by morphology. This does not bleed nasal assimilation:
9. Contour dissimulation
When the two morae of a long vowel differ tonewise, the result is a contour, which must be either HL or LH. This may require dissimulation if one of the morae has no linked tone; for example, ∅H must become LH.
10. An obligatory contour principle
Within a tone group, there's a constraint that adjacent tones must differ. There's a subtlety: within a morpheme, one and the same tone can be linked to multiple adjacent morae, which is to say, to multiple vowels. Roots such as óká stand and éáw rub are thus perfectly legal: each has two tone-bearing vowels, but just one tone. The same verbs inflected for 2p agreement are a problem, however, since the agreement marker -óó has its own tone that cannot be adjacent to the tone in the stem: neither ókáóó nor éáwóó is a legal form.
The solution isn't too strange: the tone on the second morpheme must retreat to leave a gap between it and the tone on the first morpheme.
In those two examples this is simple, since the second of the tones is multiply linked and can simply delink from its first vowel: you get óká+oó and éáw+oó---and then (by contour dissimulation) ókáòó and éáwòó. But there can be further complications.
You also get conflict when the class II 3p possessor marker éé- is prefixed to áhea campfire. But the tone on áhea is singly linked, so it cannot delink from its vowel without being entirely lost. Rather than that, it will shift to the right, and the result is ééahéa their campfire.
When the tone moves to the right, this can create a further violation of the obligatory contour principle. Such a case occurs when the class I 3p possessor marker é- is prefixed to éweó son: when the tone on the stem-initial e moves to the right, it becomes adjacent to the tone on o, creating a violation. This violation can be repaired only by creating a second gap; in this case, the tone must delink from o and try to find a host further to the right. Since it is already at the end of the word, the result is a floating tone; I will write this as éèwéoᴴ (contour dissimulation has again come into play, as has vowel merger). The floating tone might find a host if another word follows within the tone group; otherwise, it will remain unlinked and unpronounced.
There is one further sort of case to consider. See what happens when éé- gets prefixed to éàbe ko fruit: the high tone on the stem's first vowel must move to the right, but there is a separate tone already linked to the vowel to its right. The result is what you might expect: that tone must also make way, resulting here in ééheábè. And as you might also expect, the shifting of the second tone can leads to its own complications, which are resolved in the ways I have just discussed.
11. Floating tones
A tone can remain phonologically active despite not being linked to a particular mora or vowel; this is what is called a floating tone. Most floating tones result from the need to repair violations of the obligatory contour principle, as just discussed. But some morphemes simply come with floating tones, and they will also try to find a host.
When a floating tone is located at the right edge of a morpheme, everything happens as discussed above: it will link to the first vowel of any morpheme that follows within the tone group, potentially forcing some tone within that morpheme to shift further to the right.
When the floating tone is at the morpheme's left edge, as in the verbal negation marker -ᴸhro, something else can happen: the floating tone will first try to lodge on the vowel of an immediately preceding syllable. However, at most a multiply-linked tone within the preceding morpheme will contract to make way for it: no tones will actually move to the left. This can result in the floating tone seeking a host to its right, in which case things proceed as already described.
| m | n | (ŋ) | |
| p | t | k | |
| b | d | g | |
| ts | |||
| dz | |||
| f | s | h | |
| ɾ <r> | |||
| (r) | |||
| l | |||
| w |
| i | u | |
| e | o | |
| a |
There are two marked tones, high and low, as well as syllables that are unmarked for tone. Contour tones are possible on long vowels, though this is very rare within roots. Only two contours are possible, the obvious rising LH and falling HL.
I'll usually represent a high tone with an acute accent on the hosting vowel, and a low tone with a grave accent; but floating tones will get superscript ᴴ or ᴸ, as appropriate.
2. Phonotactics
Vowel sequences are common, with adjacent vowels always in hiatus. Within roots, adjacent vowels always differ in quality, and morphophonology enforces this constraint more generally.
All consonants may occur as onsets.
r, l, and w can occur in coda before a plosive, an affricate, or h, and h can occur before a plosive or affricate. Moreover, both nasals can occur before h, and a homorganic nasal before any plosive or affricate.
The voiceless plosives, ts, and r can occur geminate between vowels. (Geminate ts will be written tts. And remember that r represents /ɾ/, and rr is pronounced as [r].)
Note that the syllable constraints ensure that digraphs ts and dz will never be ambiguous.
3. Stress
So far I haven't found any reason to talk about stress.
4. Processes
There are phonological processes that are sensitive to phonological words, clitic groups, and tone groups. A phonological word includes a root and any affixes; a clitic group includes in addition any clitics; and a tone group might include multiple words in a phrase (I can't tell you the details because I don't know them yet).
5. Vowel merger
Whenever two short vowels of the same quality become adjacent within a phonological word, they will merge to form a single long vowel. If the original vowels differ in tone, the resulting long vowel will have a contour tone.
6. h-epenthesis
When two vowels of the same quality become adjacent within a clitic group, they will be separated by an epenthetic h; this process can be bled by vowel merger, just discussed.
7. Nasal assimilation
A coda nasal will assimilate in place to a following consonant other than h.
8. i-epenthesis
i will be inserted to break up any illegal consonant clusters produced by morphology. This does not bleed nasal assimilation:
9. Contour dissimulation
When the two morae of a long vowel differ tonewise, the result is a contour, which must be either HL or LH. This may require dissimulation if one of the morae has no linked tone; for example, ∅H must become LH.
10. An obligatory contour principle
Within a tone group, there's a constraint that adjacent tones must differ. There's a subtlety: within a morpheme, one and the same tone can be linked to multiple adjacent morae, which is to say, to multiple vowels. Roots such as óká stand and éáw rub are thus perfectly legal: each has two tone-bearing vowels, but just one tone. The same verbs inflected for 2p agreement are a problem, however, since the agreement marker -óó has its own tone that cannot be adjacent to the tone in the stem: neither ókáóó nor éáwóó is a legal form.
The solution isn't too strange: the tone on the second morpheme must retreat to leave a gap between it and the tone on the first morpheme.
In those two examples this is simple, since the second of the tones is multiply linked and can simply delink from its first vowel: you get óká+oó and éáw+oó---and then (by contour dissimulation) ókáòó and éáwòó. But there can be further complications.
You also get conflict when the class II 3p possessor marker éé- is prefixed to áhea campfire. But the tone on áhea is singly linked, so it cannot delink from its vowel without being entirely lost. Rather than that, it will shift to the right, and the result is ééahéa their campfire.
When the tone moves to the right, this can create a further violation of the obligatory contour principle. Such a case occurs when the class I 3p possessor marker é- is prefixed to éweó son: when the tone on the stem-initial e moves to the right, it becomes adjacent to the tone on o, creating a violation. This violation can be repaired only by creating a second gap; in this case, the tone must delink from o and try to find a host further to the right. Since it is already at the end of the word, the result is a floating tone; I will write this as éèwéoᴴ (contour dissimulation has again come into play, as has vowel merger). The floating tone might find a host if another word follows within the tone group; otherwise, it will remain unlinked and unpronounced.
There is one further sort of case to consider. See what happens when éé- gets prefixed to éàbe ko fruit: the high tone on the stem's first vowel must move to the right, but there is a separate tone already linked to the vowel to its right. The result is what you might expect: that tone must also make way, resulting here in ééheábè. And as you might also expect, the shifting of the second tone can leads to its own complications, which are resolved in the ways I have just discussed.
11. Floating tones
A tone can remain phonologically active despite not being linked to a particular mora or vowel; this is what is called a floating tone. Most floating tones result from the need to repair violations of the obligatory contour principle, as just discussed. But some morphemes simply come with floating tones, and they will also try to find a host.
When a floating tone is located at the right edge of a morpheme, everything happens as discussed above: it will link to the first vowel of any morpheme that follows within the tone group, potentially forcing some tone within that morpheme to shift further to the right.
When the floating tone is at the morpheme's left edge, as in the verbal negation marker -ᴸhro, something else can happen: the floating tone will first try to lodge on the vowel of an immediately preceding syllable. However, at most a multiply-linked tone within the preceding morpheme will contract to make way for it: no tones will actually move to the left. This can result in the floating tone seeking a host to its right, in which case things proceed as already described.
Nouns
More: show
1. Number
First you need to know the rule that non-referential nouns are always treated as singular: they may not occur in their plural form, if they have one, and they trigger singular agreement. This is so even if they occur with numbers:
That said, there are a few ways in which nouns form their plurals, and no very good general rule to predict which will be correct for any given noun.
All definite noun phrases must take a definite determiner of some sort. There are five possibilities:
3. Possession
There are three ways of marking possession, depending on the nature of the relationship.
Inalienable possession is marked with the Class I possession prefixes:
The Class II possession prefixes mark what I will call beneficiary possession. This is appropriate especially when the posssessor is expected to consume or otherwise make exhaustive use of the possessed, though these prefixes can also express more a general beneficiary relationship. Here they are:
When possession is marked by prefixes in either of these two ways, the possessor can occur either before the noun or after it. If after, it will normally follow any other modifier except possibly a preposition phrase or relative clause. If before, for phonological purposes it will count as part of the same tone group as the head noun.
Finally, a possessor can be given using the oblique preposition fi: impogá fi=tì=uéɾi Wedi's axe. I'll call this associative possession.
It's not too difficult to construct a full paradigms:
There's an open class of adjectives that can directly modify nouns:
Adjectives, along with other postnominal modifiers, normally do not count as part of the same tone group as the head noun.
5. Numbers
Numbers follow adjectives, except when the definite forms occur prenominally, as discussed above.
gí (proximal) and wòr (distal) can occur postnominally, following any numbers, so long as there is some other prenominal marker of definiteness:
These follow all other modifiers, though they can be followed by focus markers. They are usually formed simply by gapping, together with subjunctive marking on the verb:
First you need to know the rule that non-referential nouns are always treated as singular: they may not occur in their plural form, if they have one, and they trigger singular agreement. This is so even if they occur with numbers:
òwdéri pogá wekóSecond, there is a class of nouns that I will call mass nouns that are always treated as plural when referential: they'll occur in their plural forms when they have one and they'll trigger plural agreement. (So plural marking on mass nouns actually indicates referentiality.)
òwdé-rí pogá wékò
need-1s axe three
"I need three axes"
That said, there are a few ways in which nouns form their plurals, and no very good general rule to predict which will be correct for any given noun.
- An affixed high tone. If the stem's final syllable has a low tone, this will normally get replaced: toà day → toá. The affixed high tone can end up floating off to the right: ápo tree → ápo+H. (No nouns whose last vowel already has a high tone form their plural this way.)
- Echoing the vowel of the final syllable, along with any tone. This is especially common with vowel-final stems: pogá axe → pogáá.
- Suffix -ko. Thus béá person → béáko. This is especially common with animates, and also with nouns whose last vowel has a high tone.
- Suffix -koa. Thus éé snake → éékoa. This is especially common with referents that usually come in great numbers but which are rarely salient one-by-one, including many mass nouns; thus tsókò grass → tsókòkoa. Some nouns take one plural when used as genuinely plural count nouns, and -koa when used as a (referential) mass noun. Thus, áhea fire, campfire has the unsurprising plural áheá, but as a referential mass noun it becomes instead áheakoa fire.
- The -koa suffix can also be used to form nouns referring to coherent groups, such as béákoa group (of associates or companions); hunting party.
- Full reduplication forms nouns, often with a quantifying significance: toà day → toà toá everyday, all the time (but also just time), béá person → béá beá everybody. (The tone change on toà is regular, and affects all reduplicating nouns that normally use a high tone to mark the plural. The tone change on béá repairs a violation of the obligatory contour principle.)
All definite noun phrases must take a definite determiner of some sort. There are five possibilities:
- The determiner tì is used with proper names and some other nouns that have unique referents, as in tì=ekówé the sun. (I'll gloss this as PN for proper name.)
- The numbers uo one, dzì two, and wékò three can occur as determiners in their definite forms tsáw, indzì, and nékò. This is never obligatory, but will be chosen instead of using the determiner n (below) with one of the numbers in its regular (indefinite) form.
- The deictics gí (proximal) and wòr (distal) can either occur after the noun, with a prenominal definiteness marker, or can themselves occupy the position before the noun. These are often used in preference to n when the definiteness of the noun phrase is secured by a post-nominal modifier, especially a possessor or relative clause.
- Prefixes of both possessive series imply definiteness without any further marker.
- In all other cases, the proclitic n= is used, or in= before a consonant (in which case the n will assimilate in place if the consonant is not h, and a stem-initial n will become d).
imbéágagúr dé sààm dzòruNeither noun phrase in this sentence is referential, so neither can be marked plural; the first but not the second gets definite marking because of its generic significance.
n -béá -gágur dé sàà-m dzòru
DEF-person-Gagur be.3s eat-PTCP fish
"The Gagur people eat fish"
3. Possession
There are three ways of marking possession, depending on the nature of the relationship.
Inalienable possession is marked with the Class I possession prefixes:
| S | Pl | |
| 1 | a- | áá- |
| 2 | o- | owi- |
| 3 | e- | é- |
| S | Pl | |
| 1 | ado- | adóó- |
| 2 | o- | óó- |
| 3 | ee- | éé- |
Finally, a possessor can be given using the oblique preposition fi: impogá fi=tì=uéɾi Wedi's axe. I'll call this associative possession.
It's not too difficult to construct a full paradigms:
- edzóó tì=uéri Wedi's body (inalienable possession)
- eedzóó tì=uéri Wedi's meat (that she'll eat; beneficiary possession)
- gí dzóó fi=tì=uéri Wedi's meat (that she's offering at a feast, maybe; associative possession)
There's an open class of adjectives that can directly modify nouns:
- tèda wéàti a sheltered campsite
- áhea gírí a hot fire
Adjectives, along with other postnominal modifiers, normally do not count as part of the same tone group as the head noun.
5. Numbers
Numbers follow adjectives, except when the definite forms occur prenominally, as discussed above.
- pogáá wekó three axes
- éádako dzì two girls
uéri sààdze indzì dzóó6. Deictics
uéri sàà-dz -e indzì dzóó
Wedi eat-PST-3s two.DEF meat.PL
"Wedi ate the two (portions of) meat"
gí (proximal) and wòr (distal) can occur postnominally, following any numbers, so long as there is some other prenominal marker of definiteness:
- impogáá wòr those axes
- náhea gí this campfire
- wòr pogáá those axes
- gí ahéa this campfire
These follow all other modifiers, though they can be followed by focus markers. They are usually formed simply by gapping, together with subjunctive marking on the verb:
ábá máádzúú adò gíbeáko àhtéé adóódzòóIt's very common to use the proximal deictic gí as a definite determiner when the phrase's definiteness is secured by a post-nominal modifier, as here.
ábá mááde-dz -úú adò gí =béá -ko àh -téé adóó -dzóó
and.then dance-PST-1p BEN PROX=person-PL give-3p.SBJV 1p.POSS-meat.PL
"Then we danced for the ones who brought the meat"
Verbs
More: show
1. Core paradigms
There are three main verbal paradigms: nonpast, past, and subjunctive. This section will run through their forms. These register agreement with an argument of the verb, most often the subject.
Nonpast indicative verbs are marked only for agreement. Here are the markers:
Some particulars:
The past tense has the marker -dz, which becomes -ts after a coda h but which is pleasantly invariant. There is one minor irregularity: some stems ending in a high vowel have the vowel lower befre the past tense marker. (Among the example verbs, tsi know has this happen: it has past forms in tsedz-.) Besides that, verbs in the past tense take a distinct set of agreement markers:
The ó allophone of the 2s marker occurs after low tones. As discussed elsewhere, the same agreement markers occur on focus particles and on the locative copula.
Here are the example verbs conjugated in the past tense:
Finally, there are subjunctive forms. These do not distinguish nonpast from past, and except for the 1s form use the same agreement markers as the past tense. The subjunctive marker itself also varies with person and number, however; the full forms are as follows:
The shorter allophones of the 1s, 2s, and 2p markers occur after coda consonants; the high tone allphones of the 2s markers occur after a low tone.
Here are the example verbs:
2. Negation
Negation is indicated with the suffix -ᴸhro, which follows the agreement suffixes. Morphologically speaking it's straightforward.
3. The patient participle
Many verbs have what I'll call a patient participle, that can be used with various auxiliaries and in secondary predication (nonmorphological details elsewhere).
In the simplest cases, the patient participle is formed by suffixing -m to a vowel-final stem:
It would take us too far afield here to take up the question of which verbs have patient participles. You wouldn't be too far wrong if you guessed that it is verbs that take patient arguments---but there is a morphological complication. Derived causatives do not get their own patient participle even if their source verb has one. Thus we get ókóme from óká stand (intrastive) but no *òkóme corresponding to òká stand (transitive).
4. The ku- nominalisation
The most productive and regular of Bááru's nominalisations, and the one most integrated into the syntax, is that formed by prefixing the verb stem with ku-. In fact you can think of this prefix as applying to a whole verb phrase, for even in this form the verb can have its full complement of internal arguments and VP adjuncts.
It cannot have a subject, however. Instead the subject of the verb is expressed as a possessor, using the Class I possession prefixes (otherwise used for inalienable possession). Here they are:
The subject itself, if overt, can only occur before the verb, unlike with the possessor's of regular nouns.
5. The locative copula
The locative copula ts be at inflects using the same agreement suffixes as past tense verbs:
For a past or subjunctive locative predication, the verb dé be is used in place of ts.
There are three main verbal paradigms: nonpast, past, and subjunctive. This section will run through their forms. These register agreement with an argument of the verb, most often the subject.
Nonpast indicative verbs are marked only for agreement. Here are the markers:
| S | Pl | |
| 1 | -rí, -í, dí | -bí, -abí |
| 2 | -ò, -ó | -óó |
| 3 | -- | -V |
Some particulars:
- The -í allophone of the 1s ending occurs after a coda consonant, except that -dí occurs after h.
- The -abí allophone of the 1p ending occurs after a consonant and after most monosyllabic verbs (the exceptions are those ending in a `weak' i).
- The -ó allophone of the 2s ending occurs after a high tone.
- The 3p suffix is an echo vowel: it repeats the last vowel of the stem, except that if that vowel bears a low tone, it becomes high.
| dé | ko | tsi | mááde | sàà | mèàw | àh | |
| be | hit | know | dance | eat | drink | give | |
| 1s | dériᴴ | korí | tsirí | mááderí | sààrí | mèàwí | àhdí |
| 2s | déò | kóò | tsò | máádeò | sààó | mèàwó | àhó |
| 3s | dé | ko | tsi | mááde | sàà | mèàw | àh |
| 1p | déabí | koabí | tsabí | máádebí | sààbí | mèàwabí | àhabí |
| 2p | déòó | kóó | tsóó | máádeóó | sààóó | mèàwóó | àhóó |
| 3p | déé | koo | tsii | máádee | sààhá | mèàwá | àhá |
The past tense has the marker -dz, which becomes -ts after a coda h but which is pleasantly invariant. There is one minor irregularity: some stems ending in a high vowel have the vowel lower befre the past tense marker. (Among the example verbs, tsi know has this happen: it has past forms in tsedz-.) Besides that, verbs in the past tense take a distinct set of agreement markers:
| S | Pl | |
| 1 | -í | -úú |
| 2 | -ò, -ó | -óó |
| 3 | -e | -éé |
The ó allophone of the 2s marker occurs after low tones. As discussed elsewhere, the same agreement markers occur on focus particles and on the locative copula.
Here are the example verbs conjugated in the past tense:
| dé | ko | tsi | mááde | mèàw | sàà | àh | |
| be | hit | know | enjoy | see | eat | give | |
| 1s | dédzí | kodzí | tsedzí | máádedzí | mèàwdzí | sààdzí | àhtsí |
| 2s | dédzò | kodzò | tsedzò | máádedzò | mèàwdzó | sààdzó | àhtsó |
| 3s | dédze | kodze | tsedze | máádedze | mèàwdze | sààdze | àhtse |
| 1p | dédzùú | kodzúú | tsedzúú | máádedzúú | mèàwdzúú | sààdzúú | àhtsúú |
| 2p | dédzòó | kodzóó | tsedzóó | máádedzóó | mèàwdzóó | sààdzóó | àhtsóó |
| 3p | dédzèé | kodzéé | tsedzéé | máádedzéé | mèàwdzéé | sààdzéé | àhtséé |
Finally, there are subjunctive forms. These do not distinguish nonpast from past, and except for the 1s form use the same agreement markers as the past tense. The subjunctive marker itself also varies with person and number, however; the full forms are as follows:
| S | Pl | |
| 1 | -ttú, -tú | -súú |
| 2 | -ttò, -tò, -ttó, -tó | -ttóó, -tóó |
| 3 | -te | -téé |
The shorter allophones of the 1s, 2s, and 2p markers occur after coda consonants; the high tone allphones of the 2s markers occur after a low tone.
Here are the example verbs:
| dé | ko | tsi | mááde | mèàw | sàà | àh | |
| be | hit | know | enjoy | see | eat | give | |
| 1s | díttuᴴ | kottú | tsittú | máádettú | mèàwtú | sààttú | àhtú |
| 2s | díttò | kottò | tsittò | máádettò | mèàwtó | sààttó | àhtó |
| 3s | díte | kote | tsite | máádete | mèàwte | sààte | àhte |
| 1p | dísùú | kosúú | tsisúú | máádesúú | mèàwsúú | sààsúú | àhsúú |
| 2p | díttòó | kottóó | tsittóó | máádettóó | mèàwtóó | sààttóó | àhtóó |
| 3p | dítèé | kotéé | tsitéé | máádetéé | mèàwtéé | sààtéé | àhtéé |
2. Negation
Negation is indicated with the suffix -ᴸhro, which follows the agreement suffixes. Morphologically speaking it's straightforward.
3. The patient participle
Many verbs have what I'll call a patient participle, that can be used with various auxiliaries and in secondary predication (nonmorphological details elsewhere).
In the simplest cases, the patient participle is formed by suffixing -m to a vowel-final stem:
- íè see → íèm
- sàà eat → sààm
- tsi know → tsem
- óká stand → ókáme
- kóón fear → kóómbi
- éáw rub → éáwbe
It would take us too far afield here to take up the question of which verbs have patient participles. You wouldn't be too far wrong if you guessed that it is verbs that take patient arguments---but there is a morphological complication. Derived causatives do not get their own patient participle even if their source verb has one. Thus we get ókóme from óká stand (intrastive) but no *òkóme corresponding to òká stand (transitive).
4. The ku- nominalisation
The most productive and regular of Bááru's nominalisations, and the one most integrated into the syntax, is that formed by prefixing the verb stem with ku-. In fact you can think of this prefix as applying to a whole verb phrase, for even in this form the verb can have its full complement of internal arguments and VP adjuncts.
It cannot have a subject, however. Instead the subject of the verb is expressed as a possessor, using the Class I possession prefixes (otherwise used for inalienable possession). Here they are:
| S | Pl | |
| 1 | a- | áá- |
| 2 | o- | owi- |
| 3 | e- | é- |
The subject itself, if overt, can only occur before the verb, unlike with the possessor's of regular nouns.
5. The locative copula
The locative copula ts be at inflects using the same agreement suffixes as past tense verbs:
| S | Pl | |
| 1 | tsí | tsúú |
| 2 | tsò | tsóó |
| 3 | tse | tséé |
For a past or subjunctive locative predication, the verb dé be is used in place of ts.
Focus
More: show
There are four inflecting focus particles:
The focus particles take the same agreement markers as do past tense verbs (their citation form includes the default 3s agreement marker). When the focused constiuent is itself a noun phrase, the focus particle will of course agree with that noun phrase. In other cases---for example with predicate or sentence focus---it may not be obvious what if anything the particles should agree with, and even in the obvious cases focus can have consequences that you might not expect.
Let's start with the very basics. Here's an example in which the object has been focused:
Here it's the whole phrase néékoa béke live snakes that's focused.
I currently plan for heavy modifiers like relative clauses to come after the focus particle, despite logically being a part of the focused constituent:
Arbitrary constituents can be focused. Maybe the following statement would be appropriate if it had been called into question whether I was at the shrine or merely near it:
With focused adjuncts there's not much more to say in an initial sketch like this. But objects and subjects can be more interesting.
Here's a pair of examples that include the adverb ehìó repeatedly, with the object focuse in (just) the second one; you'll see that the relative position of the object and adverb differs.
The same thing happens with a focused subject: combined with a focus particle, it cannot move to its usual position before the verb. Further, the verb will no longer agree with it---as if the (agreeing) focus particle has used up its ability to control agreement. So regardless of the features of the subject, the verb will end up with default 3s agreement.
Here's an example, with an intransitive verb:
Things don't especially change in a transitive clause, though it's maybe worth mentioning that the focused subject still comes before the object:
Meanwhile, I'm still trying to sort out predicate focus and sentence focus. (Translation: I want to look some things up before making any decisions, and I'm not doing that till after this challenge is done.) Both of these will have a focus particle suffixed to the verb. For predicate focus, the focus particle will agree with the object (if there is one), for sentence focus it'll agree with the subject.
I've got myself convinced that in this set-up, predicate focus will have the same effects on word order as does object focus, but I'm more concerned about sentence focus: I want it to have the same effects on agreement and word order as does subject focus, and I'm pretty sure things tend to go that way in the languages I'm vaguely basing this on, but I can't right now see how to make it work.
Anyway, to finish this discussion, here's an example of predicate focus:
- ge also, all (exhaustive)
- le even (mirative, I guess)
- we only (contrastive)
- pe (interrogative)
The focus particles take the same agreement markers as do past tense verbs (their citation form includes the default 3s agreement marker). When the focused constiuent is itself a noun phrase, the focus particle will of course agree with that noun phrase. In other cases---for example with predicate or sentence focus---it may not be obvious what if anything the particles should agree with, and even in the obvious cases focus can have consequences that you might not expect.
Let's start with the very basics. Here's an example in which the object has been focused:
tìuéri sààdze néékoagééThe focused constituent here is a single word, with which the focus particle has thoroughly fused, phonologically speaking. I'll gloss it as a clitic, mostly because it can attach to phrases as well as to particular words:
tì=uéri sàà-dz -e n =éé -koa=g -éé
PN=Wedi eat-PST-3s DEF=snake-PL =FOC-3p
"Wedi also ate the snakes (→ in addition to the other things she ate)"
tìuéri sààdze néékoa békeléè(Aside: I don't know what's up with Wedi and snakes, but I expect I'll have fun finding out.)
tì=uéri sàà-dz -e n =éé -koa béke -ᴴ =l -éé
PN=Wedi eat-PST-3s DEF=snake-PL alive-PL=FOC-3p
"Wedi even ate the live snakes"
Here it's the whole phrase néékoa béke live snakes that's focused.
I currently plan for heavy modifiers like relative clauses to come after the focus particle, despite logically being a part of the focused constituent:
ábá máádzúú adò gíbeákogéé àhtéé adóódzòóAnother possibility would be to consistently use focus particles to mark the ends of relative clauses.
ábá mááde-dz -úú adò gí =béá -ko=g -éé àh -téé adóó -dzóó
and.then dance-PST-1p BEN PROX=person-PL=FOC-3p give-3p.SBJV 1p.POSS-meat.PL
"Then we danced for the ones who brought the meat"
Arbitrary constituents can be focused. Maybe the following statement would be appropriate if it had been called into question whether I was at the shrine or merely near it:
tsí oowe wórré(Notice the default 3s agreement on the focus particle.)
ts -í oo =w -e wórré
COP-1s LOC=FOC-3s shrine
"I was at the shrine (→ not somewhere in some other relation to the shrine)"
With focused adjuncts there's not much more to say in an initial sketch like this. But objects and subjects can be more interesting.
Here's a pair of examples that include the adverb ehìó repeatedly, with the object focuse in (just) the second one; you'll see that the relative position of the object and adverb differs.
tìuéri kodze nùtòba ehìó
tì=uéri ko -dz -e n =ùtòba ehìó
PN=Wedi hit-PST-3s DEF=stump repeatedly
"Wedi hit the stump repeatedly"
tìuéri kodza ehìó nùtòbageAdverb placement is more rigid in Bááru than it is in English, and this alternation is meant to be striking. The idea is that somehow the focus particle prevents the object from moving into its usual position in the sentence's structure.
tì=uéri ko -dz -e ehìó n =ùtòba=g -e
PN=Wedi hit-PST-3s repeatedly DEF=stump=FOC-3s
"Wedi also hit the stump repeatedly (→ in additon to the other things he hit repeatedly)"
The same thing happens with a focused subject: combined with a focus particle, it cannot move to its usual position before the verb. Further, the verb will no longer agree with it---as if the (agreeing) focus particle has used up its ability to control agreement. So regardless of the features of the subject, the verb will end up with default 3s agreement.
Here's an example, with an intransitive verb:
ábá si himáádedze béábeágèéThere's an extra morpheme in there that I'm not yet sure about: hi, which I've glossed as EXPL. I'm imagining it as having something of the function of an expletive pronoun (hence the gloss), but I'm wavering on whether to keep it. My current plan is to let it hang around a bit, and let it stay if it gets up to anything interesting.
ábá si hi -mááde-dz -e béá.bèá =g -éé
and.then also EXPL-dance-PST-3s everybody=FOC-3p
"Then everyone danced"
Things don't especially change in a transitive clause, though it's maybe worth mentioning that the focused subject still comes before the object:
hisààdze tìuérile néékoaIncidentally, I've successfully fought off the impulse to have the verb agree with the object in cases like this: I'm pretty sure that wouldn't work given other things going on in Bááru, and anyway I'm already doing something a lot like that in the main language I've been working on recently.
hi -sàà-dz -e tì=uéri=l -e n =éé -koa
EXPL-eat-PST-3s PN=Wedi=FOC-3s DEF=snake-PL
"Even Wedi ate the snakes"
Meanwhile, I'm still trying to sort out predicate focus and sentence focus. (Translation: I want to look some things up before making any decisions, and I'm not doing that till after this challenge is done.) Both of these will have a focus particle suffixed to the verb. For predicate focus, the focus particle will agree with the object (if there is one), for sentence focus it'll agree with the subject.
I've got myself convinced that in this set-up, predicate focus will have the same effects on word order as does object focus, but I'm more concerned about sentence focus: I want it to have the same effects on agreement and word order as does subject focus, and I'm pretty sure things tend to go that way in the languages I'm vaguely basing this on, but I can't right now see how to make it work.
Anyway, to finish this discussion, here's an example of predicate focus:
àtsó téfikedzíwe eA nice touch: it's going to look a bit like the verbs in sentences with predicate focus agree with their objects, though officially that's not what's going on.
àtsó téfike-dz -í =w -e e
1s watch -PST-1s=FOC-3s 3s
"I only watched"
The patient participle
More: show
This isn't worked out in any detail, but it's maybe fun enough to throw out some ideas. It's about a sort participle that you can form from many verbs with suffixes -m, -me, -bi, and -be (I gave the main morphological details when discussing verbs).
At a first pass, the verbs that have a patient participle are ones that assign thematic roles that are sufficiently patientlike. I'm thinking of patients here in contrast to themes as much as to agents: in contrast to both, patients are prototypically affected by the event reported by the verb, and that's the idea I'm mostly after here.
I want to let this idea wiggle in three directions:
The interest of these participles will depend on the uses I find for them. The motivating construction was a sort of complex negation using a stative auxiliary, now dé be, like this:
...and, that's where my brain's crashing, maybe I'll come back to this.
At a first pass, the verbs that have a patient participle are ones that assign thematic roles that are sufficiently patientlike. I'm thinking of patients here in contrast to themes as much as to agents: in contrast to both, patients are prototypically affected by the event reported by the verb, and that's the idea I'm mostly after here.
I want to let this idea wiggle in three directions:
- I want it to stretch to include the affected agents of verbs like eat and drink when they're used to describe the effect of an event on its agent. (snakes is the patient of Wedi ate snakes, but Wedi is the patient as well as the agent of Wedi ate.)
- I want it to include experiencers, regardless of whether those are realised as subjects or objects.
- I a bit want it to include the subjects of certain stative verbs, including for example sleep---though it may be enough to count the patients of inchoative verbs (fall asleep).
The interest of these participles will depend on the uses I find for them. The motivating construction was a sort of complex negation using a stative auxiliary, now dé be, like this:
àtsó dérihró sààmThis is supposed to have a stative nuance that I'm finding hard to explain.
àtsó dé-rí-ᴸhro sàà-m
1s be-1s-NEG eat-PTCP
"I am not eating"
...and, that's where my brain's crashing, maybe I'll come back to this.