Among linguists, split

“If the Romans had been obliged to learn Latin, they would never have found the time to conquer the world.” —Heinrich Heine

ergativity has long been recognized as one of the most pervasive impediments to human understanding. A particularly puzzling case is apparent in the language of Southern Quiznos, which has provoked several mutually contradicting analyses. For instance, Bloomfield (1924) suggests that split ergativity in Southern Quiznos is sensitive to animacy. Animate subjects are marked with the nominative, while inanimate subjects carry the ergative marker.

(1)	a.	Bob-o	ham	sub-by
		Bob-NOM	eat.PRS	sandwich-ACC
		‘Bob eats the sandwich.’
	b.	Sub-br	nja	philly-ø
		sandwich-ERG	need.PRS	cream cheese-ABS
		‘The sandwich needs more cream cheese.’

Bloomfield’s observation is contradicted by the following data reported by Hockett (1946):

(2)	Tja-br	ham	trsk-ø
	Charles-ERG	eat.PRS	tomato-ABS
	‘Charles eats the tomato.’

Hockett proposes that split ergativity in Southern Quiznos does not follow distinctions on the basis of animacy, but on the basis of shape. While sandwiches are usually square in the cultural setting of Southern Quiznos, cheeses come in the traditional shape of small balls. Tomatoes have a similarly round shape.

“Grammar, which knows how to lord it over kings, and with high hands makes them obey its laws.” —Jean Baptiste Poquelin Moliere

Hockett’s finding has been subsequently criticized by Mithun (1984), who tried to elicit the sentence ‘The ketchup soaked the chocolate chip cookie’, which to her surprise yielded the nominative-accusative pattern. This marking contradicts both Bloomfield’s and Hockett’s predictions.

The present paper argues that the seemingly unpredictable split-ergative pattern in Southern Quiznos falls out of three constraints that are ranked relative to each other. Based on these constraints, an optimal output is generated by the speakers of Southern Quiznos.

The first constraint is, quite in line with the proposal of Bloomfield (1924), that animate subject referents should be marked with the nominative case. This constraint will be called the ANIMACY constraint. As the data from Hockett (1946) and Mithun (1984) show, this constraint is violable. A seemingly obvious question that linguists have failed to ask is under what circumstances ANIMACY can be violated. Apparently, other constraints hold in Southern Quiznos that outrank the importance of marking animate subject referents with the nominative case, as shown in (2).

“I am the King of Rome, and above grammar.” —Sigismund

A factor that has been overlooked in much linguistic research is the relevance of color in the case marking of objects. Sandwiches, which are usually brown on the outside, receive accusative case in Southern Quiznos. By contrast, tomatoes and cream cheese are marked with the absolutive case. This suggests that object marking in Southern Quiznos is sensitive to the color hierarchy proposed in Berlin and Kay (1969) which is shown in (3).

(3)	black/white > red > green > yellow > blue > brown

Objects to the right of the color red are usually marked with the accusative case, while objects to the left are marked with the absolutive. The constraint that ensures this marking will be called the RED CROSS constraint, because the marking crosses over at the color red. RED CROSS predicts that black objects will receive absolutive marking, which is confirmed in sentences such as ‘Bob hit the chimney-sweep.’

With RED CROSS outranking ANIMACY, many of the problematic data in the literature on Southern Quiznos can be explained, but unfortunately there are still problem cases such as (4), which have been brought up by Lucy (1991).

(4)	Tja-o	ham	trsk-ky
	Charles-NOM	eat.PRS	tomato-ACC
	‘Charles eats the tomato.’

Example (4)

“ ‘Children, don’t speak so coarsely,’ said Mr. Webster, who had a vague notion that some supervision should be exercised over his daughters’ speech, and that a line should be drawn, but never knew quite when to draw it. He had allowed his daughters to use his library without restraint, and nothing is more fatal to maidenly delicacy of speech than the run of a good library.” —Robertson Davies

stands in direct contradiction to the present account, since ANIMACY appears to override RED CROSS. Is this perhaps a dialectal issue? A consultation of Lucy’s original field notes shows that the pattern is consistent across speakers of different varieties. However, the consultation also shows that the same speakers resort to the regular marking shown in (2) on other occasions. After checking the dates of the field notes, it became apparent that speakers were using the nominative-accusative pattern on Saturdays and Sundays, while the ergative-absolutive pattern was used between 9am and 5pm on regular working days. During coffee breaks and on national holidays, some speakers show marginal usage of nominative-accusative marking even at those times. We thus propose a third constraint, which will be labeled the WEEKEND constraint, which outranks both other constraints and thus enforces nominative marking on animate subjects in the spare time of speakers of Southern Quiznos.

The proposed account shows that optimality theory can straighten out even the worst things that languages come up with.

References

Berlin, B. & Kay, P. 1969. Basic Color Terms. Berkeley: University of California Press.

Bloomfield, L. 1924. Notes on Southern Quiznos. International Journal of American Linguistics 3.

Hockett, C. 1946. Quiznos revisited. Occasional papers in culinary linguistics 12 ¹/₂.

Lucy, J. 1991. Tomatoes. Shape and Color. In International Dictionary of Anthropologists. New York: Garland.

Mithun, M. 1984. The ketchup soaked the chocolate chip cookie. Language 60/4.

	How To Make A Linguistic Theory—Metalleus
	Book Review: Point’s A Grammar of the Lederhosen Tai—Enrich Barbarosa del la Boca, Ph.D
	SpecGram Vol CLII, No 1 Contents