Feature Structures

16.1 Introduction

A feature structure is a general purpose data structure which identifies and groups together individual features, each of which associates a name with one or more values. Because of the generality of feature structures, they can be used to represent many different kinds of information. Interrelations among various pieces of information, and their instantiation in markup provides a metalanguage for representing text analysis and interpretation. Moreover, this instantiation allows feature values to be of specific types, and for restrictions to be placed on the values for particular features, by means of feature system declarations, which are discussed in chapter 26 Feature System Declaration. Such restrictions provide the basis for at least partial validation of the feature-structure encodings that are used.

This chapter is organized as follows. Following this introduction, section 16.2 Elementary Feature Structures: Features with Binary Values introduces the binary feature values, and shows how elementary feature structures using features with those values may be constructed. Section 16.3 Feature, Feature-Structure and Feature-Value Libraries introduces the tags that represent libraries of features, feature structures and feature values, along with methods for pointing at features, feature structures and feature values in these libraries. Section 16.4 Symbolic, Numeric, Measurement, Rate and String Values, presents the tags for symbolic, numeric, measurement, rate, and string values. Section 16.5 Structured Values, shows how to use feature-structures themselves as values, thus enabling feature structures to be recursively defined. Section 16.6 Singleton, Set, Bag and List Collections of Values demonstrates the use of multiple values for features, for encoding set, bag, and list collections of values. Section 16.7 Alternative Features and Feature Values presents various methods for representing alternations (disjunctions) of features and feature values. Section 16.8 Boolean, Default and Uncertain Values, presents tags for boolean, default, and uncertain values, along with methods for underspecifying feature values. Section 16.9 Indirect Specification of Values Using the rel Attribute shows how to specify various logical relations, such as negation and subsumption, between the expressed values for a feature and its actual values. Finally, section 16.10 Two Illustrations, illustrates how feature structures may be linked to to text elements.

This tag set is selected as described in 3.3 Invocation of the TEI DTD; in an XML document which uses the markup described in this chapter, the document type declaration should contain the following declaration of the entity TEI.fs, or an equivalent one:

<!ENTITY % TEI.fs 'INCLUDE'>

The entire document type declaration for a document using this additional tag set together with the base tag set for prose might look like this:

<!DOCTYPE TEI.2 PUBLIC "-//TEI P4//DTD Main Document Type//EN" "tei2.dtd" [
    <!ENTITY % TEI.XML       'INCLUDE' >
    <!ENTITY % TEI.prose     'INCLUDE' >
    <!ENTITY % TEI.fs        'INCLUDE' >
 ]>

The overall document type declaration for this additional tag set has the following structure:

<!-- 16.1: Feature Structures-->
<!--
 ** Copyright 2004 TEI Consortium.
 ** See the main DTD fragment 'tei2.dtd' or the file 'COPYING' for the
 ** complete copyright notice.
-->
[declarations from 16.2: Feature structures, binary values inserted here ] 
[declarations from 16.3: Feature libraries inserted here ]
[declarations from 16.4: Symbolic, etc. values inserted here ]
[declarations from 16.6: Null values inserted here ]
[declarations from 16.7: Alternative features and feature values inserted 
here ] 
[declarations from 16.8: Boolean, default, uncertainty values inserted 
here ] 
<!-- end of 16.1-->

16.2 Elementary Feature Structures: Features with Binary Values

The fundamental elements of a feature structure system are <f> (for feature) and <fs> (for feature structure). The <fs> element has a type attribute for indicating what type of feature structure it represents, and may contain any number of <f> elements. An <f> element, in turn, has a required name attribute and any number of associated values. These may be binary, numeric, symbolic (i.e. taken from a restricted set of legal values), or string-valued, or may consist of sets, lists, or bags of binary, numeric, symbolic, or string values. Specialized values may also be given which allow partial underspecification of the feature. These possible types are all described in more detail in this and the following sections.

This section considers the special case of feature structures that contain features whose single value is one of the binary values represented by the empty elements <plus> and <minus>. The elements which are used for representing feature structures, features and the binary values, along with their descriptions and attributes, are the following.

The attributes not discussed in this section are discussed in following sections as follows: the feats and the fVal attributes in section 16.3 Feature, Feature-Structure and Feature-Value Libraries, the rel attribute in section 16.9 Indirect Specification of Values Using the rel Attribute, and the org attribute in section 16.6 Singleton, Set, Bag and List Collections of Values.

An <fs> element containing <f> elements with binary values can be straightforwardly used to encode the matrices of feature-value specifications for phonetic segments, such as the following for the English segment [s].

+---          ---+
| + consonantal  |
| - vocalic      |
| - voiced       |
| + anterior     |
| + coronal      |
| + continuant   |
| + strident     |
+---          ---+

Using the additional tag set for feature structures, this might be encoded as follows. Note that <fs> elements may have a type attribute indicating the kind of feature structure in question.

<fs type="phonological segment">
   <f name="consonantal"> <plus/>  </f>
   <f name="vocalic">     <minus/> </f>
   <f name="voiced">      <minus/> </f>
   <f name="anterior">    <plus/>  </f>
   <f name="coronal">     <plus/>  </f>
   <f name="continuant">  <plus/>  </f>
   <f name="strident">    <plus/>  </f>
</fs>

The restriction of specific features to specific types of values (e.g. the restriction of the feature ‘strident’ to the values <plus/> or <minus/>) cannot be validated by a generic SGML or XML parser (though other validation mechanisms such as XML Schemas do provide such capabilities). To enable an application program to check that only legal values for particular features appear, one may write a feature-system declaration; see chapter 26 Feature System Declaration.

Here are the formal declarations of the <fs>, <f>, <plus> and <minus> elements.

<!-- 16.2: Feature structures, binary values-->
<!ELEMENT fs %om.RR; ((f | fAlt | alt)*)> 
<!ATTLIST fs
      %a.global;
      type CDATA #IMPLIED
      feats IDREFS #IMPLIED
      rel (eq|ne|sb|ns) "sb"
      TEIform CDATA 'fs'  >
<!ELEMENT f %om.RO; ( null | ( plus | minus | any | none | dft | uncertain | 
sym | nbr | msr | rate | str | vAlt | alt | fs )* ) > 
<!ATTLIST f
      %a.global;
      name NMTOKEN #REQUIRED
      org (single|set|bag|list) #IMPLIED
      rel (eq|ne|sb|ns) "eq"
      fVal IDREFS #IMPLIED
      TEIform CDATA 'f'  >
<!ELEMENT plus %om.RO; EMPTY> 
<!ATTLIST plus
      %a.global;
      TEIform CDATA 'plus'  >
<!ELEMENT minus %om.RO; EMPTY> 
<!ATTLIST minus
      %a.global;
      TEIform CDATA 'minus'  >
<!-- end of 16.2-->

16.3 Feature, Feature-Structure and Feature-Value Libraries

As the example in the preceding section illustrates, the direct encoding of features structures can be verbose. Consequently, the effort of encoding large numbers of feature structures in this manner could be enormous, and could result in the creation of enormous files. To reduce the size and complexity of the task of encoding feature structures, one may use the feats attribute of the <fs> element to point to one or more of the features of that element. This indirect method of encoding feature structures presumes that the <f> elements are assigned unique id values, and are collected together in <fLib> elements (feature libraries). In turn, feature structures can be collected together in <fsLib> elements (feature-structure libraries). Finally, one may use the fVal attribute of the <f> element to point to its values. This indirect method of encoding feature values presumes that the value elements are assigned id specifications, and are collected together in <fvLib> elements (feature-value libraries). The elements which are used for representing feature, feature-structure and feature-value libraries, along with their descriptions and attributes, are the following.

For example, suppose a feature library for phonological feature specifications is set up as follows.

<fLib type="phonological features">
   <f id="CNS1" name="consonantal"> <plus/> </f>
   <f id="CNS0" name="consonantal"> <minus/> </f>
   <f id="VOC1" name="vocalic">     <plus/> </f>
   <f id="VOC0" name="vocalic">     <minus/> </f>
   <f id="VOI1" name="voiced">      <plus/> </f>
   <f id="VOI0" name="voiced">      <minus/> </f>
   <f id="ANT1" name="anterior">    <plus/> </f>
   <f id="ANT0" name="anterior">    <minus/> </f>
   <f id="COR1" name="coronal">     <plus/> </f>
   <f id="COR0" name="coronal">     <minus/> </f>
   <f id="CNT1" name="continuant">  <plus/> </f>
   <f id="CNT0" name="continuant">  <minus/> </f>
   <f id="STR1" name="strident">    <plus/> </f>
   <f id="STR0" name="strident">    <minus/> </f>
   <!-- ... -->
</fLib>

Then the feature structures that represent the analysis of the phonological segments (phonemes) /t/, /d/, /s/, and /z/ can be defined as follows.

<fs feats="CNS1 VOC0 VOI0 ANT1 COR1 CNT0 STR0"/>
<fs feats="CNS1 VOC0 VOI1 ANT1 COR1 CNT0 STR0"/>
<fs feats="CNS1 VOC0 VOI0 ANT1 COR1 CNT1 STR1"/>
<fs feats="CNS1 VOC0 VOI1 ANT1 COR1 CNT1 STR1"/>

The preceding are but four of the 128 logically possible fully specified phonological segments using the seven binary features listed in the feature library. Presumably not all combinations of features correspond to phonological segments (there are no strident vowels, for example). The legal combinations, however, can be collected together in a feature-structure library, with each element being given a unique id attribute, as in the following example.

<fsLib id="fsl1" type="phonological segment definitions">
   <!-- ... -->
   <fs id="T.DF" feats="CNS1 VOC0 VOI0 ANT1 COR1 CNT0 STR0"/>
   <fs id="D.DF" feats="CNS1 VOC0 VOI1 ANT1 COR1 CNT0 STR0"/>
   <fs id="S.DF" feats="CNS1 VOC0 VOI0 ANT1 COR1 CNT1 STR1"/>
   <fs id="Z.DF" feats="CNS1 VOC0 VOI1 ANT1 COR1 CNT1 STR1"/>
   <!-- ... -->
</fsLib>

Text elements can be linked to these feature structures in any of the ways described in section 15.2 Global Attributes for Simple Analyses. In the following example, a <linkGrp> element is used to link selected characters in the text ‘Caesar seized control’ to their phonological representations.

<text id='TXT1'> 
  <!-- ... -->
  <body> 
	<!-- ... -->
	<ab id='S1'>
	  <w id='S1W1'><c id='S1W1C1'>C</c>ae<c id='S1W1C2'>s</c>ar</w>
	  <w id='S1W2'><c id='S1W2C1'>s</c>ei<c id='S1W2C2'>z</c>e<c id='S1W2C3'>d</c></w>
	  <w id='S1W3'>con<c id='S1W3C1'>t</c>rol</w>.
	</ab>
	<!-- ... -->
  </body>
  <fsLib id='FSL1' type='phonological segment definitions'>
	<!-- as in previous example -->
  </fsLib>
  <linkGrp type='phonological identification of characters'
           domains='FSL1 TXT1'
           targFunc='phonological.segment character' >
	<!-- ... -->
	<link id='LT' targets='S.DF S1W3C1'/>
	<link id='LD' targets='Z.DF S1W2C3'/>
	<link id='LS' targets='S.DF S1W2C1'/>
	<link id='LZ' targets='Z.DF S1W2C2'/>
	<!-- ... -->
  </linkGrp></text>

Because of the simplicity of the binary feature values, there is no particular gain in pointing at those values rather than specifying them directly. However, the mechanism of using the fVal attribute on <f> elements is useful for representing more complex feature values, and can be illustrated using binary values. Suppose the <plus> and <minus> elements are collected together in a <fvLib>, as follows.

<fvLib type="binary values">
   <plus id="B1"/>
   <minus id="B0"/>
</fvLib>

Then the feature library presented at the beginning of this section can be represented as follows.

<fLib type="phonological features">
   <f id="CNS1" name="consonantal" fVal="B1"/>
   <f id="CNS0" name="consonantal" fVal="B0"/>
   <f id="VOC1" name="vocalic" fVal="B1"/>
   <f id="VOC0" name="vocalic" fVal="B0"/>
   <f id="VOI1" name="voiced" fVal="B1"/>
   <f id="VOI0" name="voiced" fVal="B0"/>
   <f id="ANT1" name="anterior" fVal="B1"/>
   <f id="ANT0" name="anterior" fVal="B0"/>
   <f id="COR1" name="coronal" fVal="B1"/>
   <f id="COR0" name="coronal" fVal="B0"/>
   <f id="CNT1" name="continuant" fVal="B1"/>
   <f id="CNT0" name="continuant" fVal="B0"/>
   <f id="STR1" name="strident" fVal="B1"/>
   <f id="STR0" name="strident" fVal="B0"/>
   <!-- ... -->
</fLib>

Although <fs> elements are legitimate feature values (see section 16.5 Structured Values), they are not allowed within <fvLib> elements. They should be placed in <fsLib> elements.

Here are the formal declarations of the <fLib>, <fsLib> and <fvLib> elements.

<!-- 16.3: Feature libraries-->
<!ELEMENT fLib %om.RR; ((f | fAlt)*)> 
<!ATTLIST fLib
      %a.global;
      type CDATA #IMPLIED
      TEIform CDATA 'fLib'  >
<!ELEMENT fsLib %om.RR; ((fs | vAlt)*)> 
<!ATTLIST fsLib
      %a.global;
      type CDATA #IMPLIED
      TEIform CDATA 'fsLib'  >
<!ELEMENT fvLib %om.RR; ((plus | minus | any | none | dft | uncertain | null 
| sym | nbr | msr | rate | str | vAlt)*)> 
<!ATTLIST fvLib
      %a.global;
      type CDATA #IMPLIED
      TEIform CDATA 'fvLib'  >
<!-- end of 16.3-->

16.4 Symbolic, Numeric, Measurement, Rate and String Values

In section 16.2 Elementary Feature Structures: Features with Binary Values, we defined the two empty elements <plus> and <minus> which are used to represent binary values. In this section, we define five more feature-value elements: the empty elements <sym> for expressing symbolic values, <nbr> for expressing numeric values, <msr> for expressing measurement values, and <rate> for expressing rate values; and the element <str> for expressing string values. These elements, along with their descriptions and attributes, are the following.

The <sym> element is to be used for the value of a feature when that feature can have any of a small, finite set of possible values, representable as character strings. For example, consider the problem of specifying the grammatical case, gender and number features of classical Greek noun forms. Assuming that the case feature can take on any of the five values nominative, genitive, dative, accusative and vocative; that the gender feature can take on any of the three values feminine, masculine, and neuter; and that the number feature can take on either of the values singular and plural, then the following may be used to represent the claim that the noun form θεάι goddesses has accusative case, feminine gender and plural number.

<fs type="word structure">
   <f name="case">   <sym value="accusative"/> </f>
   <f name="gender"> <sym value="feminine"/> </f>
   <f name="number"> <sym value="plural"/> </f>
</fs>

Note that instead of using a symbolic value for grammatical number, one could have named the feature singular or plural and given it an appropriate binary value, as in the following example. Whether one uses a binary or symbolic value in situations like this is largely a matter of taste.

<fs type="word structure">
   <f name="case">     <sym value="accusative"/> </f>
   <f name="gender">   <sym value="feminine"/> </f>
   <f name="singular"> <minus/> </f>
</fs>

An SGML or XML parser by itself cannot determine that particular values do or do not go with particular features; in particular, it cannot distinguish between the presumably legal encodings in the preceding two examples and the presumably illegal encoding in the following example.

<!-- *PRESUMABLY ILLEGAL* ... -->
<fs type="word structure">
   <f name="case">   <sym value="feminine"/> </f>
   <f name="gender"> <sym value="accusative"/> </f>
   <f name="number"> <minus/> </f>
</fs>

There are two ways of attempting to ensure that only legal combinations of feature names and values are used. First, if the total number of legal combinations is relatively small, one can simply list all of those combinations in <fLib> elements (together possibly with <fvLib> elements), and point to them using the feats attribute in the enclosing <fs> element. This method is suitable in the situation described above, since it requires specifying a total of only ten (5 + 3 + 2) combinations of features and values. Further, to ensure that the features are themselves combined legally into feature structures, one can put the legal feature structures inside <fsLib> elements. A total of 30 feature structures (5 × 3 × 2) is required to enumerate all the legal combinations of individual case, gender and number values in the preceding illustration. Of course, the legality of the markup requires that the feat attributes actually point at legally defined features, which an SGML or XML parser, by itself, cannot guarantee.

A more general method of attempting to ensure that only legal combinations of feature names and values are used is to provide a feature system declaration which includes a <valRange> element for each feature one uses. Here is a sample <valRange> element for the ‘case’ feature described above; for further discussion of the <valRange> element, see chapter 26 Feature System Declaration; the <vAlt> element is discussed in section 16.7 Alternative Features and Feature Values.

<!-- VALRANGE specification for CASE feature -->
<valRange>
  <vAlt>
    <sym value='nominative'/>
    <sym value='genitive'/>
    <sym value='dative'/>
    <sym value='accusative'/>
    <sym value='vocative'/>
  </vAlt>
</valRange>

Similarly, to ensure that only legal combinations of features are used as the content of feature structures, one should provide <fsConstraint> elements for each of the types of feature structure one employs. For discussion of the <fDecl> and <fsConstraint> elements, see 26 Feature System Declaration. Validation of the feature structures used in a document based on the feature-system declaration, however, requires that there be an application program that can use the information contained in the feature-system declaration.

Features with <sym>, <plus>, and <minus> values may be used to encode highly structured information such as may be obtained from precoded survey instruments. We illustrate by means of a coding scheme based on the one that is used for classifying potential printed entries in the British National Corpus. The scheme uses the following features and associated values.

medium: books and magazines; miscellaneous; written to be spoken
domain: imaginative; applied science; arts; belief and thought; commerce and finance; leisure; natural and pure science; social science; world affairs
level: high; medium; low
sampling range: beginning; middle; end; whole; whole less ten percent
date of origination: 1960–1975; 1975–1993
published (miscellaneous items only): yes; no
selection method (books and periodicals only): chosen on grounds of circulation or influence; chosen at random

A comprehensive feature library for this scheme is the following; the id specifications are those used by the British National Corpus (BNC) project: ¹³¹

<fLib type="BNC classification features">
 <f id="ca002" name="medium"><sym value="book.or.periodical"/></f>
 <f id="ca003" name="medium"><sym value="miscellaneous"/></f>
 <f id="ca004" name="medium"><sym value="written.to.be.spoken"/></f>
 <f id="ca005" name="domain"><sym value="imaginative"/></f>
 <f id="ca006" name="domain"><sym value="applied.science"/></f>
 <f id="ca007" name="domain"><sym value="arts"/></f>
 <f id="ca008" name="domain"><sym value="belief.and.thought"/></f>
 <f id="ca009" name="domain"><sym value="commerce.and.finance"/></f>
 <f id="ca00a" name="domain"><sym value="leisure"/></f>
 <f id="ca00b" name="domain"><sym value="natural.and.pure.science"/></f>
 <f id="ca00c" name="domain"><sym value="social.science"/></f>
 <f id="ca00d" name="domain"><sym value="world.affairs"/></f>
 <f id="ca00e" name="level"><sym value="high"/></f>
 <f id="ca00f" name="level"><sym value="medium"/></f>
 <f id="ca00g" name="level"><sym value="low"/></f>
 <f id="ca00h" name="sample.type"><sym value="beginning"/></f>
 <f id="ca00j" name="sample.type"><sym value="middle"/></f>
 <f id="ca00k" name="sample.type"><sym value="end"/></f>
 <f id="ca00l" name="sample.type"><sym value="whole"/></f>
 <f id="ca00m" name="sample.type"><sym value="whole.less.ten.percent"/></f>
 <f id="ca00n" name="published.between"><sym value="1960.1975"/></f>
 <f id="ca00p" name="published.between"><sym value="1975.1993"/></f>
 <f id="ca00r" name="published"><plus/></f>
 <f id="ca00s" name="published"><minus/></f>
 <f id="ca00t" name="selection.method"><sym value="principled"/></f>
 <f id="ca00u" name="selection.method"><sym value="random"/></f>
</fLib>

An entry which is a book or periodical on world affairs, medium level, sampled from the middle, published between 1975 and 1993, and selected on a principled basis could then be assigned the following feature-structure code; this code could also be placed in a feature-structure library that contains all the possible fully-specified BNC entry classifications. This library would have a total of 1620 (3 × 9 × 3 × 5 × 2 × 2) entries.

<fs id="ca2dfjpt" type="BNC classification for written documents"
    feats="ca002 ca00d ca00f ca00j ca00p ca00t"/>

The <nbr> element is to be used when the value of a feature is a number or a range of numbers. For example, suppose one wishes to encode information contained in classified advertisements for the sale or rental of real estate, such as the number of bedrooms and bathrooms in a listed property, and its advertised selling or rental price. One way of representing such information is as follows.

<fs type="real estate listing">
   <f name="number.of.bathrooms"><nbr value="2"/></f>
   <f name="number.of.bedrooms"><nbr value="3"/></f>
   <f name="monthly.rent"><nbr value="625.00"/></f>
</fs>

The information that the number of bedrooms is in the range from 3 to 5 and the monthly rent is in the range from 625.00 to 950.00 may be represented as follows, using the optional valueTo attribute.

<fs type="real estate listing">
   <f name="number.of.bedrooms"><nbr value="3" valueTo="5"/></f>
   <f name="monthly.rent"><nbr value="625.00" valueTo="950.00"/></f>
</fs>

The <nbr> (and also the <msr> and <rate> elements defined below) element also may have a type attribute to specify whether the values of the value and valueTo attributes are to be construed as integer or real numbers.

The <msr> element to be used when the value of a feature is a scalar quantity, essentially a combination of a numeric value and a symbolic value for identifying the scale on which the numeric value occurs. For example, real estate listings often provide the area (in square feet or meters) of a house or apartment and the area (in acres or hectares) of land being sold or rented. One way of representing information about such areas is as follows.

<fs type="real estate listing">
   <f name="interior.area"><msr value="2000" unit="sq.ft"/></f>
   <f name="property.area"><msr value="0.5" unit="acre"/></f>
</fs>

The value of the ‘monthly.rent’ feature in the two examples above might be more accurately analysed as a measurement rather than as a numeric value, since the amount of the rent in question is to be understood as payable in a specific currency (US or Canadian dollars, pounds sterling, euro, yen...) To make the currency scale explicit, the first example of this feature might be re-encoded as follows.

<f name="monthly.rent"><msr value="625.00" unit="USD"/></f>

The unit and value attributes of the <msr> element are both required. If the unit attribute is not needed (for example, if no confusion would result if the unit attribute is not specified), then the <nbr> element may be used to express the feature value.

The <rate> element is to be used when the value of a feature is a rate. This element has a required per attribute for expressing the interval over which the rate is measured (typically, but not necessarily, a temporal interval), and an optional unit attribute for expressing the scalar unit. For example, one might encode the wage rate of $8.25 per hour as follows.

<f name="wage.rate"><rate value="8.25" unit="USD" per="hour"/></f>

Note that the ‘monthly.rent’ feature illustrated above can be re-encoded as having a rate value, with the per attribute in this case taking the value month, as follows.

<f name="rent"><rate value="625.00" unit="USD" per="month"/></f>

To encode interest, inflation or tax rates, the unit attribute can be used to indicate that the value attribute is to be understood as a percentage. For example, an interest rate of 8.25% per year can be encoded in either of the following two ways.

<f name="interest"><rate value="8.25" unit="percent" per="year"/></f>
<f name="interest"><rate value="0.0825" per="year"/></f>

Finally, the <str> element is to be used for the value of a feature when that value is a string drawn from a very large or potentially unbounded set of possible strings of characters, so that it would be impractical or impossible to use the <sym> element. These values are expressed not as the values of the value attribute, as in the case of symbolic, numeric, measurement and rate values, but as the content of the <str> element. For example, one may encode the street address of a property in a real estate listing, as follows.

<fs type="real estate listing">
   <f name="address"><str>3418 East Third Street</str></f>
</fs>

Here are the formal declarations of the <sym>, <nbr>, <msr>, <rate>, and <str> elements.

<!-- 16.4: Symbolic, etc. values-->
<!ELEMENT sym %om.RO; EMPTY> 
<!ATTLIST sym
      %a.global;
      value CDATA #REQUIRED
      rel (eq|ne) "eq"
      TEIform CDATA 'sym'  >
<!ELEMENT nbr %om.RO; EMPTY> 
<!ATTLIST nbr
      %a.global;
      value CDATA #REQUIRED
      valueTo CDATA #IMPLIED
      rel (eq|ne|lt|le|gt|ge) "eq"
      type (int|real) #IMPLIED
      TEIform CDATA 'nbr'  >
<!ELEMENT msr %om.RO; EMPTY> 
<!ATTLIST msr
      %a.global;
      value CDATA #REQUIRED
      valueTo CDATA #IMPLIED
      unit CDATA #REQUIRED
      rel (eq|ne|lt|le|gt|ge) "eq"
      type (int|real) #IMPLIED
      TEIform CDATA 'msr'  >
<!ELEMENT rate %om.RO; EMPTY> 
<!ATTLIST rate
      %a.global;
      value CDATA #REQUIRED
      valueTo CDATA #IMPLIED
      unit CDATA #IMPLIED
      per CDATA #REQUIRED
      rel (eq|ne|gt|ge|lt|le) "eq"
      type (int|real) #IMPLIED
      TEIform CDATA 'rate'  >
<!ELEMENT str %om.RR; (#PCDATA)> 
<!ATTLIST str
      %a.global;
      rel (eq|ne|sb|ns|lt|le|gt|ge) "eq"
      TEIform CDATA 'str'  >
<!-- end of 16.4-->

16.5 Structured Values

Features may have structured values as well; these values are represented by either the <fs> element, or the fVal attribute on the <f> element, which can point to an <fs> element. Since an <fs> or a pointer to an <fs> is permitted to occur as a value of an <f>, recursion is possible. For example, an <fs> element may contain or point to an <f> element, which may contain or point to an <fs> element, which may contain or point to an <f> element, and so on. To illustrate the use of structured values, consider the following simple model of a personal record, consisting of a person's name, date of birth, place of birth, and sex. Each personal record is a <fs type='personal record'> tag, consisting of the corresponding four features, three of which take structured values, as in the following example.

<fs type="personal record">
  <f name="full.name">
    <fs type="name record">
      <f name="first.name"> <str>Kathleen</str> </f>
      <f name="middle.name"> <str>Anne</str> </f>
      <f name="surname"> <str>Barnett</str> </f>
    </fs>
  </f>
  <f name="date.of.birth">
    <fs type="date record">
      <f name="year"> <nbr value="1968"/> </f>
      <f name="month"> <nbr value="4"/> </f>
      <f name="day"> <nbr value="17"/> </f>
    </fs>
  </f>
  <f name="place.of.birth">
    <fs type="place record">
      <f name="city"> <str>Austin</str> </f>
      <f name="state"> <sym value="TX"/> </f>
    </fs>
  </f>
  <f name="sex"> <sym value="female"/> </f>
</fs>

Now suppose that feature-structure libraries are maintained for name records and place records. Further suppose that the feature structure representing the name record in the previous example has an id attribute with the value nkab027", while the feature structure representing the place record has an id attribute whose value is txaustin.¹³² Then the preceding example could also be encoded as follows. (An identifier is also provided for the personal record.)

<fs id="pkab027" type="personal record">
   <f name="full.name" fVal="nkab027"/>
   <f name="date.of.birth">
      <fs type="date record">
         <f name="year"> <nbr value="1968"/> </f>
         <f name="month"> <nbr value="4"/> </f>
         <f name="day"> <nbr value="17"/> </f>
      </fs>
   </f>
   <f name="place.of.birth" fVal="txaustin"/>
   <f name="sex"> <sym value="female"/> </f>
</fs>

This representation could be simplified further if a feature library is maintained for the year, month, day and sex features, so that the feats attribute may be used as follows.

<fs id="pkab027" type="personal record" feats="sxf">
   <f name="full.name" fVal="nkab027"/>
   <f name="date.of.birth"><fs type="date record" feats="y1968 m04 d17"/></f>
   <f name="place.of.birth" fVal="txaustin"/>
</fs>

Next, suppose that a feature-structure library is also maintained for personal records, and that the library also contains records for the parents of the individual identified in the previous example. Suppose that the father is identified as pmfb009 and the mother as parn002. Then the personal-record feature structure could be easily augmented to include pointers to the parents, as follows.

<fs id="pkab027" type="personal record" feats="sxf">
   <f name="full.name" fVal="nkab027"/>
   <f name="date.of.birth"><fs type="date record" feats="y1968 m04 d17"/></f>
   <f name="place.of.birth" fVal="austintx"/>
   <f name="mother" fVal="parn002"/>
   <f name="father" fVal="pmfb009"/>
</fs>

If the personal records identified as parn002 and pmfb009 also contain information about the parents of those individuals, then from the present record, one would have access to that individual's grandparents as well.

Assuming that personal records of the sort described in this section are being maintained in association with text files, the records can be linked to those texts in any of the ways described in chapter 14 Linking, Segmentation, and Alignment, provided that identifiers are added for appropriate features, as in the following illustration.

<text id="bfile"><body>
    <div id="tkab027" type="birth certificate">
      <p><name id="t1kab027" type="person">Kathleen Anne Barnett</name> 
        was born at <time id="t1t0659">6:59 a.m.</time> on
        <date id="t1d680417">April 17, 1968</date> in 
        <name id="t1setonhsp" type="org">Seton Hospital</name> in 
        <name id="t1txaustin" type="place">Austin</name> to 
        <seg id="s1">Mr.</seg> and <seg id="s2">Mrs.</seg> 
        <name id="t1mfb009" type="person">Michael F. Barnett</name> 
        of <name id="t1sansabatx" type="place">San Saba</name>.  
      </p>
      <!-- ... --> 
      <join id="t1arn002" targets="s2 t1mfb009"/> 
      <join id="t2mfb009" targets="s1 t1mfb009"/> 
      <!-- ... -->
    </div></body>
  <fsLib id="prec" type="personal records">
    <fs id="pkab027" type="personal record" feats="sxf">
      <f name="full.name" fVal="nkab027"/>
      <f id="dkab027" name="date.of.birth">
        <fs type="date record" feats="y1968 m04 d17"/>
      </f>
      <f id="bkab027" name="place.of.birth" fVal="txaustin"/>
      <f id="mkab027" name="mother" fVal="parn002"/>
      <f id="fkab027" name="father" fVal="pmfb009"/>
    </fs></fsLib>
  <linkGrp type="record verification" domains="bfile prec" targFunc="source goal">
    <link targets="t1kab027 nkab027"/>
    <link targets="t1d680417 dkab027"/>
    <link targets="t1txaustin bkab027"/>
    <link targets="t1arn002 mkab027"/>
    <link targets="t2mfb009 fkab027"/>
  </linkGrp>
</text>

16.6 Singleton, Set, Bag and List Collections of Values

In the discussion to this point, we have assumed that features have exactly one simple value. However, for many purposes, it is useful to be able to consider the values of certain features to be organized in more complex ways, for example as sets, bags (or multisets), or lists. Accordingly, we provide for four different ways in which feature values may be organized, namely as singletons, sets, bags and lists. We do so by means of an org attribute on the <f> element, which takes on one of the designated values single, set, bag, and list. A feature whose value is organized as a singleton is understood as having exactly one simple value. If more than one value is specified for it, we assume that only the first one is considered to be its true value. A feature whose value is organized as a set, bag or list may have any positive number of values as its content. In a set, items are ordered, and may not be repeated. In a bag, items are not ordered, and may repeat. In a list, items are ordered and may repeat. Sets and bags are thus distinguished from lists in that the order in which the values are specified does not matter for the former, but does matter for the latter, while sets are distinguished from bags and lists in that repetitions of values do not count for the former but do count for the latter.¹³³

No default value for the org attribute is declared in the DTD; however, a default value for that attribute can be declared for particular features in the feature-system declaration; see chapter 26 Feature System Declaration. Note that if only one value is specified for a given <f> element, the set, bag and list values of the org are all essentially equivalent to the singleton value, so the omission of the org attribute for such a feature is not problematic.¹³⁴

To illustrate the use of the org attribute, suppose that the illustration of personal records from the previous section is extended to include pointers to an individual's siblings. Suppose also that the individual identified as <fs id="pkab027"> has siblings identified as <fs id="panb005">, <fs id="pmfb010"> and <fs id="pzrb001"> in the personal records library. Then we may extend the personal record for <fs id="pkab027"> as follows.

<fs id="pkab027" type="personal record" feats="sxf">
   <f name="full.name" fVal="nkab027"/>
   <f name="date.of.birth">
      <fs type="date record" feats="y1988 m04 d17"/>
   </f>
   <f name="place.of.birth" fVal="austintx"/>
   <f name="mother" fVal="parn002"/>
   <f name="father" fVal="pmfb009"/>
   <f name="siblings" org="set" fVal="panb005 pmfb010 pzrb001"/>
</fs>

A more elaborate illustration of the use of the org attribute is the the following <f name="career" org="list"> element which may be added to the personal records of an individual to record the job career of that individual. The feature structures which constitute the value of this feature document the jobs which the individual has held in the order in which they were held. Note that a list has been embedded within a list by means of intervening <fs type="employment record"> and <f name="promotion.history"> elements.

<f name="career" org="list">
  <fs type="employment record">
    <f name="employer"><str>Safeway Stores</str></f>
    <f name="hiring.information">
      <fs type="hire structure">
        <f name="hire.date"><fs type="date structure" feats="y1988 m06"/></f>
        <f name="job.title"><sym value="stocker"/></f>
        <f name="wage"><rate value="6.00" per="hour"/></f>
        <f name="hours.worked"><rate value="40" per="week"/></f>
        <f name="status.code" fVal="sc4a"/>
      </fs>
    </f>
    <f name="promotion.history" org="list">
      <fs type="promotion record">
        <f name="date"><fs type="date structure" feats="y1988 m12"/></f>
        <f name="job.title"><sym value="cashier"/></f>
        <f name="wage"><rate value="7.00" per="hour"/></f>
        <f name="hours.worked"><rate value="40" per="week"/></f>
        <f name="status.code" fVal="sc4a"/>
      </fs>
      <fs type="promotion record">
        <f name="date"><fs type="date structure" feats="y1990 m02"/></f>
        <f name="job.title"><sym value="supervisor"/></f>
        <f name="salary"><rate value="18000" per="year"/></f>
        <f name="status.code" fVal="sc3c"/>
      </fs>
    </f>
    <f name="termination.information">
      <fs type="termination structure">
        <f name="termination.date"><fs type="date structure" feats="y1991 m04"/></f>
        <f name="status.code" fVal="sc3c"/>
        <f name="reason.for.termination"><sym value="laid.off"/></f>
      </fs>
    </f>
  </fs>
  <fs type="employment record">
    <!-- ... -->
  </fs>
  <!-- ... -->
</f>

The information contained in such features may be linked to textual references in the usual way. The <f name="status.code"> feature has been included to show how evaluative or interpretive information can be included along with information gleaned from textual records. The example presumes that the status code values are maintained in a designated <fvLib>.

Features with values organized as sets, bags or lists can sometimes be used instead of features organized as singletons, whose values are individual feature structures. For example, consider the following encoding of the English verb form ‘sinks’, which contains an ‘agreement’ feature whose value is a feature structure which contains ‘person’ and ‘number’ features with symbolic values.

<fs type="word structure">
  <!-- ... -->
  <f name="word.class"> <sym value="verb"/> </f>
  <f name="tense"> <sym value="present"/> </f>
  <f name="agreement">
    <fs type="agreement structure">
      <f name="person"> <sym value="third"/> </f>
      <f name="number"> <sym value="singular"/> </f>
    </fs>
  </f>
  <!-- ... -->
</fs>

If one does not care about the names of the features contained within the ‘agreement’ feature structure, the containing <f name="agreement"> element can be given an org attribute with the value set, and the contained <fs> element, together with the person and number feature elements it contained, can be eliminated, as follows.

<fs type="word structure">
  <!-- ... -->
  <f name="word.class"> <sym value="verb"/> </f>
  <f name="tense"> <sym value="present"/> </f>
  <f name="agreement" org="set"><sym value="third"/><sym value="singular"/></f>
  <!-- ... -->
</fs>

The encoding in the preceding example presumes that the <fDecl> element for the ‘agreement’ feature would look something like the following; for further details, see chapter 26 Feature System Declaration.

<fDecl name="agreement" org='set'>
  <!-- ... -->
  <vRange>
    <vAlt>
      <sym value='first'/>
      <sym value='second'/>
      <sym value='third'/>
      </vAlt>
    <vAlt>
      <sym value='singular'/>
      <sym value='plural'/>
      </vAlt>
    </vRange>
  <!-- ... -->
  </fDecl>

The set, bag or list which has no members is known as the null (or empty) set, bag or list. To refer to it, the <null> element is provided; its description and attributes are as follows.

So, for example, to indicate that the individual identified above by the <fs id="pkab027"> element has no siblings, we may specify the ‘siblings’ feature as follows.

<f name="siblings" org="set"> <null/> </f>

The <null> element when used with a feature organized as a singleton is a semantic error; however, its appearance as a value for such a feature cannot be flagged by SGML or XML parsers. The <null> element, when it appears as a feature value, must be the only value.

Here is the formal declarations of the <null> element.

<!-- 16.6: Null values-->
<!ELEMENT null %om.RO; EMPTY> 
<!ATTLIST null
      %a.global;
      TEIform CDATA 'null'  >
<!-- end of 16.6-->

16.7 Alternative Features and Feature Values

In this section, two methods of representing the alternation (ambiguity or uncertainty) of features and feature values are presented. The first of these methods is to be used for nonsystematic or sporadic markup of alternation of individual features or values; it makes use of the special-purpose <fAlt> and <vAlt> elements. The other is to be used for systematic markup of alternation and for the alternation of groups of features or values; it makes use of the general-purpose <alt> element introduced in section 14.8 Alternation. The <fAlt> and <vAlt> elements have the following description and attributes.

To illustrate the use of the <fAlt> element to represent the alternation of features, suppose one is uncertain whether a particular real estate advertisement describes a house with two bedrooms or with two bathrooms. This uncertainty can be represented as follows.

<fs type="real estate listing">
   <fAlt>
      <f name="number.of.bathrooms" > <nbr value="2"/> </f>
      <f name="number.of.bedrooms"> <nbr value="2"/> </f>
   </fAlt>
</fs>

This representation leaves unspecified whether or not the alternation is mutually exclusive (i.e. whether having two bathrooms excludes the possibility of having two bedrooms and vice versa). To make this aspect of the alternation explicit, one can specify a value for the mutExcl attribute, as follows.

<fs type="real estate listing">
   <fAlt mutExcl="N">
      <f name="number.of.bathrooms" > <nbr value="2" /> </f>
      <f name="number.of.bedrooms" > <nbr value="2" /> </f>
   </fAlt>
</fs>

The <fAlt> element can also be used to represent uncertainty about whether the number of bathrooms is two or three, as follows; note that the attribute value mutExcl="Y" can be inferred for the <fAlt> element in this example.

<fs type="real estate listing">
   <fAlt>
      <f name="number.of.bathrooms" > <nbr value="2" /> </f>
      <f name="number.of.bathrooms" > <nbr value="3" /> </f>
   </fAlt>
</fs>

Since the ‘number.of.bathrooms’ feature in this example can be factored out of the alternation, a <vAlt> element could be used in place of it to represent the alternation of the feature values more simply, as follows:

<fs type="real estate listing">
   <f name="number.of.bathrooms" >
      <vAlt>
         <nbr value="2" />
         <nbr value="3" />
      </vAlt>
   </f>
</fs>

The <fAlt> and <vAlt> elements can also be used to indicate certain alternations among values of features organized as sets, bags or lists. For example, suppose one uses a <f name="extras" org="set"> element in feature structures for real estate listings to represent items that are mentioned to enhance a property's sales value, such as whether it has a pool or a good view. Now suppose for a particular listing, the extras include an alarm system and a fenced-in yard, and either a pool or a jacuzzi (but not both). This situation could be represented, using the <vAlt> element, as follows.

<fs type="real estate listing">
   <!-- ... -->
   <f name="extras" org="set" >
      <str>alarm system</str>
      <str>fenced-in yard</str>
      <vAlt mutExcl="Y">
         <str>pool</str>
         <str>jacuzzi</str>
      </vAlt>
   </f>
   <!-- ... -->
</fs>

Now suppose the situation is like the preceding except that one is also uncertain whether the property has an alarm system or a fenced-in yard, or possibly both. This can be represented as follows.

<fs type="real estate listing">
   <!-- ... -->
   <f name="extras" org="set" >
      <vAlt mutExcl="N">
         <str>alarm system</str>
         <str>fenced-in yard</str>
      </vAlt>
      <vAlt mutExcl="Y">
         <str>pool</str>
         <str>jacuzzi</str>
      </vAlt>
   </f>
   <!-- ... -->
</fs>

Finally, suppose that the listing specifies that the property has a finished basement, and that it also has either an alarm system and a pool or a fenced-in yard and a jacuzzi. This situation cannot be represented using the <vAlt> element, because the alternation holds between subsets of two values each. It can, however, be represented using the <fAlt> element, as follows; note that the <str> element with the value finished basement element must be repeated.

<fs type="real estate listing">
   <!-- ... -->
   <fAlt mutExcl="Y">
      <f name="extras" org="set" >
         <str>finished basement</str>
         <str>alarm system</str>
         <str>pool</str>
      </f>
      <f name="extras" org="set" >
         <str>finished basement</str>
         <str>fenced-in yard</str>
         <str>jacuzzi</str>
      </f>
   </fAlt>
   <!-- ... -->
</fs>

If a large number of ambiguities or uncertainties involving a relatively small number of features and values need to be represented, it is recommended that the general-purpose <alt> element discussed in section 14.8 Alternation be used, rather than the special-purpose <fAlt> and <vAlt> elements. The use of the <alt> element avoids the need to explictly represent the alternating elements more than once.

For example, suppose one has set up a <fsLib> element containing feature structures representing the morphological structures of classical Greek inflected words, along with collections of individual features and feature values, encoded by <fLib> and <fvLib> elements as appropriate. The following example shows how one might then represent the morphological structure of a feminine gender, accusative case, plural number noun form in classical Greek, such as ‘θεάι’ goddesses discussed in section 16.4 Symbolic, Numeric, Measurement, Rate and String Values:

<fsLib type="noun structures">
   <!-- ... -->
   <!-- plural accusative feminine noun -->
   <fs id="wngfkanp" type="noun structure" feats="wn gf ka np"/>
   <!-- ... --> </fsLib>

<fLib type="morphological features">
   <f id="wn" name="word.class"  fVal="nn"/>
   <!-- ... -->
   <f id="gf" name="gender"  fVal="fe"/>
   <!-- ... -->
   <f id="ka" name="case"  fVal="ac"/>
   <!-- ... -->
   <f id="np" name="number"  fVal="pl"/>
   <!-- ... --> </fLib>

<fvLib type="morphological feature values">
   <!-- ... -->
   <sym id="nn" value="noun" />
   <!-- ... -->
   <sym id="fe" value="feminine" />
   <!-- ... -->
   <sym id="ac" value="accusative" />
   <!-- ... -->
   <sym id="pl" value="plural" />
   <!-- ... --> </fvLib>

Now consider the noun form ‘θεάί’ goddesses, which is analyzable as a feminine plural noun form in either the nominative or the vocative case. We may represent this ambiguity by adding the following entries to the <fsLib>, <fLib>, and <fvLib> elements in the preceding example; assume that appropriate entries for unambiguous nominative and vocative case forms have already been entered.

<!-- Add the following to the feature-structure library  -->
   <!-- plural nominative-or-vocative feminine noun -->
   <fs id="wngfknvnp" type="noun structure" feats="wn gf knv np"/>
<!-- Add the following to the feature library  -->
   <!-- CASE='nominative' or vocative -->
   <f id="knv" name="case"  fVal="novo"/>
<!-- Add the following to the feature value library -->
   <!-- nominative or vocative -->
   <alt id="novo" targets="no vo"/>

If the <fvLib> element is not used, and specifications for particular feature values are entered as content of the <f> elements in the <fLib> element, then the ambiguity can be represented as follows.

<fsLib type="noun structures">
   <!-- ... -->
   <!-- plural nominative-or-vocative feminine noun -->
   <fs id="wngfknvnp" type="noun structure" feats="wn gf knv np"/>
   <!-- ... -->
</fsLib>
<fLib type="morphological features">
   <!-- ... -->
   <f id="kn" name="case" >
      <sym value="nominative" />
      <!-- ... -->
   </f>
   <f id="kv" name="case" >
      <sym value="vocative" />
      <!-- ... -->
      <alt id="knv" targets="kn kv"/>
      <!-- ... -->
   </f>
</fLib>

The <alt> element together with the <join> element can, unlike the <fAlt> and <vAlt> elements, be used to express alternations between sets of features. An example of such an alternation is found in certain feminine gender Greek noun forms ending in -ας, such as πείρας attempt(s), which may be analyzed as having either genitive case and singular number features or accusative case and plural number features, as follows (again, assuming the existence of other elements and identifier attributes for simple features and values).

<!-- Add the following to the feature structure library -->
   <!-- feminine noun, either genitive singular or accusative plural -->
   <fs id="wngfkg.nska.np" type="noun structure" feats="wn gf kg.nska.np"/>
<!-- Add the following to the feature library  -->
   <join id="kg.ns" targets="kg ns"/><!-- genitive singular -->
   <join id="ka.np" targets="ka np"/><!-- accusative plural -->
   <!-- alternation:  gen. sg. or acc. plural -->
   <alt id="kg.nska.np" targets="kg.ns ka.np"/>

Here are the formal declarations of the <fAlt> and <vAlt> elements.

<!-- 16.7: Alternative features and feature values-->
<!ELEMENT fAlt %om.RR; ((f | fs | fAlt), (f | fs | fAlt)+)> 
<!ATTLIST fAlt
      %a.global;
      mutExcl (Y|N) #IMPLIED
      TEIform CDATA 'fAlt'  >
<!ELEMENT vAlt %om.RR; ((plus | minus | any | none | dft | uncertain | null | 
sym | nbr | msr | rate | str | vAlt | fs), 
(plus | minus | any | none | dft | uncertain | null | 
sym | nbr | msr | rate | str | vAlt | fs)+) > 
<!ATTLIST vAlt
      %a.global;
      mutExcl (Y|N) #IMPLIED
      TEIform CDATA 'vAlt'  >
<!-- end of 16.7-->

16.8 Boolean, Default and Uncertain Values

In this section we define four special empty elements used as feature values: the boolean value elements <any> and <none>, the <dft> element, and the <uncertain> element.

The boolean value elements are used to indicate whether the features they are associated with have values. The element <any> corresponds to the boolean value true (i.e., that the feature it is associated with has a value — not the same as the binary value plus), and the element <none> corresponds to the boolean value false (i.e., that the feature it is associated with has no value). The <dft> element is used to indicate that the feature it is associated with has its default value in the feature structure in which it appears. Finally, the <uncertain> element may be used to indicate uncertainty about what value, if any, its associated feature has; it is equivalent to the alternation of the <any> and <none> elements. To indicate uncertainty about which of the possible legal values a particular feature has, one should use the <any> element.

The descriptions and attributes of these elements are as follows.

The values <null> and <none> are distinct. The former is to be used with a feature organized as a set, bag, or list to indicate that its value is the null set, bag, or list in a particular feature structure. The latter is to be used with such a feature to indicate that it has no value in a particular feature structure.

The boolean values <any> and <none> are also distinct from the binary values <plus> and <minus>. The latter pair are specific possible values for features, whereas the former pair represent ranges of possible values, not specific possible values, for features. For example, suppose that the <valRange> element for the ‘auxiliary’ feature is declared as follows in the feature structure declaration, so that either boolean value is legal.

<vRange><vAlt><plus/><minus/></vAlt></vRange>

Given this <vRange>, then the following pair of specifications is distinct:

<f name="auxiliary"><plus/></f> 
<f name="auxiliary"><any/></f>

In this situation, the <any> element is equivalent to the alternation of the <plus> and <minus> values.

Given the same <vRange>, then the following pair of specifications is also distinct.

<f name="auxiliary"><minus/></f> 
<f name="auxiliary"><none/></f>

The <none> element is equivalent to the negation of the alternation of the <plus> and <minus> elements.

However, if the auxiliary feature is declared to take only the <plus> value, then the specifications below are equivalent:

<f name="auxiliary"><plus/></f> 
<f name="auxiliary"><any/></f>

If the auxiliary feature is declared to take only the <plus> value, then the specifications below are not equivalent; in fact, the specification is invalid.

<!-- invalid! -->
<f name="auxiliary"><minus/></f>
<f name="auxiliary"><none/></f>

It is even possible to declare that a particular feature can never have values, as follows for the ‘impossible’ feature:

<vRange><null/></vRange>

In this case, the following specifications are equivalent.

<f name="impossible"><any/></f>
<f name="impossible"><none/></f>

The elements <any> and <dft> are also designed to be used in conjunction with the <fDecl> and <valDefault> elements in the feature system declaration discussed in chapter 26 Feature System Declaration. First, consider the <any> element, and suppose that the <vRange> element in the <fDecl> element for the ‘gender’ feature is specified as follows.

<vRange>
  <vAlt>
    <sym value='feminine'/>
    <sym value='masculine'/>
    <sym value='neuter'/>
    </vAlt>
  </vRange>

Then the following two representations are equivalent.

<f name="gender"> <any/> </f>
<f name="gender">
   <vAlt>
      <sym value="feminine"/>
      <sym value="masculine"/>
      <sym value="neuter"/>
   </vAlt>
</f>

Second, consider the <dft> element, and suppose that the default value for the ‘gender’ feature (as specified by the <valDefault> element of its <fDecl> element) is feminine. Then the following three representations are equivalent; note that if an <f> element appears without content and without a valid fVal attribute, then it is equivalent to the same element with the <dft> element as its content.

<f name="gender"/>

<f name="gender"> <dft/> </f>

<f name="gender"> <sym value="feminine"/> </f>

Using the <any> and <dft> elements, together with an <fDecl> element for the corresponding feature in the feature system declaration, provides a method for underspecifying the value of that feature. The <any> element means that the associated feature has a legal value but what value it has is not specified. The <dft> element means that the associated feature has the value which the encoder has declared is the normal value of the feature.

The boolean elements <any> and <none> also have specific uses within <fsConstraints> and <fDecl> elements in feature system declarations, as described in chapter 26 Feature System Declaration. For example, the element <any> can appear as the value of a feature contained within an <fs> of a particular type which appears in the <cond> element of an <fsConstraints> element, to indicate that the feature must appear in feature structures of the designated type (i.e., that it is obligatory) and that when it does appear, it may appear with any of its legal values. Similarly, <none> can appear in this way to specify that the feature cannot be present in feature structures of the indicated type (i.e., that it is obligatorily absent from such feature structures). All other features that are declared to have values are understood to be optional in such feature structures.

For example, the following may appear as part of the <fsConstraints> of a feature system declaration to indicate that an ‘agreement structure’ feature structure must contain a legal ‘number’ feature, but must not contain a ‘category’ feature.

<cond> <fs type='agreement structure'></fs>
   <then/><fs>
           <f name='number'><any/></f>
           <f name='category'><none/></f>
         </fs>
</cond>

Further constraints can be imposed on a feature structure of a particular type in the <vRange> elements of features which take feature structures of that type as values. For example, suppose that verb and adjective agreement in German are represented by feature structures of the following sorts, in which verb forms agree in person and number with their subjects and adjective forms agree in gender, case, and number with their subjects.

<fs type="verb structure">
   <!-- ... -->
   <f name="verbAgreement">
      <fs type="agreement structure">
         <f name="person"> <sym value="first"/> </f>
         <f name="number"> <sym value="plural"/> </f>
      </fs>
   </f>
   <!-- ... -->
</fs>
<fs type="adjective structure">
   <!-- ... -->
   <f name="adjagreement">
      <fs type="agreement structure">
         <f name="gender"> <sym value="feminine"/> </f>
         <f name="case"> <sym value="accusative"/> </f>
         <f name="number"> <sym value="plural"/> </f>
      </fs>
   </f>
   <!-- ... -->
</fs>

In order to ensure that an ‘agreement structure’ feature structure which appears as the value of a ‘verbAgreement’ feature may be specified for any person and number feature, but for no gender and case feature, we may provide a <vRange> element for the ‘verbAgreement’ feature as follows.

<vRange>
  <fs type='agreement structure'>
    <f name='person'><any/></f>
    <f name='case'><none/></f>
    <f name='gender'><none/></f>
    <f name='number'><any/></f>
    </fs>
  </vRange>

Similarly, to ensure that an ‘agreement structure’ feature structure which appears as the value of a ‘adjAgreement’ feature may be specified for any case, gender, and number feature, but for no person feature, we may provide a <vRange> element for the ‘adjAgreement’ feature as follows.

<vRange>
  <fs type='agreement structure'>
    <f name='person'><none/></f>
    <f name='case'><any/></f>
    <f name='gender'><any/></f>
    <f name='number'><any/></f>
    </fs>
  </vRange>

The combination of declarations like these and the principle of subsumption discussed in section 16.9 Indirect Specification of Values Using the rel Attribute, allows feature structures to be underspecified in text markup. For example, to indicate that a given adjective inflection feature (tagged <f name="adjInflection">) is a feature structure (tagged <fs type="inflection structure">) specifying plural number and any gender and case, we can omit the elements for gender and case on the <fs> element, as follows.

<f name="adjinflection">
   <fs type="inflection structure">
      <f name="number"> <sym value="plural"/> </f>
   </fs>
</f>

When supplied as the value of a ‘verbInflection’ feature, the same feature structure would be interpreted as an inflection structure specifying plural number and any person.

If an optional feature is not specified in a feature-structure value, then it is assumed to occur with the <uncertain> value. For further discussion, see section 16.9 Indirect Specification of Values Using the rel Attribute.

Here are the formal declarations of the <any>, <none>, <dft>, and <uncertain> elements.

<!-- 16.8: Boolean, default, uncertainty values-->
<!ELEMENT any %om.RO; EMPTY> 
<!ATTLIST any
      %a.global;
      TEIform CDATA 'any'  >
<!ELEMENT none %om.RO; EMPTY> 
<!ATTLIST none
      %a.global;
      TEIform CDATA 'none'  >
<!ELEMENT dft %om.RO; EMPTY> 
<!ATTLIST dft
      %a.global;
      TEIform CDATA 'dft'  >
<!ELEMENT uncertain %om.RO; EMPTY> 
<!ATTLIST uncertain
      %a.global;
      TEIform CDATA 'uncertain'  >
<!-- end of 16.8-->

16.9 Indirect Specification of Values Using the rel Attribute

The rel attribute is provided for the feature value elements <sym>, <nbr>, <msr>, <rate>, <str>, <fs>, and <default> (but not <plus>, <minus>, <null>, <vAlt>, <any>, <none>, and <uncertain>). This attribute may be used for specifying which of various logical relations the given value has to the actual value of the feature. For all value elements for which the rel attribute is defined, except for <fs>, the default value for that attribute is eq, which means that the actual value is equal (or identical) to the given value. Accordingly, the following representations are both interpreted to mean that the value of the ‘case’ feature is the <sym value="genitive"> element.

<f name="case"> <sym value="genitive"/> </f>

<f name="case"> <sym rel="eq" value="genitive"/> </f>

16.9.1 The Not-Equals Relation

The rel attribute can also be specified as having the value ne, which means that the associated feature has a value which is not equal to the given value. For example, the value <nbr rel="ne" value="1"> in the following example denotes any numeric value other than 1 for the feature ‘number.of.bathrooms’.

<f name="number.of.bathrooms"> <nbr value="1" rel="ne"/> </f>

If an <fDecl> element has been provided which defines the legal values for the associated feature, then the value ne can be given a positive interpretation. For example, suppose that the <vRange> element is declared in the <fDecl> element for the ‘case’ feature as follows.

<vRange>
  <vAlt>
    <sym value='nominative'/>
    <sym value='genitive'/>
    <sym value='dative'/>
    <sym value='accusative'/>
    <sym value='vocative'/>
    </vAlt>
  </vRange>

Suppose also that the ‘case’ feature is declared as obligatory in a particular feature structure. Then the following specifications are equivalent in that structure.

<f name="case"> <sym value="genitive" rel="ne"/> </f>

<f name="case">
   <vAlt>
      <sym value="nominative"/>
      <sym value="dative"/>
      <sym value="accusative"/>
      <sym value="vocative"/>
   </vAlt>
</f>

That is, when the rel attribute occurs with the value ne in the value of an obligatory feature in a feature structure, the actual value of that feature may be any of its legal values other than the specified value.

On the other hand, if the ‘case’ feature is declared as optional in a particular feature structure, then the following specifications are equivalent in that structure.

<f name="case"> <sym value="genitive" rel="ne"/> </f>

<f name="case">
   <vAlt>
      <sym value="nominative"/>
      <sym value="dative"/>
      <sym value="accusative"/>
      <sym value="vocative"/>
      <none/>
   </vAlt>
</f>

That is, when the rel attribute has the value ne in the value of an optional feature in a feature structure, the actual value of that feature may be any of its legal values other than the specified value, or <none>.

If the rel attribute is specified with the value ne for a <nbr>, <msr>, or <rate> element for which the valueTo attribute is also specified, then the actual range may be any range distinct from that given. For example, the following means that the number of bathrooms is a range distinct from 3 to 5 (e.g., 3 to 4, 3 to 6, 4 to 5, 4 to 6, 0 to 2, etc.).

<f name="number.of.bathrooms"> <nbr value="3" valueTo="5" rel="ne"/> </f>

16.9.2 Other Inequality Relations

For the elements <nbr>, <msr>, <rate>, and <str>, the rel attribute may also take on the following values; the use of these values for the <str> element presumes that a particular character and string ordering (or sorting) convention is understood.

lt: The actual value or range is any legal value or range less than the specified value or range.
le: The actual value or range is any legal value or range less than or equal to the specified value or range.
gt: The actual value or range is any legal value or range greater than the specified value or range.
ge: The actual value or range is any legal value or range greater than or equal to the specified value or range.

These attribute values may be used as shown in the following examples. The first states that the number of bedrooms is less than 5; the second that an illegal speed is any speed greater than 65 miles per hour; the third that a lot size is in a range which is less than or equal to the range of from 5 to 10 acres;¹³⁵ the fourth that the last name is any string greater than the empty string (i.e., any nonempty string, given normal string-ordering conventions); and the fifth that for a feature whose value is a list of two strings, the first precedes the string ‘M’ and the second is the string ‘M’, or any string following it.

<f name="number.of.bedrooms"> <nbr value="5" rel="lt"/> </f>

<f name="illegal.speed"> <rate value="65" unit="miles" per="hour" rel="gt"/> </f>

<f name="lot.size"> <msr value="5" valueTo="10" unit="acre" rel="le"/> </f>

<f name="last.name"> <str rel="gt"/> </f>

<f name="pairs" org="list"> <str rel="lt">M</str> <str rel="ge">M</str> </f>

16.9.3 Subsumption and Non-subsumption Relations

When the rel attribute is given the values sb or ns, the markup expresses the claim that the value given subsumes, or does not subsume, the actual value for the feature in question.

On the <str> element, these values are used to specify that the string value given in the <str> element is or is not a substring of the actual value of the feature. The first example below specifies that the actual feature value may be any string at all (since the empty string is a substring of every string), the second that it might be any string in which the string ‘the’ occurs as a substring, and the third that it might be any string in which the string ‘the’ does not occur as a substring.

<str rel="sb"/>

<str rel="sb">the</str>

<str rel="ns">the</str>

On the <fs> element, the attribute values sb and ns indicate that the given feature structure does or does not legally subsume the actual feature structure. By definition, one feature structure subsumes another if the second feature structure is identical to the first or contains more information than the first. The default value for the rel attribute of the <fs> element is sb. The subsumption of feature structures is illustrated by the following four examples; suppose that the ‘person’ and ‘number’ features are either optional or obligatory in these <fs type="agreement structure"> example elements.

<fs id="p3ns" type="agreement structure">
   <f name="person"> <sym value="third"/> </f>
   <f name="number"> <sym value="singular"/> </f>
</fs> <!-- third person singular -->

<fs id="p3nx" type="agreement structure">
   <f name="person"> <sym value="third"/> </f>
</fs> <!-- third person -->

<fs id="pxns" type="agreement structure">
   <f name="number"> <sym value="singular"/> </f>
</fs> <!-- singular -->

<fs id="pxnx" type="agreement structure"/> <!-- -->

The fourth example, pxnx, subsumes all four of the examples, since each contains at least as much information as does feature structure pxnx. Conversely, the first example, p3ns, subsumes only itself. Finally, the second and third examples, identified as p3nx and pxns attributes, subsume themselves and the first feature structure, but not each other.

If both person and number are obligatory features of agreement structure elements, then the last three elements in the preceding list have the same interpretation as their counterparts in the following list.

<fs id="p3na" type="agreement structure">
   <f name="person"> <sym value="third"/> </f>
   <f name="number"> <any/> </f>
</fs> <!-- third person -->
<fs id="pans" type="agreement structure" >
   <f name="person"> <any/> </f>
   <f name="number"> <sym value="singular"/> </f>
</fs> <!-- singular -->
<fs id="pana" type="agreement structure" >
   <f name="person"> <any/> </f>
   <f name="number"> <any/> </f>
</fs> <!-- -->

On the other hand, if both person and number are optional features of agreement structures, then those three elements have the same interpretation as their counterparts in the following list.

<fs id="p3nu" type="agreement structure">
   <f name="person"> <sym value="third"/> </f>
   <f name="number"> <uncertain/> </f>
</fs> <!-- 3d person -->
<fs id="puns" type="agreement structure">
   <f name="person"> <uncertain/> </f>
   <f name="number"> <sym value="singular"/> </f>
</fs> <!-- singular -->
<fs id="punu" type="agreement structure">
   <f name="person"> <uncertain/> </f>
   <f name="number"> <uncertain/> </f>
</fs> <!-- -->

That is, if an optional feature is omitted from a feature-structure representation, then that feature may have any of its legal values or the value <uncertain>.

The value sb is chosen as the default value for the rel attribute of the <fs> element, because it provides for the most economical means for underspecifying them. One situation in which it may be preferable to specify <fs rel="eq"> is when the feature structure has many optional features and it is known that none of them occurs.

The specification <fs rel="ns"> is used to denote the feature structures that the specified feature structure does not subsume. This provides a handy way of saying that a certain combination of features is not present, for example the combination of third person and singular number, as in the agreement structure of the English verb form ‘sink’, understood as a present tense verb form. The following example expresses the claim that third-person and singular-number features are not both present in the agreement feature, but makes no further claim about what is present.

<f name="agreement">
   <fs id="np3ns" type="agreement structure" rel="ns">
      <f name="person"> <sym value="third"/> </f>
      <f name="number"> <sym value="singular"/> </f>
   </fs>
</f>

In most real situations, of course, one can infer, from the range of possible values for person and number, what the remaining possibilities are. Suppose, for example, that in the relevant feature system declaration, the features ‘person’ and ‘number’ are given the following <vRange> elements:

<vRange><!-- for the PERSON feature  -->
  <vAlt>
    <sym value='first'/>
    <sym value='second'/>
    <sym value='third'/>
    </vAlt>
  </vRange>
<vRange><!-- for the NUMBER feature  -->
  <vAlt>
    <sym value='singular'/>
    <sym value='plural'/>
    </vAlt>
  </vRange>

Suppose, further, that the person and number features are obligatory in feature structures of the type agreement structure. Then the element <fs id="NP3NS"> above is equivalent to the following alternation; the features whose value is <any> may be omitted, since they are implied by the default value of sb for the rel attribute in the enclosing <fs> elements.

<vAlt id="p12na-panp">
   <fs id="p12na" type="agreement structure">
      <f name="person">
         <vAlt> <sym value="first"/> <sym value="second"/> </vAlt>
      </f>
      <f name="number"> <any/> </f>
   </fs>
   <fs id="panp" type="agreement structure">
      <f name="person"> <any/> </f>
      <f name="number"> <sym value="plural"/> </f>
   </fs>
</vAlt>

If, on the other hand, the person and number features were optional in feature structures of type agreement structure, then the interpretation of an underspecified feature structure will change. The element <fs id="NP3NS"> given above is then equivalent to the following alternation; the features whose value is <uncertain> may be omitted as they are implied by the default subsumption relation holding between the structure given and the actual structure.

<vAlt id="p120nu-punp0">
   <fs id="p120nu" type="agreement structure">
      <f name="person">
         <vAlt> <sym value="first"/> <sym value="second"/> <none/> </vAlt>
      </f>
      <f name="number"> <uncertain/> </f>
   </fs>
   <fs id="punp0" type="agreement structure">
      <f name="person"> <uncertain/> </f>
      <f name="number">
         <vAlt> <sym value="plural"/> <none/> </vAlt>
      </f>
   </fs>
</vAlt>

16.9.4 Relations Holding with Sets, Bags, and Lists

The rel attribute is also provided for the <f> element, but is designed to be used with that element only when its org attribute (see section 16.6 Singleton, Set, Bag and List Collections of Values) is set, bag, or list. When associated with the <f> element, the rel attribute may take on any of the following four values: eq, ne, sb, and ns. The default value is eq. Consider first the use of the rel attribute with the <f> element when the given value of the feature is <null>.

<f name="extras" org="set"> <null/> </f>

<f name="extras" org="set" rel="ne"> <null/> </f>

<f name="extras" org="set" rel="sb"> <null/> </f>

<f name="extras" org="set" rel="ns"> <null/> </f>

The first example states that the ‘extras’ feature has the null set as its value. The second example states that the ‘extras’ feature is a set which is not equal to the null set. That is, its actual value might be any non-null set. The third example states that the ‘extras’ feature has as its value a set of which the null set is a subset; that is to say, any set at all, including the null set. Note that this is not equivalent to the following, which states that the extras feature has as its value a single element which is any legal value for the ‘extras’ feature, including for example a <str> element containing the value pool.

<f name="extras" org="set"> <any/> </f>

Finally, the fourth example states that the ‘extras’ feature has as its value a set of which the null set is not a subset. Since the null set is a subset of every set, the fourth example in effect claims that the ‘extras’ feature has no legal value; it is thus equivalent to the following, which states directly that the ‘extras’ feature has no value.

<f name="extras" org="set"> <none/> </f>

Consider next the use of the rel attribute with the <f> element when the given value of the feature is a single <str> element with the content pool:

<f name="extras" org="set"> <str>pool</str> </f>

<f name="extras" org="set" rel="ne"> <str>pool</str> </f>

<f name="extras" org="set" rel="sb"> <str>pool</str> </f>

<f name="extras" org="set" rel="ns"> <str>pool</str> </f>

The first example states that the value of the ‘extras’ feature is a set consisting of a single member, namely a <str> element containing the value pool. The second example states that the ‘extras’ feature has as its value a set which is not equal to the set consisting of this particular member. It could, however, be a two-membered set, one of whose members is some other value. This example is thus not equivalent to the following, which states that the ‘extras’ feature has as its value a set comprising a single member other than a <str> element with the content pool:

<f name="extras" org="set"> <str rel="ne">pool</str> </f>

The third example states that the ‘extras’ feature has as its value any set of which the set consisting of the single member specified is a subset (i.e., any set which contains the element <str> with the value pool, and possibly others). Finally, the fourth example states that the ‘extras’ feature has as its value any set which does not contain this element as a member.

16.9.5 Varieties of Subsumption and Non-subsumption

The rel values sb and ns have different meanings depending on whether they occur within a <str>, <fs> or <f> element. However, the use of a common name for the value reflects a fundamental similarity in those meanings. For example, the value sb can be used in all three elements to indicate that the actual value is any string, any feature structure, or any set, bag or list, as follows. In the second example below, the rel attribute has not been specified, since it has the value sb by default on <fs> elements.

<str rel="sb"></str>

<fs></fs>

<f name="..." org="set" rel="sb"> <null/> </f>

<f name="..." org="bag" rel="sb"> <null/> </f>

<f name="..." org="list" rel="sb"> <null/> </f>

Because the value sb is not defined for the attribute rel on the <nbr>, <msr> and <rate> elements, the indication that a value may be any number, measure or rate is sometimes not quite as simple. Here is one way of specifying any positive or negative integer numeric value.¹³⁶

<vAlt>
   <nbr value="0" rel="gt" type="int"/>
   <nbr value="0" rel="le" type="int"/>
</vAlt>

The value ns also is understood in similar ways in the different elements in which it may occur. Above in this section, the equivalence of the following representations under certain conditions was shown (the id attributes and the redundant features with <any/> values have been omitted).

<f name="agreement">
   <fs type="agreement structure" rel="ns">
      <f name="person"> <sym value="third"/> </f>
      <f name="number"> <sym value="singular"/> </f>
   </fs>
</f>
<f name="agreement">
   <vAlt>
      <fs type="agreement structure">
         <f name="person">
            <vAlt> <sym value="first"/> <sym value="second"/> </vAlt>
         </f>
      </fs>
      <fs type="agreement structure">
         <f name="number"> <sym value="plural"/> </f>
      </fs>
   </vAlt>
</f>

The value ns has an analogous meaning when the value in question is a set rather than a feature structure. Recast in such terms, the equivalence above still holds good:

<f name="agreement" org="set" rel="ns">
   <sym value="third"/>
   <sym value="singular"/>
</f>
<f name="agreement" org="set" rel="sb">
   <vAlt>
      <vAlt> <sym value="first"/> <sym value="second"/> </vAlt>
      <sym value="plural"/>
   </vAlt>
</f>

16.10 Two Illustrations

In this section, we present two illustrations of how to associate feature structures and their components with textual elements. Both are taken from the same example text, an article called Memoirs of a Dog Shrink that appeared in the popular magazine Dogs Today in August 1991, and was also included in the British National Corpus (BNC). The first illustration associates the text with a structure that represents a significant portion of the information contained in the text. The second marks up the grammatical structure of the orthographic words and certain other comparable units in the text. Here is the text, with markup provided down to the level of <s> elements. The n attribute values are taken from the BNC markup; the id attribute values have been added for purposes of these illustrations.

<div1 id="dt91mds" type="article">
  <head rend="italic"><s id="mds01" n="00732">Memoirs of a Dog Shrink</s></head>
  <head rend="italic" type="sub">
    <s id="mds04" n="00735">Cartoonist Russell Jones takes a ramble
      through Peter Neville's files</s> </head>
  <list>
    <item><s id="mds05" n="00736">Case number: 72</s></item>
    <item><s id="mds06" n="00737">Name: Jessie</s></item>
    <item><s id="mds07" n="00738">Breed: Collie</s></item>
    <item><s id="mds08" n="00739">Problem: Light bulb phobia</s></item>
  </list>
  <p><s id="mds09" n="00740">Jess the collie was a laid-back sort of
      hound who spent most of his life stretched out on a fireside rug
      in his large Surrey home.</s></p>
  <p><s id="mds10" n="00741">The closest he came to exercise was to
      open one eye every so often, if someone entered the room, or to
      open both eyes, smile, and wag his tail as he'd done on one
      occasion when confronted by a housebreaker!</s></p>
  <p><s id="mds11" n="00742">This extremely lazy lifestyle was one long
      yawn from dawn to dusk.</s>
    <s id="mds12" n="00743">Only the odd bouts of involuntary
      twitching in his sleep reassured his owner that Jess was still
      safe and sound in the land of the living!</s></p>
  <p><s id="mds13" n="00744">One winter night, as the mutt twitched
      away in front of the fire, his mind somewhere between
      Basingstoke and the twilight zone, a 100-watt light bulb in the
      standard lamp above his head suddenly exploded without
      warning!</s></p>
  <p><s id="mds14" n="00745">According to his owner, who witnessed the
      spectacle, Jessie rose gracefully toward the ceiling like a
      space shuttle and, after lingering in mid-air for what seemed an
      eternity, crashed to the floor and fled the house with a speed
      and agility the owner found quite amazing.</s></p>
  <p><s id="mds15" n="00746">Jessie did not return home for several hours.</s>
    <s id="mds16" n="00747">When he eventually did show up, it
      was obvious to all that he was a changed dog!</s>
    <s id="mds17" n="00748">What plodded through the front door was
      not the lovable, lazy hound who had once lived there but a
      grim-faced light bulb serial killer!</s></p>
  <p><s id="mds18" n="00749">Within seconds of his return, Jessie
      launched a vicious attack on a table lamp, popping the bulb and
      wrecking the shade before charging into the lounge.</s>
    <s id="mds19" n="00750">There, in a frenzy of violence, he reduced
      the standard lamp to a table lamp in 10 seconds flat!</s></p>
  <p><s id="mds20" n="00751">After a room-to-room chase lasting
      several minutes, during which every lamp in the house was turned
      to sawdust, the dog was finally caught and wrestled to the
      ground.</s></p>
  <p><s id="mds21" n="00752">With his house plunged into darkness,
      Jessie's owner sought my help.</s></p>
  <div2 type="part">
    <head> <s id="mds22" n="00753">SIMPLE SOLUTION</s> </head>
    <p><s id="mds23" n="00754">When I first saw the dog, it was quite
        obvious he'd been deeply affected by the explosion and had
        developed a 100-watt phobia for light bulbs!</s></p>
    <p><s id="mds24" n="00755">By placing his feeding bowl closer each
        day to a table lamp the dog gradually learned to live with his
        enemy.</s>
      <s id="mds25" n="00756">Within a couple of weeks, his killer
        instincts had disappeared and he was back where he belonged
        &mdash; twitching away peacefully on the fireside rug.</s></p>
  </div2>
</div1>

The first illustration is based on the observation that from the example text, it is possible to infer a fairly extensive medical history for the dog described in it. Suppose that we have a definition of a feature structure that represents a canine medical history. Then we can fill in feature values in that history from the text, and prepare a <linkGrp> element that specifies the links between the text segments and the various features specified in the feature structure. Here is a hypothetical example of such a filled-in feature structure and associated link group.

<fs id="j

Text Encoding Initiative

The XML Version of the TEI Guidelines

16 Feature Structures