In English, you have subject > direct objects > indirect objects > obliques. In Sayula, you have primary objects > secondary objects. How does this work out?

Consider a simple transitive sentence:

I tore the book. Here ‘the book’ is the patient.

I read the book. Here ‘the book’ is the theme.

Now consider a ditransitive sentence:

I gave the book to the boy. ‘The book’ is the theme, and ‘the boy’ is the recipient.

(Please note that this depends upon a fairly conservative and partially confused theory of thematic roles. But think of ‘patients’ as something that undergo change, ‘themes’ as things that are manipulated but not changed. You already know what recipients are.)

In a primary/secondary language, the recipient in the ditransitive gets treated the same way as the theme in the simple transitive. My notes on all of this aren’t terribly clear. It’s worth checking out Matt Dreyer’s 1977 article about primary/secondary and direct/indirect objects.

Sayula doesn’t really do passives other than the word for ‘eclipse’ (lit, ‘it gets swallowed’), which it uses the passive marker is ji for. There are also some other reflexive idioms associated with things that happen to you that you can’t help, like sweating, being thirsty, sneezing.

The way you say, ‘the one who comes’ in Sayula is mimpay, min+p+ay, and since you’ve got an /n/ by a /p/, the /n/à/m/.

Now, by rights, the /p/ should voice—it’s between two voiced sounds, right? But there’s actually a word boundary at work here: mimp#ay, so the /p/ stays a /p/.

(This happens for all the clitics, too. ‘He definitely comes’ = mimpama7

MesoAmerica is famous for using body parts in fun ways, so you get ‘head’ used for ‘up’ and ‘stomach’ used for ‘in’. Then these (maybe) get grammaticalized as prepositions.

Sayula names body parts with other body parts—they have a bit of an impoverished system. So the same word jo:t, is used for heart and liver. (Maybe Spanish has just taken over?)

My favorites, though:

Finger = ‘nose of the hand’

Elbow = ‘rear end of the arm’

There are four types of verbs in Sayula Popoluca:

vi = true intransitives (‘I sleep’)

vt1 = allow agent intransitives, like ‘eat’ in English (‘I eat’ means that I’m doing the eating, not that I’m being eaten)

vt2 = allow patient intransitives, like ‘burn’ in English (‘I burn the paper’ is okay, but if you say ‘I burn’ then you’re talking about the burning happening to you (not that you’re burning something).

vt3 = true transitives

In other words, if vt1 or vt2 occur with a single argument, that argument is understood to be the agent or patient, respectively.

English has relatively few vt1’s and vt2’s, MesoAmerica has a lot.

In Sayula, kew7 ('cook') is vi. There is no agent and the patient is obligatory (it's what gets cooked). To have an agent, you have to add a causative morpheme. This shows that you can't just look at the basic meaning to determine what class a verb is.

For fun, think about the intransitive English verb, 'dream'. (You can only maybe have a cognate object: 'I dream a dream'.) But in Ojibwe, bwaaN is a vt3. So it's a true transitive there, where you have to have both the agent and the patient. If you just want to say, 'I dreamt', you have to use the antipassive morpheme that itself means 'about something I'm not specifying'.

There are a bunch of quasi-auxiliary uses for some verbs, especially verbs of motion and starting/stopping verbs.

nix = ‘go’

7oy = ‘go and come back’

min = ‘come’

chu:chiy = ‘start, begin’

k ix = ‘end, finish, stop’

When these are working in an auxiliary way, person markers don’t go where you think they would.

(1)               chú:chiw ti yòxwátka

Ø aux+aspect person number Ø (subordinate present tense)

‘I started to work.’

(2)               ti chú:chiw yòxwátpay

‘I started working.’

It’s very weird that (2) has the person in the middle, which is why it’s worth paying attention to. The auxiliary carries the aspect, the lexical verb carries the person.

[ In Mayan, auxiliaries don't have any explicit aspect, so it isn't as obvious that the things you expect in one place are in two. ]

'Want' is like this in Sayula, too, which is like equi-NP deletion:

'I want to go' = [I want [I go]] but you cross out that interior I.

In Sayula: [I want [I go]] crosses off that first I.

I'm unclear on whether this may happen with Verb+preposition, too.

Speaking of these auxiliaries, Zoque languages don't have a future tense like Sayula. They use 'go' or another auxiliary to do the future. So where did Mixe's funny future comes from? Perhaps Sayula's future wa7n comes from 'want/like', wamp.

-ó7k is an ‘augmentative’ affix that you can add to adjectives and verbs. It’s basically ‘very’. But it seems to come from the root for ‘die’, so it has the look of something that should be compounded. Sayula speakers don’t recognize the connection between pò7pó7k (‘it’s very white’) and 7ó7k (‘to die’), but Rich is pretty sure that’s the origin. Hence, a better gloss:

pò7pó7k = ‘white to die for’

Some of us may remember that participles are a form of the verb that can make complex tenses, like “loving” and “loved”. In English these can also be used as adjectives (http://encarta.msn.com/encnet/features/dictionary/DictionaryResults.aspx?refid=1861723966).

There are two kinds of participles in Sayula.

-ay is a bit like ‘-ing’ in English. It has a sort of present tense sense.

-ik is a ‘past passive’ like ‘-en’ in English.

That means that if you get –p­ (which indicates an incompletive action) and –ay together, you have a ‘doer’: yóxwátpay, which is literally “work+make+incompletive+relative” = ‘worker’. You see –pay all of the time in Sayula.

I think –ik is a little less common and is usually part of the stem. Here’s one I just found in the dictionary: yó7kpénik, which is literally “neck+smash+past.passive” = ‘strangled’.

General rules

Obstruents: C_{1 [+obsruent]} à j / _C₁

Non-obstruents: C_1[-obsruent] à Ø / _C₁

J deletion: j à Ø / _x

Metathesis: CjV_[-stress] à jCV

J cluster simplification: j à Ø / _CC (see above)

Special rules

ppàp

nnàny

Vw [+lab] à V: [+lab]

nà [a place] / __ [a place] in close construction

(see 9/14/05 handout for more)

• For 2/1 and 3/2 (7ix) you need a pronoun or context to disambiguate.
• You pluralize with –ka- or -kix-.
• Note that the “northeast” has /n/’s and the “southwest” has /x/’s. This suggests an inverse relationship, so that 3/1 vs. 1/3 is really a matter of /x/ vs. /n/.
• In other systems it is typical that the third person is involved in such weirdness (I think).
• It is frequently the case that either 1/2 or 2/1 is odd.

 In Sayula, these change if the verb is subordinate. (The trick for translators is to figure out when something looks independent but is subordinate.)

                        Independent                 Subordinate

Incompletive                 -p                     -Ø

Completive                   -w                    -j

Future                          -áj                    -wá7n

• If you see 7ix- and the verb has an independent aspect marker, then you have an independent relationship of 2/1 (you hit me) or 3/2 (he hit you) (not 2/3 (you hit him), which would have an aspect from the subordinate column).
• If you see tix- and an independent aspect marker, then you have an independent relationship of 3/1 (he hit me) (not a 1/3 (I hit him).
• If you see ti-, then you have 1/2 (I hit you) and need to look at the aspect marker to tell about independent or subordinate status.
• If you have tin- then you have an independent relationship of 1/3 (I hit him).
• If you have 7in-, then you have an independent clause with a 2/3 (you hit him) relationship.
• If you have 7i- then you have an independent clause with 3/3 (he hit him).
• If you have 7igi- then you have a 3/3 relationship (he hit him) but you have to look at the aspect marker to determine whether it’s an independent clause or a subordinate one.

First, second, third, inclusive, exclusive, plural

                        Singular            Plural

1^st exclusive      ti-                     -ka

1^st inclusive       na-                   -ka

2^nd                 mi-                   -ka

3^rd                 Ø                     -ka

As with French and languages all over the world, the indefinite (“one”) is expressed in the first person inclusive. This link between first person inclusive and indefinite is worth thinking further about, though I don’t have any ideas myself.

Sayula is an aspectual language. So that means that instead of past/present/future, you get:

• incompleted
• completed
• future

Metathesis (the swapping of sounds) is sensitive to stress:

(1)               wàyját

(2)               wàjyat

Here is an interesting thing to try to work out in Optimality Theory:

            (3)        tíjk       ‘house’

            (4)        tíkm      ‘in the house’

The apparent rule is j à Ø / __CC, but there are two words that retain their /j/’s.

            (5)        tzajp     ‘sky’

            (6)        ni7jy     ‘name’

The way to solve this is to look at the Protowords. Some of them used to have word final vowels and some of them didn’t.

(7)               *tik

(8)               *kama

This is important because all syllable-final obstruents got aspiration in Mixe.

(9)               *tijk (added because the /k/ is final)

(10)           *kama (no /j/ because there was a vowel at the end)

This was further generalized so that not only did syllable-final obstruents get aspiration, but other consonants did, too.

 (11)      *jin à jijn

So there are two ways to deal with this:

Terry’s historical way: insert the /j/ for words that end in consonants and then delete the vowel for words that end in vowels—that way they won’t have /j/’s.

a.       *tikà*tijk=*tijk

b.      *kama=*kamaà*kam

Rich’s lexicon way: mark words like ‘sky’ and ‘name’ as exceptional. Then just have an underlying j and say that jCC à CC. Rich doesn’t want synchronic facts to need to refer to history.

What's the root?

            (1)        tijk       ‘house’

            (2)        tikm      ‘on the house’

Historically, the /j/ was inserted but it’s less clear what’s happening synchronically. That’s a matter of some controversy.

Definitions

Synchronic linguistics look at language at a particular point in time; diachronic linguistics (aka historical linguistics) looks at language over time. In the case of the /j/ above, the problem is that, sure, in the past there wasn't a /j/, then there was and now there is, but how does it function in the language right now? Speakers don't know that it wasn't there and now it is--yet they treat it in special ways and can make distinctions between words that historically never had the /j/ and still don't and those that do have the /j/ but didn't a few hundred years ago:

There are at least three degrees of closeness for compounding. Rich uses "=", "-" and "." to indicate these degrees.

When you have greater degree of closeness, the sounds affect each other:

• n=p, n=k, n=t (loose compound)
• n-b, n-g, n-d (medium)
• m.b, ŋ.g, n.d (tight)

 Note that n-g contrasts with ŋ.g. You can’t tell the difference between n-d and n.d.

You can get three types of vowels in Sayula. 

• V (short)
• V: (long)
• V7 where the glottal stop acts as part of the vowel (there's also V7 where the glottal stop is a separate consonant that you hear as a creaky vowel--check up on this)

People who are writing or transcribing often get the glottal stops wrong (for example, in the text I worked on there were lots of “i7” that should have been “7i”).

The actual vowels are:

i           i           u

e                      o

            a 

As for consonants:

• /x/ is not a “sh”, it’s a retroflex…or at least it used to be. It’s the only native sibilant.
• /k/ voices ‘like crazy’
• /Ctz/ often comes out as Cs for many speakers
• /s/ indicates a loan word and/or a sound-symbolic word (like whisper)
• /l/ is fairly rare, though native
• /r/ is somewhat rare, but some borrowed /d/’s are turned into /r/’s
• /rr/ is almost always a borrowing or a sound symbolic word
• Enye contrasts with /ny/, which is a native phoneme—Rich Rhodes isn’t convinced that people have gotten these right 

There’s a tendency to have a weak breathiness at the ends of words. That’s carried over into the Spanish these folks speak.

You don’t really get *CCC in a morpheme except for with a glottal stop, for example ta7kx, ‘shed light on’.

There are some other morpheme limitations: *NC, *jC, *xC, *Cx, *Ctz. (NB: My notes make it unclear whether these are allowed or unallowed—see handout from 9/7/05.)

Definitions

C-Consonant

V-Vowel