Abstract
Grammars in Optimality Theory can be characterized by sets of Elementary Ranking Conditions (ERCs). Antimatroids are structures that arose initially in the study of lattices. In this paper we prove that antimatroids and consistent ERC sets have the same formal structures. We do so by defining two functions Antimat and RCErc, Antimat being a function from consistent sets of ERCs to antimatroids and RCErc a function from antimatroids to ERC sets. We then show that these functions are inverses of each other and that both maintain the structural properties of ERC sets and antimatroids. This establishes that antimatroids and consistent ERC sets have the same formal structure, allowing linguists to import from the sizable work done on antimatroids any and all results.
Similar content being viewed by others
Notes
A lattice is a partially ordered set in which every pair of elements has a meet and a join (a greatest lower bound and least upper bound).
Note that here we define MChain as a function from a total order to a subset of the power set lattice on the constraints, even though our ultimate goal is to define MChain from consistent ERC sets to a subset of the power set lattice. We will in subsequent sections use this total order MChain to define our ultimate ERC set MChain.
As Gaja Jarosz has pointed out (p.c.) this encoding may include other total orders (cf. the three total orders abcd, abdc, and bacd—when embedded in the power set lattice, the total order badc is also encoded). It is only when the orders are ERC-set representable that the encoding is conservative; that is, no other orders are included in their power-set representation. We will show this in latter sections of this paper.
An ERC with only one W and multiple Ls corresponds to a set of simple ERCs. E.g. \(\langle\mathrm{W}, \mathrm{e}, \mathrm{L}, \mathrm{L}\rangle = \{\langle \mathrm{W}, \mathrm{e}, \mathrm{L}, \mathrm{e}\rangle, \langle\mathrm{W},\mathrm{e}, \mathrm{e}, \mathrm{L}\rangle\}\).
This definition differs slightly from the literature (having the property of augmentation) but is equivalent in the systems investigated here.
It should also be obvious that if given any set of ERCs \(E\) then by defining the constraint set to be the ground set \(G\), and MChain(\(E\)) to be \(F\) defines a set system—this immediacy comes from the triviality of the definition of a set system, any subset of the power set, not to be equated with the more structured accessible set system.
There are numerous equivalent definitions of antimatroids. See Dietrich (1987) for a number of variants.
This antimatroid is equivalent to \(\mathit{MChain}(\langle\mathrm{W},\mathrm{e},\mathrm{e}, \mathrm{L}\rangle)\) depicted in Fig. above.
Recall that the set of constraints marked with a W in ERC \(\alpha\) is denoted \(\mathrm{W}(\alpha)\), the set of constraints marked L is \(\mathrm{L}(\alpha)\), and those marked e, \(\mathrm{e}(\alpha)\).
This antimatroid is equivalent to \(\mathit{MChain}(\langle\mathrm{W},\mathrm{W},\mathrm{L},\mathrm{L}\rangle)\).
Freedom and rootedness do not exhaust the outcomes of trace production—a trace can be neither free nor a rooted circuit—see \(\{a, b, d\}\) with \(\mathit{Antimat}(\mathrm{W}, \mathrm{e}, \mathrm{e}, \mathrm{L}\rangle)\).
References
Brasoveanu, Adrian, and Alan Prince. 2011. Ranking and necessity: the Fusional Reduction Algorithm. Natural Language and Linguistic Theory 29: 3–70.
Dietrich, Brenda. 1987. A circuit set characterization of antimatroids. Journal of Combinatorial Theory. Series B 43: 314–321.
Dietrich, Brenda. 1989. Matroids and antimatroids—a survey. Discrete Mathematics 78: 223–237.
Dilworth, Robert. 1940. Lattices with unique irreducible decompositions. Annals of Mathematics 41: 771–777.
Doignon, Jean-Paul, and Jean-Claude Falmagne. 1999. Knowledge spaces. Berlin: Springer.
Korte, Bernhard, and László Lovász. 1984. Greedoids, a structural framework for the greedy algorithm. In Progress in combinatorial optimization, The Silver Jubilee Conference on Combinatorics, ed. W. R. Pulleyblank. Waterloo, 1982, 221–243. New York: Academic Press.
Korte, Bernhard, László Lovász, and Rainer Schrader. 1991. Greedoids, 19–43. Berlin: Springer.
McCarthy, John. 2008. Doing optimality theory: Applying theory to data. Oxford: Blackwell Publishing.
Merchant, Nazarré. 2008. Discovering underlying forms: Contrast pairs and ranking. Ph.D. Dissertation, Rutgers University, New Brunswick, NJ.
Merchant, Nazarré. 2011. Learning ranking information from unspecified overt forms using the join. In Chicago Linguistics Society (CLS), Vol. 44. Chicago: Chicago Linguistics Society.
Merchant, Nazarré, and Bruce Tesar. 2005/2008. Learning underlying forms by searching restricted lexical subspaces. In Chicago Linguistics Society, Vol. 41, 33–47. Chicago: Chicago Linguistics Society.
Monjardet, Bernard. 1985. A use for frequently rediscovering a concept. Order 1: 415–417.
Parmar, Aarati. 2003. Some mathematical structures underlying efficient planning. In AAAI Spring Symposium on Logical Formalization of Commonsense Reasoning.
Prince, Alan. 2002. Entailed ranking arguments. ROA-500. http://roa.rutgers.edu/.
Prince, Alan. 2015, forthcoming. On the equivalence of antimatroids and ERC sets. Ms. Rutgers University.
Prince, Alan, and Paul Smolensky. 1993/2004. Optimality theory: Constraint interaction in generative grammar. Oxford: Blackwell Publishers. Technical Report, Rutgers University Center for Cognitive Science and Computer Science Department, University of Colorado at Boulder (1993).
Riggle, Jason. 2009. The complexity of ranking hypotheses in optimality theory. Computational Linguistics 35(1): 47–59.
Riggle, Jason, Max Bane, James Kirby, and John Sylak. 2008. Multilingual learning with parameter co-occurrence clustering. The North East Linguistic Society (NELS) 39.
Tesar, Bruce, and Paul Smolensky. 1998. Learnability in optimality theory. Linguistic Inquiry 29: 229–268.
Tesar, Bruce, and Paul Smolensky. 2000. Learnability in optimality theory. Cambridge: MIT Press.
Acknowledgements
Much thanks goes to Alan Prince for regular discussions about this paper, discussions which greatly facilitated its completion. We are also appreciative of comments by Gaja Jarosz and three anonymous reviewers who suggested numerous improvements. The provenance of errors and miscommunications in the paper falls solely at the feet of its two authors.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Merchant, N., Riggle, J. OT grammars, beyond partial orders: ERC sets and antimatroids. Nat Lang Linguist Theory 34, 241–269 (2016). https://doi.org/10.1007/s11049-015-9297-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11049-015-9297-5