Documentation for the Grammar Matrix Customization Person Library

Introduction

This document explains how to fill out the Person page of the Grammar Matrix Customization questionnaire and presents background information on the person 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 Person Library

The standard reference for the Person Library and its implementations is Drellishak 2009. The full reference and .bib entry can be found here.

Options

The Person page in Matrix Customization questionnaire is divided into two sections, allowing you to define person values in the first section and to describe the interaction of person with number in the second section.

Note that feature values defined on the Lexicon and Morphology pages may affect the hierarchy of person. Underspecifications in the lexicon or morphosyntax will automatically be integrated in the hierarchy of person: you do not need to define these explicitly on the Person page.

In the first section you are asked to define person values existing in your language. The possible person values are: first (referring to the speaker), second (the person spoken to), third (everybody else), and fourth person (its definition varies from language to language). Grammar Matrix Customization system requires you to choose from the following options:

Please choose the option that best describes the person system in your language. The person library allows you to specify the range of values for the feature [PERSON] (or [PERNUM]) that will be available elsewhere in the customization system and in the resulting grammar (see more on that in the Analyses section below).

Please note that like gender, in terms of filling out the Grammar Matrix questionnaire, we are dealing with person only insofar as it is expressed grammatically, i.e. through some form of agreement. If your language does not express person through agreement, please choose the option "None".

The next section of the Matrix Customization questionnaire deals with the absence/presence of the subtypes of the first person in your language. Depending on your answers in this section, your grammar will use either a feature for PERSON or PERNUM for PERSON and NUMBER (more details in section Analyses below). You are provided with the following options in this section:

If your language does not distinguish subtypes for the first person, please select "none" in this section and move to the next page of the questionnaire. If your language has different forms for the first person, please select options "inclusive and exclusive" or "other." If you choose one of these options, you will also need to enter the subtypes of the first person on the bottom of the page. If you choose "inclusive and exclusive" option, a dropdown menu will appear to the right of this option with the number values you have defined on the Number page of the questionnaire.

When defining inclusive/exclusive or "other" distinctions of first person, the grammar will contain an attribute [PERNUM], rather than two separate features for person and number. All defined person and number values are cross-classified under this feature (e.g. [PERNUM 3sg] rather than [PERSON 3, NUMBER sg]).

English is an example of a language with two distinct forms for the first person of the verb to be, i.e. 1sg "am" and 1pl "are" and therefore requires creation of subtypes for first person. Even though English typically makes this distinction for 3sg versus all other forms, it can be advantageous to create this distinction for first person (see explanation above).

For more information on how to use the questionnaire to define [PERSON] values in your language, please refer to the Person page in Matrix Customization.

For more analysis and implementation of the person hierarchy in your language please refer to the Analyses section below.

Motivation

Depending on the presence (absence) of the subtypes in the first person in your language, Grammar Matrix will use either two separate features [PERSON] and [NUMBER] or a single [PERNUM] feature. This approach allows Grammar Matrix customization system to account accordingly both for the languages with the distinct form for each combination of person and number and for the languages with merged person and number values.

For example, [PERNUM] provides a more suitable manner to capture person and number agreement in English as suggested by Flickinger 2000. In particular, it can model the standard distinction between 3rd-sg and non-3rdsg in a straight-forward manner. With two distinct features, you would need two forms for each standard English finite verb in present tense non-3rd-sg: one for 1st, 2nd and 3rd person plural and one for 1st and 2nd singular. The combined feature allows you to group person-number values according to necessity: furthermore distinguishing 1sg for the verb to be and 2nd person, underspecified number for you.

For more information on using customized number hierarchy for your target language, please refer to Drellishak 2009, Section 5.3.4.

Analyses

Depending on the choices you make on the Person page in Matrix Customization, your starter grammar will either include the feature [ PER person ] or [ PERNUM pernum ]. If your language has a distinct form for each person and number combination and does not have subtypes for first person (i.e. if you chose option "none" in the second section of the person questionnaire page), then your grammar will use a separate feature [PER person] for PERSON. If your language is inclusive/exclusive or minimal/augmented and has subtypes for first person, then your grammar will use the feature [PERNUM] instead, in order to account for the merged person and number hierarchy.

Note that like GENDER and NUMBER, PERSON is a feature of the nominal INDEX. Making PERSON a feature of semantic INDEX instead of syntactic HEAD allows for correct semantic representation of person in your grammar and avoiding spurious ambiguity during generation.

If your grammar uses feature PERNUM, you will be able to define its values later in the questionnaire on the Other Features page.

The values you assign to PERSON feature will be available to use later in the customization system, especially the Morphology and Lexicon pages. The answers you provide on this page will determine the values available on the Lexicon page for the PERSON feature. PERSON will also appear as one of the features of the lexical rule types on Morphology page.

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

The analyses implemented in the person library are described in Drellishak 2009, Section 5.3.

PERSON is a feature often involved in agreement phenomena. These are handled through the lexicon and morphology libraries. The analyses of agreement implemented in the Grammar Matrix are described in Drellishak 2009, Section 5.2.

Note that feature values defined on the Lexicon and Morphology pages may affect the hierarchy of person. Underspecifications in the lexicon or morphosyntax will automatically be integrated in the hierarchy of person: you do not need to define these explicitly on the Person page.

Trivia

English does not distinguish inclusive 1st person from exclusive. The following quote from a speech made by the Mayor of Frankfurt am Main (in English) in an attempt to get people to trust politicians and bankers again shows why this distinction can be useful (Die Zeit, 6. Dezember 2012, p.25):

>>There is no us and them, or better: us and you. There is only us.<<

Upcoming Work

References

Drellishak, Scott. 2009. Widespread but Not Universal: Improving the Typological Coverage of the Grammar Matrix. PhD thesis, University of Washington.

bibtex:

@phdthesis{Drellishak:09,
author = {Scott Drellishak},
year = {2009},
title = {Widespread but Not Universal: Improving the Typological Coverage of the {G}rammar {M}atrix},
school = {University of Washington}
}

Flickinger, Dan. 2000. On Building a More Efficient Grammar by Exploiting Types. Natural Language Engineering, 6(1), 15–28.

bibtex:

@article{Flickinger:00,
author = {Dan Flickinger},
year = {2000},
title = {On Building a More Efficient Grammar by Exploiting Types. Natural Language Engineering},
journal = {Natural Language Engineering},
volume = {6},
pages = {15-28},
}

MatrixDoc/Person (last edited 2013-01-13 22:23:19 by AntskeFokkens)

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