Datasets:

WINGNUS
/

ACL-OCL

	{
	"paper_id": "J13-3005",
	"header": {
	"generated_with": "S2ORC 1.0.0",
	"date_generated": "2023-01-19T02:18:20.319891Z"
	},
	"title": "XMG: eXtensible MetaGrammar",
	"authors": [
	{
	"first": "Beno\u00eet",
	"middle": [],
	"last": "Crabb\u00e9",
	"suffix": "",
	"affiliation": {
	"laboratory": "",
	"institution": "Universit\u00e9",
	"location": {
	"addrLine": "Paris Diderot-Paris 7, Case 7003, 2, F",
	"postCode": "75205",
	"settlement": "Paris Cedex 13",
	"country": "France"
	}
	},
	"email": "bcrabbe@linguist.jussieu.fr."
	},
	{
	"first": "Denys",
	"middle": [],
	"last": "Duchier",
	"suffix": "",
	"affiliation": {
	"laboratory": "",
	"institution": "Universit\u00e9",
	"location": {
	"addrLine": "Paris Diderot-Paris 7, Case 7003, 2, F",
	"postCode": "75205",
	"settlement": "Paris Cedex 13",
	"country": "France"
	}
	},
	"email": "denys.duchier@univ-orleans.fr."
	},
	{
	"first": "Claire",
	"middle": [],
	"last": "Gardent",
	"suffix": "",
	"affiliation": {
	"laboratory": "",
	"institution": "Universit\u00e9",
	"location": {
	"addrLine": "Paris Diderot-Paris 7, Case 7003, 2, F",
	"postCode": "75205",
	"settlement": "Paris Cedex 13",
	"country": "France"
	}
	},
	"email": "gardent@loria.fr.\u2021"
	},
	{
	"first": "Joseph",
	"middle": [],
	"last": "Le Roux",
	"suffix": "",
	"affiliation": {},
	"email": "leroux@univ-paris13.fr.\u00a7"
	}
	],
	"year": "",
	"venue": null,
	"identifiers": {},
	"abstract": "In this article, we introduce eXtensible MetaGrammar (XMG), a framework for specifying tree-based grammars such as Feature-Based Lexicalized Tree-Adjoining Grammars (FB-LTAG) and Interaction Grammars (IG). We argue that XMG displays three features that facilitate both grammar writing and a fast prototyping of tree-based grammars. Firstly, XMG is fully declarative. For instance, it permits a declarative treatment of diathesis that markedly departs from the procedural lexical rules often used to specify tree-based grammars. Secondly, the XMG language has a high notational expressivity in that it supports multiple linguistic dimensions, inheritance, and a sophisticated treatment of identifiers. Thirdly, XMG is extensible in that its computational architecture facilitates the extension to other linguistic formalisms. We explain how this architecture naturally supports the design of three linguistic formalisms, namely, FB-LTAG, IG, and Multi-Component Tree-Adjoining Grammar (MC-TAG). We further show how it permits a straightforward integration of additional mechanisms such as linguistic and formal principles. To further illustrate the declarativity, notational expressivity, and extensibility of XMG, we describe the methodology used to specify an FB-LTAG for French augmented with a",
	"pdf_parse": {
	"paper_id": "J13-3005",
	"_pdf_hash": "",
	"abstract": [
	{
	"text": "In this article, we introduce eXtensible MetaGrammar (XMG), a framework for specifying tree-based grammars such as Feature-Based Lexicalized Tree-Adjoining Grammars (FB-LTAG) and Interaction Grammars (IG). We argue that XMG displays three features that facilitate both grammar writing and a fast prototyping of tree-based grammars. Firstly, XMG is fully declarative. For instance, it permits a declarative treatment of diathesis that markedly departs from the procedural lexical rules often used to specify tree-based grammars. Secondly, the XMG language has a high notational expressivity in that it supports multiple linguistic dimensions, inheritance, and a sophisticated treatment of identifiers. Thirdly, XMG is extensible in that its computational architecture facilitates the extension to other linguistic formalisms. We explain how this architecture naturally supports the design of three linguistic formalisms, namely, FB-LTAG, IG, and Multi-Component Tree-Adjoining Grammar (MC-TAG). We further show how it permits a straightforward integration of additional mechanisms such as linguistic and formal principles. To further illustrate the declarativity, notational expressivity, and extensibility of XMG, we describe the methodology used to specify an FB-LTAG for French augmented with a",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Abstract",
	"sec_num": null
	}
	],
	"body_text": [
	{
	"text": "In the late 1980s and early 1990s, many grammar engineering environments were developed to support the specification of large computational grammars for natural language. One may, for instance, cite XLE (Kaplan and Newman 1997) for specifying Lexical-Functional Grammars (LFG), LKB (Copestake and Flickinger 2000) for specifying Head-driven Phrase Structure Grammars (HPSG), and DOTCCG (Baldridge et al. 2007) for specifying Combinatory Categorial Grammars (CCG). Concretely, such environments usually rely on (i) a formal language used to describe a target computational grammar, and (ii) a processor for this language, which aims at generating the actual described grammar (and potentially at checking it, e.g., by feeding it to a parser).",
	"cite_spans": [
	{
	"start": 199,
	"end": 202,
	"text": "XLE",
	"ref_id": null
	},
	{
	"start": 203,
	"end": 227,
	"text": "(Kaplan and Newman 1997)",
	"ref_id": "BIBREF36"
	},
	{
	"start": 278,
	"end": 281,
	"text": "LKB",
	"ref_id": null
	},
	{
	"start": 282,
	"end": 313,
	"text": "(Copestake and Flickinger 2000)",
	"ref_id": "BIBREF18"
	},
	{
	"start": 386,
	"end": 409,
	"text": "(Baldridge et al. 2007)",
	"ref_id": "BIBREF4"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "Although these environments were tailored for specific grammar formalisms, they share a number of features. Firstly, they are expressive enough to characterize subsets of natural language. Following Shieber (1984) , we call this feature weak completeness. Secondly, they are notationally expressive enough to relatively easily formalize important theoretical notions. Thirdly, they are rigorous, that is, the semantics of their underlying language is well defined and understood. Additionally, for an environment to be useful in practice, it should be simple to use (by a linguist), and make it possible to detect errors in the described target grammar.",
	"cite_spans": [
	{
	"start": 199,
	"end": 213,
	"text": "Shieber (1984)",
	"ref_id": "BIBREF49"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "If we consider a particular type of computational grammar, namely, tree-based grammars-that is, grammars where the basic units are trees (or tree descriptions) of arbitrary depth, such as Tree-Adjoining Grammar (TAG; Joshi, Levy, and Takahashi 1975) , D-Tree Grammar (DTG; Rambow, Vijay-Shanker, and Weir 1995) , Tree Description Grammars (TDG; Kallmeyer 1999) or Interaction Grammars (IG; Perrier 2000)environments sharing all of the listed features are lacking. As we shall see in Section 7 of this article, there have been some proposals for grammar engineering environments for tree-based grammar (e.g., Candito 1996; Xia, Palmer, and Vijay-Shanker 1999 , but these lack notational expressivity. This is partly due to the fact that tree-based formalisms offer an extended domain of locality where one can encode constraints between remote syntactic constituents. If one wants to define such constraints while giving a modular and incremental specification of the grammar, one needs a high level of notational expressivity, as we shall see throughout the article (and especially in Section 4).",
	"cite_spans": [
	{
	"start": 217,
	"end": 249,
	"text": "Joshi, Levy, and Takahashi 1975)",
	"ref_id": "BIBREF30"
	},
	{
	"start": 273,
	"end": 310,
	"text": "Rambow, Vijay-Shanker, and Weir 1995)",
	"ref_id": "BIBREF47"
	},
	{
	"start": 345,
	"end": 360,
	"text": "Kallmeyer 1999)",
	"ref_id": "BIBREF31"
	},
	{
	"start": 608,
	"end": 621,
	"text": "Candito 1996;",
	"ref_id": "BIBREF12"
	},
	{
	"start": 622,
	"end": 657,
	"text": "Xia, Palmer, and Vijay-Shanker 1999",
	"ref_id": "BIBREF52"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "In this article, we present XMG (eXtensible MetaGrammar), a framework for specifying tree-based grammars. Focusing mostly on Feature-Based Lexicalized Tree-Adjoining Grammars (FB-LTAG) (but using Interaction Grammars [IG] and Multi-Component Tree-Adjoining Grammars [MC-TAG] to illustrate flexibility), we argue that XMG departs from other existing computational frameworks for designing tree-based grammars in three main ways:",
	"cite_spans": [
	{
	"start": 217,
	"end": 221,
	"text": "[IG]",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "r First, XMG is a declarative language. In other words, grammaticality is defined in an order-independent fashion by a set of well-formedness constraints rather than by procedures. In particular, XMG permits a fully declarative treatment of diathesis that markedly departs from the procedural rules (called meta-rules or lexical rules) previously used to specify tree-based grammars.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "r Second, XMG is notationally expressive. The XMG language supports full disjunction and conjunction of grammatical units, a modular treatment of multiple linguistic dimensions, multiple inheritance of units, and a sophisticated treatment of identifiers. We illustrate XMG's notational expressivity by showing (i) how it facilitates the modeling of the tree fragment hierarchies required to specify tree-based grammars and (ii) how it permits a natural modeling of the syntax/semantics interface between semantic representations and syntactic trees as can be used in FB-LTAG.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "r Third, XMG is extensible in that its computational architecture facilitates (i) the integration of an arbitrary number of linguistic dimensions (syntax, semantics, etc.), (ii) the modeling of different grammar formalisms (FB-LTAG, MC-TAG, IG), and (iii) the specification of general linguistic principles (e.g., clitic ordering in French).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "The article is structured as follows. Section 2 starts by giving a brief introduction to FB-LTAG, the grammar formalism we used to illustrate most of XMG's features. The next three sections then go on to discuss and illustrate XMG's three main featuresnamely, declarativity, notational expressivity, and flexibility. In Section 3, we focus on declarativity and show how XMG's generalized disjunction permits a declarative encoding of diathesis. We then contrast the XMG approach with the procedural methods previously resorted to for specifying FB-LTAG. Section 4 addresses notational expressivity. We present the syntax of XMG and show how the sophisticated identifier handling it supports or permits a natural treatment (i) of identifiers in tree based hierarchies and (ii) of the unification-based syntax/semantics interface often used in FB-LTAG. In Section 5, we concentrate on extensibility. We first describe the operational semantics of XMG and the architecture of the XMG compiler. We then show how these facilitate the adaptation of the basic XMG language to (i) different grammar formalisms (IG, MC-TAG, FB-LTAG), (ii) the integration of specific linguistic principles such as clitic ordering constraints, and (iii) the specification of an arbitrary number of linguistic dimensions. In Section 6, we illustrate the usage of XMG by presenting an XMG specification for the verbal fragment of a large scale FB-LTAG for French augmented with a unification-based semantics. We also briefly describe the various other treebased grammars implemented using XMG. Section 7 discusses the limitations of other approaches to the formal specification of tree-based grammars, and Section 8 concludes with pointers for further research.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Introduction",
	"sec_num": "1."
	},
	{
	"text": "A Tree-Adjoining Grammar (TAG) consists of a set of auxiliary or initial elementary trees and of two tree composition operations, namely, substitution and adjunction. Initial trees are trees whose leaves are either substitution nodes (marked with \u2193) or terminal symbols (words). Auxiliary trees are distinguished by a foot node (marked with ) whose category must be the same as that of the root node. Substitution inserts a tree onto a substitution node of some other tree and adjunction inserts an auxiliary tree into a tree. Figure 1 shows a toy TAG generating the sentence Marie a vu Jean 'Mary has seen John' and sketches its derivation. 1 Among existing variants of TAG, one commonly used in practice is Lexicalized FB-LTAG (Vijay-Shanker and Joshi 1988) . A lexicalized TAG is such that each elementary tree has at least one leaf labeled with a lexical item (word), whereas in an FB-LTAG, tree nodes are additionally decorated with two feature structures (called top and bottom). These feature structures are unified during derivation as follows. On substitution, the top features of the substitution node are unified with the top features of the root node of the tree being substituted in. On adjunction, the top features of the root of the auxiliary tree are unified with the top features of the node where adjunction takes place; and the bottom features of the foot node of the auxiliary tree are unified with the bottom features of the node where adjunction takes place. At the end of a derivation, the top and bottom feature structures of all nodes in the derived tree are unified.",
	"cite_spans": [
	{
	"start": 642,
	"end": 643,
	"text": "1",
	"ref_id": null
	},
	{
	"start": 729,
	"end": 759,
	"text": "(Vijay-Shanker and Joshi 1988)",
	"ref_id": "BIBREF50"
	}
	],
	"ref_spans": [
	{
	"start": 527,
	"end": 535,
	"text": "Figure 1",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Tree-Adjoining Grammar",
	"sec_num": "2."
	},
	{
	"text": "Implementation of Tree-Adjoining Grammars. Most existing implementations of TAGs follow the three-layer architecture adopted for the XTAG grammar (XTAG Research Group 2001), a feature-based lexicalized TAG for English. Thus the grammar consists of (i) a set of so-called tree schemas (i.e., elementary trees having a leaf node labeled with a referring to where to anchor lexical items 2 ), (ii) a morphological lexicon associating words with lemmas, and (iii) a syntactic lexicon associating lemmas with tree schemas (these are gathered into families according to syntactic properties, such as the subcategorization frame for verbs). Figure 2 shows some of the tree schemas associated with transitive verbs in the XTAG grammar. The tree corresponds (a) to a declarative sentence, (b) to a WH-question on the subject, (c) to a passive clause with a BY-agent, and (d) to a passive clause with a WH-object. As can be seen, each tree schema contains an anchor node (marked with ). During parsing this anchor node can be replaced by any word morphologically related to a lemma listed in the syntactic lexicon as anchoring the transitive tree family.",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 634,
	"end": 642,
	"text": "Figure 2",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Tree-Adjoining Grammar",
	"sec_num": "2."
	},
	{
	"text": "This concept of tree family allows us to share structural information (tree schemas) between words having common syntactic properties (e.g., sub-categorization frames). There still remains a large redundancy within the grammar because many elementary tree schemas share common subtrees (large coverage TAGs usually consist of hundreds, sometimes thousands, of tree schemas). An important issue when specifying",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Tree-Adjoining Grammar",
	"sec_num": "2."
	},
	{
	"text": "(a) (b) S r NP 0 \u2193 VP V NP 1 \u2193 S q NP 0 \u2193 wh + [] S r NP NA VP V NP 1 \u2193 (c) (d) S r [] mode 3 NP 1 \u2193 VP mode 3 \u23a1 \u23a3 passive 1 mode 2 \u23a4 \u23a6 V \u23a1 \u23a3 passive 1 + mode 2 ppart \u23a4 \u23a6 [] PP P by NP 0 \u2193 S q NP 1 \u2193 wh + [] S r [] mode 3 NP NA VP mode 3 \u23a1 \u23a3 passive 1 mode 2 \u23a4 \u23a6 V \u23a1 \u23a3 passive 1 + mode 2 ppart \u23a4 \u23a6 [] PP P by NP 0 \u2193",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Tree-Adjoining Grammar",
	"sec_num": "2."
	},
	{
	"text": "Some tree schemas for English transitive verbs.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 2",
	"sec_num": null
	},
	{
	"text": "such grammars is thus structure sharing. Being able to share structural information is necessary not only for a faster grammar development, but also for an easier grammar maintenance (modifications to be applied to the tree schemas would be restricted to shared structures). In the next section, we will see how XMG declarativity can be efficiently used to factorize TAGs. In addition, Section 4 will show how XMG notational expressivity facilitates the specification of another commonly used tree sharing device, namely, inheritance hierarchies of tree fragments.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 2",
	"sec_num": null
	},
	{
	"text": "Extending TAG with a Unification-Based Semantics. To extend FB-LTAG with a compositional semantics, Gardent and Kallmeyer (2003) propose to associate each elementary tree with a flat semantic representation. For instance, in Figure 3 , the trees 3 for John, runs, and often are associated with the semantics l 0 :name(j,john), l 1 :run(e,s), and l 2 :often(x), respectively. Importantly, the arguments of semantic functors are represented by unification variables which occur both in the semantic representation of this functor and on some nodes of the associated syntactic tree. Thus in Figure 3 , the semantic index s occurring in the semantic representation of runs also occurs on the subject substitution node of the associated elementary tree. The value of semantic arguments is then determined by the unifications resulting from adjunction and substitution. For instance, the semantic index s in the tree for runs is unified during substitution with the semantic index j labeling the root node of the tree for John. As a result, the semantics of John often runs is {l 0 :name(j,john), l 1 :run(e,j), l 2 :often(e)}. Gardent and Kallmeyer's (2003) proposal was applied to various semantic phenomena (Kallmeyer and Romero 2004a , 2004b , 2008 . Its implementation, however,",
	"cite_spans": [
	{
	"start": 100,
	"end": 128,
	"text": "Gardent and Kallmeyer (2003)",
	"ref_id": "BIBREF27"
	},
	{
	"start": 1122,
	"end": 1152,
	"text": "Gardent and Kallmeyer's (2003)",
	"ref_id": "BIBREF27"
	},
	{
	"start": 1204,
	"end": 1231,
	"text": "(Kallmeyer and Romero 2004a",
	"ref_id": "BIBREF35"
	},
	{
	"start": 1232,
	"end": 1239,
	"text": ", 2004b",
	"ref_id": "BIBREF35"
	},
	{
	"start": 1240,
	"end": 1246,
	"text": ", 2008",
	"ref_id": null
	}
	],
	"ref_spans": [
	{
	"start": 225,
	"end": 233,
	"text": "Figure 3",
	"ref_id": null
	},
	{
	"start": 588,
	"end": 596,
	"text": "Figure 3",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Figure 2",
	"sec_num": null
	},
	{
	"text": "NP j John l 0 :name(j,john) S g NP\u2193 s VP g f V f e runs l 1 :run(e,s) VP x often VP* x l 2 :often(x)",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 2",
	"sec_num": null
	},
	{
	"text": "\u21d2 l 0 :name(j,john), l 1 :run(e,j), l 2 :often(e)",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 2",
	"sec_num": null
	},
	{
	"text": "A toy lexicalized FTAG with unification-based semantics (l 0 , l 1 , l 2 , e, and j are constants and s, f, g, x are unification variables).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 3",
	"sec_num": null
	},
	{
	"text": "relies on having a computational framework that associates syntactic trees with flat semantic formulae while allowing for shared variables between trees and formulae. In the following sections, we will show how XMG notational expressivity makes it possible to specify an FB-LTAG equipped with a unification-based semantics.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 3",
	"sec_num": null
	},
	{
	"text": "In this section, we show how a phenomenon which is often handled in a procedural way by existing approaches can be provided with a declarative specification in XMG. Concretely, we show how XMG supports a declarative account of diathesis that avoids the drawbacks of lexical rules (e.g., information erasing). We start by presenting the lexical rule approach. We then contrast it with the XMG account.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Declarativity",
	"sec_num": "3."
	},
	{
	"text": "Following Flickinger (1987) , redundancy among grammatical descriptions is often handled using two devices: an inheritance hierarchy and a set of lexical rules. Whereas the inheritance hierarchy permits us to encode the sharing of common substructures, lexical rules (sometimes called meta-rules) permit us to capture relationships between trees by deriving new trees from already specified ones. For instance, passive trees will be derived from active ones. Although Flickinger's (1987) approach was developed for HPSGs, several similar approaches have been put forward for FB-LTAG (Vijay-Shanker and Schabes 1992; Becker 1993; Evans, Gazdar, and Weir 1995; XTAG Research Group 2001) . One important drawback of these approaches, however, is that they are procedural in that the order in which lexical rules apply matters. For instance, consider again the set of trees given in Figure 2 . In the meta-rule representation scheme adopted by Becker (1993) , the base tree (a) would be specified in the inheritance hierarchy grouping all base trees, and the derived trees (b, c, d) would be generated by applying one or more meta-rules on this base tree. Figure 4 sketches these meta-rules. The left-hand side of the meta-rule is a matching pattern replaced with the right-hand side of the meta-rule in the newly generated tree. Symbol \"?\" denotes a meta-variable whose matching subtree in the input is substituted in place of the variable in the output tree. Given these, the tree family in Figure 2 is generated as follows: (b) and (c) are generated by application to the base tree (a) of the Wh-Subject and Passive meta-rules, respectively. Further, (d) is generated by applying first, the Wh-Subject meta-rule and second, the Passive meta-rule to the base tree. ",
	"cite_spans": [
	{
	"start": 10,
	"end": 27,
	"text": "Flickinger (1987)",
	"ref_id": "BIBREF25"
	},
	{
	"start": 468,
	"end": 487,
	"text": "Flickinger's (1987)",
	"ref_id": "BIBREF25"
	},
	{
	"start": 583,
	"end": 615,
	"text": "(Vijay-Shanker and Schabes 1992;",
	"ref_id": "BIBREF50"
	},
	{
	"start": 616,
	"end": 628,
	"text": "Becker 1993;",
	"ref_id": "BIBREF7"
	},
	{
	"start": 629,
	"end": 658,
	"text": "Evans, Gazdar, and Weir 1995;",
	"ref_id": null
	},
	{
	"start": 659,
	"end": 684,
	"text": "XTAG Research Group 2001)",
	"ref_id": null
	},
	{
	"start": 940,
	"end": 953,
	"text": "Becker (1993)",
	"ref_id": "BIBREF7"
	}
	],
	"ref_spans": [
	{
	"start": 879,
	"end": 887,
	"text": "Figure 2",
	"ref_id": null
	},
	{
	"start": 1152,
	"end": 1160,
	"text": "Figure 4",
	"ref_id": null
	},
	{
	"start": 1489,
	"end": 1497,
	"text": "Figure 2",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Capturing Diathesis Using Lexical Rules",
	"sec_num": "3.1"
	},
	{
	"text": "Simplified meta-rules for passive and wh-subject extraction.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 4",
	"sec_num": null
	},
	{
	"text": "More generally a meta-rule is a procedural device that, given a tree instance, generates a new tree instance by adding, suppressing (hence possibly substituting) information in grammatical units. Prolo (2002) defines a set of meta-rules that can be used to specify a large FB-LTAG for English. Given an ordered set of meta-rules, however, there is no guarantee that the trees they derive are linguistically appropriate and that the derivation process terminates. Thus, to ensure termination and consistency, Prolo needs to additionally provide rule ordering schemes (expressed as automata).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 4",
	"sec_num": null
	},
	{
	"text": "XMG provides an alternative account for describing tree sets such as that of Figure 2 without lexical rules and without the related ordering constraints. In essence, the approach consists of enumerating trees by combining tree fragments using conjunction and disjunction.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Capturing Diathesis Using Disjunction",
	"sec_num": "3.2"
	},
	{
	"text": "More specifically, the tree set given in Figure 2 can be generated by combining some of the tree fragments sketched in Figure 5 using the following conjunctions and disjunctions: 4",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 41,
	"end": 49,
	"text": "Figure 2",
	"ref_id": null
	},
	{
	"start": 119,
	"end": 127,
	"text": "Figure 5",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "XMG: Capturing Diathesis Using Disjunction",
	"sec_num": "3.2"
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "\u2192 CanonicalSubject \u2228 Wh-NP-Subject (1) ActiveTransitiveVerb \u2192 Subject \u2227 ActiveVerb \u2227 CanonicalObject (2) PassiveTransitiveVerb \u2192 Subject \u2227 PassiveVerb \u2227 CanonicalByObject (3) TransitiveVerb \u2192 ActiveTransitiveVerb \u2228 PassiveTransitiveVerb",
	"eq_num": "(4)"
	}
	],
	"section": "Subject",
	"sec_num": null
	},
	{
	"text": "The first clause (Subject) groups together two subtrees representing the possible realizations of a subject (canonical and wh). The next two clauses define a tree set for active and passive transitive verbs, respectively. The last clause defines the TransitiveVerb family as a disjunction of the two verb forms (passive or active). In sum, the TransitiveVerb clause defines the tree set sketched in Figure 2 as a disjunction of conjunctions of tree fragments.",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 399,
	"end": 407,
	"text": "Figure 2",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Subject",
	"sec_num": null
	},
	{
	"text": "One of the issues of meta-rules reported by Prolo (2002) is the handling of feature equations. For a number of cases (including subject relativization in passive trees),",
	"cite_spans": [
	{
	"start": 44,
	"end": 56,
	"text": "Prolo (2002)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Subject",
	"sec_num": null
	},
	{
	"text": "Canonical Subject \u2192 Wh-NP-Subject \u2192 Canonical Object \u2192 Wh-NP-Object \u2192 S r NP\u2193 VP S q NP\u2193 wh + [] S r NP NA VP VP V NP\u2193 S q NP\u2193 wh + [] S r VP NP NA Canonical By Object \u2192 Wh By Object \u2192 Active Verb \u2192 Passive Verb \u2192 VP V PP P by NP\u2193 VP PP P by NP\u2193 V S r VP V S r [] mode 3 VP mode 3 \u23a1 \u23a3 passive 1 mode 2 \u23a4 \u23a6 V \u23a1 \u23a3 passive 1 + mode 2 ppart \u23a4 \u23a6 [] Figure 5",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Subject",
	"sec_num": null
	},
	{
	"text": "Tree fragments.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Subject",
	"sec_num": null
	},
	{
	"text": "ad hoc meta-rules are needed, for a unified tree transformation cannot be defined. In a declarative approach such as the one here, dealing with feature equations can be done relatively easily. Let us imagine that we now want to extend the trees of Figure 2 with feature equations for subject-number agreement. We can for instance do so by defining the following tree fragment (the dashed line indicates that the VP node can be a descendant, not only a daughter, of the S node): 5",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 248,
	"end": 256,
	"text": "Figure 2",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Subject",
	"sec_num": null
	},
	{
	"text": "\u2192 S NP\u2193 num 1 num 1 VP num 1 num 1",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SubjAgreement",
	"sec_num": null
	},
	{
	"text": "Then we extend the definition of Subject as follows:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SubjAgreement",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "Subject \u2192 SubjAgreement \u2227 ( CanonicalSubject \u2228 Wh-NP-Subject )",
	"eq_num": "(5)"
	}
	],
	"section": "SubjAgreement",
	"sec_num": null
	},
	{
	"text": "If we want to get further with the description of transitive verbs, for instance by taking into account wh-objects and by-objects, this can be done as follows. We first define the elementary fragments Wh-NP-Object and Wh-By-Object (see Figure 5 ), and then define the following additional combinations: 6",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 236,
	"end": 244,
	"text": "Figure 5",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "SubjAgreement",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "\u2192 CanonicalSubject \u2227 ActiveVerb \u2227 Wh-Np-Object (6) PassiveTransitiveVerb \u2192 CanonicalSubject \u2227 PassiveVerb \u2227 Wh-By-Object",
	"eq_num": "(7)"
	}
	],
	"section": "ActiveTransitiveVerb",
	"sec_num": null
	},
	{
	"text": "Evans, Gazdar, and Weir (1995) argue for the necessity of using lexical rules for grammatical description based on two arguments: (i) morphology is irregular and has to be handled by a non-monotonic device and (ii) erasing rules such as the agentless passive (John eats an apple / An apple is eaten ) are needed to erase an argument from the canonical base tree. Neither of these arguments holds here, however: The first argument because we describe tree schema hence lexical and morphological issues are ruled out; the second because agentless passive and, more generally, argument erasing constructions can simply be defined by an additional clause such as:",
	"cite_spans": [
	{
	"start": 7,
	"end": 30,
	"text": "Gazdar, and Weir (1995)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "ActiveTransitiveVerb",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "AgentlessPassiveTransitiveVerb \u2192 Subject \u2227 PassiveVerb",
	"eq_num": "(8)"
	}
	],
	"section": "ActiveTransitiveVerb",
	"sec_num": null
	},
	{
	"text": "To summarize, using a declarative language to specify a tree-based grammar offers an adequate level of control on the structures being described while avoiding having to deal with ordering and termination issues. It facilitates grammar design and maintenance, by providing an abstract view on grammar trees, uniquely made of monotonic (no information removal) combinations of tree fragments.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "ActiveTransitiveVerb",
	"sec_num": null
	},
	{
	"text": "We now focus on notational expressivity and show how XMG supports a direct encoding of (i) distinct linguistic dimensions (here syntax, semantics and the syntax/ semantics interface) and (ii) the various types of coreferences 7 that arise in the development of tree-based grammars.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Notational Expressivity",
	"sec_num": "4."
	},
	{
	"text": "The syntax of the XMG language can be formally defined as follows.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Notational Expressivity",
	"sec_num": "4."
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "Class ::= Name C 1 ,...,C k x 1 ,...,x n \u2192 Content (9) Content ::= SYN, SEM, DYN \| Name \| Content \u2227 Content \| Content \u2228 Content (10) SYN ::= n 1 \u2192 n 2 \| n 1 \u2192 + n 2 \| n 1 \u2192 * n 2 \| n 1 \u227a n 2 \| n 1 \u227a + n 2 \| n 1 \u227a * n 2 \| n 1 [f 1 : v 1 ,..., f k : v k ] \| n 1 (c 1 : cv 1 ,..., c l : cv l ) \| n 1 = n 2 \| x = C i .y \| n 1 (c 1 : cv 1 ,..., c l : cv l ) [f 1 : v 1 ,..., f k : v k ] \| SYN \u2227 SYN (11) SEM ::= l i : p(E 1 ,...,E n ) \| l i \u2264 h j \| SEM \u2227 SEM (12) DYN ::= f 1 : v 1 ,...,f n : v n",
	"eq_num": "(13)"
	}
	],
	"section": "Notational Expressivity",
	"sec_num": "4."
	},
	{
	"text": "Here and in what follows, we use the following notational conventions. C i denote variables over class names; x i , x, and y are variables ranging over tree nodes or feature values; n i refer to node variables; f, f i are features and v, v i and feature values (constants or variables); l i , h j , p, and E i are variables over semantic labels, semantic holes, predicates, and predicate arguments in flat semantic formulae, respectively. 8 [ ] are used to associate a node variable with some feature constraint. ( ) are used to associate a node variable with some property constraint (e.g., node colors, see Section 5). c i and cv i denote a property constraint and a property constraint value, respectively. C i .y denotes the y variable declared in class C i and = is unification; \u227a and \u2192 denote linear precedence and immediate dominance relations between nodes. Finally, +, * represent the transitive and transitive-reflexive closure of a relation, respectively.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Notational Expressivity",
	"sec_num": "4."
	},
	{
	"text": "The first two clauses of the formal definition here specify XMG classes and how they combine. The next three clauses define the languages supported for describing three linguistic dimensions, namely, syntax (SYN), semantics (SEM), and the syntax/semantics interface (called DYN for dynamic interface). We now discuss each of these in more detail starting bottom-up with the three linguistic dimensions and ending with the control language that permits us to combine basic linguistic units into bigger ones.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Notational Expressivity",
	"sec_num": "4."
	},
	{
	"text": "The XMG formalism for syntax (copied here for convenience) is a tree description logic similar to that proposed by Vijay-Shanker and Schabes (1992) and Rogers and Vijay-Shanker (1994) to describe tree-based grammars.",
	"cite_spans": [
	{
	"start": 115,
	"end": 147,
	"text": "Vijay-Shanker and Schabes (1992)",
	"ref_id": "BIBREF50"
	},
	{
	"start": 152,
	"end": 183,
	"text": "Rogers and Vijay-Shanker (1994)",
	"ref_id": "BIBREF48"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SYN.",
	"sec_num": null
	},
	{
	"text": ":",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SYN",
	"sec_num": null
	},
	{
	"text": ":= n 1 \u2192 n 2 \| n 1 \u2192 + n 2 \| n 1 \u2192 * n 2 \| n 1 \u227a n 2 \| n 1 \u227a + n 2 \| n 1 \u227a * n 2 \| n 1 [f 1 : v 1 ,..., f k : v k ] \| n 1 (c 1 : cv 1 ,..., c l : cv l ) \| n 1 = n 2 \| x = C i .y \| n 1 (c 1 : cv 1 ,..., c l : cv l ) [f 1 : v 1 ,..., f k : v k ] \| SYN \u2227 SYN",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SYN",
	"sec_num": null
	},
	{
	"text": "It includes tree node variables, feature names, feature values, and feature variables. Tree node variables can be related by equality (node identification), precedence (immediate or non-immediate), and dominance (immediate or non-immediate). Tree nodes can also be labeled with feature structures of depth 2, that is, sets of feature/value pairs where feature values are either variables, constants (e.g., syntactic category), or non-recursive feature structure (e.g., top and bottom feature structures).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SYN",
	"sec_num": null
	},
	{
	"text": "Here is a graphical illustration of how tree logic formulae can be used to describe tree fragments: The depicted tree fragment is a model satisfying the given formula.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SYN",
	"sec_num": null
	},
	{
	"text": "n 1 \u2192 n 2 \u2227 n 1 \u2192 n 3 \u2227 n 2 \u227a n 3 \u2227 n 1 [cat : S] \u2227 n 2 (mark : subst) [cat : NP] \u2227 n 3 [cat : VP] S",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SYN",
	"sec_num": null
	},
	{
	"text": "One distinguishing feature of the XMG tree language is the introduction of node constraints (n 1 (c : cv)) that generalize Muskens and Krahmer's (1998) use of positive and negative node markings. Concretely, node constraints are attribute-value matrices, which contain information to be used when solving tree descriptions to produce grammar trees. In other words, node constraints are used to further restrict the set of models satisfying a tree description. As an example of node constraint, consider node annotations in FB-LTAG (foot node, substitution node, null-adjunction, etc.). Such annotations can be used as node constraints to allow the description solver to apply well-formedness constraints (e.g., there is at most one foot node).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "NP\u2193 VP",
	"sec_num": null
	},
	{
	"text": "Another interesting feature of XMG concerns the inclusion of the dot operator, which permits us to identify variables across classes in cases where name sharing cannot be resorted to. When a variable y is declared in a class C, the latter being instantiated within a class D, y can be accessed from D by C.y (the identifier y still being available in D's namespace).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "NP\u2193 VP",
	"sec_num": null
	},
	{
	"text": "The semantic dimension supports a direct encoding of the flat semantic formulae used by Gardent and Kallmeyer (2003) :",
	"cite_spans": [
	{
	"start": 88,
	"end": 116,
	"text": "Gardent and Kallmeyer (2003)",
	"ref_id": "BIBREF27"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "SEM ::= l i : p(E 1 ,...,E n ) \| l i \u2264 h j \| SEM \u2227 SEM where l i : p(E 1 ,..., E n )",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "represents a predicate p with label l i and arguments E 1 ,..., E n and l i \u2264 h j is a scope constraint between label l i and scope h j . Expressions (predicate arguments E i ) can refer to semantic holes, constants (atomic values), or unification variables (written x, y hereafter). For instance, the following flat semantic formula can be used to underspecify the meaning of the sentence \"Every dog chases a cat\":",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "l 0 : \u2200(x, h 1 , h 2 ) \u2227 l 1 \u2264 h 1 \u2227 l 1 : Dog(x) \u2227 l 2 \u2264 h 2 \u2227 l 2 : Chase(x, y) \u2227 l 3 : \u2203(y, h 3 , h 4 ) \u2227 l 4 \u2264 h 3 \u2227 l 4 : Cat(y) \u2227 l 2 \u2264 h 4",
	"eq_num": "(14)"
	}
	],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "This formula denotes the following two first-order logic formulae, thereby describing the two possibles readings of this sentence. 9",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "l 0 : \u2200(x, l 1 , l 3 ) \u2227 l 1 : Dog(x) \u2227 l 2 : Chase(x, y) \u2227 l 3 : \u2203(y, l 4 , l 2 ) \u2227 l 4 : Cat(y) (15) l 0 : \u2200(x, l 1 , l 2 ) \u2227 l 1 : Dog(x) \u2227 l 2 : Chase(x, y) \u2227 l 3 : \u2203(y, l 4 , l 0 ) \u2227 l 4 : Cat(y)",
	"eq_num": "(16)"
	}
	],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "DYN. The DYN dimension generalizes Kinyon's hypertag (Kinyon 2000) which is unified whenever two tree fragments are combined. Similarly, in XMG the DYN dimension is a feature structure that is unified whenever two XMG classes are combined through inheritance or through conjunction (see the discussion on XMG control language, subsequently). For instance, the following constraints ensure a coreference between the index I occurring in the syntactic dimension and the argument X occurring in the semantic dimension (index subject and arg 1 are feature names, and E, I, X, and V local unification variables).",
	"cite_spans": [
	{
	"start": 53,
	"end": 66,
	"text": "(Kinyon 2000)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "C 1 \u2192 Node [idx : I] \u2227 index subject : I (17) C 2 \u2192 L : P(E) \u2227 L : Theta 1 (E, X) \u2227 arg 1 : X (18) SubjectArg 1 \u2192 C 1 \u2227 C 2 \u2227 index subject : V, arg 1 : V",
	"eq_num": "(19)"
	}
	],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "More generally, the DYN dimension permits us to unify nodes and feature values that belong to distinct classes and dimensions, and are thus often not related within the inheritance hierarchy. As we shall see in Section 6, the DYN dimension permits a modular account of the syntax/semantics interface in which linking constraints can be stipulated separately and reused to specify the various diatheses. In other words, the DYN feature structure allows us to extend the scope of some specific variables so that they can be unified with variables (or values) introduced in some other classes of the metagrammar. This concept of scope extension can be compared with that of hook in Copestake, Lascarides, and Flickinger (2001) .",
	"cite_spans": [
	{
	"start": 679,
	"end": 723,
	"text": "Copestake, Lascarides, and Flickinger (2001)",
	"ref_id": "BIBREF19"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "Control language. The linguistic units (named Content here) defined by the linguist can be abstracted and combined as follows:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "Class ::= Name C 1 ,...,C k x 1 ,...,x n \u2192 Content Content ::= SYN, SEM, DYN \| Name \| Content \u2227 Content \| Content \u2228 Content",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "The first clause states that the linguistic information encoded in Content is abstracted in a class named Name and that this class inherits classes C 1 ,..., C k and exports variables x 1 ,..., x n . That is, XMG allows for abstraction, inheritance, and variable exports. By default, variables (referring to nodes and feature values) are local to a class. Export statements extend the scope of a variable to all sub-classes, however. An exported variable can also be accessed from outside its class in case of class instantiation (using the dot operator introduced earlier in this section). The second clause states that an XMG class consists of a syntactic, a semantic, and a dynamic description (each of them possibly empty), and that XMG classes can be combined by conjunction and disjunction and reused through class instantiation. The notation SYN, SEM, DYN represents simultaneous contributions (possibly empty) to all three dimensions. 10 The XMG control language differs from other frameworks used to specify treebased grammars (Vijay-Shanker and Schabes 1992; Xia et al. 1998; Candito 1999) in two main ways. First, it supports generalized conjunctions and disjunctions of classes. As shown in Section 3, this permits us, inter alia, a declarative treatment of diathesis.",
	"cite_spans": [
	{
	"start": 943,
	"end": 945,
	"text": "10",
	"ref_id": null
	},
	{
	"start": 1069,
	"end": 1085,
	"text": "Xia et al. 1998;",
	"ref_id": null
	},
	{
	"start": 1086,
	"end": 1099,
	"text": "Candito 1999)",
	"ref_id": "BIBREF13"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "Second, it allows for both local and exported variables. As mentioned in Section 3, a common way to share structure within a tree-based grammar is to define an inheritance hierarchy of either tree fragments (Evans, Gazdar, and Weir 1995) or tree descriptions (Vijay-Shanker and Schabes 1992; Candito 1996; Xia 2001) . When considering an FB-LTAG augmented with unification semantics, the hierarchy will additionally contain semantic representations and/or tuples made of tree fragments and semantic representations. In all cases, the question arises of how to handle identifiers across classes and, more specifically, how to share them.",
	"cite_spans": [
	{
	"start": 292,
	"end": 305,
	"text": "Candito 1996;",
	"ref_id": "BIBREF12"
	},
	{
	"start": 306,
	"end": 315,
	"text": "Xia 2001)",
	"ref_id": "BIBREF51"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "In Candito's (1996) approach, tree nodes are referred to using constants so that multiple occurrences of the same node constant refer to the same node. As pointed out in Gardent and Parmentier (2006) , global names have several non-trivial shortcomings. First, they complicate grammar writing in that the grammar writer must remember the names used and their intended interpretation. Second, they fail to support multiple uses of the same class within one class. For instance, in French, some verbs sub-categorize for two prepositional phrases (PP). A natural way of deriving the tree for such verbs would be to combine a verbal tree fragment with two instances of a PP fragment. If, however, the nodes in the PP fragment are labeled with global names, then the two occurrences of these nodes will be identified thereby blocking the production of the appropriate tree. 11 A less restrictive treatment of identifiers is proposed by Vijay-Shanker and Schabes (1992), where each tree description can be associated with a set of declared node variables and subsets of these node variables can be referred to by descriptions in the hierarchy that inherit from the description in which these node variables were declared. For instance, if entity A in the hierarchy declares such a special node variable X and B inherits from A, then X can be referred to in B using the notation A.X. 12 XMG generalizes Vijay-Shanker and Schabes's (1992) approach by integrating an export mechanism that can be used to extend the scope of a given identifier (node or feature value variable) to classes that inherit from the exporting class. Thus if class B inherits from class A and class A exports variable X, then X is visible in B and its reuse forces identity. If B inherits from several classes and two (or more) of these inherited classes export the same variable name X, then X is not directly visible from B. It can be accessed though using the dot operator. First A is identified with a local variable (e.g., T = A), then T.X can be used to refer to the variable X exported by A.",
	"cite_spans": [
	{
	"start": 3,
	"end": 19,
	"text": "Candito's (1996)",
	"ref_id": "BIBREF12"
	},
	{
	"start": 170,
	"end": 199,
	"text": "Gardent and Parmentier (2006)",
	"ref_id": "BIBREF28"
	},
	{
	"start": 869,
	"end": 871,
	"text": "11",
	"ref_id": null
	},
	{
	"start": 1377,
	"end": 1379,
	"text": "12",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "To summarize, XMG allows for local variables to be exported to sub-classes as well as for prefixed variables-that is, variables that are prefixed (using the dot operator) with a reference to the class in which they are declared. In this way, the pitfalls introduced by global names are avoided while providing enough expressivity to handle variable coreference (via the definition of variable namespaces). Section 6 will further illustrate the use of the various coreference devices made available by XMG showing how they concretely facilitate grammar writing.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "Let us finally illustrate variable handling with XMG in the example of Figure 2 . Recall that we define the trees of Figure 2 as the conjunctions and disjunctions of some tree fragments of Figure 5 , such as:",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 71,
	"end": 79,
	"text": "Figure 2",
	"ref_id": null
	},
	{
	"start": 117,
	"end": 125,
	"text": "Figure 2",
	"ref_id": null
	},
	{
	"start": 189,
	"end": 197,
	"text": "Figure 5",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "Subject \u2192 SubjAgreement \u2227 ( CanonicalSubject \u2228 Wh-NP-Subject )",
	"eq_num": "(20)"
	}
	],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "CanonicalSubject can be defined as a tree description formula as follows (only variables n 2 and n 3 are exported):",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "CanonicalSubject n 2 ,n 3 \u2192 n 1 \u2192 n 2 \u2227 n 1 [cat : S] \u2227 n 2 (mark : subst) [cat : NP] \u2227 n 1 \u2192 n 3 \u2227 n 3 [cat : VP] \u2227 n 2 \u227a n 3",
	"eq_num": "(21)"
	}
	],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "The class Wh-NP-Subject is defined accordingly (i.e., by means of a slightly more complex tree description formula using the n 2 and n 3 variable identifiers to refer to the nodes involved in subject agreement). The class SubjAgreement is defined slightly differently (we do not impose any tree relation between the node concerned with number agreement): [[top : [num : x] ",
	"cite_spans": [
	{
	"start": 355,
	"end": 372,
	"text": "[[top : [num : x]",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "SubjAgreement n 1 ,n 2 \u2192 n 1",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "We can then explicitly control the way the fragments combine as follows:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "Subject \u2192 C 1 = SubjAgreement n 1 ,n 2 \u2227 C 2 = ( CanonicalSubject n 2 ,n 3 \u2228 Wh-NP-Subject n 2 ,n 3 ) \u2227 C 1 .n 1 = C 2 .n 2 \u2227 C 1 .n 2 = C 2 .n 3",
	"eq_num": "(23)"
	}
	],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "In this example, we see how to constrain, via variable export and unification, some given syntactic nodes to be labeled with feature structures defined somewhere else in the metagrammar. We use XMG's flexible management of variable scope to deal with node coreference. Compared with previous approaches on metagrammars such as those of Candito (1996) , Xia (2001) , having the possibility of handling neither only global nor only local variables, offers a high level of expressivity along with a precise control on the structures being described.",
	"cite_spans": [
	{
	"start": 336,
	"end": 350,
	"text": "Candito (1996)",
	"ref_id": "BIBREF12"
	},
	{
	"start": 353,
	"end": 363,
	"text": "Xia (2001)",
	"ref_id": "BIBREF51"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEM.",
	"sec_num": null
	},
	{
	"text": "A third distinguishing feature of XMG is extensibility. XMG is extensible in that (i) dimensions can be added and (ii) each dimension can be associated with its own interpreter. In order to support an arbitrary number of dimensions, XMG relies on a device permitting the accumulation of an arbitrary number of types of literals, namely, Extensible Definite Clause Grammar (EDCG) (Van Roy 1990). Once literals are accumulated according to their type (i.e., each type of literals is accumulated separately), they can be fed to dedicated interpreters. Because each of these sets of literals represents formulas of a description language, these interpreters are solvers whose role is to compute models satisfying the accumulated formulas. Via this concept of separated dimensions, XMG allows us (i) to describe different levels of language (not only syntax, but also semantics and potentially morphology, 13 etc.), and (ii) to define linguistic principles (well-formedness constraints to be applied on the structures being described). These principles depend either on the dimension (e.g., scope constraints in flat semantics), the target formalism (e.g. cooccurrence predicatearguments in FB-LTAG), or the natural language (e.g., clitic ordering in Romance languages) being described.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Extensibility",
	"sec_num": "5."
	},
	{
	"text": "In what follows, we start by showing how XMG handles dimensions independently from each other introducing EDCG (Section 5.1). We then summarize the architecture of the XMG system (Section 5.2). We finally show how different solvers can be used to implement various constraints on each of these dimensions (Section 5.3). In particular, we discuss three kinds of extensions implemented in XMG: extension to several grammar formalisms, integration of explicit linguistic generalizations, and inclusion of color-based node marking to facilitate grammar writing.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Extensibility",
	"sec_num": "5."
	},
	{
	"text": "Accumulating (tree) descriptions. First, let us notice that XMG is nothing other than a logic language\u00e0 la Prolog ",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "= Description \| Name \| Content \u2228 Content \| Content \u2227 Content Axiom ::= Name",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "Recall that the descriptions handled by XMG are in fact tuples of the form SYN, SEM, DYN . An XMG class can thus describe, in a non-exclusive way, any of these three levels of description. If one wants to add another level of description (i.e., another dimension), one needs to extend the arity of this tuple. Before discussing this, let us first see how such tuples are processed by XMG.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "As mentioned earlier, XMG's control language is comparable to Horn clauses. A common way to represent Horn clauses is by using Definite Clause Grammar (DCG) (Pereira and Warren 1980) . Concretely, a DCG is a rewriting system (namely, a context-free grammar), where the symbols of the rewriting rules are equipped with pairs of unification variables (these are usually called difference list or accumulator) (Blackburn, Bos, and Striegnitz 2006, page 100) . As an illustration, consider the following toy example.",
	"cite_spans": [
	{
	"start": 157,
	"end": 182,
	"text": "(Pereira and Warren 1980)",
	"ref_id": "BIBREF43"
	},
	{
	"start": 407,
	"end": 454,
	"text": "(Blackburn, Bos, and Striegnitz 2006, page 100)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "s --> np,vp. np --> det,n. vp --> v,np. vp --> v. det --> [the]. det --> [a]. n --> [cat]. n --> [mouse]. v --> [eats].",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "The string language described by this DCG can be obtained by submitting the query s(X,[]) where X is a unification variable to be bound with lists of facts (these being the sentences belonging to the string language). As we can easily see, this language contains the sentences \"a cat eats,\" \"the cat eats,\" \"a mouse eats,\" \"the mouse eats,\" \"a cat eats a mouse,\" \"a mouse eats a cat,\" and so on.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "Similarly, we can represent XMG classes as DCG clauses. For instance, the combinations of syntactic fragments given in relations (1)-(4) can be rewritten as DCG clauses as follows:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "subject --> canonicalSubject. subject --> whNpSubject. activeTransitiveVerb --> subject, activeVerb, canonicalObject. passiveTransitiveVerb --> subject, passiveVerb, canonicalByObject. transitiveVerb --> activeTransitiveVerb. transitiveVerb --> passiveTransitiveVerb.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "Disjunctions (e.g., the subject specification) translate to multiple clauses with identical heads and conjunctions (e.g., activeTransitiveVerb) to a clause body.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "In our case, the terminal symbols of the underlying DCG are not just facts, but tuples of descriptions. In other words, the DCG clause whose head is canonicalSubject is associated with a tuple of the following form (the dots have to be replaced with adequate descriptions, these can contain unification variables, whose scope is by default local to the clause):",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "canonicalSubject --> [desc(syn(...),sem(...),dyn(...))].",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "In order to allow for an extension of XMG to an arbitrary number of dimensions, instead of compiling XMG classes into a DCG whose accumulator stores tuples with a fixed arity, these classes are compiled into an EDCG (Van Roy 1990) . EDCG are DCG with multiple accumulators. In XMG, each dimension is thus allocated a dedicated accumulator in the underlying EDCG. Note that although the content of the various dimensions is accumulated separately, dimensions may nevertheless share information either via local unification variables (if the XMG class defines several dimensions locally), via exported unification variables (in case of class instantiation or inheritance), or via the shared unification variables supported by the DYN dimension.",
	"cite_spans": [
	{
	"start": 211,
	"end": 230,
	"text": "EDCG (Van Roy 1990)",
	"ref_id": null
	},
	{
	"start": 357,
	"end": 362,
	"text": "EDCG.",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "At the end of the EDCG execution, we obtain, for each axiom of the metagrammar (i.e., for each class name to be valuated), a list of description formulas per accumulator. These lists are grouped together into a tuple of lists of the following form (N is the number of dimensions, and consequently of accumulators):",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "desc(accu1(L1),accu2(L2), ... ,accuN(LN))",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "Each element (i.e., list Li) of such a tuple is a complete description of a given dimension, where shared variables have been unified (via unification with backtracking).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "Solving (tree) descriptions. As illustrated earlier, interpreting XMG's control language in terms of an EDCG yields tuples whose arity is the number of dimensions defined by the linguist, that is, triples of the form SYN, SEM, DYN if syntax, semantics, and the dynamic interface are described.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "For each dimension D, XMG includes a constraint solver S D that computes the set of minimal models",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "M D = S D (d D ) satisfying the description (d D ) of that dimension.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "In other words, each dimension is interpreted separately by a specific solver. For instance, the syntactic dimension is handled by a tree description solver that produces, for a given tree description, the set of trees satisfying that description, whereas the solver for the semantic dimension simply outputs the flat semantic representation (list of semantic literals) built by the EDCG through accumulation.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "Note that, although solvers are distinct, the models computed in each dimension may nonetheless be coupled through shared variables. In that case, these variables can constrain the models computed by the respective solvers. For instance, shared variables can be used for the syntactic tree description solver to be parametrized by some value coming from the semantic input description. Note that the output of the solving process is a Cartesian product of the sets of minimal models of each solver. As a consequence, the worst case complexity of metagrammar compilation is that of the various solvers associated with relevant dimensions.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "In addition to having separate solvers for each dimension, the constraint-solving approach used in XMG permits us to modularize a given solver by combining different principles. Each such principle enforces specific constraints on the models satisfying the description of a given dimension. For instance, for the syntactic dimension of an FB-LTAG, a set of principles is used to enforce that the structures produced by the compiler are trees, and that these conform to the FB-LTAG formalism (e.g., there is no tree having two foot nodes).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "XMG: Accumulating and Interpreting an Arbitrary Number of Descriptions",
	"sec_num": "5.1"
	},
	{
	"text": "The XMG compiler 14 consists of the following three modules:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Architecture",
	"sec_num": "5.2"
	},
	{
	"text": "r A compiler that parses XMG's concrete syntax and compiles XMG classes into clauses of an EDCG.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Architecture",
	"sec_num": "5.2"
	},
	{
	"text": "r A virtual machine (VM), which interprets EDCG. This VM performs the accumulation of dimensions along with scope management and identifiers resolution. This VM is basically a unification engine equipped with backtracking, and which is extended to support EDCG. Although its architecture is inspired by the Warren Abstract Machine (A\u00eft-Kaci 1991), it uses structure-sharing to represent and unify prolog terms, and, given a query on a class, processes the conjunctions, disjunctions, inheritance, and export statements related to that class to produce its full definition, namely, a tree description for the SYN dimension, a flat semantic formula for the SEM dimension, and a feature structure for the DYN dimension.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Architecture",
	"sec_num": "5.2"
	},
	{
	"text": "r A constraint-solving phase that produces for each dimension the minimal models satisfying the input description as unfolded by the preceding two steps.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Architecture",
	"sec_num": "5.2"
	},
	{
	"text": "As already mentioned, the first part is extensible in that new linguistic dimensions can be added by specifying additional dedicated accumulators to the underlying EDCG. The second part is a unification engine that interprets EDCG while performing both term unification and polarized unification (i.e., unification of polarized feature structures, as defined by Perrier [2000] , and discussed in Section 5.3.1). This extended unification is the reason why XMG does not merely recourse to an existing Prolog engine to process EDCG, but relies on a specific VM instead.",
	"cite_spans": [
	{
	"start": 362,
	"end": 376,
	"text": "Perrier [2000]",
	"ref_id": "BIBREF45"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Architecture",
	"sec_num": "5.2"
	},
	{
	"text": "The third part is completely modular in that various constraint solvers can be plugged in depending on the requirements set by the dimensions used, and the chosen grammatical framework. For instance, the SYN dimension is solved in terms of tree models, and the SEM dimension is solved in terms of underspecified flat semantic formulae (i.e., the input semantics remains untouched modulo the unification of its shared variables).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Architecture",
	"sec_num": "5.2"
	},
	{
	"text": "Importantly, these additional solvers can be \"turned on/off\" (via a primitive of the XMG language) so that, for instance, the same processor can be used to compile an XMG specification for an FB-LTAG using linguistic principles such as those defined in the next section (i.e., clitic ordering principle) or not.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Architecture",
	"sec_num": "5.2"
	},
	{
	"text": "We now show (i) how the modular architecture of the XMG compiler permits us to specify grammars for several tree-based linguistic formalisms; (ii) how it can be extended to enforce language specific constraints on the syntactic trees; and (iii) how additional formal constraints (namely node marking) can be integrated to simplify node identifications (and consequently grammar writing).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Three Extensions of XMG",
	"sec_num": "5.3"
	},
	{
	"text": "Partition of the nodes of tree models. MC-TAG, and IG: Producing Trees, Tree Sets, or Tree Descriptions. XMG integrates a generic tree solver that computes minimal tree models from tree description logic formulae built on the language SYN introduced in Section 4. This solver integrates the dominance solving technique proposed by Duchier and Niehren (2000) and can be summarized as follows. A minimal tree model is described in terms of the relative positions of its nodes. For each node n in a minimal tree model T, the set of all the nodes of T can be partitioned in five subsets, depending on their position relative to n. Hence, for each node variable n appearing in a tree description, it is first associated with an integer (called node id). We then define the five sets of node ids (i.e., sets of integers) Down n , Up n , Left n , Right n , and Eq n referring to the ids of the nodes located below, above, on the left, on the right, or identified with n, respectively (see Figure 6 ). Note that we require that these sets are a partition of all node ids.",
	"cite_spans": [
	{
	"start": 39,
	"end": 85,
	"text": "MC-TAG, and IG: Producing Trees, Tree Sets, or",
	"ref_id": null
	},
	{
	"start": 331,
	"end": 357,
	"text": "Duchier and Niehren (2000)",
	"ref_id": "BIBREF22"
	}
	],
	"ref_spans": [
	{
	"start": 982,
	"end": 990,
	"text": "Figure 6",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Figure 6",
	"sec_num": null
	},
	{
	"text": "Using this set-based representation of a model, we translate each node relation from the input formula (built on the tree description language introduced in Section 4) into constraints on the sets of node ids that must hold in a valid model. For instance, the sub-formula n 1 \u227a + n 2 , which states that node n 1 strictly precedes node n 2 , is translated into:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "TAG,",
	"sec_num": "5.3.1"
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "n 1 \u227a + n 2 \u2261 EqDown n 1 \u2286 Left n 2 \u2227 EqDown n 2 \u2286 Right n 1 \u2227 Right n 2 \u2286 Right n 1 \u2227 Left n 1 \u2286 Left n 2",
	"eq_num": "(24)"
	}
	],
	"section": "TAG,",
	"sec_num": "5.3.1"
	},
	{
	"text": "where 15 EqDown x = Eq x Down x for x \u2208 {n 1 , n 2 }. In other words, in a valid minimal tree model, the set of nodes below or equal to n 1 is included in the set of nodes (strictly) on the left of n 2 , the set of nodes below or equal to n 2 is included in the set of nodes (strictly) on the right of n 1 , the set of nodes on the right of n 2 is included in the set of nodes on the right of n 1 , and finally the set of nodes on the left of n 1 is included in the set of nodes on the left of n 2 . Once all input relations are translated into set constraints, the solver uses standard Constraint Satisfaction techniques (e.g., a first-fail exploration of the search tree) to find a set of consistent partitions. Finally, the nodes of the models are obtained by considering nodes with distinct Eq n .",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "TAG,",
	"sec_num": "5.3.1"
	},
	{
	"text": "To support the specification of FB-LTAG trees, the XMG compiler extends the generic tree solver described here with a set of constraints ensuring that the trees are well-formed TAG trees. In effect, these constraints require the trees to be linear ordered trees with appropriate decorations. Each node must be labeled with a syntactic category. Leaf nodes are either terminal, foot, or substitution nodes. There is at most one foot node per tree and the category of the foot node must be identical to that of the root node. Finally, each tree must have at least one leaf node that is an anchor. MCTAG tree sets. Where FB-LTAG consists of trees, MC-TAG (Weir 1988) consists of sets of trees. To support the specification of MC-TAG, the sole extension needed concerns node variables that are not dominated by any other node variable in the tree description. Whereas for FB-LTAG, these are taken to denote either the same root node or nodes that are connected to some other node (i.e., uniqueness of the root), for MC-TAG they can be treated as distinct nodes, thereby allowing for models that are sets of trees rather than trees (Parmentier et al. 2007) . In other words, the only modification brought to the tree description solver is that, in MC-TAG mode, it does not enforce the uniqueness of a root node in a model. IG polarized tree descriptions. IG (Perrier 2000) consist of tree descriptions whose node variables are labeled with polarized feature structures. A polarized feature structure is a set of polarized feature triples (f, p, v) where f and v are standard features and feature values, respectively, and p is a polarity value in {\u2192, \u2190, =, \u2248}. Polarities are used to guide parsing in that a valid derivation structure must neutralize polarities.",
	"cite_spans": [
	{
	"start": 1127,
	"end": 1151,
	"text": "(Parmentier et al. 2007)",
	"ref_id": "BIBREF42"
	},
	{
	"start": 1353,
	"end": 1367,
	"text": "(Perrier 2000)",
	"ref_id": "BIBREF45"
	},
	{
	"start": 1533,
	"end": 1542,
	"text": "(f, p, v)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "FB-LTAG trees.",
	"sec_num": null
	},
	{
	"text": "To support an XMG encoding of IG, two extensions are introduced, namely, (i) the ability to output tree descriptions rather than trees, and (ii) the ability to write polarized feature structures. The first extension is trivially realized by specifying a description solver that ensures that any output description has at least one tree model. For the second point, the SYN language is extended to define polarized feature structures and the unification engine to support unification of polarized features (for instance, a \u2192 feature will unify with a neutral (=) feature to yield a \u2192 polarized feature value triple).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "FB-LTAG trees.",
	"sec_num": null
	},
	{
	"text": "The Case of Clitics. XMG can be extended to support specific constraints on tree descriptions (e.g., constraints on node linear order), which make it possible to describe linguistic-dependent phenomena, such as, for instance, clitic ordering in French, at a meta-level (i.e., within the metagrammar).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Adding Specific Linguistic Constraints:",
	"sec_num": "5.3.2"
	},
	{
	"text": "According to Perlmutter (1970) , clitics are subject to two hard constraints. First, they appear in front of the verb in a fixed order according to their rank (Examples 25a and 25b). 16 Second, two different clitics in front of the verb cannot have the same rank (Example 25c).",
	"cite_spans": [
	{
	"start": 13,
	"end": 30,
	"text": "Perlmutter (1970)",
	"ref_id": "BIBREF44"
	},
	{
	"start": 183,
	"end": 185,
	"text": "16",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Adding Specific Linguistic Constraints:",
	"sec_num": "5.3.2"
	},
	{
	"text": "(25) a. Jean le 3 lui 4 donne.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Adding Specific Linguistic Constraints:",
	"sec_num": "5.3.2"
	},
	{
	"text": "'John gives it to him.' b. Jean lui 4 le 3 donne. 'John gives to him it.' c. Jean le 3 la 3 donne. 'John gives it it.'",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Adding Specific Linguistic Constraints:",
	"sec_num": "5.3.2"
	},
	{
	"text": "S N\u2193 \u227a + V' \u2227 V' Cl\u2193 3 \u227a + V \u2227 V' Cl\u2193 4 \u227a + V \u2227 S V' V \u21d2 S N\u2193 V' Cl\u2193 3 Cl\u2193 4 V S N\u2193 V' Cl\u2193 4 Cl\u2193 3 V",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Adding Specific Linguistic Constraints:",
	"sec_num": "5.3.2"
	},
	{
	"text": "Clitic ordering in French.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "To support a direct encoding of Perlmutter's observation, XMG includes both a node uniqueness principle and a node ordering principle. The latter allows us to label nodes with some property (let us call it rank) whose value is an integer (for instance, one can define a node as n 1 (rank : 2)[cat : Cl]). When solving tree descriptions, XMG further requires that in a valid tree model, (i) there are no two nodes with the same rank and (ii) sibling nodes labeled with a rank are linearly ordered according to their rank.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "Accordingly, in the French grammar of Crabb\u00e9 (2005) , each node labeled with a clitic category is also labeled with a numerical node property representing its rank. 17 XMG ordering principle then ensures that the ill-formed tree crossed out in Figure 7 is not produced. Note that in Figure 7 , every type of clitic is defined locally (i.e., in a separate class), and that the interactions between these local definitions are handled by XMG using this rank principle, to produce only one valid description (pictured to the right of the arrow).",
	"cite_spans": [
	{
	"start": 38,
	"end": 51,
	"text": "Crabb\u00e9 (2005)",
	"ref_id": null
	}
	],
	"ref_spans": [
	{
	"start": 244,
	"end": 252,
	"text": "Figure 7",
	"ref_id": null
	},
	{
	"start": 283,
	"end": 291,
	"text": "Figure 7",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "That is, XMG ordering constraints permit a simple, declarative encoding of the interaction between clitics. This again contrasts with systems based on lexical rules. As noted by Perlmutter (1970) , if clitics are assumed to be moved by transformations, then the order in which lexical rules apply this movement must be specified.",
	"cite_spans": [
	{
	"start": 178,
	"end": 195,
	"text": "Perlmutter (1970)",
	"ref_id": "BIBREF44"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "To implement the uniqueness principle, one needs to express the fact that in a valid model \u03c6, there is only one node having a given property p (i.e., a parameter of the constraint, here the value of the rank node property). This can be done by introducing, for each node n of the description, a Boolean variable p n indicating whether the node denoting n in the model has this property or not (i.e., are there two nodes of identical rank?). Then, if we call V \u03c6 p the set of integers referring to nodes having the property p in a model, we have:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "p n \u2261 (Eq n \u2229 V \u03c6 p ) = \u2205.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "Finally, if we represent p n being true with 1 and p n being false with 0, 18 and we sum p n for each n in the model, we have that in a valid model this sum is strictly lower than 2: n\u2208\u03c6 p n < 2.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "To implement the ordering principle, one needs to express the fact that in a valid model \u03c6, two sibling nodes n 1 and n 2 having a given property p of type integer and of values p 1 and p 2 , respectively, are such that the linear precedence between these nodes conform to the natural order between p 1 and p 2 . This can be done by first introducing, for each pair of nodes n, m of the description, a Boolean variable b n,m indicating whether they have the same ancestors: b n,m \u2261 (Up n \u2229 Up m ) = (Up n \u222a Up m ). For each pair of nodes that do so, we check whether they both have the property p,",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "17 Recall that node properties are features whose values are used by the tree description solver in order to restrict the set of valid models. These properties may not appear in the trees produced from the input metagrammar. For instance, the rank property is not part of the FB-LTAG formalism, and thus does not appear in the FB-LTAG elementary trees produced by XMG. 18 These integer representations are usually called reified constraints. and if this is the case, we add to the input description a strict precedence constraint on these nodes according to their respective values of the property p: (Crabb\u00e9 and Duchier 2004) , reminiscent of the polarity-based node identification first proposed by Muskens and Krahmer (1998) and later used by Duchier and Thater (1999) and Perrier (2000) . Such a mechanism offers an alternative to explicit node identification using equations between node variables. The idea is to label node variables with a color property, whose value (either red, black, or white) can trigger node identifications. This mechanism is another parameter of the tree solver. When in use, the valid tree models must satisfy some color constraints, namely, they must only have red or black nodes (no remaining white nodes; these have to be identified with some black nodes). As shown in the following table, node identification must observe the following constraints: A white node must be identified with a black node; a red node cannot be identified with any other node; and a black node may be identified with one or more white nodes. 20",
	"cite_spans": [
	{
	"start": 601,
	"end": 626,
	"text": "(Crabb\u00e9 and Duchier 2004)",
	"ref_id": "BIBREF20"
	},
	{
	"start": 701,
	"end": 727,
	"text": "Muskens and Krahmer (1998)",
	"ref_id": null
	},
	{
	"start": 746,
	"end": 771,
	"text": "Duchier and Thater (1999)",
	"ref_id": "BIBREF24"
	},
	{
	"start": 776,
	"end": 790,
	"text": "Perrier (2000)",
	"ref_id": "BIBREF45"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "19 b n,m \u2227 (p n < p m ) \u21d2 n \u227a + m (26) b n,m \u2227 (p m < p n ) \u21d2 m \u227a + n",
	"eq_num": "("
	}
	],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "\u2022 B \u2022 R \u2022 W \u22a5 \u2022 B \u22a5 \u22a5 \u2022 B \u22a5 \u2022 R \u22a5 \u22a5 \u22a5 \u22a5 \u2022 W \u2022 B \u22a5 \u2022 W \u22a5 \u22a5 \u22a5 \u22a5 \u22a5 \u22a5",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "We now briefly describe how the constraint solver sketched in Section 5.3.1 was extended to support colors. As mentioned previously, in valid models all white nodes are identified with a black node (at most one black node per white node). Consequently, there is a bijection from the red and black nodes of the tree description to the nodes of the model. In order to take this bijection into account, we add a node variable RB n to the five sets already associated with a node variable n from Section 5.1. RB n denotes either n if n is a black or red node, or the black node identified with n if n is a white node. Note that all the node variables must be colored: the set of node variables in a tree description can then be partitioned into three sets: Red, Black, and White. Basically, we know that, for all nodes n, RB n \u2208 Eq n (this is what the bijection is about). Again we translate color information into constraints on node sets (these constraints help the generic tree solver by reducing the ambiguity for the Eq n sets):",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "n \u2208 Red \u21d2 (n = RB n ) \u2227 (Eq n = {n}) (28) n \u2208 Black \u21d2 (n = RB n ) \u2227 (Eq n \\{n} \u2286 White) (29) n \u2208 White \u21d2 (RB n \u2208 Black) \u2227 (Eq n \u2229 Black = {RB n })",
	"eq_num": "(30)"
	}
	],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "Node coloring offers an alternative to complex namespace management. The main advantage of this particular identification mechanism is its economy: Not only is there no longer any need to remember node identifiers, there is in fact no need to choose a name for node variables.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "It is worth stressing that the XMG node identification process is reduced to a constraint-solving problem and so it is not a sequential process. Thus the criticisms leveled by Wintner (2007, 2009) against non-associative constraints on node unification do not apply.",
	"cite_spans": [
	{
	"start": 176,
	"end": 196,
	"text": "Wintner (2007, 2009)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "Briefly, in their work, Wintner (2007, 2009) showed that any polarity-based tree description formalism is not associative. In other words, when describing trees in terms of combinations of polarized structures, the order in which the structures are combined matters (i.e., the output structures depend on the combination order). This feature makes such formalisms not appropriate for a modular and collaborative grammar engineering, such as that of Cohen-Sygal and Wintner (2011) for Unification Grammar.",
	"cite_spans": [
	{
	"start": 24,
	"end": 44,
	"text": "Wintner (2007, 2009)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "In the XMG case, when using node colors, the tree description solver does not rely on any specific fragment combination order. It computes all possible combination orders. In this context, the grammar designer cannot think in terms of sequences of node identifications. This would lead to tree overgeneration.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "Again, it is important to remember that tree solving computes any valid tree model, independently of any specific sequence of node identifications (all valid node identifications are computed). In this context, non-associativity of color-based node identification is not an issue, but rather a feature, as it allows for a compact description of a large number of node identifications (and thus of tree structures).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 7",
	"sec_num": null
	},
	{
	"text": "In this section, we first provide a detailed example showing how XMG can be used to specify the verbal trees of a large FB-LTAG for French extended with unification-based semantics. We then give a brief description of several large-and middle-scale grammars that were implemented using XMG.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Writing Grammars with XMG",
	"sec_num": "6."
	},
	{
	"text": "We now outline the XMG specification for the verbal trees of SEMTAG, a large FB-LTAG for French. This specification further illustrates how the various features of XMG (e.g., combined use of disjunction and conjunction, node colors) permit us to specify compact and declarative grammar descriptions. We first discuss the syntactic dimension (SYN). We then go on to show how the semantic dimension (SEM) and the syntax/semantic interface (DYN) are specified.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "SEMTAG: A large FB-LTAG for French Covering Syntax and Semantics",
	"sec_num": "6.1"
	},
	{
	"text": "The methodology used to implement the verbal fragment of SEMTAG can be summarized as follows. First, tree fragments are defined that represent either a possible realization of a verb argument or a possible realization of the verb. The verbal elementary TAG trees of SEMTAG are then defined by appropriately combining these tree fragments.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "The Syntactic Dimension.",
	"sec_num": "6.1.1"
	},
	{
	"text": "To maximize structure sharing, we work with four levels of abstraction. First, basic tree fragments describing verb or verb argument realizations are defined. Second, grammatical functions are defined as disjunctions of argument realizations. Third, verbal diathesis alternatives are defined as conjunctions of verb realizations and grammatical",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "The Syntactic Dimension.",
	"sec_num": "6.1.1"
	},
	{
	"text": "CanonSubj \u2192 S\u2022 W N\u2193\u2022 R V\u2022 W CanonObj \u2192 S\u2022 W V\u2022 W N\u2193\u2022 R CanonIndirObj \u2192 S\u2022 W V\u2022 W PP\u2022 R P\u2022 R a\u2022 R N\u2193\u2022 R CanonByObj \u2192 S\u2022 W V\u2022 W PP\u2022 R P\u2022 R par\u2022 R N\u2193\u2022 R RelatSubj \u2192 N\u2022 R N \u2022 R S\u2022 W N\u2193\u2022 R V\u2022 W WhObj \u2192 S\u2022 R N\u2193\u2022 R S\u2022 W V\u2022 W WhByObj \u2192 S\u2022 R PP\u2022 R P\u2022 R par\u2022 R N\u2193\u2022 R S\u2022 W WhIndirObj \u2192 S\u2022 R PP\u2022 R P\u2022 R a\u2022 R N\u2193\u2022 R S\u2022 W ActiveVerbForm\u2192 S\u2022 B V \u2022 B PassiveVerbForm\u2192 S\u2022 B V\u2022 B V\u2193\u2022 B V \u2022 B",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "The Syntactic Dimension.",
	"sec_num": "6.1.1"
	},
	{
	"text": "Elementary tree fragments used as building blocks of the grammar (nodes are colored to control their identification when blocks are combined).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 8",
	"sec_num": null
	},
	{
	"text": "functions. Fourth, diathesis alternatives are gathered into tree families. In the next paragraphs, we explain each of these levels in more detail.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 8",
	"sec_num": null
	},
	{
	"text": "Tree fragments. Tree fragments are the basic building blocks used to define SEMTAG. These are the units that are shared and reused in the definition of many elementary trees. For instance, the fragment for a canonical subject will be used by all FB-LTAG elementary trees involving a canonical subject. As mentioned earlier, to specify the verbal elementary trees of SEMTAG, we begin by defining tree fragments which describe the possible syntactic realizations of the verb arguments and of the verb itself. Figure 8 provides some illustrative examples of these fragments. Here and in the following, we omit the feature structures decorating the trees to facilitate reading. 21 To further factorize information and facilitate grammar maintenance, the basic tree fragments are organized in an inheritance hierarchy. 22 Figure 9 shows a partial view of this hierarchy illustrating how the tree fragments for argument realization depicted in Figure 8 are organized to maximize the sharing of common information. The hierarchy classifies the verbal arguments depicted in Figure 8 into four categories:",
	"cite_spans": [
	{
	"start": 674,
	"end": 676,
	"text": "21",
	"ref_id": null
	}
	],
	"ref_spans": [
	{
	"start": 507,
	"end": 515,
	"text": "Figure 8",
	"ref_id": null
	},
	{
	"start": 817,
	"end": 825,
	"text": "Figure 9",
	"ref_id": null
	},
	{
	"start": 938,
	"end": 946,
	"text": "Figure 8",
	"ref_id": null
	},
	{
	"start": 1066,
	"end": 1074,
	"text": "Figure 8",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Figure 8",
	"sec_num": null
	},
	{
	"text": "1.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 8",
	"sec_num": null
	},
	{
	"text": "The canonical subject is a noun realized in front of the verb.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 8",
	"sec_num": null
	},
	{
	"text": "Canonical complements occur after the verb. The canonical object is a noun phrase whereas prepositional complements are introduced by specific prepositions, namely,\u00e0 for the canonical indirect object and par for the canonical by object.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "2.",
	"sec_num": null
	},
	{
	"text": "Wh-arguments (or questioned arguments) occur in front of a sentence headed by a verb. A Wh-object is an extracted noun whereas questioned prepositional objects are extracted prepositional phrases that are introduced by a specific preposition.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Finally, the relativized subject is a relative pronoun realized in front of the sentence. Extracted subjects in French cannot be realized at an unbounded distance from the predicate.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "4.",
	"sec_num": null
	},
	{
	"text": "The second level of abstraction uses syntactic function names such as Subject and Object to group together alternative ways in which a given syntactic function can be realized. For instance, if we make the simplifying assumption that the possible argument realizations are limited to those given in Figure 8 , the Subject, Object, ByObject, and IndirectObject classes would be defined as follows. 23",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 299,
	"end": 307,
	"text": "Figure 8",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Syntactic functions.",
	"sec_num": null
	},
	{
	"text": "Subject \u2192 CanonSubj \u2228 RelatSubj (31) Object \u2192 CanonObj \u2228 WhObj (32) ByObject \u2192 CanonByObj \u2228 WhByObj (33) IndirectObject \u2192 CanonIndirObj \u2228 WhIndirObj (34)",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Syntactic functions.",
	"sec_num": null
	},
	{
	"text": "That is, we define the Subject class as an abstraction for talking about the set of tree fragments that represent the possible realizations of a subject argument-namely, in Similarly, tree families for non-verbal predicates (adjectives, nouns) can be defined using the abstraction over grammatical functions defined for verbs. For instance, the examples in (44a-44b) can be captured using the adjectival trees defined in Equations (46) and (47), respectively, where Subject extends the definition of subject given above with a Wh-subject, PredAdj combines a subject tree fragment with a tree fragment describing a predicative adjective, and PredAdjAObj extends a PredAdj tree fragment with a canonical a-object. 44 ",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Syntactic functions.",
	"sec_num": null
	},
	{
	"text": "Subject \u2192 CanonSubj \u2228 RelatSubj \u2228 WhSubj (45) PredAdj \u2192 Subject \u2227 AdjectivalForm (46) PredAdjAObj \u2192 PredAdj \u2227 CanonAObj (47) 6.1.2",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Syntactic functions.",
	"sec_num": null
	},
	{
	"text": "The Semantic Dimension and the Syntax/Semantic Interface. We now show how to extend the XMG specification presented in the previous section to integrate a unification-based compositional semantics. Three main changes need to be carried out:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Syntactic functions.",
	"sec_num": null
	},
	{
	"text": "1. Each elementary tree must be associated with a semantic formula. This is done using the SEM dimension.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Syntactic functions.",
	"sec_num": null
	},
	{
	"text": "The nodes of elementary trees must be labeled with the appropriate semantic indices. This involves introducing the correct attribute-value pair in the correct feature structure (top or bottom) on the appropriate node.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "2.",
	"sec_num": null
	},
	{
	"text": "Syntax and semantics need to be synchronized-that is, variable sharing between semantic formulae and tree indices need to be enforced. To this end we use the DYN dimension.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Informing the semantic dimension. To associate each elementary tree with a formula representing the meaning of the words potentially anchoring that tree, we use the SEM dimension to specify a semantic schema. For instance, the TransitiveFamily class defined in Equation (42) for verbs taking two nominal arguments is extended as follows:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "TransitiveFamily \u2192 TransitiveDiathesis \u2227 BinaryRel (48)",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "where TransitiveDiathesis is the XMG class defined in Equation (39) to describe the set of trees associated with transitive verbs and BinaryRel the class describing the following semantic schema:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "L : P(E) \u2227 L : Theta 1 (E, X) \u2227 L : Theta 2 (E, Y)",
	"eq_num": "(49)"
	}
	],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "In this semantic schema, P, Theta 1 , and Theta 2 are unification variables that become ground when the tree is anchored with a specific word. For instance, P, Theta 1 , and Theta 2 are instantiated to eat, agent, and patient, respectively, when the anchor is ate (these pieces of information-predicate, thematic roles-are associated with lemmas, located in the syntactic lexicon, and unified with adequate semantic variables via anchoring equations). Further, X, Y, E, L are unification variables representing semantic arguments. As illustrated in Figure 3 , these become ground during (or after) derivation as a side effect of the substitutions and adjunctions taking place when trees are combined. It is worth noting that by combining semantic schemas with diathesis classes, one such specification assigns the specified semantic schema to many trees, namely, all the trees described by the corresponding diathesis class. In this way, the assignment of semantic formulae to trees is relatively economical. Indeed in SEMTAG, roughly 6,000 trees are assigned a semantic schema using a total of 75 schema calls.",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 549,
	"end": 557,
	"text": "Figure 3",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Co-indexing trees and formulae indices. Assuming that tree nodes are appropriately decorated with semantic indices by the specification scheme described in the next paragraph, we now show how to enforce the correct mapping between syntactic and semantic arguments. This is done in two steps. First, we define a set of interface constraints of the form index F : V, arg i : V which are used to enforce the identification of the semantic index (index F ) labeling a given tree node with grammatical function F (e.g., F := subject) with the index (arg i ) representing the i-th argument in a semantic schema. For instance, the following constraints ensure a subject/arg 1 mapping, that is, a coreference between the index labeling a subject node and the index representing the first argument of a semantic schema:",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "C 1 \u2192 Node [idx : I] \u2227 index subject : I C 2 \u2192 L : P(E) \u2227 L : Theta 1 (E, X) \u2227 arg 1 : X SubjectArg 1 \u2192 C 1 \u2227 C 2 \u2227 index subject : V, arg 1 : V (50)",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Given such interface constraints, we refine the diathesis definitions so as to ensure the correct bindings. For instance, the specification in Equation (38) is modified to: ) and the passive diathesis is specified as:",
	"cite_spans": [
	{
	"start": 173,
	"end": 174,
	"text": ")",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "TransitiveDiathesis \u2192 TransitiveActive \u2228 TransitivePassive TransitiveActive \u2192 (SubjectArg 1 \u2227 ObjectArg 2 \u2227 Subject \u2227 ActiveVerbForm \u2227 Object)",
	"eq_num": "(51"
	}
	],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "TransitivePassive \u2192 (SubjectArg 2 \u2227 ByObjectArg 1 \u2227 Subject \u2227 PassiveVerbForm \u2227 ByObject)",
	"eq_num": "(52)"
	}
	],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Labeling tree nodes with semantic indices. This scheme relies on the assumption that tree nodes are appropriately labeled with semantic indices (e.g., the subject node must be labeled with a semantic index) and that these indices are appropriately named (arg 1 must denote the parameter representing the first argument of a binary relation and index subject the value of the index feature on a subject node). As suggested by , a complete semantic labeling of a TAG with the semantic features necessary to enrich this TAG with the unification-based compositional semantics sketched in the previous section can be obtained by applying the following labeling principles: 26",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Argument labeling: In trees associated with semantic functors, each argument node is labeled with a semantic index 27 named after the grammatical function of the argument node (e.g., index subject for a subject node).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Controller/Controllee: In trees associated with control verbs, the semantic index of the controller is identified with the value of the controlled index occurring on the sentential argument node.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Anchor projection: The anchor node projects its index up to its maximal projection.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Foot projection: A foot node projects its index up to the root. 28",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "As we shall now see, XMG permits a fairly direct encoding of these principles. The Argument Labeling principle states that, in the tree associated with a syntactic functor (e.g., a verb), each node representing a syntactic argument (e.g., the subject node) should be labeled with a semantic index named after the grammatical function of that node (e.g., index subject ). 29 To specify this labeling, we define for each grammatical function Function \u2208 {Subject, Object, ByObject, IndirectObject, Additionally, in the tree fragments describing the possible realizations of the grammatical functions, the (exported) variable denoting the argument node is systematically named ArgNode. Finally, we modify the specification of the realizations of the grammatical functions to import the appropriate semantic class and identify ArgNode and FunctionNode. For instance, the Subject specification given above is changed to: ",
	"cite_spans": [
	{
	"start": 371,
	"end": 373,
	"text": "29",
	"ref_id": null
	},
	{
	"start": 461,
	"end": 494,
	"text": "Object, ByObject, IndirectObject,",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "EQUATION",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [
	{
	"start": 0,
	"end": 8,
	"text": "EQUATION",
	"ref_id": "EQREF",
	"raw_str": "Subject \u2192 SubjectSem \u2227 ArgNode = SubjectNode \u2227 (CanonSubj \u2228 RelatSubj \u2228 WhSubj)",
	"eq_num": "(54)"
	}
	],
	"section": "3.",
	"sec_num": null
	},
	{
	"text": "Anchor/Foot projection.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 10",
	"sec_num": null
	},
	{
	"text": "As a result, all ArgNode nodes in the tree descriptions associated with a subject realization are labeled with an index feature I whose global name is index subject . Value sharing between the semantic index of the controller (e.g., the subject of the control verb) and that of the controllee (e.g., the empty subject of the infinitival complement) is enforced using linking constraints between the semantic index labeling the controller node and that labeling the sentential argument node of the control verb. Control verb definitions then import the appropriate (object or subject control) linking constraint.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Figure 10",
	"sec_num": null
	},
	{
	"text": "The anchor (respectively, foot) projection principle stipulates the projection of semantic indices from the anchor (respectively, foot) node up to the maximal projection (respectively, root). Concretely, this means that the top and bottom features of the nodes located on this path between the anchor (respectively, foot) and the maximal projection (respectively, root) all include an index feature whose value is shared between adjacent nodes (see variables E i in Figure 10) . 30 Once the top and bottom structures are unified, so are the semantic indices along this path (modulo expected adjunctions realized on the projection).",
	"cite_spans": [
	{
	"start": 479,
	"end": 481,
	"text": "30",
	"ref_id": null
	}
	],
	"ref_spans": [
	{
	"start": 466,
	"end": 476,
	"text": "Figure 10)",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Figure 10",
	"sec_num": null
	},
	{
	"text": "To implement these principles, we define a set of anchor projection classes {Depth1, Depth2, Depth3} as illustrated in Figure 10 . We then \"glue\" these projection skeletons onto the relevant syntactic trees by importing the skeletons in the syntactic tree description and explicitly identifying the anchor node of the semantic projection classes with the anchor or foot node of these syntactic tree descriptions. Because the models must be trees, the nodes dominating the anchor node of the projection class will deterministically be identified with those dominating the anchor or foot node of the trees being combined with. For instance, for verbs, the class specifying the verbal spine (e.g., ActiveVerbForm, see Figure 10 ) equates the anchor node of the verbal spine with that of the projection skeleton. As a result, the verb projects its index up to the root.",
	"cite_spans": [],
	"ref_spans": [
	{
	"start": 119,
	"end": 128,
	"text": "Figure 10",
	"ref_id": null
	},
	{
	"start": 715,
	"end": 724,
	"text": "Figure 10",
	"ref_id": null
	}
	],
	"eq_spans": [],
	"section": "Figure 10",
	"sec_num": null
	},
	{
	"text": "As mentioned previously, SEMTAG is a large FB-LTAG for French equipped with semantics (Gardent 2008) ; it extends the purely syntactic FTAG of Crabb\u00e9 (2005) with a unification based compositional semantics as described by Gardent and Kallmeyer (2003) . 31 The syntactic FTAG in essence implements Abeill\u00e9's (2002) verbal sub-categorization frames. For each of these frames, FTAG defines a set of argument alternations (active, passive, middle, neuter, reflexivization, impersonal, passive impersonal) and of argument realizations (cliticization, extraction, omission, permutations, etc.) possible for this frame. Predicative (adjectival, nominal, and prepositional) and light verb constructions are also covered as well as some common sub-categorizing noun and adjective constructions. Basic descriptions are provided for the remaining constructions namely, adverbs, determiners, and prepositions. FTAG and SEMTAG were both evaluated on the Test Suite for Natural Language Processing (TSNLP) (Lehmann et al. 1996) , using a lexicon designed specifically on the test suite, hence reducing lexical ambiguity (Crabb\u00e9 2005; . This test suite focuses on difficult syntactical phenomena, providing grammatical and ungrammatical sentences. These competence grammars accept 76% of the grammatical items, reject 83% of the ungrammatical items, and have an average ambiguity of 1.64 parses per sentence. To give an idea of the compilation time, under architectures made of a 2-Ghz processor with 1 Gb of RAM, it takes XMG 10 minutes to compile the whole SEMTAG (recall that there is no semantic description solving, hence the compilation times between FTAG and SEMTAG do not differ). 32 Note that SEMTAG can be used for assigning semantic representations to sentences when combined with an FB-LTAG parser and a semantic construction module as described by Parmentier (2005, 2007) . 33 Conversely, it can be used to verbalize the meaning denoted by a given semantic representation when coupled with the GenI surface realizer described by Gardent and Kow (2007) .",
	"cite_spans": [
	{
	"start": 86,
	"end": 100,
	"text": "(Gardent 2008)",
	"ref_id": "BIBREF27"
	},
	{
	"start": 143,
	"end": 156,
	"text": "Crabb\u00e9 (2005)",
	"ref_id": null
	},
	{
	"start": 222,
	"end": 250,
	"text": "Gardent and Kallmeyer (2003)",
	"ref_id": "BIBREF27"
	},
	{
	"start": 253,
	"end": 255,
	"text": "31",
	"ref_id": null
	},
	{
	"start": 297,
	"end": 313,
	"text": "Abeill\u00e9's (2002)",
	"ref_id": "BIBREF0"
	},
	{
	"start": 992,
	"end": 1013,
	"text": "(Lehmann et al. 1996)",
	"ref_id": "BIBREF40"
	},
	{
	"start": 1106,
	"end": 1119,
	"text": "(Crabb\u00e9 2005;",
	"ref_id": null
	},
	{
	"start": 1674,
	"end": 1676,
	"text": "32",
	"ref_id": null
	},
	{
	"start": 1846,
	"end": 1869,
	"text": "Parmentier (2005, 2007)",
	"ref_id": null
	},
	{
	"start": 1872,
	"end": 1874,
	"text": "33",
	"ref_id": null
	},
	{
	"start": 2027,
	"end": 2049,
	"text": "Gardent and Kow (2007)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Some Figures About SEMTAG.",
	"sec_num": "6.1.3"
	},
	{
	"text": "XMG has been used mainly to design FB-LTAG and IG for French or English. More recently, it has also been used to design a FB-LTAG for Vietnamese and a TreeTuple MC-TAG for German. We now briefly describe each of these resources.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Other Grammars Designed with XMG",
	"sec_num": "6.2"
	},
	{
	"text": "SemXTAG. The English grammar, SEMXTAG (Alahverdzhieva 2008) , reimplements the FB-LTAG developed for English at the University of Pennsylvania (XTAG Research Group 2001) and extends it with a unification-based semantics. It contains 1,017 trees and covers the syntactic fragment of XTAG, namely, auxiliaries, copula, raising and small clause constructions, topicalization, relative clauses, infinitives, gerunds, passives, adjuncts, ditransitives (and datives), ergatives, it-clefts, wh-clefts, PRO constructions, noun-noun modification, extraposition, determiner sequences, genitives, negation, noun-verb contractions, sentential adjuncts, imperatives, and resultatives. The grammar was tested on a handbuilt test-suite of 998 sentences illustrating the various syntactic constructions meant to be covered by the grammar. All sentences in the test suite can be parsed using the grammar.",
	"cite_spans": [
	{
	"start": 38,
	"end": 59,
	"text": "(Alahverdzhieva 2008)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Other Grammars Designed with XMG",
	"sec_num": "6.2"
	},
	{
	"text": "FrenchIG. The extended XMG framework was used to design a core IG for French consisting of 2,059 tree descriptions compiled out of 448 classes (Perrier 2007) . The resulting grammar is lexicalized, and its coverage was evaluated using the previously mentioned TSNLP. The French IG accepts 88% of the grammatical sentences and rejects 85% of the ungrammatical sentences, although the current version of the French IG does not yet cover all the syntactic phenomena presented in the test suite (for example, causative and superlative constructions).",
	"cite_spans": [
	{
	"start": 143,
	"end": 157,
	"text": "(Perrier 2007)",
	"ref_id": "BIBREF46"
	},
	{
	"start": 406,
	"end": 415,
	"text": "French IG",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Other Grammars Designed with XMG",
	"sec_num": "6.2"
	},
	{
	"text": "Vietnamese TAG. The XMG language was used by Le Hong, N'Guyen, and Roussanaly (2008) to produce a core FB-LTAG for Vietnamese. Their work is rather a proof of concept than a large-scale implementation. They focused on Vietnamese's categorization frames, and were able to produce a TAG covering the following frames: intransitive (tree family N0V), transitive with a nominal complement (N0VN1), transitive with a clausal complement (N0VS1), transitive with modal complement (N0V0V1), ditransitive (N0VN1N2), ditransitive with a preposition (N0VN1ON2), ditransitive with a verbal complement (N0V0N1V1), ditransitive with an adjectival complement (N0VN1A), movement verbs with a nominal complement (N0V0V1N1), movement verbs with an adjectival complement (N0V0AV1), and movement ditransitive (N0V0N1V1N2).",
	"cite_spans": [
	{
	"start": 45,
	"end": 84,
	"text": "Le Hong, N'Guyen, and Roussanaly (2008)",
	"ref_id": "BIBREF39"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Other Grammars Designed with XMG",
	"sec_num": "6.2"
	},
	{
	"text": "GerTT. Another XMG-based grammar corresponds to the German MC-TAG of . This grammar, called GerTT, is in fact an MC-TAG with Tree Tuples (Lichte 2007) . This variant of MCTAG has been designed to model free word order phenomena. This is done by imposing node sharing constraints on MCTAG derivations (Kallmeyer 2005) . GerTT covers phenomena such as scrambling, coherent constructions, relative clauses, embedded questions, copula verbs, complementized sentences, verbs with various sub-categorization frames, nouns, prepositions, determiners, adjectives, and partly includes semantics. It is made of 103 tree tuples, compiled from 109 classes.",
	"cite_spans": [
	{
	"start": 137,
	"end": 150,
	"text": "(Lichte 2007)",
	"ref_id": "BIBREF41"
	},
	{
	"start": 300,
	"end": 316,
	"text": "(Kallmeyer 2005)",
	"ref_id": "BIBREF32"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Other Grammars Designed with XMG",
	"sec_num": "6.2"
	},
	{
	"text": "We now compare XMG with existing environments for designing tree-based grammars and briefly report on the grammars designed with these systems.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Related Work",
	"sec_num": "7."
	},
	{
	"text": "Candito's Metagrammar Compiler. The concept of metagrammar was introduced by Candito (1996) . In her paper, Candito presented a compiler for abstract specifications of FB-LTAG trees (the so-called metagrammars). Such specifications are based on three dimensions, each of them being encoded in a separate inheritance hierarchy of linguistic descriptions. Dimension 1 describes canonical sub-categorization frames (e.g., transitive), the Dimension 2 describes redistributions of syntactic functions (e.g., active to passive), and Dimension 3 the tree descriptions corresponding to the realizations of the syntactic functions defined in Dimension 2. This three-dimensional metagrammatical description is then processed by a compiler to compute FB-LTAG tree schemas. In essence, these tree schemas are produced by associating a canonical sub-categorization frame (Dimension 1) with a compatible redistribution schema (Dimension 2), and with exactly one function realization (Dimension 3) for each function required by the sub-categorization frame. Candito's (1996 Candito's ( , 1999 approach improves on previous proposals by Vijay-Shanker and Schabes (1992) and Evans, Gazdar, and Weir (1995) in that it provides a linguistically principled basis for structuring the inheritance hierarchy. As shown in Section 6.1, the XMG definition of SEMTAG uses similar principles. Candito's approach differs, however, from the XMG account in several important ways:",
	"cite_spans": [
	{
	"start": 77,
	"end": 91,
	"text": "Candito (1996)",
	"ref_id": "BIBREF12"
	},
	{
	"start": 1044,
	"end": 1059,
	"text": "Candito's (1996",
	"ref_id": "BIBREF12"
	},
	{
	"start": 1060,
	"end": 1078,
	"text": "Candito's ( , 1999",
	"ref_id": "BIBREF13"
	},
	{
	"start": 1140,
	"end": 1154,
	"text": "Schabes (1992)",
	"ref_id": "BIBREF50"
	},
	{
	"start": 1159,
	"end": 1189,
	"text": "Evans, Gazdar, and Weir (1995)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "r Much of the linguistic knowledge used to determine which classes to combine is hard-coded in the compiler (unlike in XMG, there is no explicit control on class combinations). In other words, there is no clear separation between the linguistic knowledge needed to specify a high-level FB-LTAG description and the algorithm used to compile an actual FB-LTAG from this description. This makes grammar extension and maintenance by linguists extremely difficult.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "r As in Schabes (1992) Evans, Gazdar, and Weir (1995) , the linguistic description is non-monotonic in that some erasing classes are used to remove information introduced by other dimensions (e.g., agentless passive).",
	"cite_spans": [
	{
	"start": 8,
	"end": 41,
	"text": "Schabes (1992) Evans, Gazdar, and",
	"ref_id": null
	},
	{
	"start": 42,
	"end": 53,
	"text": "Weir (1995)",
	"ref_id": "BIBREF47"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "r The approach fails to provide an easy means to state exceptions. These are usually encoded in the compiling algorithm.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "r The tree description language used to specify classes in Dimension 3 relies on global node variables. Thus, two variables with identical names introduced in different classes are expected to refer to the same tree node. As argued in Section 4, this makes it hard to design large-scale metagrammars.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "The LexOrg system. An approach similar to Candito's was presented by Xia et al. (1998) , Xia (2001) , and Xia, Palmer, and Vijay-Shanker (2005, 2010) . As in Candito's approach, a TAG abstract specification relies on a three-dimensional description made of, namely, sub-categorization frames, blocks, and lexical redistribution rules. To compile this specification into a TAG, the system selects a canonical sub-categorization frame, and applies some lexical redistribution rules to derive new frames and finally select blocks corresponding to the resulting frames. These blocks contain tree descriptions using the logic of Rogers and Vijay-Shanker (1994) .",
	"cite_spans": [
	{
	"start": 69,
	"end": 86,
	"text": "Xia et al. (1998)",
	"ref_id": null
	},
	{
	"start": 89,
	"end": 99,
	"text": "Xia (2001)",
	"ref_id": "BIBREF51"
	},
	{
	"start": 102,
	"end": 105,
	"text": "and",
	"ref_id": null
	},
	{
	"start": 106,
	"end": 122,
	"text": "Xia, Palmer, and",
	"ref_id": "BIBREF52"
	},
	{
	"start": 123,
	"end": 149,
	"text": "Vijay-Shanker (2005, 2010)",
	"ref_id": "BIBREF50"
	},
	{
	"start": 624,
	"end": 655,
	"text": "Rogers and Vijay-Shanker (1994)",
	"ref_id": "BIBREF48"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "LexOrg suffers from similar limitations as Candito's compiler. Much of the linguistic knowledge is embedded in the compiling algorithm, making it difficult for linguists to extend the grammar description and to handle exceptions. Unlike in Candito's framework, the tree description language uses local node variables and lets the tree description solver determine node identifications. Although this avoids having to memorize node names, this requires that the descriptions be constrained enough to impose the required node identifications and prevent the unwanted ones. In practice, this again complicates grammar writing. In contrast, XMG provides an intermediate solution which, by combining local variables with export declarations, avoids having to memorize too many node variable names (only those local to the relevant sub-hierarchy need memorizing) while allowing for explicit node identification.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "The Metagrammar Compiler of Gaiffe, Crabb\u00e9, and Roussanaly. Gaiffe, Crabb\u00e9, and Roussanaly (2002) proposed a compiler for FB-LTAG that aims to remedy both the lack of a clear separation between linguistic information and compilation algorithm, and the lack of explicit control on the class combinations prevalent in Candito (1996) , Xia et al. (1998) , and Xia (2001) . In their approach, the linguistic specification consists of a single inheritance hierarchy of classes, each class containing a tree description. The description logic used is similar to Candito's. That is, global node names are used. To trigger class combinations, classes are labeled with two types of information: needs and resources. The compiler selects all final classes of the hierarchy, performs all possible combinations, and only keeps those combinations that neutralize the stated needs and resources. The tree descriptions contained in these neutral combinations are then solved to produce the expected trees.",
	"cite_spans": [
	{
	"start": 28,
	"end": 97,
	"text": "Gaiffe, Crabb\u00e9, and Roussanaly. Gaiffe, Crabb\u00e9, and Roussanaly (2002)",
	"ref_id": "BIBREF26"
	},
	{
	"start": 316,
	"end": 330,
	"text": "Candito (1996)",
	"ref_id": "BIBREF12"
	},
	{
	"start": 333,
	"end": 350,
	"text": "Xia et al. (1998)",
	"ref_id": null
	},
	{
	"start": 357,
	"end": 367,
	"text": "Xia (2001)",
	"ref_id": "BIBREF51"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "Although this approach implements a clear separation between linguistic information and compilation algorithm, the fully automatic derivation of FB-LTAG trees from the inheritance hierarchy makes it difficult in practice to control overgeneration. In contrast, XMG's explicit definitions of class combinations by conjunction, disjunction, and inheritance makes it easier to control the tree set that will be generated by the compiler from the grammar specification. Additionally, the issues raised by global variables remain (no way to instantiate twice a given class, and cumbersome definition of variables in large metagrammars).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "The MGCOMP System. More recently, Villemonte de la Clergerie (2005, 2010) proposed a compiler for FB-LTAG that aims at preserving a high degree of factorization in both the abstract grammar specification and the grammar which is compiled from it. Thus, the MGCOMP system does not compute FB-LTAG elementary trees, but factorized trees.",
	"cite_spans": [
	{
	"start": 61,
	"end": 67,
	"text": "(2005,",
	"ref_id": null
	},
	{
	"start": 68,
	"end": 73,
	"text": "2010)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "In MGCOMP, like in Gaiffe, Crabb\u00e9, and Roussanaly's (2002) approach, a metagrammar consists of a single hierarchy of classes. The classes are labeled with needs and resources, and final classes of the hierarchy are combined to compute tree descriptions. The main differences with Gaiffe, Crabb\u00e9, and Roussanaly (2002) , lies in the fact that (i) a description can include new factorizing operators, such as repetition (Kleene-star operator), shuffling (interleaving of nodes), optionality, and disjunctions; and (ii) it offers namespaces to specify the scope of variables. MGCOMP's extended tree descriptions are not completely solved by the compiler. Rather, it compiles underspecified trees (also called factorized trees). With this approach, a large grammar is much smaller in terms of number of grammatical structures than a classical FB-LTAG. As a result, the grammars it compiles are only compatible with the DyALog parsing environment (Villemonte de La Clergerie 2005). And, because the linguist designs factorized trees and not actual TAG trees, debugging the metagrammar becomes harder.",
	"cite_spans": [
	{
	"start": 19,
	"end": 58,
	"text": "Gaiffe, Crabb\u00e9, and Roussanaly's (2002)",
	"ref_id": "BIBREF26"
	},
	{
	"start": 280,
	"end": 317,
	"text": "Gaiffe, Crabb\u00e9, and Roussanaly (2002)",
	"ref_id": "BIBREF26"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Environments for Designing Tree-Based Grammars",
	"sec_num": "7.1"
	},
	{
	"text": "Candito's system has been used by Candito (1999) herself to design a core FB-LTAG for French and Italian, and later by Barrier (2006) to design a FB-LTAG for adjectives in French. Xia's system (LexOrg) has been used to semi-automatically generate XTAG (Xia 2001) . De La Clergerie's system (MGCOMP) has been used to design a grammar for French named FRMG (FRench MetaGrammar) (Villemonte de la Clergerie 2010). FRMG makes use of MGCOMP's factorizing operators (e.g., shuffling operator), thus producing not sensu stricto a FB-LTAG, but a factorized FB-LTAG. FRMG is freely available, contains 207 factorized trees (having optional branches, etc.) built from 279 metagrammatical classes, and covers 95% of the TSNLP.",
	"cite_spans": [
	{
	"start": 34,
	"end": 48,
	"text": "Candito (1999)",
	"ref_id": "BIBREF13"
	},
	{
	"start": 119,
	"end": 133,
	"text": "Barrier (2006)",
	"ref_id": "BIBREF6"
	},
	{
	"start": 252,
	"end": 262,
	"text": "(Xia 2001)",
	"ref_id": "BIBREF51"
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Resources Built Using Candito, Xia, and De La Clergerie's Systems",
	"sec_num": "7.2"
	},
	{
	"text": "In this article, we presented the eXtensible MetaGrammar framework and argued that, contrary to other existing grammar writing environments for tree-based grammar, XMG is declarative, extensible, and notationally expressive. We believe that these features make XMG particularly appropriate for a fast prototyping of the kind of deep tree-based grammars that are used in applications requiring high precision in grammar modeling (e.g., language teaching, man/machine dialogue systems, data-to-text generation).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Conclusion",
	"sec_num": "8."
	},
	{
	"text": "The XMG language is documented on-line, and its compiler is open source software, freely available under the terms of the GPL-compliant CeCILL license. 34 Many grammars designed with XMG (FB-LTAG and IG for French and English, TT-MCTAG for German) are also open-source and available on-line. 35 Future research will focus on extensibility. So far, XMG has been used to design treebased grammars for different languages. We plan to extend XMG to handle other types of formalisms 36 such as dependency grammars, and to support dimensions other than syntax and semantics such as for instance, phonology or morphology. As mentioned here, XMG offers a modular architecture, making it possible to extend it relatively easily. Nonetheless, in its current state, such extensions imply modifying XMG's code. We are exploring new extensions of the formalism, which would allow the linguist to dynamically define her/his metagrammar formalism (e.g., which principles or descriptions to use) depending on the target formalism.",
	"cite_spans": [
	{
	"start": 152,
	"end": 154,
	"text": "34",
	"ref_id": null
	},
	{
	"start": 207,
	"end": 247,
	"text": "French and English, TT-MCTAG for German)",
	"ref_id": null
	},
	{
	"start": 292,
	"end": 294,
	"text": "35",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Conclusion",
	"sec_num": "8."
	},
	{
	"text": "Another interesting question concerns cross-language grammar engineering. So far, the metagrammar allows for dealing with structural redundancy. As pointed out by Kinyon et al. (2006) , a metagrammar can be used to capture generalizations across languages and is surely worth further investigating.",
	"cite_spans": [
	{
	"start": 163,
	"end": 183,
	"text": "Kinyon et al. (2006)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Conclusion",
	"sec_num": "8."
	},
	{
	"text": "Finally, we plan to extend XMG with features borrowed from Integrated Development Environments (IDE) for programming languages. Designing a grammar is, in some respect, similar to programming an application. Grammar environments should benefit from the same tools as those used for the development of applications (incremental compilation, debugger, etc.). Xia, Fei, Martha Palmer, and K. Vijay-Shanker. 2010 ",
	"cite_spans": [
	{
	"start": 357,
	"end": 408,
	"text": "Xia, Fei, Martha Palmer, and K. Vijay-Shanker. 2010",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Conclusion",
	"sec_num": "8."
	},
	{
	"text": "The elementary trees displayed in this article conform toAbeill\u00e9 (2002), that is, we reject the use of a VP constituent in French. 2 As mentioned earlier, we describe lexicalized TAG, thus every tree schema has to contain at least one anchor (node labeled ).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "C x /C x abbreviate a node with category C and a top/bottom feature structure including the feature-value pair { index : x}.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "For now, let us consider that the tree fragments are combined in order to produce minimal trees by merging nodes whose categories (and features) unify. In the next section, we will see how to precisely control node identification using either node variables or node constraints.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "Note that in XMG, it is not mandatory to define any tree structure inside SubjAgreement. We could define independent NP and VP nodes, and associate them with variables, say n 1 and n 2 . n 1 and n 2 would then be exported and reused directly in the classes CanonicalSubject and Wh-NP-Subject, respectively. 6 Note that these clauses only consider canonical subjects to avoid having both a Wh-subject and a Wh-object. This is not entirely satisfactory, as we would prefer to define a single abstraction over objects (as was done for subjects) and use it wherever possible. There would then be another mechanism to capture this exception and cause the invalid combination to fail (that is, the resulting tree description not to have any model). Such a mechanism exists in XMG, and is called linguistic principle (see Section 5).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "By coreference, we mean the sharing of information between distinct elementary fragments of the grammar specification. 8 SeeGardent and Kallmeyer (2003) for a detailed introduction to flat semantics.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "For more details on the interpretation of flat semantics and on its association with a grammar of natural language, seeGardent (2008).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "Although formally precise, this notation can be cumbersome. In the interest of legibility we adopt throughout the convention that SYN stands for SYN, , , SEM for , SEM, , and DYN for , , DYN . 11 An analogous situation may arise in English with ditransitive verbs requiring two direct objects.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "In fact, the notation used by Vijay-Shanker and Schabes (1992) is attr:X with attr an attribute variable ranging over a finite set of attributes, to indicate special node variables that scope outside their class; and attr(A) to refer to such variables from outside the entity in which they were declared. We use a different notation here to enforce consistency with the XMG notation.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "Recently, XMG has been used to describe the morphology of verbs in Ikota, a Bantu language spoken in Gabon.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "The XMG compiler is open source software released under the terms of the CeCILL GPL-compliant licence. See http://sourcesup.renater.fr/xmg.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "represents disjoint union.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "In (Examples 25a-c), the numbers on the clitics indicate their rank.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "In fact, rather than adding strict precedence constraints to the tree description, we directly add to the solver their equivalent set constraints on Eq, Up,Left, Right, Down, introduced earlier. 20 In other words, node colors can be seen as information on node saturation.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "See Crabb\u00e9 (2005) for a complete description of SEMTAG tree fragments, including feature structures. 22 Recall from Section 4 that inheritance is used to share namespaces. Thus, (node or feature) variables introduced in a given class C can be directly reused in the sub-classes of C.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "Note that, when these abstractions will be combined to describe for instance transitive verbs, the combination of WhObj with WhByObj will be ruled out by using a uniqueness principle such as introduced in Section 5.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "It is often argued that a language of grammatical representation must be equipped with an \"erasing device\" like lexical rules because of phenomena such as the passive without agent. In this framework it turns out that this kind of device is not needed because we do not grant any special status to base trees.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "Note that number and gender agreements are dealt with using coreferences between features labeling syntactic nodes, seeCrabb\u00e9 (2005).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "The principles required to handle quantification are omitted. We refer the reader to for a more extensive presentation of how semantics is implemented using XMG. 27 For simplicity, we only mention indices. To be complete, however, labels should also be used. 28 The foot projection principle only applies to foot nodes that are not argument nodes (i.e., to modifiee nodes). 29 In other words, this argument labeling principle defines an explicit and normalized reference to any realization of a semantic argument. Following FB-LTAG predicate-argument co-occurrence principle(Abeill\u00e9, Candito, and Kinyon 1999), we know that any elementary tree includes a leaf node for each realized semantic argument of its anchor. This principle thus holds in any FB-LTAG. Its implementation, however, is closely related to the architecture of the metagrammar; here we benefit from the fact that verbal arguments are described in dedicated classes to reach a high degree of factorization.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	},
	{
	"text": "As a comparison, about one hour was needed byCandito's (1999) compiler to produce a French FB-LTAG containing about 1,000 tree schemas. 33 As an alternative way to parse FB-LTAG grammars equipped with flat semantics such as those produced by XMG, one can use the T\u00fcbingen Linguistic Parsing Architecture (TuLiPA)(Kallmeyer et al. 2010).",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "",
	"sec_num": null
	}
	],
	"back_matter": [
	{
	"text": "We are grateful to the three anonymous reviewers for their valuable comments. Any remaining errors are ours.",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "Acknowledgments",
	"sec_num": null
	},
	{
	"text": "our restricted example, canonical and relativized subject. Thus, the simplified Subject class defined in Equation (31) characterizes contexts such as the following: (35) a. Jean mange.(canonical subject) 'John eats.' b. Le gar\u00e7on qui mange (relativized subject) 'The boy who eats' Similarly, the IndirectObject class abstracts over the realization of an argument introduced by the preposition\u00e0 to the right of the verb (CanonIndirObj) or realized in extracted position (possibly realized at an unbounded distance from the predicate) as illustrated by the following examples: (36) a. Jean parle\u00e0 Marie.(canonical indirect object) 'John talks to Mary.' b.\u00c0 qui Jean parle-t-il ?(wh indirect object) 'To whom is John talking ?' c.\u00c0 qui Pierre croit-il que Jean parle ?(wh indirect object) 'To whom Peter thinks that John talks ?'This way of grouping tree fragments is reminiscent of the informal classification of French syntactic functions presented by Iordanskaja and Mel'\u010duk (2009) whereby each syntactic function is associated with a set of possible syntactic constructions.Diathesis alternations. In this third level, we take advantage of the abstractions defined in the previous level to represent diathesis alternations. Again, we are interested here in describing alternatives. Diathesis alternations are those alternations of mapping between arguments and syntactic functions such as for instance the active/passive alternation. In a diathesis alternation, the actual form of the verb constrains the way predicate arguments are realized in syntax. Thus, in the following example, it is considered that both Examples (37a) and (37b) are alternative realizations of a predicate argument structure such as send(John, a letter).(37) a. Jean envoie une lettre.'John sends a letter.' b. Une lettre est envoy\u00e9e par Jean.'A letter is sent by John.'The active/passive diathesis alternation captures the fact that if the verb is in the active form, its two arguments are realized by a subject and an object whereas if the verb is in the passive form, then the arguments consist of a subject and a by-object.Finally a traditional case of \"erasing,\" 24 such as the agentless passive (or passive without agent) can be expressed in our language by adding an additional alternative where the by-object or agentive complement is not expressed. Thus Equation 39is an augmentation of (38) where we have added the agentless passive alternative (indicated in boldface).This methodology can be further augmented to implement an actual linking in the manner of Bresnan and Zaenen (1990) . For the so-called erasing cases, one can map the \"erased\" predicative argument to an empty realization in syntax. We refer the reader to Crabb\u00e9 (2005) for further details.Tree families. Finally, tree families are defined-that is, sets of trees capturing alternative realizations of a given verb type (i.e., sub-categorization frame). Continuing with the simplified example presented so far, we can for instance define the tree family for verbs taking a nominal subject, a nominal object, and an indirect nominal object (i.e., ditransitive verbs) as follows:The trees generated for such a family will, among others, handle the following contexts: 25 It is straightforward to extend the grammar with new families. Thus, for instance, Equation (42) shows how to define the transitive family (for verbs taking a nominal subject and a nominal object) and Equation (43), the intransitive one (alternatives of a verb sub-categorizing for a nominal subject).",
	"cite_spans": [
	{
	"start": 165,
	"end": 169,
	"text": "(35)",
	"ref_id": null
	},
	{
	"start": 575,
	"end": 579,
	"text": "(36)",
	"ref_id": "BIBREF9"
	},
	{
	"start": 951,
	"end": 981,
	"text": "Iordanskaja and Mel'\u010duk (2009)",
	"ref_id": null
	},
	{
	"start": 2545,
	"end": 2570,
	"text": "Bresnan and Zaenen (1990)",
	"ref_id": "BIBREF11"
	},
	{
	"start": 2710,
	"end": 2723,
	"text": "Crabb\u00e9 (2005)",
	"ref_id": null
	}
	],
	"ref_spans": [],
	"eq_spans": [],
	"section": "annex",
	"sec_num": null
	},
	{
	"text": "(42)",
	"cite_spans": [],
	"ref_spans": [],
	"eq_spans": [],
	"section": "TransitiveFamily \u2192 TransitiveDiathesis",
	"sec_num": null
	}
	],
	"bib_entries": {
	"BIBREF0": {
	"ref_id": "b0",
	"title": "Une grammaire\u00e9lectronique du fran\u00e7ais",
	"authors": [
	{
	"first": "A",
	"middle": [],
	"last": "Abeill\u00e9",
	"suffix": ""
	}
	],
	"year": 2002,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Abeill\u00e9, A. 2002. Une grammaire\u00e9lectronique du fran\u00e7ais. CNRS Editions.",
	"links": null
	},
	"BIBREF1": {
	"ref_id": "b1",
	"title": "Ftag: current status and parsing scheme",
	"authors": [
	{
	"first": "A",
	"middle": [],
	"last": "Abeill\u00e9",
	"suffix": ""
	},
	{
	"first": "M",
	"middle": [],
	"last": "Candito",
	"suffix": ""
	},
	{
	"first": "A",
	"middle": [],
	"last": "Kinyon",
	"suffix": ""
	}
	],
	"year": 1999,
	"venue": "Proceedings of Vextal '99",
	"volume": "",
	"issue": "",
	"pages": "283--292",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Abeill\u00e9, A., M. Candito, and A. Kinyon. 1999. Ftag: current status and parsing scheme. In Proceedings of Vextal '99, pages 283-292, Venice.",
	"links": null
	},
	"BIBREF2": {
	"ref_id": "b2",
	"title": "Warren's Abstract Machine: A Tutorial Reconstruction",
	"authors": [
	{
	"first": "Hassan",
	"middle": [],
	"last": "A\u00eft-Kaci",
	"suffix": ""
	}
	],
	"year": 1991,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "A\u00eft-Kaci, Hassan. 1991. Warren's Abstract Machine: A Tutorial Reconstruction. MIT Press, Cambridge, MA.",
	"links": null
	},
	"BIBREF4": {
	"ref_id": "b4",
	"title": "DotCCG and VisCCG: Wiki and programming paradigms for improved grammar engineering with OpenCCG",
	"authors": [
	{
	"first": "Jason",
	"middle": [],
	"last": "Baldridge",
	"suffix": ""
	},
	{
	"first": "Sudipta",
	"middle": [],
	"last": "Chatterjee",
	"suffix": ""
	},
	{
	"first": "Alexis",
	"middle": [],
	"last": "Palmer",
	"suffix": ""
	},
	{
	"first": "Ben",
	"middle": [],
	"last": "Wing",
	"suffix": ""
	}
	],
	"year": 2007,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Baldridge, Jason, Sudipta Chatterjee, Alexis Palmer, and Ben Wing. 2007. DotCCG and VisCCG: Wiki and programming paradigms for improved grammar engineering with OpenCCG. In Tracy Holloway King and Emily M.",
	"links": null
	},
	"BIBREF5": {
	"ref_id": "b5",
	"title": "Proceedings of the Grammar Engineering Across Framework Workshop (GEAF 07). CSLI",
	"authors": [
	{
	"first": "",
	"middle": [],
	"last": "Bender",
	"suffix": ""
	}
	],
	"year": null,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "5--25",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Bender, editors, Proceedings of the Grammar Engineering Across Framework Workshop (GEAF 07). CSLI, Stanford, CA, pages 5-25.",
	"links": null
	},
	"BIBREF6": {
	"ref_id": "b6",
	"title": "Une m\u00e9tagrammaire pour les noms pr\u00e9dicatifs du fran\u00e7ais : d\u00e9veloppement et exp\u00e9rimentations pour les grammaires TAG",
	"authors": [
	{
	"first": "S\u00e9bastien",
	"middle": [],
	"last": "Barrier",
	"suffix": ""
	}
	],
	"year": 2006,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Barrier, S\u00e9bastien. 2006. Une m\u00e9tagrammaire pour les noms pr\u00e9dicatifs du fran\u00e7ais : d\u00e9veloppement et exp\u00e9rimentations pour les grammaires TAG. Ph.D. thesis, Universit\u00e9 Paris 7.",
	"links": null
	},
	"BIBREF7": {
	"ref_id": "b7",
	"title": "HyTAG: A New Type of Tree Adjoining Grammars for Hybrid Syntactic Representation of Free Word Order Language",
	"authors": [
	{
	"first": "Tilman",
	"middle": [],
	"last": "Becker",
	"suffix": ""
	}
	],
	"year": 1993,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Becker, Tilman. 1993. HyTAG: A New Type of Tree Adjoining Grammars for Hybrid Syntactic Representation of Free Word Order Language. Ph.D. thesis, Universit\u00e4t des Saarlandes.",
	"links": null
	},
	"BIBREF9": {
	"ref_id": "b9",
	"title": "Preliminary work on cross-framework grammar engineering has been realized by Cl\u00e9ment and Kinyon (2003), who used Gaiffe et al.'s compiler to produce both a TAG and a LFG from a given metagrammar",
	"authors": [],
	"year": null,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Preliminary work on cross-framework grammar engineering has been realized by Cl\u00e9ment and Kinyon (2003), who used Gaiffe et al.'s compiler to produce both a TAG and a LFG from a given metagrammar.",
	"links": null
	},
	"BIBREF10": {
	"ref_id": "b10",
	"title": "Learn Prolog Now!",
	"authors": [
	{
	"first": "Patrick",
	"middle": [],
	"last": "Blackburn",
	"suffix": ""
	},
	{
	"first": "Johan",
	"middle": [],
	"last": "Bos",
	"suffix": ""
	},
	{
	"first": "Kristina",
	"middle": [],
	"last": "Striegnitz",
	"suffix": ""
	}
	],
	"year": 2006,
	"venue": "Texts in Computing",
	"volume": "7",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Blackburn, Patrick, Johan Bos, and Kristina Striegnitz. 2006. Learn Prolog Now!, volume 7 of Texts in Computing. College Publications, London.",
	"links": null
	},
	"BIBREF11": {
	"ref_id": "b11",
	"title": "Deep unaccusitivity in LFG",
	"authors": [
	{
	"first": "Joan",
	"middle": [],
	"last": "Bresnan",
	"suffix": ""
	},
	{
	"first": "Annie",
	"middle": [],
	"last": "Zaenen",
	"suffix": ""
	}
	],
	"year": 1990,
	"venue": "Grammatical Relations: A Cross-Theoretical Perspective. CSLI publications",
	"volume": "",
	"issue": "",
	"pages": "45--57",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Bresnan, Joan and Annie Zaenen. 1990. Deep unaccusitivity in LFG. In K. Dziwirek, P. Farell, and E. Mejias-Bikandi, editors, Grammatical Relations: A Cross-Theoretical Perspective. CSLI publications, Stanford, CA, pages 45-57.",
	"links": null
	},
	"BIBREF12": {
	"ref_id": "b12",
	"title": "A principle-based hierarchical representation of LTAGs",
	"authors": [
	{
	"first": "Marie",
	"middle": [],
	"last": "Candito",
	"suffix": ""
	}
	],
	"year": 1996,
	"venue": "Proceedings of the 16th International Conference on Computational Linguistics (COLING'96)",
	"volume": "",
	"issue": "",
	"pages": "194--199",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Candito, Marie. 1996. A principle-based hierarchical representation of LTAGs. In Proceedings of the 16th International Conference on Computational Linguistics (COLING'96), pages 194-199, Copenhagen.",
	"links": null
	},
	"BIBREF13": {
	"ref_id": "b13",
	"title": "Repr\u00e9sentation modulaire et param\u00e9trable de grammaires\u00e9lectroniques lexicalis\u00e9es : application au fran\u00e7ais et\u00e0 l'italien",
	"authors": [
	{
	"first": "Marie",
	"middle": [],
	"last": "Candito",
	"suffix": ""
	}
	],
	"year": 1999,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Candito, Marie. 1999. Repr\u00e9sentation modulaire et param\u00e9trable de grammaires\u00e9lectroniques lexicalis\u00e9es : application au fran\u00e7ais et\u00e0 l'italien. Ph.D. thesis, Universit\u00e9 Paris 7.",
	"links": null
	},
	"BIBREF14": {
	"ref_id": "b14",
	"title": "Generating parallel multilingual lfg-tag grammars from a metagrammar",
	"authors": [
	{
	"first": "Lionel",
	"middle": [],
	"last": "Cl\u00e9ment",
	"suffix": ""
	},
	{
	"first": "Alexandra",
	"middle": [],
	"last": "Kinyon",
	"suffix": ""
	}
	],
	"year": 2003,
	"venue": "Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics",
	"volume": "",
	"issue": "",
	"pages": "184--191",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Cl\u00e9ment, Lionel and Alexandra Kinyon. 2003. Generating parallel multilingual lfg-tag grammars from a metagrammar. In Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, pages 184-191, Sapporo.",
	"links": null
	},
	"BIBREF15": {
	"ref_id": "b15",
	"title": "The Non-Associativity of Polarized Tree-Based Grammars",
	"authors": [
	{
	"first": "Yael",
	"middle": [],
	"last": "Cohen-Sygal",
	"suffix": ""
	},
	{
	"first": "Shuly",
	"middle": [],
	"last": "Wintner",
	"suffix": ""
	}
	],
	"year": 2007,
	"venue": "Proceedings of the Eighth International Conference on Intelligent Text Processing and Computational Linguistics (CICLing-2007)",
	"volume": "",
	"issue": "",
	"pages": "208--217",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Cohen-Sygal, Yael and Shuly Wintner. 2007. The Non-Associativity of Polarized Tree-Based Grammars. In Proceedings of the Eighth International Conference on Intelligent Text Processing and Computational Linguistics (CICLing-2007), pages 208-217, Mexico City.",
	"links": null
	},
	"BIBREF16": {
	"ref_id": "b16",
	"title": "Associative grammar combination operators for tree-based grammars",
	"authors": [
	{
	"first": "Yael",
	"middle": [],
	"last": "Cohen-Sygal",
	"suffix": ""
	},
	{
	"first": "Shuly",
	"middle": [],
	"last": "Wintner",
	"suffix": ""
	}
	],
	"year": 2009,
	"venue": "Journal of Logic, Language and Information",
	"volume": "18",
	"issue": "3",
	"pages": "293--316",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Cohen-Sygal, Yael and Shuly Wintner. 2009. Associative grammar combination operators for tree-based grammars. Journal of Logic, Language and Information, 18(3):293-316.",
	"links": null
	},
	"BIBREF17": {
	"ref_id": "b17",
	"title": "Towards modular development of typed unification grammars",
	"authors": [
	{
	"first": "Yael",
	"middle": [],
	"last": "Cohen-Sygal",
	"suffix": ""
	},
	{
	"first": "Shuly",
	"middle": [],
	"last": "Wintner",
	"suffix": ""
	}
	],
	"year": 2011,
	"venue": "Computational Linguistics",
	"volume": "37",
	"issue": "1",
	"pages": "29--74",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Cohen-Sygal, Yael and Shuly Wintner. 2011. Towards modular development of typed unification grammars. Computational Linguistics, 37(1):29-74.",
	"links": null
	},
	"BIBREF18": {
	"ref_id": "b18",
	"title": "An open-source grammar development environment and broad-coverage English grammar using HPSG",
	"authors": [
	{
	"first": "Ann",
	"middle": [],
	"last": "Copestake",
	"suffix": ""
	},
	{
	"first": "Dan",
	"middle": [],
	"last": "Flickinger",
	"suffix": ""
	}
	],
	"year": 2000,
	"venue": "Proceedings of the Second Conference on Language Resources and Evaluation (LREC-2000)",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Copestake, Ann and Dan Flickinger. 2000. An open-source grammar development environment and broad-coverage English grammar using HPSG. In Proceedings of the Second Conference on Language Resources and Evaluation (LREC-2000), Athens.",
	"links": null
	},
	"BIBREF19": {
	"ref_id": "b19",
	"title": "Repr\u00e9sentation informatique de grammaires fortement lexicalis\u00e9es : Application\u00e0 la grammaire d'arbres adjoints",
	"authors": [
	{
	"first": "Ann",
	"middle": [],
	"last": "Copestake",
	"suffix": ""
	},
	{
	"first": "Alex",
	"middle": [],
	"last": "Lascarides",
	"suffix": ""
	},
	{
	"first": "Dan",
	"middle": [],
	"last": "Flickinger",
	"suffix": ""
	}
	],
	"year": 2001,
	"venue": "Proceedings of 39th Annual Meeting of the Association for Computational Linguistics",
	"volume": "",
	"issue": "",
	"pages": "140--147",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Copestake, Ann, Alex Lascarides, and Dan Flickinger. 2001. An algebra for semantic construction in constraint-based grammars. In Proceedings of 39th Annual Meeting of the Association for Computational Linguistics, pages 140-147, Toulouse. Crabb\u00e9, Benoit. 2005. Repr\u00e9sentation informatique de grammaires fortement lexicalis\u00e9es : Application\u00e0 la grammaire d'arbres adjoints. Ph.D. thesis, Universit\u00e9 Nancy 2.",
	"links": null
	},
	"BIBREF20": {
	"ref_id": "b20",
	"title": "Metagrammar redux",
	"authors": [
	{
	"first": "Beno\u00eet",
	"middle": [],
	"last": "Crabb\u00e9",
	"suffix": ""
	},
	{
	"first": "Denys",
	"middle": [],
	"last": "Duchier",
	"suffix": ""
	}
	],
	"year": 2004,
	"venue": "Proceedings of the Workshop on Constraint Solving for Language Processing",
	"volume": "",
	"issue": "",
	"pages": "32--47",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Crabb\u00e9, Beno\u00eet and Denys Duchier. 2004. Metagrammar redux. In Proceedings of the Workshop on Constraint Solving for Language Processing (CSLP 2004), pages 32-47, Copenhagen.",
	"links": null
	},
	"BIBREF21": {
	"ref_id": "b21",
	"title": "Describing morphologically-rich languages using metagrammars: A look at verbs in Ikota",
	"authors": [
	{
	"first": "Denys",
	"middle": [],
	"last": "Duchier",
	"suffix": ""
	},
	{
	"first": "Yannick",
	"middle": [],
	"last": "Brunelle Magnana Ekoukou",
	"suffix": ""
	},
	{
	"first": "Simon",
	"middle": [],
	"last": "Parmentier",
	"suffix": ""
	},
	{
	"first": "Emmanuel",
	"middle": [],
	"last": "Petitjean",
	"suffix": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Schang",
	"suffix": ""
	}
	],
	"year": 2012,
	"venue": "8th SALTMIL Workshop on Minority Languages and 4th Workshop on African Language Technology, International Conference on Language Resources and Evaluation, LREC 2012",
	"volume": "",
	"issue": "",
	"pages": "55--60",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Duchier, Denys, Brunelle Magnana Ekoukou, Yannick Parmentier, Simon Petitjean, and Emmanuel Schang. 2012. Describing morphologically-rich languages using metagrammars: A look at verbs in Ikota. In Workshop on \"Language Technology for Normalisation of Less-resourced Languages,\" 8th SALTMIL Workshop on Minority Languages and 4th Workshop on African Language Technology, International Conference on Language Resources and Evaluation, LREC 2012, pages 55-60, Istanbul.",
	"links": null
	},
	"BIBREF22": {
	"ref_id": "b22",
	"title": "Dominance constraints with set operators",
	"authors": [
	{
	"first": "Denys",
	"middle": [],
	"last": "Duchier",
	"suffix": ""
	},
	{
	"first": "Joachim",
	"middle": [],
	"last": "Niehren",
	"suffix": ""
	}
	],
	"year": 2000,
	"venue": "Proceedings of the First International Conference on Computational Logic",
	"volume": "1861",
	"issue": "",
	"pages": "326--341",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Duchier, Denys and Joachim Niehren. 2000. Dominance constraints with set operators. In John W. Lloyd, Ver\u00f3nica Dahl, Ulrich Furbach, Manfred Kerber, Kung-Kiu Lau, Catuscia Palamidessi, Lu\u00eds Moniz Pereira, Yehoshua Sagiv, and Peter J. Stuckey, editors, Proceedings of the First International Conference on Computational Logic, volume 1861 of Lecture Notes in Computer Science. Springer, Berlin, pages 326-341.",
	"links": null
	},
	"BIBREF23": {
	"ref_id": "b23",
	"title": "Metagrammars as logic programs",
	"authors": [
	{
	"first": "Denys",
	"middle": [],
	"last": "Duchier",
	"suffix": ""
	},
	{
	"first": "Yannick",
	"middle": [],
	"last": "Parmentier",
	"suffix": ""
	},
	{
	"first": "Simon",
	"middle": [],
	"last": "Petitjean",
	"suffix": ""
	}
	],
	"year": 2012,
	"venue": "International Conference on Logical Aspects of Computational Linguistics",
	"volume": "",
	"issue": "",
	"pages": "1--4",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Duchier, Denys, Yannick Parmentier, and Simon Petitjean. 2012. Metagrammars as logic programs. In International Conference on Logical Aspects of Computational Linguistics (LACL 2012). Proceedings of the Demo Session, pages 1-4, Nantes.",
	"links": null
	},
	"BIBREF24": {
	"ref_id": "b24",
	"title": "Encoding lexicalized tree adjoining grammars with a nonmonotonic inheritance hierarchy",
	"authors": [
	{
	"first": "Denys",
	"middle": [],
	"last": "Duchier",
	"suffix": ""
	},
	{
	"first": "Stefan",
	"middle": [],
	"last": "Thater",
	"suffix": ""
	}
	],
	"year": 1995,
	"venue": "Proceedings of the Sixth International Workshop on Natural Language Understanding and Logic Programming (NLULP'99)",
	"volume": "",
	"issue": "",
	"pages": "77--84",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Duchier, Denys and Stefan Thater. 1999. Parsing with tree descriptions: A constraint-based approach. In Proceedings of the Sixth International Workshop on Natural Language Understanding and Logic Programming (NLULP'99), pages 17-32, Las Cruces, NM. Evans, Roger, Gerald Gazdar, and David Weir. 1995. Encoding lexicalized tree adjoining grammars with a nonmonotonic inheritance hierarchy. In Proceedings of the 33rd Annual Meeting of the Association for Computational Linguistics, pages 77-84, Cambridge, MA.",
	"links": null
	},
	"BIBREF25": {
	"ref_id": "b25",
	"title": "Lexical Rules in the Hierarchical Lexicon",
	"authors": [
	{
	"first": "Daniel",
	"middle": [],
	"last": "Flickinger",
	"suffix": ""
	}
	],
	"year": 1987,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Flickinger, Daniel. 1987. Lexical Rules in the Hierarchical Lexicon. Ph.D. thesis, Stanford University.",
	"links": null
	},
	"BIBREF26": {
	"ref_id": "b26",
	"title": "Tree adjoining grammar, semantic calculi and labelling invariants",
	"authors": [
	{
	"first": "Bertrand",
	"middle": [],
	"last": "Gaiffe",
	"suffix": ""
	},
	{
	"first": "Beno\u00eet",
	"middle": [],
	"last": "Crabb\u00e9",
	"suffix": ""
	},
	{
	"first": "Azim",
	"middle": [],
	"last": "Roussanaly",
	"suffix": ""
	}
	],
	"year": 2002,
	"venue": "Proceedings of the Sixth International Workshop on Tree Adjoining Grammars and Related Frameworks (TAG+6)",
	"volume": "",
	"issue": "",
	"pages": "101--108",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Gaiffe, Bertrand, Beno\u00eet Crabb\u00e9, and Azim Roussanaly. 2002. A new metagrammar compiler. In Proceedings of the Sixth International Workshop on Tree Adjoining Grammars and Related Frameworks (TAG+6), pages 101-108, Venice. Gardent, Claire. 2007. Tree adjoining grammar, semantic calculi and labelling invariants. In Proceedings of the International Workshop on Computational Semantics (IWCS), Tilburg.",
	"links": null
	},
	"BIBREF27": {
	"ref_id": "b27",
	"title": "Integrating a unification-based semantics in a large scale lexicalised tree adjoininig grammar for French",
	"authors": [
	{
	"first": "Claire",
	"middle": [],
	"last": "Gardent",
	"suffix": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Manchester",
	"suffix": ""
	},
	{
	"first": "Claire",
	"middle": [],
	"last": "Gardent",
	"suffix": ""
	},
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	}
	],
	"year": 2003,
	"venue": "Proceedings of the 10th Conference of the European Chapter of the Association for Computational Linguistics",
	"volume": "",
	"issue": "",
	"pages": "328--335",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Gardent, Claire. 2008. Integrating a unification-based semantics in a large scale lexicalised tree adjoininig grammar for French. In Proceedings of the 22nd International Conference on Computational Linguistics (COLING'08), pages 249-256, Manchester. Gardent, Claire and Laura Kallmeyer. 2003. Semantic construction in feature-based tree adjoining grammar. In Proceedings of the 10th Conference of the European Chapter of the Association for Computational Linguistics, pages 123-130, Budapest. Gardent, Claire and Eric Kow. 2007. A symbolic approach to near-deterministic surface realisation using tree adjoining grammar. In 45th Annual Meeting of the Association for Computational Linguistics, pages 328-335, Prague.",
	"links": null
	},
	"BIBREF28": {
	"ref_id": "b28",
	"title": "Large scale semantic construction for tree adjoining grammars",
	"authors": [
	{
	"first": "Claire",
	"middle": [],
	"last": "Gardent",
	"suffix": ""
	},
	{
	"first": "Yannick",
	"middle": [],
	"last": "Parmentier",
	"suffix": ""
	}
	],
	"year": 2005,
	"venue": "Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics (COLING/ACL 2006) Main Conference Poster Sessions",
	"volume": "",
	"issue": "",
	"pages": "247--254",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Gardent, Claire and Yannick Parmentier. 2005. Large scale semantic construction for tree adjoining grammars. In Proceedings of the Fifth International Conference on Logical Aspects of Computational Linguistics (LACL'05), pages 131-146, Bordeaux. Gardent, Claire and Yannick Parmentier. 2006. Coreference Handling in XMG. In Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics (COLING/ACL 2006) Main Conference Poster Sessions, pages 247-254, Sydney.",
	"links": null
	},
	"BIBREF29": {
	"ref_id": "b29",
	"title": "Establishing an inventory of surfacesyntactic relations: valence-controlled surface-dependents of the verb in French",
	"authors": [
	{
	"first": "Claire",
	"middle": [],
	"last": "Gardent",
	"suffix": ""
	},
	{
	"first": "Yannick",
	"middle": [],
	"last": "Parmentier",
	"suffix": ""
	}
	],
	"year": 2007,
	"venue": "Proceedings of the 45th Annual Meeting of the Association for Computational Linguistics Companion Volume Proceedings of the Demo and Poster Sessions",
	"volume": "",
	"issue": "",
	"pages": "151--234",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Gardent, Claire and Yannick Parmentier. 2007. SemTAG: A platform for specifying tree adjoining grammars and performing TAG-based semantic construction. In Proceedings of the 45th Annual Meeting of the Association for Computational Linguistics Companion Volume Proceedings of the Demo and Poster Sessions, pages 13-16, Prague. Iordanskaja, Lidija and Igor Mel'\u010duk, 2009. Establishing an inventory of surface- syntactic relations: valence-controlled surface-dependents of the verb in French. In A. Polgu\u00e8re and I. A. Mel'duk, editors, Dependency in Linguistic Description. John Benjamins, Amsterdam, pages 151-234.",
	"links": null
	},
	"BIBREF30": {
	"ref_id": "b30",
	"title": "Tree adjunct grammars",
	"authors": [
	{
	"first": "Aravind",
	"middle": [
	"K"
	],
	"last": "Joshi",
	"suffix": ""
	},
	{
	"first": "Leon",
	"middle": [
	"S"
	],
	"last": "Levy",
	"suffix": ""
	},
	{
	"first": "Masako",
	"middle": [],
	"last": "Takahashi",
	"suffix": ""
	}
	],
	"year": 1975,
	"venue": "Journal of Computer and System Sciences",
	"volume": "10",
	"issue": "1",
	"pages": "136--163",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Joshi, Aravind K., Leon S. Levy, and Masako Takahashi. 1975. Tree adjunct grammars. Journal of Computer and System Sciences, 10(1):136-163.",
	"links": null
	},
	"BIBREF31": {
	"ref_id": "b31",
	"title": "Tree Description Grammars and Underspecified Representations",
	"authors": [
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	}
	],
	"year": 1999,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Kallmeyer, Laura. 1999. Tree Description Grammars and Underspecified Representations. Ph.D. thesis, Universit\u00e4t T\u00fcbingen.",
	"links": null
	},
	"BIBREF32": {
	"ref_id": "b32",
	"title": "Tree-local multicomponent tree-adjoining grammars with shared nodes",
	"authors": [
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	}
	],
	"year": 2005,
	"venue": "Computational Linguistics",
	"volume": "31",
	"issue": "2",
	"pages": "187--226",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Kallmeyer, Laura. 2005. Tree-local multicomponent tree-adjoining grammars with shared nodes. Computational Linguistics, 31(2):187-226.",
	"links": null
	},
	"BIBREF33": {
	"ref_id": "b33",
	"title": "Developing a TT-MCTAG for German with an RCG-based parser",
	"authors": [
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	},
	{
	"first": "Timm",
	"middle": [],
	"last": "Lichte",
	"suffix": ""
	},
	{
	"first": "Wolfgang",
	"middle": [],
	"last": "Maier",
	"suffix": ""
	},
	{
	"first": "Yannick",
	"middle": [],
	"last": "Parmentier",
	"suffix": ""
	},
	{
	"first": "Johannes",
	"middle": [],
	"last": "Dellert",
	"suffix": ""
	}
	],
	"year": 2008,
	"venue": "Proceedings of the Sixth Language Resources and Evaluation Conference (LREC)",
	"volume": "",
	"issue": "",
	"pages": "782--789",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Kallmeyer, Laura, Timm Lichte, Wolfgang Maier, Yannick Parmentier, and Johannes Dellert. 2008. Developing a TT-MCTAG for German with an RCG-based parser. In Proceedings of the Sixth Language Resources and Evaluation Conference (LREC), pages 782-789, Marrakech.",
	"links": null
	},
	"BIBREF34": {
	"ref_id": "b34",
	"title": "TuLiPA-Parsing extensions of TAG with range concatenation grammars",
	"authors": [
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	},
	{
	"first": "Wolfgang",
	"middle": [],
	"last": "Maier",
	"suffix": ""
	},
	{
	"first": "Yannick",
	"middle": [],
	"last": "Parmentier",
	"suffix": ""
	},
	{
	"first": "Johannes",
	"middle": [],
	"last": "Dellert",
	"suffix": ""
	}
	],
	"year": 2010,
	"venue": "Bulletin of the Polish Academy of Sciences: Technical Sciences",
	"volume": "58",
	"issue": "3",
	"pages": "377--392",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Kallmeyer, Laura, Wolfgang Maier, Yannick Parmentier, and Johannes Dellert. 2010. TuLiPA-Parsing extensions of TAG with range concatenation grammars. Bulletin of the Polish Academy of Sciences: Technical Sciences, 58(3):377-392.",
	"links": null
	},
	"BIBREF35": {
	"ref_id": "b35",
	"title": "Scope and situation binding in LTAG using semantic unification",
	"authors": [
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	},
	{
	"first": "Maribel",
	"middle": [],
	"last": "Romero",
	"suffix": ""
	},
	{
	"first": ";",
	"middle": [],
	"last": "Vancouver",
	"suffix": ""
	},
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	},
	{
	"first": "Maribel",
	"middle": [],
	"last": "Romero",
	"suffix": ""
	}
	],
	"year": 2004,
	"venue": "Proceedings of 7th International Workshop on Tree-Adjoining Grammar and Related Formalisms (TAG+7)",
	"volume": "6",
	"issue": "",
	"pages": "3--52",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Kallmeyer, Laura and Maribel Romero. 2004a. LTAG semantics for questions. In Proceedings of 7th International Workshop on Tree-Adjoining Grammar and Related Formalisms (TAG+7), pages 186-193, Vancouver. Kallmeyer, Laura and Maribel Romero. 2004b. LTAG semantics with semantic unification. In Proceedings of 7th International Workshop on Tree-Adjoining Grammar and Related Formalisms (TAG+7), page 155-162, Vancouver. Kallmeyer, Laura and Maribel Romero. 2008. Scope and situation binding in LTAG using semantic unification. Research on Language and Computation, 6(1):3-52.",
	"links": null
	},
	"BIBREF36": {
	"ref_id": "b36",
	"title": "Lexical resource reconciliation in the Xerox linguistic environment",
	"authors": [
	{
	"first": "Ronald",
	"middle": [],
	"last": "Kaplan",
	"suffix": ""
	},
	{
	"first": "Paula",
	"middle": [],
	"last": "Newman",
	"suffix": ""
	}
	],
	"year": 1997,
	"venue": "Proceedings of the ACL Workshop on Computational Environments for Grammar Development and Linguistic Engineering",
	"volume": "",
	"issue": "",
	"pages": "446--452",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Kaplan, Ronald and Paula Newman. 1997. Lexical resource reconciliation in the Xerox linguistic environment. In Proceedings of the ACL Workshop on Computational Environments for Grammar Development and Linguistic Engineering, pages 54-61, Madrid. Kinyon, Alexandra. 2000. Hypertags. In Proceedings of the 18th International Conference on Computational Linguistics (COLING'00), pages 446-452, Saarbr\u00fccken.",
	"links": null
	},
	"BIBREF38": {
	"ref_id": "b38",
	"title": "The metagrammar goes multilingual: A cross-linguistic look at the v2-phenomenon",
	"authors": [
	{
	"first": "",
	"middle": [],
	"last": "Joshi",
	"suffix": ""
	}
	],
	"year": 2006,
	"venue": "Proceedings of the Eighth International Workshop on Tree Adjoining Grammar and Related Formalisms",
	"volume": "",
	"issue": "",
	"pages": "17--24",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Joshi. 2006. The metagrammar goes multilingual: A cross-linguistic look at the v2-phenomenon. In Proceedings of the Eighth International Workshop on Tree Adjoining Grammar and Related Formalisms, pages 17-24, Sydney.",
	"links": null
	},
	"BIBREF39": {
	"ref_id": "b39",
	"title": "A metagrammar for Vietnamese",
	"authors": [
	{
	"first": "Le",
	"middle": [],
	"last": "Hong",
	"suffix": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Phuong",
	"suffix": ""
	},
	{
	"first": "Azim",
	"middle": [],
	"last": "Thi-Min-Huyen N'guyen",
	"suffix": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Roussanaly",
	"suffix": ""
	}
	],
	"year": 2008,
	"venue": "Proceedings of the 9th International Workshop on Tree-Adjoining Grammar and Related Formalisms (TAG+9)",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Le Hong, Phuong, Thi-Min-Huyen N'Guyen, and Azim Roussanaly. 2008. A metagrammar for Vietnamese. In Proceedings of the 9th International Workshop on Tree-Adjoining Grammar and Related Formalisms (TAG+9), T\u00fcbingen.",
	"links": null
	},
	"BIBREF40": {
	"ref_id": "b40",
	"title": "TSNLP-Test suites for natural language processing",
	"authors": [
	{
	"first": "Sabine",
	"middle": [],
	"last": "Lehmann",
	"suffix": ""
	},
	{
	"first": "Stephan",
	"middle": [],
	"last": "Oepen",
	"suffix": ""
	},
	{
	"first": "Sylvie",
	"middle": [],
	"last": "Regnier-Prost",
	"suffix": ""
	},
	{
	"first": "Klaus",
	"middle": [],
	"last": "Netter",
	"suffix": ""
	},
	{
	"first": "Veronika",
	"middle": [],
	"last": "Lux",
	"suffix": ""
	},
	{
	"first": "Judith",
	"middle": [],
	"last": "Klein",
	"suffix": ""
	},
	{
	"first": "Kirsten",
	"middle": [],
	"last": "Falkedal",
	"suffix": ""
	},
	{
	"first": "Frederik",
	"middle": [],
	"last": "Fouvry",
	"suffix": ""
	},
	{
	"first": "Dominique",
	"middle": [],
	"last": "Estival",
	"suffix": ""
	},
	{
	"first": "Eva",
	"middle": [],
	"last": "Dauphin",
	"suffix": ""
	},
	{
	"first": "Herv\u00e9",
	"middle": [],
	"last": "Compagnion",
	"suffix": ""
	},
	{
	"first": "Judith",
	"middle": [],
	"last": "Baur",
	"suffix": ""
	},
	{
	"first": "Lorna",
	"middle": [],
	"last": "Balkan",
	"suffix": ""
	},
	{
	"first": "Doug",
	"middle": [],
	"last": "Arnold",
	"suffix": ""
	}
	],
	"year": 1996,
	"venue": "Proceedings of the 16th International Conference on Computational Linguistics (COLING'96)",
	"volume": "",
	"issue": "",
	"pages": "711--716",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Lehmann, Sabine, Stephan Oepen, Sylvie Regnier-Prost, Klaus Netter, Veronika Lux, Judith Klein, Kirsten Falkedal, Frederik Fouvry, Dominique Estival, Eva Dauphin, Herv\u00e9 Compagnion, Judith Baur, Lorna Balkan, and Doug Arnold. 1996. TSNLP- Test suites for natural language processing. In Proceedings of the 16th International Conference on Computational Linguistics (COLING'96), pages 711-716, Copenhagen.",
	"links": null
	},
	"BIBREF41": {
	"ref_id": "b41",
	"title": "SemTAG: une plate-forme pour le calcul s\u00e9mantique\u00e0 partir de Grammaires d'Arbres Adjoints",
	"authors": [
	{
	"first": "Timm",
	"middle": [],
	"last": "Lichte",
	"suffix": ""
	}
	],
	"year": 1998,
	"venue": "Fourth International Workshop on Tree Adjoining Grammars and Related Frameworks",
	"volume": "",
	"issue": "",
	"pages": "112--115",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Lichte, Timm. 2007. An MCTAG with tuples for coherent constructions in German. In Proceedings of the 12th Conference on Formal Grammar (FG 2007), 12 pages, Dublin. Muskens, Reinhard and Emiel Krahmer. 1998. Description theory, LTAGs and Underspecified Semantics. In Fourth International Workshop on Tree Adjoining Grammars and Related Frameworks, pages 112-115, Philadelphia, PA. Parmentier, Yannick. 2007. SemTAG: une plate-forme pour le calcul s\u00e9mantique\u00e0 partir de Grammaires d'Arbres Adjoints. Ph.D. thesis, Universit\u00e9 Henri Poincar\u00e9 -Nancy.",
	"links": null
	},
	"BIBREF42": {
	"ref_id": "b42",
	"title": "XMG: eXtending MetaGrammars to MCTAG",
	"authors": [
	{
	"first": "Yannick",
	"middle": [],
	"last": "Parmentier",
	"suffix": ""
	},
	{
	"first": "Laura",
	"middle": [],
	"last": "Kallmeyer",
	"suffix": ""
	},
	{
	"first": "Timm",
	"middle": [],
	"last": "Lichte",
	"suffix": ""
	},
	{
	"first": "Wolfgang",
	"middle": [],
	"last": "Maier",
	"suffix": ""
	}
	],
	"year": 2007,
	"venue": "Proceedings of the Workshop on High-Level Syntactic Formalisms, 14 th Conference on Natural Language Processing (TALN'2007)",
	"volume": "",
	"issue": "",
	"pages": "473--482",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Parmentier, Yannick, Laura Kallmeyer, Timm Lichte, and Wolfgang Maier. 2007. XMG: eXtending MetaGrammars to MCTAG. In Proceedings of the Workshop on High-Level Syntactic Formalisms, 14 th Conference on Natural Language Processing (TALN'2007), pages 473-482, Toulouse.",
	"links": null
	},
	"BIBREF43": {
	"ref_id": "b43",
	"title": "Definite clause grammars for language analysis-A survey of the formalism and a comparison to augmented transition networks",
	"authors": [
	{
	"first": "Fernando",
	"middle": [],
	"last": "Pereira",
	"suffix": ""
	},
	{
	"first": "David",
	"middle": [],
	"last": "Warren",
	"suffix": ""
	}
	],
	"year": 1980,
	"venue": "Artificial Intelligence",
	"volume": "13",
	"issue": "",
	"pages": "231--278",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Pereira, Fernando and David Warren. 1980. Definite clause grammars for language analysis-A survey of the formalism and a comparison to augmented transition networks. Artificial Intelligence, 13:231-278.",
	"links": null
	},
	"BIBREF44": {
	"ref_id": "b44",
	"title": "Surface structure constraints in syntax",
	"authors": [
	{
	"first": "David",
	"middle": [],
	"last": "Perlmutter",
	"suffix": ""
	}
	],
	"year": 1970,
	"venue": "Linguistic Inquiry",
	"volume": "1",
	"issue": "",
	"pages": "187--255",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Perlmutter, David. 1970. Surface structure constraints in syntax. Linguistic Inquiry, 1:187-255.",
	"links": null
	},
	"BIBREF45": {
	"ref_id": "b45",
	"title": "Interaction grammars",
	"authors": [
	{
	"first": "Guy",
	"middle": [],
	"last": "Perrier",
	"suffix": ""
	}
	],
	"year": 2000,
	"venue": "Proceedings of the 18th International Conference on Computational Linguistics (COLING 2000)",
	"volume": "",
	"issue": "",
	"pages": "600--606",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Perrier, Guy. 2000. Interaction grammars. In Proceedings of the 18th International Conference on Computational Linguistics (COLING 2000), pages 600-606, Saarbr\u00fccken.",
	"links": null
	},
	"BIBREF46": {
	"ref_id": "b46",
	"title": "Generating the XTAG English grammar using metarules",
	"authors": [
	{
	"first": "Guy",
	"middle": [],
	"last": "Perrier",
	"suffix": ""
	}
	],
	"year": 2002,
	"venue": "Proceedings of the 19th International Conference on Computational Linguistics (COLING'2002)",
	"volume": "",
	"issue": "",
	"pages": "814--820",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Perrier, Guy. 2007. A French interaction grammar. In Proceedings of the 6th Conference on Recent Advances in Natural Language Processing (RANLP 2007), pages 463-467, Borovets. Prolo, Carlos A. 2002. Generating the XTAG English grammar using metarules. In Proceedings of the 19th International Conference on Computational Linguistics (COLING'2002), pages 814-820, Taipei.",
	"links": null
	},
	"BIBREF47": {
	"ref_id": "b47",
	"title": "D-tree grammars",
	"authors": [
	{
	"first": "Owen",
	"middle": [],
	"last": "Rambow",
	"suffix": ""
	},
	{
	"first": "K",
	"middle": [],
	"last": "Vijay-Shanker",
	"suffix": ""
	},
	{
	"first": "David",
	"middle": [],
	"last": "Weir",
	"suffix": ""
	}
	],
	"year": 1995,
	"venue": "Proceedings of the 33th Meeting of the Association for Computational Linguistics",
	"volume": "",
	"issue": "",
	"pages": "151--158",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Rambow, Owen, K. Vijay-Shanker, and David Weir. 1995. D-tree grammars. In Proceedings of the 33th Meeting of the Association for Computational Linguistics, pages 151-158, Cambridge, MA.",
	"links": null
	},
	"BIBREF48": {
	"ref_id": "b48",
	"title": "Obtaining trees from their descriptions: An application to tree-adjoining grammars",
	"authors": [
	{
	"first": "James",
	"middle": [],
	"last": "Rogers",
	"suffix": ""
	},
	{
	"first": "K",
	"middle": [],
	"last": "Vijay-Shanker",
	"suffix": ""
	}
	],
	"year": 1994,
	"venue": "Computational Intelligence",
	"volume": "10",
	"issue": "",
	"pages": "401--421",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Rogers, James and K. Vijay-Shanker. 1994. Obtaining trees from their descriptions: An application to tree-adjoining grammars. Computational Intelligence, 10:401-421.",
	"links": null
	},
	"BIBREF49": {
	"ref_id": "b49",
	"title": "Extended DCG notation: A tool for applicative programming in prolog",
	"authors": [
	{
	"first": "Stuart",
	"middle": [
	"M"
	],
	"last": "Shieber",
	"suffix": ""
	}
	],
	"year": 1984,
	"venue": "Proceedings of the Tenth International Conference on Computational Linguistics",
	"volume": "",
	"issue": "",
	"pages": "362--366",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Shieber, Stuart M. 1984. The design of a computer language for linguistic information. In Proceedings of the Tenth International Conference on Computational Linguistics, pages 362-366, Stanford, CA. Van Roy, Peter. 1990. Extended DCG notation: A tool for applicative programming in prolog. Technical Report UCB/CSD 90/583, University of California, Berkeley.",
	"links": null
	},
	"BIBREF50": {
	"ref_id": "b50",
	"title": "DyALog: a tabular logic programming based environment for NLP",
	"authors": [
	{
	"first": "K",
	"middle": [],
	"last": "Vijay-Shanker",
	"suffix": ""
	},
	{
	"first": "K",
	"middle": [],
	"last": "Aravind",
	"suffix": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Joshi",
	"suffix": ""
	},
	{
	"first": "K",
	"middle": [],
	"last": "Vijay-Shanker",
	"suffix": ""
	},
	{
	"first": "Yves",
	"middle": [],
	"last": "Schabes",
	"suffix": ""
	}
	],
	"year": 1988,
	"venue": "Proceedings of the 10th International Workshop on Tree-Adjoining Grammar and Related Formalisms (TAG+10)",
	"volume": "",
	"issue": "",
	"pages": "111--118",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Vijay-Shanker, K. and Aravind K. Joshi. 1988. Feature structures based tree adjoining grammars. In Proceedings of the 12th Conference on Computational Linguistics (COLING'88), pages 714-719, Budapest. Vijay-Shanker, K. and Yves Schabes. 1992. Structure sharing in lexicalized tree adjoining grammars. In Proceedings of the 14th International Conference on Computational Linguistics (COLING'92), pages 205-212, Nantes. Villemonte de La Clergerie,\u00c9ric. 2005. DyALog: a tabular logic programming based environment for NLP. In Proceedings of 2nd International Workshop on Constraint Solving and Language Processing (CSLP'05), pages 18-33, Barcelona. Villemonte de la Clergerie,\u00c9ric. 2010. Building factorized TAGs with meta-grammars. In Proceedings of the 10th International Workshop on Tree-Adjoining Grammar and Related Formalisms (TAG+10), pages 111-118, New Haven, CT. Weir, David J. 1988. Characterizing Mildly Context-Sensitive Grammar Formalisms. Ph.D. thesis, University of Pennsylvania.",
	"links": null
	},
	"BIBREF51": {
	"ref_id": "b51",
	"title": "Automatic Grammar Generation from Two Different Perspectives",
	"authors": [
	{
	"first": "Fei",
	"middle": [],
	"last": "Xia",
	"suffix": ""
	}
	],
	"year": 2001,
	"venue": "",
	"volume": "",
	"issue": "",
	"pages": "",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Xia, Fei. 2001. Automatic Grammar Generation from Two Different Perspectives. Ph.D. thesis, University of Pennsylvania.",
	"links": null
	},
	"BIBREF52": {
	"ref_id": "b52",
	"title": "Automatically generating tree adjoining grammars from abstract specifications",
	"authors": [
	{
	"first": "Fei",
	"middle": [],
	"last": "Xia",
	"suffix": ""
	},
	{
	"first": "Martha",
	"middle": [],
	"last": "Palmer",
	"suffix": ""
	},
	{
	"first": "K",
	"middle": [],
	"last": "Vijay-Shanker",
	"suffix": ""
	},
	{
	"first": ";",
	"middle": [],
	"last": "Xia",
	"suffix": ""
	},
	{
	"first": "Martha",
	"middle": [],
	"last": "Fei",
	"suffix": ""
	},
	{
	"first": "K",
	"middle": [],
	"last": "Palmer",
	"suffix": ""
	},
	{
	"first": "",
	"middle": [],
	"last": "Vijay-Shanker",
	"suffix": ""
	}
	],
	"year": 1999,
	"venue": "Proceedings of the 5th Natural Language Processing Pacific Rim Symposium (NLPRS-99)",
	"volume": "21",
	"issue": "",
	"pages": "246--287",
	"other_ids": {},
	"num": null,
	"urls": [],
	"raw_text": "Xia, Fei, Martha Palmer, and K. Vijay- Shanker. 1999. Toward semi-automating grammar development. In Proceedings of the 5th Natural Language Processing Pacific Rim Symposium (NLPRS-99), pages 96-101, Beijing. Xia, Fei, Martha Palmer, and K. Vijay- Shanker. 2005. Automatically generating tree adjoining grammars from abstract specifications. Journal of Computational Intelligence, 21(3):246-287.",
	"links": null
	}
	},
	"ref_entries": {
	"FIGREF0": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "Sample derivation of Marie a vu Jean 'Mary has seen John' in a TAG."
	},
	"FIGREF1": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": "Organization of elementary fragments in an inheritance hierarchy."
	},
	"FIGREF2": {
	"uris": null,
	"type_str": "figure",
	"num": null,
	"text": ". . . }, a semantic class FunctionSem which associates with an (exported) node variable called FunctionNode the feature value pair [index : I] and a DYN constraint of the form index Function : I . For instance, the class SubjectSem associates the node SubjectNode with the feature value pair [index : I] and the DYN constraint index subject : I . SubjectSem \u2192 SubjectNode [index : I] \u2227 index subject : I (53)"
	},
	"TABREF2": {
	"num": null,
	"html": null,
	"content": "<table><tr><td>Logic Program</td><td>XMG Metagrammar</td></tr><tr><td>Clause ::= Head \u2192 Body</td><td>Class ::= Name \u2192 Content</td></tr><tr><td>Body ::= Fact \| Head \|</td><td>Content ::</td></tr><tr><td>Body \u2228 Body \|</td><td/></tr><tr><td>Body \u2227 Body</td><td/></tr><tr><td>Query ::= Head</td><td/></tr></table>",
	"type_str": "table",
	"text": "). More precisely, an XMG specification is a collection of Horn clauses, which contribute a declarative description of what a computational tree grammar is."
	},
	"TABREF6": {
	"num": null,
	"html": null,
	"content": "<table><tr><td>E 2 E 1</td><td>for [bot : [index :</td></tr></table>",
	"type_str": "table",
	"text": "proposal for an FB-LTAG-based modeling of French syntax. FTAG contains around 6,000 elementary trees built from 293 XMG classes and covers some 40 basic 30 For sake of brevity, we write E 1 ] top : [index :E 2 ]].refers to the anchor / foot. 31 FTAG and SEMTAG are freely available under the terms of the GPL-compliant CeCILL license, the former at https://sourcesup.renater.fr/scm/viewvc.php/trunk/METAGRAMMARS/FrenchTAG/?root=xmg, and the latter on request."
	},
	"TABREF7": {
	"num": null,
	"html": null,
	"content": "<table/>",
	"type_str": "table",
	"text": ". Developing tree-adjoining grammars with lexical descriptions. In Srinivas Bangalore and Aravind Joshi, editors, Supertagging: Using Complex Lexical Descriptions in Natural Language Processing. MIT Press, Cambridge, MA, pages 73-110. Xia, Fei, Martha Palmer, K. Vijay-Shanker, and Joseph Rosenzweig. 1998. Consistent grammar development using partial-tree descriptions for LTAGs. In Proceedings of the 4th International Workshop on Tree Adjoining Grammar and Related Formalisms (TAG+ 1998), pages 180-183, Philadelphia, PA. XTAG Research Group. 2001. A lexicalized tree adjoining grammar for English. Technical Report IRCS-01-03, IRCS, University of Pennsylvania."
	}
	}
	}
	}