Documentation for the Grammar Matrix Customization Tense, Aspect and Mood Library

Introduction

This document explains how to fill out the Tense, Aspect, and Mood page of the Grammar Matrix Customization questionnaire and presents background information on the Tense, Aspect and Mood (TAM) library of the Grammar Matrix Customization System (Bender et al., 2002; Bender and Flickinger, 2005; Bender et al., 2010). General instructions on using the questionnaire can be found here.

Citing the Tense, Aspect, and Mood Library

The standard reference for the Tense, Aspect, and Mood Library and its implementations is Poulson 2011. The full reference and .bib entry can be found here.

Options

The Tense, Aspect, and Mood page allows you to specify a range of values for each of these features in your language either by choosing pre-existing values or creating your own values. This page is divided into two sections: (1). Semantic Features (semantic features of Tense/Aspect/Mood) and (2). Syntactic Feature (verb forms if your grammar contains auxiliary verbs).

The Tense, Aspect, and Mood library allows you to specify the range of values for the features [TENSE tense], [ASPECT aspect], [SITUATION situation] , [MOOD mood], and [FORM form] that will be available elsewhere in the customization system and in the resulting grammar (see more on that in the Analyses section below). The values you define for TENSE, ASPECT, SITUATION, MOOD, and FORM on this page will determine the values available on the Lexicon and Morphology pages. Note that feature values defined on the lexicon and morphology libraries may affect the hierarchy of Tense, Aspect, and Mood. Underspecifications in the lexicon or morphosyntax will automatically be integrated in the hierarchy of Tense, Aspect, and Mood: you do not need to define these explicitly on the Tense, Aspect, and Mood page.

Semantic Features

This section describes semantic features of Tense, Aspect, and Mood. For both TENSE and ASPECT features, user can select among common values and/or define additional ones. MOOD values are purely user-defined.

1. Tense

In this section please define TENSE values and their subtypes (if applicable) for your language. You can either choose from TENSE values pre-defined for you by the Matrix Customization system or you can create your own TENSE values.

You are provided with the pre-defined TENSE values, which cover the most common tense hierarchies across the languages. If these pre-defined values provided below are sufficient to describe the tense system in your language, please choose "Select among common hierarchy elements" option and select from the values below as applicable to your language. For each option you select, please add a subtype if applicable.

If the pre-defined tense values provided above are not sufficient to describe the tense hierarchy in your language, you can build your own tense hierarchy. If your language requires a deeper tense hierarchy, please choose the option "Build your own TENSE hierarchy" on the bottom of TENSE section and enter all tense types and their supertypes existing in your language. If you choose to build your own tense hierarchy, the number of values that you can enter for tense types and their subtypes is unlimited.

Motivation for build your own hierarchy

This option is especially useful for the languages with deeper tense hierarchies, such as Amerindian language Kiksht. Kiksht tense system is divided into non-past and past, with the past tense being subdivided into remote (further subdivided into early-remote and late-remote tenses), far (further subdivided into early-far and late-far tenses), middle, and near tenses (Comrie 1985). Clearly, languages like Kiksht with deep tense hierarchies require a user to build a more elaborate tense hierarchy than the one already provided by Matrix customization system.

2. Aspect

In this section please define ASPECT values existing in your language. As with tense, you can either choose from the aspect values pre-defined for you by the Matrix Customization system or you can create your own aspect hierarchy. This section is divided into two parts: Viewpoint aspect and Situation aspect.

3. Mood

This section is designed to help you define the Mood and Modality system in general as applicable to your language. Modality roughly describes the opinions or attitudes of the speaker, with most common values being Subjunctive and Indicative. If your language has only these two subcategories of mood, please choose the pre-defined MOOD values available in Matrix Customization system by checking the "Create a hierarchy consisting of just the values subjunctive and indicative as subtypes of mood" option.

Otherwise, you can either skip this section entirely or build your own Mood hierarchy by adding mood types and their supertypes as applicable to your language. Please note that if you choose the option of building your own Mood hierarchy, this will override the pre-defined values binary Subjunctive/Indicative option above.

4.Additional features

If you need to define additional features for Tense, Aspect, and Mood in your language, such as for the arbitrary or quasi-semantic verb classes, e.g. French verbs in terms of their auxiliary selection (for more information about these verb classes please refer to Section 5.4 in Poulson 2011), you can define these features on the Other Features questionnaire page.

Syntactic Feature

The only syntactic feature this section deals with is the verb form, distinguishing finite verbs from nonfinite verbs. The main difference between these two verb forms is that finite verbs can head stand-alone clauses, while nonfinite verb forms cannot.

If you have indicated on the Word Order questionnaire page that your language contains auxiliaries, your starter grammar will have finite and nonfinite values of FORM by default.

If your language does not use auxiliary verbs, but makes a syntactic distinction between finite and nonfinite verbs, please check the corresponding option: "My language has no auxiliaries but does make a syntactically relevant finite/non-finite distinction."

Often languages have subtypes of nonfinite verbs. For example, English has Infinitive, Past Participle, and Present participle as subtypes of nonfinite verbs. If applicable to your language, please add the subtypes of finite and/or nonfinite verbs.

If your language does not make any syntactic distinction between finite and nonfinite verbs, you can skip this section entirely and move to the next page Other Features.

If you need to define additional syntactic features, you can define these features on the Other Features questionnaire page.

For more information on how to use the questionnaire to define values for Tense, Aspect, and Mood in your language, please refer to the Tense, Aspect, and Mood page in Matrix Customization.

For more information on the implementation of Tense, Aspect, and Mood features in Grammar Matrix please refer to the Section Analyses.

Motivation

Unlike some well-understood grammatical phenomena that can be covered by pre-defined analyses in Matrix libraries, Tense, Aspect, and Mood are more complicated phenomena with varying terminology and values/hierarchies that vary tremendously from language to language. A library with pre-defined analyses is not sufficient for such a complicated grammatical phenomenon. To allow user to have more flexibility for coverage of Tense, Aspect and Mood, Matrix Customization system uses a meta-modeling approach, enabling the user to define tense/aspect/mood values themselves instead of relying on a library with limited pre-defined analyses. For more information on the meta-modeling approach and its use in Matrix Customization system please refer to Section 4.1 of Poulson 2011.

Analyses

When you define tense, aspect, and mood hierarchies, your starter grammar will include the features [TENSE tense], [ASPECT aspect], [SITUATION situation], and [MOOD mood], with values based on the choices you make on the Tense, Aspect, and Mood customization page. TENSE, ASPECT, SITUATION, and MOOD are features of semantic INDEX. On the other hand, FORM is a feature of syntactic HEAD and can be used to constrain auxiliary complement forms. Below are snippets of code related to Tense, Aspect, and Mood (TAM) in matrix.tdl and in the your_language_name.tdl files (given no additional information comes from the lexicon and morphology libraries).

Code from matrix.tdl file:

tense := sort.
aspect := sort.
mood := sort.

tam := avm &
  [ TENSE tense,
    ASPECT aspect,
    MOOD mood ]. 

Below is an example of what the relevant code would look like in your_language_name.tdl file, if you were working on a language that contains only perfective and imperfective viewpoint aspects, covered by pre-defined option in Matrix Customization system:

perfective := aspect.
imperfective := aspect.

Below is an example of what the relevant code would look like in your_language_name.tdl file, if you were working on a language that has a deeper hierarchy system, with the imperfective aspect further subdivided into habitual and continuous aspects, with continuous aspect further subdivided into nonprogressive and progressive aspects, in which case you would have to build your own viewpoint aspect hierarchy:

perfective := aspect.
imperfective := aspect.
habitual := imperfective.
continuous := imperfective.
nonprogresssive := continuous.
progressive := continuous.

The choices you make on the Tense, Aspect, and Mood customization page will also affect the lexical types (your_language_name.tdl file), lexical rules (lrules.tdl file), and inflectional rules (irules.tdl file).

Note that feature values defined on the lexicon and morphology libraries may affect the hierarchies of Tense, Aspect, and Mood. Underspecifications in the lexicon or morphosyntax will automatically be integrated in the hierarchies of Tense, Aspect, and Mood: you do not need to define these explicitly on the Tense, Aspect, and Mood page.

Upcoming Work

The main direction of the future work in the Tense, Aspect, and Mood implementation in the Matrix Customization system is to extend the overall meta-modeling coverage. Although quite flexible, the meta-modeling approach currently cannot accommodate all analyses. In particular, the process of defining lexical types and type hierarchies needs more flexibility, both in terms of expanding the questionnaire and coverage of grammatical phenomena.

For example, as noted in the situation aspect section of the questionnaire, the coverage of situation aspect needs to be expanded. At present, the coverage of situation aspect in the Matrix customization system is limited only to the aspect expressed by inherent lexical qualities or overt morphological marking. More work is required in order to implement the compositionally derived situation aspect in the Matrix Customization system, which is not covered yet. Another possible enhancement of the situation aspect section is to provide a user with the pre-defined hierarchy of situation aspect features (such as dynamic, durative, and telic). These and other enhancements to the Tense, Aspect, and Mood section will help provide more flexibility and more coverage across languages.

For a more detailed discussion about future work on expanding the coverage of the Matrix Customization system please refer to Section 5.7 in Poulson 2011.

References

Comrie, B. (1985). Tense. Cambridge [Cambridgeshire]; New York: Cambridge University Press.

Poulson, L. (2011). Meta-modeling of Tense and Aspect in a Cross-linguistic Grammar Engineering Platform. UW Working Papers in Linguistics, 28.

Smith, C. S. (1991). The parameter of aspect. Dordrecht; Boston: Kluwer Academic Publishers, 1st ed.

Smith, C. S. (1997). The parameter of aspect. Dordrecht; Boston: Kluwer, 2nd ed.

Vendler, Z. (1957). Verbs and times. The Philosophical Review, 66(2), 143-160. (Reprinted in a revised version in Vendler, 1967).

MatrixDoc/TenseAspectMood (last edited 2013-02-22 22:42:53 by VaryaGracheva)

(The DELPH-IN infrastructure is hosted at the University of Oslo)