Ergativity in Kaqchikel: Are subjects easier to understand?

As a dust storm blows in from the Sahara desert, it strikes me as a good time to stay indoors and write up an overview of the question which I hope to begin to answer with field work.

How the dust manages to get from the Sahara to Guatemala. Besides bringing a pall of opacity over the landscape, it also blows in a host of foreign bacteria.   

The project I am pursuing here in Patzun is building off the experiment conducted by - among others - Pedro (Dr. Mateo Pedro) and Masha (Dr. Polinsky), published as Ergativity and the complexity of extraction: a view from Mayan (2014.)

For a brief overview of ergativity, see my previous post: "A Dip into Syntax: Ergativity" - I'll be refering to it below. 

In their experiment, they used ergativity to test for a) the subject processing advantage (SPA) and b) whether case and agreement have equal weight in licensing grammatical function. I'll be focusing on the part a) portion. 

Much as its name suggests, SPA is a theory proposing that processing subjects is easier than processing objects. For instance, in:

1) The girl who lifted the boy was smiling vs.

2) The girl who the boy lifted was smiling

Sentence 1 on the left and 2 on the right. This and below pictures courtesy of Pedro Mateo Pedro.

Here, 1) which involves subject extraction is easier to process/comprehend than 2) which involves object extraction. This seems to support SPA, as subject processing is noticeably easier.

However, as I mentioned in my earlier post about ergativity, English (and other accusative languages) are not the best evidence for this. Since case/agreement and grammatical function cannot be separated in these languages, it is not clear whether subject processing is easier because of the subject postion (aka grammatical function) or the nominative case of the subject (case/agreement) the paper turns to ergative languages. 

The study looks at Ch'ol and Q'anjob'al which are both morphologically ergative languages. In these languages, the subject position (grammatical function) can be separated from case/agreement. This is because, in ergative languages, the subjects of intransitives (S) have absolutive case while the subjects of transitives (A) have ergative case.

Remember: "Her gallops" vs. "He hugged her"

Therefore, ergative languages are ideal for testing whether the subject position - independent of case/agreement - is inherently easier to process.

And in fact, the results of the study showed that in both Ch'ol (morphologically ergative) and Q'anjob'al (morphologically and syntactically ergative) SPA was strongly supported, regardless of case.

I plan to replicate the basic model of the study, albeit in Kaqchikel.

Now, let's talk about the methods of my pilot experiment (based on the methods of the Ch'ol/Q'anjob'al study (Clemens et al 2014.) 

For both languages, the researchers looked at 4 types of sentences with 4 types of gaps (below examples in Kaqchikel):

1. Intransitive (S - absolutive subject - gap)
1. 'Where is the squirrel that jumped [next to the rabbit]?'
2. Ambiguous Transitive (A/O - ergative subject/absolutive object - gap)

1. a) 'Where is the chicken that ate the snake?' OR b) 'Where is the chicken that the snake ate?'

3. Biased Subject Transitive (A - ergative subject - gap)
1. Where is the boy that picked the beans?
4. Biased Object Transitive (O - absolutive subject - gap)
1. Where is the tomato that the girl bought?

Each type of sentence is present in both perfective and progressive forms. The nature of Mayan language is such that both 2a and 2b are actually the same sentence, just with an ambiguous interpretation depending on whether they are parsed as having a subject or object in the position of the initial argument. 

Left: 2a; Right: 2b
All the pictures in this post and  in the experiment were drawn by Daniel Pedro Mateo.

The sentences in 3 and 4 could perhaps be interpreted as ambiguous, but since only one argument has animacy, they are semantically biased. For instance, the implausibility of 'Where is the tomato that bought the girl?' prevents the alternate interpretation of 4. 

Native speakers will be presented with a recording of several instances of each of the 4 types of sentences alongside a picture identification task, such as the ones pictured below. 

1. Intransitive: "Where is the squirrel that jumped next to the rabbit?"

2. Ambiguous Transitive: "Where is the dog that licked the cat?" OR "Where is the dog that the cat licked?"

3. Subject Biased Transitive: "Where is the boy who bought tomatoes?" 

4. Object Biased Transitive: "Where are the flowers that the boy picked?"

Hypothesis to support the SPA:

1. Both the gaps for intransitive, absolutive subject (S) and the transitive, ergative subject (A) should be processed faster and easier than the transitive object (O) regardless of case. 

2. Additionally, in the ambiguous sentences, the subject initial interpretation (for instance, 2a) should be favored over object initial (2b.) 

The results of the Ch'ol and Q'anjob'al experiment study supported the above. I hope to find out whether Kaqchikel will as well!

Things to Note: Kaqchikel differs from Ch'ol and Q'anjob'al in 2 major ways.

1. While Ch'ol has no syntactic ergativity, and Q'anjob'al has high syntactic ergativity - where no ergative arguments may be (A-bar) extracted, Kaqchikel falls somewhere in between. In Kaqchikel, certain constructions (wh-questions) prohibit ergative extraction but in others (relative clauses) it is permitted, and actually more prevalent (Stiebels, 2006.)
2. Kaqchikel shows a different case alignment than Ch'ol/Q'anjob'al in the non-perfective aspect (Imanishi, 2014.)

Whether these differences will be reflected in my results is something I'll be keeping an eye out for. Especially the second difference:

These charts show the switch in absolutive/ergative alignment from perfective to non-perfective aspect. Note that in non-perfective aspect, the alignment changes to a nominative-accusative style. 

There you have it - the basics of the corner of Kaqchikel grammar I'm hoping to explore. I'm hoping to start running the study soon, stay tuned for how that goes!

Kaqchikel word of the day: kinäq' (KEE-nuq') - beans. A crucial word to know, since beans in all their forms are one of the pillars of Guatemalan cuisine. The ä symbol represents a lax vowel, which appear in Kaqchikel in correspondence to each of the five tense vowels (a, e, i, o, u).

Works Cited

1. Clemens L.E., Coon J., Mateo Pedro P., Morgan A.M., Polinsky M., Tandet G., Wagers

M. 2014. Ergativity and the complexity of extraction: A view from Mayan.

Natural Language and Linguistic Theory.

2. Imanishi, Yusuke. 2014. When ergative is default: A view from Mayan. Proceedings of the 32nd West Coast Conference on Formal Linguistics.

3. Stiebels, Barbara. 2006. Agent focus in Mayan languages. Natural Language &

Linguistic Theory, 24(2), 501-570.
4. Heaton, Raina. 2015. The status of syntactic ergativity in Kaqchikel. Society for the Study of Indigenous Languages of the Americas.