A traveller's report in Buruya Nzaysa

[Cedh]

I translated a fairly long text into Buruya Nzaysa recently. Originally, I had envisioned to use it as the starting text for one of the next Conlang Relays, but it's too long for that (previous relays have usually started with texts of about a quarter the size), so I'm posting it here as an annotated translation instead, and I also recommend it as a great translation challenge for all ambitious conlangers.

The text is a slightly adapted version of Marco Polo's description of alligators in Yunnan (~1298). I'm splitting it into four sections to try and avoid a TL;DR reaction on your part. Here's section one:

Nzɔ ɛma bura

Sah ni tsə tavlɛ nzɔ ɛma bura roni, ño u rudi mvomu ri ta’oh rabɛ lu mɛsə ada pɛmo, o ño u rifə mvomu ri wɛ’oh owa lu ivrɛ ome ñugɛ oltəña. Tiya podɔ ovla rɛ sah emodo ño u lamu o mɛgɔ kwə ntsa muda.

Ɛ’i’ɔxa rumɛ rɛ saxa ɔ rɔ kəlu ah lamu modo, o rɛ saxa ɔ raxolbo yɛni ah asi ñolu. Saxa mɛsu lo əmo ɔ ñe bitsi mvɔ mvunɛ, dal saxa ño lo xol ayru tsɛga ɔ munanɔ ño u tali. Sə lo əmo ayru ɔ tətsɔ ntsa, o saxa lo soldu ɔ suni ah ulpi sopsə. Sə lo mab ɔ ñavra mvomu ri esa ñalta leda ɔ ru sewe inzɔ, o sa ɔ tselɔ ño u silvo olda. Nte’a, esah nzɔ mvɔwa ño u kili o rifə kwə muda xutsɔ olah ɔra maldɔ o əña kwə u nzɔ payu o lesəgɔ.

The great serpents

In this province are found great serpents of such vast size as to strike fear into those who see them, and so hideous that the very account of them must excite the wonder of those to hear it. I will tell you how long and big they are.

You may be assured that some of them are ten paces in length, and in bulk they are equal to a great cask. They have two short forelegs near the head, but for foot nothing but claws like those of a hawk. The head is very big, and the eyes are bigger than apples. The mouth is large enough to swallow a man whole, and is garnished with great pointed teeth. And in short they are so fierce-looking and so hideously ugly that every man and beast must stand in fear and trembling of them.

Well, let's start analyzing the text sentence by sentence:

Sah ni tsə tavlɛ nzɔ ɛma bura roni,
NULL.AUX-3PL in.3 that.ACC far_away_region TOP.NOM serpent notable dwell
In that region dwell great serpents,

ño u rudi mvomu ri ta’oh rabɛ lu mɛsə ada pɛmo,
as.3 INDEF.NOM giant enough_for.3 SUB.NOM INCH.AUX-2PL during.3 DEF.ACC meet.VN to.3 fear
so much like giants that you begin to be afraid when you encounter them,

o ño u rifə mvomu ri wɛ’oh owa lu ivrɛ ome ñugɛ oltəña.
and as.3 INDEF.NOM scary enough_for.3 SUB.NOM FUT.AUX-2PL after.3 DEF.ACC story from.1 about.3 be_fascinated
and so scary that you will be fascinated after my story about them.

The very first sentence of the text already posed a significant challenge, twice: How does Buruya Nzaysa express a conditional clause that is not dependent on another clause, but on the degree to which an attributive noun fits its referent? In the English version of the text that I translated from, the structure in question was expressed in two different ways: 'of such vast size as to...' (using a preposition + infinitive construction dependent on a noun that is itself the object of a preposition), and 'so hideous that...' (using a complement clause dependent on a predicate adjective). Buruya Nzaysa does not distinguish between nouns and adjectives, so I decided that these constructions could be seen as logically equivalent, and would both be expressed in the same way here. But how? At first I decided to coin an adverb mvɔwa 'so much, to such an extent' (from Ndak Ta mbopm 'enough'), to which a normal complement clause would be added. Then I realized that complement clauses act like nouns, and nouns can't be subordinated to adverbs directly, so I added the genitive/circumstantial preposition u in between: mvɔwa u ri.... This felt semantically natural ('enough of it that...'), but it had two minor disadvantages: Firstly, this use of u relies on an older benefactive/purposive sense that has otherwise been lost, and secondly the precise sequence of words is a bit clumsy to pronounce for a probably fairly common construction. The solution I found was to treat the first part of the phrase as a single phonological word in the parent language already - mbopm wau, which would regularly evolve into mvomu in Buruya Nzaysa. Since this is an ordinary preposition, it can simply take a complement clause as its object.

Another thing to note is the parallel structure of the two conditional complement clauses, with two nominalized verbs as the objects of prepositions, and with each of these two nominalized verbs expressing their own patient/theme as a stranded preposition with an implicit 3rd person object: rabɛ lu mɛsə ada, lit. 'during the meeting with it', and owa lu ivrɛ ome ñugɛ, lit. 'after the telling by me about it'.

Tiya podɔ ovla rɛ sah emodo ño u lamu o mɛgɔ kwə ntsa muda.
INT.AUX-1SG>3 for.2 describe SUB.ACC NULL.COP-3PL how_much as.3 INDEF.NOM long and mighty with.3 very look
I'm going to describe to you how much they appear very long and powerful.

The second sentence, the second significant challenge. (The rest of the text proved easier.) Once again, we have an issue of degree, but this time directly subordinated to a matrix verb in a structure equivalent to indirect speech, although there's a slight semantic difference to the latter. The question here is: How does the language deal with interrogative content clauses? In English, the interrogative pronoun 'how much' is used as a subordinator directly, but this is not possible in Buruya Nzaysa. Instead, emodo 'how much' is an adverb modifying the verb of a complement clause, which is itself used as the direct object of the matrix verb: 'describe that it looks how much like...'

Within the complement clause, we also find a grouped noun phrase where two nouns (both of them semantically more like adjectives, by the way) are coordinated by using a single indefinite article, the conjunction o 'and', and a stranded preposition kwə 'with', so that the phrase in question is literally 'like a long one and a mighty one with it'.

Ɛ’i’ɔxa rumɛ rɛ saxa ɔ rɔ kəlu ah lamu modo,
OPT.AUX-2PL>3 know SUB.ACC NULL.AUX-3PL>3 INDEF.ACC ten step of.3 long count
You will want to know that they measure ten paces in length,

o rɛ saxa ɔ raxolbo yɛni ah asi ñolu.
and SUB.ACC NULL.AUX-3PL>3 INDEF.ACC cask wine of.3 thickness match
and that they match a barrel of wine in thickness.

There's not so much to say about this sentence. One of the more interesting points is that numerals can't function as determiners, so the quantification of the length of the serpent requires an additional indefinite article. In the second clause we can see that comparison is typically expressed verbally, here stating equality by using ñolu 'match'. (I've written more on comparative structures in Buruya Nzaysa here.)

New words:
kəlu (n.) 'step, yard (as a measurement of distance)'. Etymology: Ndak Ta kenglau 'walking space'.
raxolbo (n.) 'cask, barrel'. Etymology: Ndak Ta ragolbu 'that which is filled'.

Saxa mɛsu lo əmo ɔ ñe bitsi mvɔ mvunɛ,
NULL.AUX-3PL>3 near.3 DEF.NOM head INDEF.ACC two leg short hold
Near the head they have two short legs,

dal saxa ño lo xol ayru tsɛga ɔ munanɔ ño u tali.
but NULL.AUX-3PL>3 as.3 DEF.NOM foot of.3.ANIM only INDEF.ACC claw as.3 INDEF.NOM hawk
but as their feet only claws like a hawk.

Another example of a numeral preceded by an article in ɔ ñe bitsi mvɔ 'two short legs'. The second half of the sentence consists of a verb-gapping construction. The content verb would have been the same as in the first clause, and so it can get left out. However, the auxiliary verb can't be omitted; the lack of an overt content verb causes it to refer back to the verb of the previous clause.

New words:
munanɔ (n.) 'claw'. A relatively recent compound of mu 'skin' and nanɔ 'cut', literally 'skin-cutter'.
tali (n.) 'hawk, falcon'. Borrowed from Miwan utalju, which literally means 'discoverer'. A likely reason for the loss of the initial /u/ is that it may have been interpreted as part of the article (DEF.ACC lu and/or INDEF.NOM u); also, unstressed initial vowels are rare in Buruya Nzaysa anyway.

Sə lo əmo ayru ɔ tətsɔ ntsa,
NULL.COP-3SG DEF.NOM head of.3.ANIM INDEF.ACC big very
Their head is very big,

o saxa lo soldu ɔ suni ah ulpi sopsə.
and NULL.AUX-3PL>3 DEF.NOM eyes INDEF.ACC apple of.3 size defeat
and the eyes surpass apples in size.

The beginning of this sentence provides an example of a nominal predicate, which is formed in Buruya Nzaysa by inflecting the auxiliary intransitively, but adding the predicate as an additional noun phrase in the accusative case. The auxiliary then functions like a copula. (Note that the only structural difference to the verb-gapping construction in the previous sentence is whether the auxiliary carries transitive or intransitive agreement.)

In the second clause we see another comparison, this time expressing superiority by using the verb sopsə 'defeat, exceed, surpass'.

New words:
ulpi (n.) 'size'. Etymology: Ndak Ta ula pai 'rise big'. Buruya Nzaysa tends to use quality words both as attributes and as abstract nouns (an example is asi, which normally means 'thick, obese' but which I've glossed as 'thickness' above), but some other words denoting qualities refer primarily to the bearer of the quality and can't be used in the abstract sense (an example is rudi 'giant' in the first sentence of this text), and yet others refer primarily to the abstraction. The latter type, of which ulpi is one of the most prominent instances, is fairly rare overall though.

Sə lo mab ɔ ñavra mvomu ri esa ñalta leda ɔ ru sewe inzɔ,
NULL.COP-3SG DEF.NOM mouth INDEF.ACC large enough_for.3 SUB.NOM EMPH.AUX-3SG>3 indeed completely INDEF.ACC man tall swallow
The mouth is so large that it can actually swallow a tall man whole,

o sa ɔ tselɔ ño u silvo olda.
and NULL.AUX-3SG>3 INDEF.ACC tooth as.3 INDEF.NOM dagger be_armed_with
and it is armed with teeth like daggers.

The first part of this sentence contains another 'so much that...' construction with a complement clause as the object of the preposition mvomu, as described earlier. Note the emphatic auxiliary esa, which highlights the unusual abilities of the serpent.

New words:
inzɔ (v.) 'swallow'. Borrowed from Delta Naidda ina så 'swallow down'.
silvo (n.) 'dagger'. Borrowed from Miwan sīlvo 'cut'; this word was at first used verbally like its source (with a semantic shift to 'pierce, stab'), but the ending /-vo/ quickly became associated with the homophonous native agent nominalizer, and so silvo eventually acquired nominal semantics, with a new verb sili being back-formed to take over the verbal meaning.
olda (v.) 'be armed, be equipped with'. Etymology: Ndak Ta oldas 'should fight'. This verb, originally a mostly intransitive formation based on an old obligative mood form, can now also be used transitively with the weapon as its direct object. It is still mostly limited to the semantic field of fighting though; if you wanted to talk about e.g. being equipped with a bag to carry things, you would have to use a more generic verb like mvunɛ 'have, hold, own' instead.

Nte’a, esah nzɔ mvɔwa ño u kili o rifə kwə muda
truly, EMPH.AUX-3PL TOP.NOM so_much as.3 INDEF.NOM dangerous and ugly with.3 look
They are truly so fierce-looking and ugly

xutsɔ olah ɔra maldɔ o əña kwə u nzɔ payu o lesəgɔ.
therefore OBL.AUX-3PL all person and animal with.3 from.3 TOP.NOM worry and tremble
that all people and animals must be worried and trembling because of them.

Where the English translation has an adjectival predicate, Buruya Nzaysa uses a prepositional phrase with the essive preposition ño 'as, like' as an adjunct to the verb muda 'look, appear, resemble'. Within the prepositional phrase, we see a grouped noun phrase again, combining two separate characterizations with the conjunction o 'and' and a stranded preposition kwə.

New words:
mvɔwa (adv.) 'so much, to such an extent, sufficiently'. Etymology: Ndak Ta mbopm 'enough'. This is the "original" version of what I discussed in connection with the first sentence of this text. In most instances I eventually replaced it with the prepositional variant mvomu, but mvɔwa remains a valid word, and it's used here as an adverbial qualifier, with the resulting consequence introduced using the conjunction xutsɔ 'so, therefore'.
kili (n.) 'wild, feral, fierce, dangerous'. Borrowed from Delta Naidda kïlye. Its native cognate is the verb kəye 'threaten, confront, challenge, provoke, upset', about which I wrote a word of the day article back in December 2012.

[Cedh]

This is the second part of an annotated translation of a traveller's report on a type of "great serpents" into Buruya Nzaysa, based on a real-world account by the famous explorer Marco Polo (~1298). Here's the relevant section of the text:

Tsonah rabɛ lu ɛte gɔla lu ɛnalu əno mpu lo rəlse, dal tah rabɛ lu tolbɔ ada lu sade oskə ogu rɛ saxa ɔ tola dəsmoh, o saxa ɔra əña rɔmaxa mɛsɛ alme nzugi. O pɔwah nzɔ ɛma ɔ ani ntsa, ni wəyaxa wɛtuldɔ lu nzəwo ayru u tilɛ ga u əña sədə kunə o kili nimə́luyə o wəyaxa lu əñali ayru nzugi, ntɛ ri maxa lo mɛwɛnomɔ xa esə ɛru. Pɔwaxa ñalta lu ñavo ŋkə same rapsə, ni wəyaxa nzɔ waru skɛga əño o tul; o mah tsa ntɛvɔ́ve esə rɛvle.

By day they live underground because of the great heat, but in the evening they go out in order to search for food, and they devour every animal they can catch. And when these serpents are very hungry, sometimes they will seek out the lairs of tigers or other large wild beasts, and devour their cubs, without the parents being able to prevent it. Indeed if they catch the big ones themselves they devour them too; they can make no resistance.

And here's a detailed sentence-by-sentence analysis:

Tsonah rabɛ lu ɛte gɔla lu ɛnalu əno mpu lo rəlse,
HAB.AUX-3PL during.3 DEF.ACC day under.3 DEF.ACC ground stay due_to.3 DEF.NOM hot_weather
During the day they stay under the ground because of the heat,

dal tah rabɛ lu tolbɔ ada lu sade oskə ogu rɛ saxa ɔ tola dəsmoh,
but INCH.AUX-3PL during.3 DEF.ACC evening to.3 DEF.ACC outside.3 come so_as_to.3 SUB.ACC NULL.AUX-3PL>3 INDEF.ACC meal search_for
but in the evening they come out in order to search for food,

o saxa ɔra əña rɔmaxa mɛsɛ alme nzugi.
and NULL.AUX-3PL>3 all animal REL.AUX-3PL>3 meet without_exception devour
and they devour every single animal that they come across.

There are a few things to note in this sentence. The first of these is that the word mpu 'because', normally a conjunction, is used as a preposition here. It's the only one of Buruya Nzaysa's core conjunctions that can be used in this way. As its argument it has the noun rəlse here, which is interesting too because of its semantics: It's a monolexemic word that refers to hot weather, not necessarily with sunshine but always with a high temperature that makes people feel slightly uncomfortable.

In the second line we find an instance of a nominalized preposition, something that Buruya Nzaysa does quite frequently, especially with temporal prepositions. Here, it's done with a locative preposition though: ada lu sade 'to the outside', which is literally more like 'to the out (of it)'.

In the third line we see the first relative clause in the text - a fairly minimalistic one, rɔmaxa mɛsɛ 'which they meet'. In spoken Buruya Nzaysa, relative clauses are formed like normal main clauses without an overt subject, using the special auxiliary verb rɔma- without any other indication of subordination. This construction derives from reanalysis of a former relative pronoun due to syntactic and phonological similarity with other auxiliaries. In the more conservative registers of the language, the old construction is still in use; here it would have the form rɔma wəru mɛsə REL 3PL.ANIM.NOM meet.VN.

There are also lots of new words here:
gɔla (pp.) 'under, below'. Etymology: Ndak Ta ob nggolang 'at the foot of' (parallel to Fáralo wægól and Ndok Aisô k'eulag).
ɛnalu (n.) 'ground, floor, bottom'. Etymology: Ndak Ta ainalau 'downward place'.
ogu (pp.) 'in order to, for the purpose of, in preparation of'. Etymology: Ndak Ta ob gau 'on the road to'.
alme (adv.) 'invariably, without exception'. Etymology: Ndak Ta al mi 'without none'.
nzugi (v.) 'devour, eat greedily (usually said of animals); use up'. Borrowed from Miwan zugīr 'swallow'. The initial fricative ended up prenasalized mostly because Buruya Nzaysa does not have non-prenasalized voiced /z/. However, it seems very likely that a slightly older version of Miwan did have word-initial prenasalized obstruents, at least as clusters, and quite possibly there was something like that in the ancestor of this word too. In any case, the name of the language itself clearly derives from earlier *mbiw (from Proto-Eigə-Isthmus *ʔum-pew 'of the people').

O pɔwah nzɔ ɛma ɔ ani ntsa,
and COND.COP-3PL TOP.NOM serpent INDEF.ACC hungry very
And if these serpents are very hungry,

ni wəyaxa wɛtuldɔ lu nzəwo ayru u tilɛ ga u əña sədə kunə o kili nimə́luyə
if/then FUT.AUX-3PL>3 even DEF.ACC shelter of.3.ANIM INDEF.NOM tiger or INDEF.NOM animal other tough and wild approach
they will even approach the shelter of a tiger or another strong and wild animal

o wəyaxa lu əñali ayru nzugi,
and FUT.AUX-3PL>3 DEF.ACC cub of.3.ANIM devour
and they will devour their cubs,

ntɛ ri maxa lo mɛwɛnomɔ xa esə ɛru.
with.3 SUB.NOM NEG.AUX-3PL>3 DEF.NOM parent this.ACC EMPH.AUX prevent
with the parents not being able to prevent it.

This sentence exhibits a typical if/then-construction. In English, we use a pair of conjunctions to express a conditional event, with 'if' at the beginning of the protasis and 'then' at the beginning of the apodosis. In Buruya Nzaysa, there's only a single conjunction ni, which is placed between the two clauses. In addition, one of the clauses must carry the conditional auxiliary pɔ-. Usually, the conditional auxiliary is placed in the protasis, as we see here, but that's not necessarily so; it mostly depends on which clause has the stronger semantic need to use a different auxiliary. It's very likely that things are more complicated than this though; I think I'll have to write about conditionals in more detail someday.

There are two other notable things in here, both of them in the last line. One of these is the word mɛwɛnomɔ 'parents', which is a dvandva compound of mɛwɛ 'father' and omɔ 'mother'; the linking /-n-/ is a residue of the conjunction in the Ndak Ta phrase mebwe on omo. The second thing is that the sentence-final complement clause contains two auxiliaries, in initial position the inflected negative auxiliary maxa, and then in pre-final position the uninflected emphatic potential auxiliary esə, which acts as an adverbial modifier to the verb, highlighting the fact that the clause is not about actual prevention, but more fundamentally about the ability to prevent. Together, ma- and esə- provide the semantics of 'not able to'.

New words:
tilɛ (n.) 'tiger'. Borrowed from an unknown language of the Peninsular family; ultimately from Proto-Peninsular *ktiçi-laha (cf. Kibülʌiṅ ttīlʌ̄, Merneha cirilah, Vylessa ktīlḗ, Lotoka kila).

Pɔwaxa ñalta lu ñavo ŋkə same rapsə,
COND.AUX-3PL>3 indeed DEF.ACC adult same bring_down instead.3
Actually, if they catch the adults instead,

ni wəyaxa nzɔ waru skɛga əño o tul;
if/then FUT.AUX-3PL>3 TOP.NOM 3PL.ANIM.ACC likewise kill and eat
they will kill and eat them just the same;

o mah tsa ntɛvɔ́ve esə rɛvle.
and NEG.AUX-3PL that.NOM in_vain EMPH.AUX resist
and they can't resist at all.

We see another if/then construction here, but it's syntactically parallel to the previous one so it's more interesting to talk about other things, for instance the deictic use of the attributive noun ŋkə 'same, identical, -self' (referring back to the same parents mentioned in the previous clause), or the similar anaphoric use of the topic marker nzɔ (referring back to the serpents, the main topic of the whole story) to disambiguate the two different 3rd person plural participants. Another detail I would like to draw attention to is that I changed the part of speech of rapsə for this text. The word used to be a conjunction, but it couldn't be used as a conjunction here because it has no overt object clause, so I redefined it as a preposition. Prepositions in Buruya Nzaysa are interpreted to have an implicit object when none is present overtly, and both here and on the only occasion that I've used the word before, a reading as 'instead of that' works out quite well.

New words:
same (v.) 'bring down, hunt down, slay, prey upon (usually said of animals)'. Etymology: Ndak Ta sapmi 'force'.
ntɛvɔ́ve (adv.) 'in vain; (in negated clauses) not at all'. This word is the gerund of bɔve 'fail', itself borrowed from Delta Naidda båve. The Ndak Ta etymon bambi survives natively in Buruya Nzaysa only in the verb əbabe 'be lucky', derived from a futilitive mood form erbambi 'accidentally succeed' (lit. 'fail to fail').

[Cedh]

Here's the third part of the traveller's report:

Sah nzɔ ɛma ɔ marob mvomu ri rabɛ lu amɔ’a ayru ogu rɛ saxa ɔ sudusa ga sɔmɛ mɛsu dəsmoh, rɔmaxa desuga rabɛ lu mɔromə nzɛ, tsonaxa lo ntəwa ayru ɔ tewalu ñavra ni lu buli tsugə, ño ri pexa ɔ raxolbo yɛni tətsɔ mvəbo tselu ɔdɛ.

Ɛ’i’ɔxa rumɛ rɛ rolah ewitsa nzo ɛma xo’a. Pɔwaxa lo maldɔ ɔ tewalu ŋkə olna’a, ni wəyaxa o tselu ɔ esevo mvoñu ogu rɛ esaxa lu ɛma same mpu wə mvəbo xa nolɔ ŋkə nzɔ oskə bu. Saxa lo xo’avo ɔ ɛde ɔpsoga ni lu buli ñawe, on o lu əmo ah tsa ɔ ñəbo nanɔvo tsiba, o saxa ntɛga ɛxɛ ntɛ u pɛlbɛ ntaskɔda xutsɔ ɔdɔwə lo ɛma rugɛ o maña tsa ena. Rabɛ rɛ sə ada tselu oskə bu wə wa lu ñəbo otɛ ntɛ u ñalɔ́nzi mvomu ri ɔdɔwa tsa leda lu mvusmo ayru nanɔ, owa lu mɔdɛ ada lu ntəwa ayru, o ri ɔdɔwə ntise nzɔ ɛma no. O ntɛga ta lo adavo nzəxrə, o wəyaxa lo xo’avo rumɛ rɛ ɔdɔwə lo ɛma ɔ omva, o wəya ada tsa oskə o lɛysa.

The weight of these serpents is so great that when they travel in search of food or drink, as they do by night, the tail makes a great furrow in the soil as if a full ton of liquor had been dragged along.

Now the way in which they are caught is this: The huntsmen take them by certain traps which they set in the track over which the serpent has passed, knowing that the beast will come back the same way. They plant a stake deep in the ground and fix on the head of this a sharp blade made like a razor, and then they cover the whole with sand so that the serpent cannot see it. On coming to the spot, the beast strikes against the blade with such force that it enters his breast and rives him up to the navel, so that he dies on the spot. And the crows on seeing the brute dead begin to caw, and then the huntsmen know that the serpent is dead and come in search of him.

Let's analyse this section in detail too:

Sah nzɔ ɛma ɔ marob mvomu ri rabɛ lu amɔ’a ayru
NULL.COP-3PL TOP.NOM serpent INDEF.ACC heavy enough_for.3 SUB.NOM during.3 DEF.ACC travel.VN of.3.ANIM
These serpents are so heavy that when they are traveling

ogu rɛ saxa ɔ sudusa ga sɔmɛ mɛsu dəsmoh,
so_as_to.3 SUB.ACC NULL.AUX-3PL>3 INDEF.ACC meat or water near.3 search_for
in order to search for meat or water,

rɔmaxa desuga rabɛ lu mɔromə nzɛ,
REL.AUX-3PL>3 usually during.3 DEF.ACC night do
which they usually do at night,

I have mentioned grouped noun phrases before, both earlier in this text and in a blogpost from a few months ago, but all of the examples discussed there were of the additive type. To recap: Two or more nouns can be combined into a single noun phrase by using a single determiner, the conjunction o 'and', and a stranded comitative preposition kwə at the end of the group, e.g. ɔ kwamɛdi o nzuma kwə INDEF.ACC tea and coffee with.3 'both tea and coffee'. In this sentence we see that the above pattern can also be used with the conjunction ga 'or', which however requires a different stranded preposition, mɛsu 'near, next to': ɔ sudusa ga sɔmɛ mɛsu 'meat or water' (lit. 'a meat or a water near it'). And there's more: mutually exclusive alternatives can be grouped with bɔ’a ... rapsə 'either ... or' (lit. 'X xor Y instead of it'), and counterfactual alternatives can be grouped with ləh ... ala 'neither ... nor' (lit. 'X nor Y without it').

Another thing to note is the relative clause in the final line, which is much longer than the one we've seen before, and whose antecedent, the nominalized verb amɔ’a (ayru) '(their) traveling', appears quite a bit earlier in the sentence.

tsonaxa lo ntəwa ayru ɔ tewalu ñavra ni lu buli tsugə,
HAB.AUX-3SG>3 DEF.NOM tail of.3.ANIM INDEF.ACC furrow wide into.3 DEF.ACC soil press
their tail leaves a wide furrow in the soil,

ño ri pexa ɔ raxolbo yɛni tətsɔ mvəbo tselu ɔdɛ.
as.3 SUB.NOM COND.AUX-1PL>3 INDEF.ACC cask wine big through.3 that_place pull
as if we had pulled a large barrel of wine along that way.

What's most interesting here is Buruya Nzaysa's equivalent of an as-if-construction, a simile which compares a real event to an unrelated hypothetical event. The relevant grammatical structure is built with the preposition ño 'as, like', the complementizer ri, and a subclause introduced by the conditional auxiliary pɔ-. In the above example, the result is literally something like 'as that we might pull a barrel of wine'. (Also note the first person plural subject, which can be interpreted as an impersonal dummy because the identity of the agent is completely irrelevant here. The English translation of Marco Polo's original text, quoted above, contains a passive ['had been dragged along'], but Buruya Nzaysa tends to avoid a passive wherever possible, using an expletive plural subject instead. All three persons are possible; the first person is used here because a third person might be confused with the actual protagonists of the story, and a second person would make even less sense because the listener was definitely not involved in what is being described.)

New words:
tewalu (n.) 'furrow, trench, groove, notch, indentation'. Borrowed from Delta Naidda tewalo, lit. 'tiller-place'; the locative suffix /-lo/ was replaced with its native equivalent /-lu/.

Ɛ’i’ɔxa rumɛ rɛ rolah ewitsa nzo ɛma xo’a.
OPT.AUX-2PL>3 know SUB.ACC HAB.PASS.AUX-3PL how TOP.ACC serpent hunt
You will want to know how these serpents are hunted.

This short sentence starts off with the set presentational phrase ɛ’i’oxa rumɛ rɛ... 'you will want to know that...', which we have already encountered earlier in the text, and which is a frequent device in storytelling to start a new chapter of the narrative.

More central to the grammar of the language, however, is what we can observe in the subordinated clause, namely a passive voice construction. I have said above that the passive tends to be avoided in Buruya Nzaysa, and up to now I didn't think there would even be a passive voice, but in this particular sentence, a circumlocution with a dummy subject simply didn't feel right. I wanted to focus on the serpents entirely, and after some debate I decided that a passive voice seems to suit the language quite well after all. Partly because the parent language Ndak Ta had a passive, partly because the two most closely related sister languages Delta Naidda and Ndok Aisô also have a passive (although it's an innovation in Naidda), and partly because Buruya Nzaysa already had a class of "ergative verbs" which basically work like passives anyway. It seemed logical to extend the rules applying to this special class of verbs to make them usable with regular verbs too.

This is done by using a special set of auxiliaries, which line up only partially with the auxiliaries that are available in the active voice. The semantically most basic one, sp-/spɛ-, is derived from the Ndak Ta ergative verb ispe 'feel, experience'; most of the others descend from passive voice forms of the Ndak Ta copula with different mood prefixes. The habitual passive auxiliary rol- that we see here is a case in point, it contains a reflex of the habitive mood prefix ru-. There are also auxiliaries for a resultative passive (ɔdɔl-, derived from Ndak Ta oto 'come' like its active voice counterpart), for a potential passive (gal-, containing the probabilitive/permissive mood prefix bwa-), for a conditional passive (pɔl-, containing the conditional mood prefix pâu-), and for a negative passive (ml-/mal-, containing the negative prefix m-).

Other new words:
ewitsa (pron.) 'how, in what way'. Etymology: Ndak Ta iwa etsn 'which method'.

Pɔwaxa lo maldɔ ɔ tewalu ŋkə olna’a,
COND.AUX-3PL>3 DEF.NOM person INDEF.ACC furrow same find
If the people find such a furrow,

ni wəyaxa o tselu ɔ esevo mvoñu
if/then FUT.AUX-3PL>3 at.3 that_place INDEF.ACC trap establish
they will set up a trap at that place

ogu rɛ esaxa lu ɛma same
so_as_to SUB.ACC EMPH.AUX-3PL>3 DEF.ACC serpent bring_down
in order to bring down the serpent

mpu wə mvəbo xa nolɔ ŋkə nzɔ oskə bu.
because FUT.AUX-3SG through.3 this.ACC path same TOP.NOM come again
because it will come back along the same path.

There's not too much remarkable stuff going on here, except for the use of attributive ŋkə 'same, identical, -self' in combination with an indefinite article to give the meaning 'such a..., one of that kind'. It's also worth pointing out that the auxiliary in the third line is the emphatic potential esaxa instead of the null auxiliary saxa; this choice highlights that the trap provides the hunters with the ability to bring down the serpent, which they wouldn't have without it.

New words:
esevo (n.) 'trap'. Etymology: Ndak Ta isibu 'that which acts suddenly'.

Saxa lo xo’avo ɔ ɛde ɔpsoga ni lu buli ñawe,
NULL.AUX-3PL>3 DEF.NOM hunter INDEF.ACC pole deep-ADV in.3 DEF.ACC soil put
The hunters place a pole deep in the ground,

on o lu əmo ah tsa ɔ ñəbo nanɔvo tsiba,
and at.3 DEF.ACC head of.3 that.NOM INDEF.ACC blade sharp attach
and to the head of that they attach a sharp blade,

o saxa ntɛga ɛxɛ ntɛ u pɛlbɛ ntaskɔda
and NULL.AUX-3PL>3 then everything with.3 INDEF.NOM sand wrap
and then they cover everything with sand

xutsɔ ɔdɔwə lo ɛma rugɛ o maña tsa ena.
therefore RES.AUX-3SG DEF.NOM serpent be_misguided and NEG.AUX-3SG<3 that.NOM see
so that the serpent will be tricked and cannot see it.

In this portion of the text we find two words which are not new, but semantically interesting. The first of these is ntaskɔda 'wrap', derived from the Ndak Ta phrase ntatsn kota 'cover all around'. The semantic range of this verb is a lot broader than what the English gloss suggests. The core meaning is the same - 'wrap (e.g. in cloth)' -, but the Buruya Nzaysa word can also be used in other situations where the object is getting enclosed by something else, for instance 'cover with sand' as we see in the text, 'pack (into a box, e.g. for storage or transport)', or 'surround (with military forces)'. When used reflexively, the verb usually means 'get dressed', but it can also be used for 'tuck oneself in (under a blanket)'. There's also a back-derived noun ntaskɔ, motivated by reanalysis of the final /-da/ as the homophonous inchoative/causative suffix, which takes its meaning 'respectable' from the reflexive sense (via 'dressed').

The second semantically interesting word is rugɛ 'be misguided, get tricked into sth.', which is a loan of Ndok Aisô rugeu 'be naïve'. In the source language, this verb implies that it's (at least partly) the subject's own fault if something goes wrong, because s/he could and should have known better. The thing that goes wrong can basically be anything, a malicious action by somebody else, an unfortunate side effect of another action, or a natural event. In Buruya Nzaysa, the meaning has shifted slightly but significantly: The Buruya Nzaysa word implies that something detrimental to the subject must happen, and that this event is always the purposeful action of a human adversary, and thus usually outside of the subject's control. It thus fits perfectly for the serpent running into the hunters' hidden trap.

There's also a morphosyntactically interesting detail in here: ena 'see, notice' belongs to the above-mentioned class of "ergative verbs", and therefore requires the auxiliary to take on a special "inverse" transitive agreement suffix, which we can see in maña; note that the arrow in the gloss is reversed in order to indicate that the subject part of the agreement refers to the experiencer. (The semantics are notable too: seeing is conceived of as being caused by that which is visible, and so the "agentive" nominative pronoun tsa 'that' in this clause refers to the trap, not to the serpent.)

New words:
xo’avo (n.) 'hunter'. Agent nominalization of xo’a 'hunt, chase'.
ɔpsoga (adv.) 'deeply, solidly'. Adverbialization of ɔpso 'deep'.
nanɔvo (n.) 'sharp'. Agent nominalization of nanɔ 'cut', with semantic drift towards a more attributive meaning.
tsiba (v.) 'attach, fix, put together'. Borrowed from Fáralo čiəba 'tie, bind'.

Rabɛ rɛ sə ada tselu oskə bu
during.3 SUB.ACC NULL.AUX-3SG to.3 that_place come again
When it comes back to that place,

wə wa lu ñəbo otɛ ntɛ u ñalɔ́nzi
FUT.AUX-3SG against.3 DEF.ACC blade run with.3 INDEF.NOM speed
it will run against the blade with speed

mvomu ri ɔdɔwa tsa leda lu mvusmo ayru nanɔ,
enough_for.3 SUB.NOM RES.AUX-3SG>3 that.NOM completely DEF.ACC belly of.3.ANIM cut
such that it will cut open its belly completely,

owa lu mɔdɛ ada lu ntəwa ayru,
from.3 DEF.ACC heart to.3 DEF.ACC tail of.3.ANIM
from the heart to the tail,

o ri ɔdɔwə ntise nzɔ ɛma no.
and SUB.NOM RES.AUX-3SG immediately TOP.NOM serpent die
and the serpent will die immediately.

Most of this part is fairly straightforward, but there's one syntactic detail that deserves attention. The main clause here is wə wa lu ñəbo otɛ 'it will run against the blade', extendedy by the prepositional phrase ntɛ u ñalɔ́nzi 'with speed'. To the object of this preposition, the noun ñalɔ́nzi 'speed', yet another prepositional phrase is subordinated, starting with mvomu 'enough for'. This prepositional phrase contains not just one, but two objects - two complement clauses that describe two events that are both caused by the content of the main clause, and also intricately connected (the serpent's death as described in the second complement clause is of course the result of it cutting open its belly as described in the first complement clause). Without the last complementizer ri (i.e. with a subordination structure similar to that in English, where the last clause is simply coordinated to the rest of the sentence), the serpent's death would be interpreted as an independent event that simply occurs on returning to the place, which of course doesn't fit the intended semantics.

New words:
ñalɔ́nzi (n.) 'speed, velocity, acceleration'. Abstraction of lɔnzi 'quick, fast', itself borrowed from Delta Naidda lånzi.

O ntɛga ta lo adavo nzəxrə,
and then INCH.AUX-3PL DEF.NOM crow shout,
And then the crows will start to shout,

o wəyaxa lo xo’avo rumɛ rɛ ɔdɔwə lo ɛma ɔ omva,
and FUT.AUX-3PL>3 DEF.NOM hunter know SUB.ACC RES.COP-3SG DEF.NOM serpent INDEF.ACC dead_body
and the hunters will know that the serpent is dead

o wəya ada tsa oskə o lɛysa.
and FUT.AUX-3PL to.3 that.NOM come and collect
and come back to collect its body.

This sentence gives an example of how Buruya Nzaysa, which lacks a formal distinction between nouns and adjectives, uses words with strongly nominal semantics where English would use an adjectival predicate. We can see this in the phrasing of 'is dead' as a nominal predicate 'has become a corpse'. There's a welcome syntactic side effect to this, because the noun omva can be referenced by the deictic pronoun tsa in the next line, and it can also be used as the implicit object of the verb lɛysa 'collect', enabling the latter to appear in an abbreviated serial verb construction without a separate transitive auxiliary.

New words:
adavo (n.) 'crow, raven'. Etymology: Ndak Ta adwabu 'the dark one'.

[Cedh]

This is the fourth and last part of the traveller's report. The final portion of the text reads as follows:

Puysaxa ntɛga rɛ saxa lu smɛnɔ owa lu ni ah lo ɛma sañawe, mpu esə xə ɔ orono rafa ntsa. Dənɛrɛ pɔlə ɔ ru u ɔ tsə ñalvo pɛnə, o pɔwa kɛ’u xagala mvɔ tsɛga u xə orono savla ni ɔdɔwə ntise ɔ laste. Ntɛxavu, pɔwə ɔ asa ntɛ u ñu’o steyə mvotsɛsə, o pɔwa skɛga kɛ’u xagala savla, ni ɔdɔwə uvo lo dəpse nzəwoga məlu. Saxa oma lu sudusa ah nzɔ ɛma puh lo tola ayru pewada, mpu esə xə salɔ, o saxa lo maldɔ xa dəñɛ ntsa.

Those who take them proceed to extract the gall from the inside, for it makes a most precious medicine. Thus if a person is bitten by a mad dog, and they give him but a small pennyweight of this medicine to drink, he is cured in a moment. Again if a woman is hard in labour they give her just such another dose and she is delivered at once. They also take the flesh of this serpent, for it is excellent eating, and the people are very fond of it.

As with the three previous sections of the text, here's a detailed analysis with interlinear glosses and translation notes:

Puysaxa ntɛga rɛ saxa lu smɛnɔ owa lu ni ah lo ɛma sañawe,
CONT.AUX-3PL>3 then SUB.ACC NULL.AUX-3PL>3 DEF.ACC gall from.3 DEF.ACC in.3 of.3 DEF.NOM serpent extract,
Then they proceed to extract the gall from the inside of the serpent,

mpu esə xə ɔ orono rafa ntsa.
because EMPH.COP-3SG this.NOM INDEF.ACC medicine valuable very
because it is a very precious medicine.

The notion of 'proceed' is expressed by a continuative auxiliary puysa-; note that this is placed in a separate matrix clause because otherwise it would be interpreted as a continuation of the action described here, i.e. extracting the gall. There is also another instance of a nominalized preposition in this sentence: owa lu ni 'from the inside', or more literally 'from the in (of it)'.

New words:
smɛnɔ (n.) 'gall'. Etymology: Ndak Ta saungwe eno 'bitter water'.
sañawe (v.) 'take out, retrieve, extract'. Etymology: Ndak Ta sa nggawi 'put out of'.

Dənɛrɛ pɔlə ɔ ru u ɔ tsə ñalvo pɛnə,
namely COND.PASS.AUX-3SG INDEF.ACC man from.3 INDEF.NOM dog berserk bite
For if a man is bitten by a rabid dog,

o pɔwa kɛ’u xagala mvɔ tsɛga u xə orono savla
and COND.AUX-3SG>3 one_of portion small only from.3 this.NOM medicine drink
and if he drinks only a small portion of this medicine,

ni ɔdɔwə ntise ɔ laste.
if/then RES.COP-3SG immediately INDEF.ACC healthy
he will be cured immediately.

There are several interesting things here. The first of these is the introductory adverb dənɛrɛ. I've glossed it as 'namely' above, and that's probably the English word that sums up best what it means, but there's more to this word. It is typically used to introduce a reason, an indirect cause, or a circumstantial factor leading up to something else, but it can also be used to adduce evidence for something or to present additional information, often with connotations of surprise or peculiarity. It shifts the focus of discourse to the clause it appears in, but at the same time it presents the information as rather incidental, similar to the English construction 'by the way'. Syntactically, it patterns like a sentential adverb most of the time, which means that it can occur in all four positions that are immediately adjacent either to the content verb at the end of a clause, or to the auxiliary at the beginning of a clause. When it is placed sentence-initially, however, it can also function much like a conjunction, connecting two statements by declaring that a certain type of relationship exists between them. It is therefore no wonder that the free English translation of the above passage uses the conjunction 'for' as a translation equivalent for dənɛrɛ.

Then there's the fact that we have a complex if/then construction with two different conditions, both of which must be fulfilled for the resolution to happen. Syntactically, this is relatively simple - the two conditions are simply connected with the conjunction o 'and' -, but it has some interesting consequences for the auxiliary verb in the first conditional clause. As I have said earlier, passives tend to be avoided in Buruya Nzaysa, but here the two conditional clauses need to have the same subject, and thus a passive (expressed by choosing the conditional passive auxiliary pɔlə) is required in the first condition.

Also worth noting are the partitive quantifier kɛ’u, which indicates that only one element out of a larger amount is referenced, and the phrasing of 'is cured' as a nominal predicate 'is a healthy one', parallel to how the serpent's death was expressed at the end of the previous section.

New words:
dənɛrɛ (adv.) 'namely, specifically, in fact'. Etymology: Ndak Ta [âk lu] dene raing 'with the reason that'.
ñalvo (n.) 'berserk; one who is mad with rage'. Agentive derivation from ñalə 'rage, frenzy'; the word stem is ultimately a loan from Fáralo ŋalər 'excessive'.
xagala (n.) 'portion, fraction, part'. Etymology: Ndak Ta gabbwala 'that which is shared'.

Ntɛxavu, pɔwə ɔ asa ntɛ u ñu’o steyə mvotsɛsə,
furthermore, COND.AUX-3SG INDEF.NOM woman with.3 INDEF.NOM pain PROG.AUX give_birth
Also, if a woman is hard in labour,

o pɔwa skɛga kɛ’u xagala savla,
and COND.AUX-3SG>3 likewise one_of portion drink
and if she likewise drinks a portion,

ni ɔdɔwə uvo lo dəpse nzəwoga məlu.
if/then RES.AUX-3SG soon DEF.NOM child safely arrive
the child will soon arrive safely.

There are no new words here, but once more we see an if/then construction with two conditions, parallel to the previous sentence. Another detail worth noting is how the progressive auxiliary steyə is used adverbially by being added in pre-final position in the first line, expressing imperfective aspect. And finally there's a slightly idiomatic construction in the last line, ɔdɔwə [...] lo dəpse nzəwoga məlu 'the child will arrive safely', which is based on the cognitive metaphor LIFE IS A JOURNEY.

Saxa oma lu sudusa ah nzɔ ɛma puh lo tola ayru pewada,
NULL.AUX-3PL>3 additionally DEF.ACC meat of.3 TOP.NOM serpent for.3 DEF.NOM meal of.3.ANIM use
They also use the flesh of the serpent for their meals,

mpu esə xə salɔ,
because EMPH.AUX-3SG this.NOM be_delicious
because it is delicious in taste,

o saxa lo maldɔ xa dəñɛ ntsa.
and NULL.AUX-3PL>3 DEF.NOM person this.ACC enjoy very
and the people are very fond of it.

The final sentence of the text is rather straightforward. The two most notable things are that the emphatic potential auxiliary esə, which we've encountered several times as a marker of capability or possibility, is used here simply to emphasize a fact, and that the demonstrative pronoun xə/xa must appear twice, inflected for different cases, because of the different syntactic role that the serpent's flesh plays in those two subclauses. One might want to consider this repetition a way of avoiding a passive, but since passivization in Buruya Nzaysa usually doesn't change the case of the subject, a passive version of the last clause without an overt demonstrative is in fact impossible. This is because case in passive constructions is assigned semantically, so that the serpent's flesh would still have to appear in the accusative, and thus must be referenced overtly because this case is not the same as the case it has in the previous clause.

New words:
oma (adv.) 'also, additionally'. Etymology: Ndak Ta ongwâ 'and' (object-gapping conjunction).
pewada (v.) 'use (for a purpose)'. Etymology: Ndak Ta pipm ta 'useful-DYNAMIC'.



I know this series of posts has been a lot to read, and it's not always easy to give constructive feedback on something that's already quite elaborate, but I hope some of you have something to say about it. Comments? Critique? Questions or remarks about specific details in the language? How did you like the presentation? Any suggestions for a different format? Were my explanations easy to follow? Would you like to read something similar again? Do you have suggestions for other texts to translate?