index.html 149 KB

Raw Blame History Permalink

<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
<head>
  <meta content=
  "HTML Tidy for Linux/x86 (vers 1 September 2005), see www.w3.org"
  name="generator" />
  <meta http-equiv="Content-Type" content=
  "text/html; charset=utf-8" />

  <title>Pronunciation Lexicon Specification (PLS) Version
  1.0</title>
<style type="text/css">
/*<![CDATA[*/
.diff-add   {  }
.diff-del   { display: none }
.diff-chg   {  }
.diff-eval  { background-color: #99FFFF; }
.diff-off   { display: none }
.issues { font-weight: bold; color: green }
.ipa { font-family: "Lucida Sans Unicode", monospace; }

div.issue {
  border-width: thin;
  border-style: solid;
  border-color: maroon;
  background-color: #FFEECC;
  color: maroon;
  width: 95%; padding: 0.5em; }

div.issue h4 {
  margin-top: 0;
  font-weight: bold;
  color: maroon;
}

pre {
   font-family: monospace;
   font-weight: bold;
   white-space: pre;
   background-color: rgb(204,204,255);
   padding: 0.5em;
   border: none;
   margin-left: 0;
   width: 90%;
}

pre.xml {
   font-family: "Lucida Sans Unicode", monospace;
   background-color: rgb(204,204,255);
   border: solid black thin;
}
code {
    color: green;
    font-family: monospace;
    font-weight: bold;
}

.RFC2119 {
  text-transform: lowercase;
  font-style: italic;
  font-weight: bold;
}

    /*]]>*/
</style>
  <link rel="stylesheet" type="text/css" href=
  "http://www.w3.org/StyleSheets/TR/W3C-REC" />
</head>

<body>
  <div class="head">
    <p><a href="http://www.w3.org/"><img height="48" alt="W3C" src=
    "http://www.w3.org/Icons/w3c_home" width="72" /></a></p>

    <h1 class="notoc" id="name">Pronunciation Lexicon Specification
    (PLS) Version 1.0</h1>

    <h2 class="notoc" id="date">W3C Recommendation 14 October 2008</h2>

    <dl>
      <dt>This version:</dt>

      <dd><a href=
      "http://www.w3.org/TR/2008/REC-pronunciation-lexicon-20081014/">
      http://www.w3.org/TR/2008/REC-pronunciation-lexicon-20081014/</a></dd>

      <dt>Latest version:</dt>

      <dd><a href=
      "http://www.w3.org/TR/pronunciation-lexicon/">http://www.w3.org/TR/pronunciation-lexicon/</a></dd>

      <dt>Previous version:</dt>

      <dd><a href=
      "http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20080818/">
      http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20080818/</a></dd>

      <dt>Editor:</dt>

      <dd>Paolo Baggia, Loquendo</dd>

      <dt>Authors:</dt>

      <dd>Paul Bagshaw, France Telecom</dd>

      <dd>Daniel C. Burnett, Voxeo</dd>

      <dd>Jerry Carter, Nuance</dd>

      <dd>Frank Scahill, BT <i>(until 10 October 2001)</i></dd>
    </dl>

    <p>Please refer to the
    <a href="http://www.w3.org/2008/10/pronunciation-lexicon-errata.html">
    <strong>errata</strong></a>
    for this document, which may include some normative
    corrections.</p>

    <p>See also
    <a href="http://www.w3.org/2003/03/Translations/byTechnology?technology=pronunciation-lexicon">
    <strong>translations</strong></a>.</p>

    <p class="copyright"><a href=
    "http://www.w3.org/Consortium/Legal/ipr-notice#Copyright">Copyright</a>
    © 2008 <a href="http://www.w3.org/"><acronym title=
    "World Wide Web Consortium">W3C</acronym></a><sup>®</sup>
    (<a href="http://www.csail.mit.edu/"><acronym title=
    "Massachusetts Institute of Technology">MIT</acronym></a>,
    <a href="http://www.ercim.org/"><acronym title=
    "European Research Consortium for Informatics and Mathematics">ERCIM</acronym></a>,
    <a href="http://www.keio.ac.jp/">Keio</a>), All Rights
    Reserved. W3C <a href=
    "http://www.w3.org/Consortium/Legal/ipr-notice#Legal_Disclaimer">
    liability</a>, <a href=
    "http://www.w3.org/Consortium/Legal/ipr-notice#W3C_Trademarks">trademark</a>
    and <a href=
    "http://www.w3.org/Consortium/Legal/copyright-documents">document
    use</a> rules apply.</p>
  </div>
  <hr title="Separator for header" />

  <div>
    <h2 class="notoc" id="abstract">Abstract</h2>

    <p>This document defines the syntax for specifying <a href=
    "#term-Pron-Lexicon">pronunciation lexicons</a> to be used by
    <a href="#term-ASR">Automatic Speech Recognition</a> and
    <a href="#term-TTS">Speech Synthesis</a> engines in voice
    browser applications.</p>
  </div>

  <h2 id="Status">Status of this Document</h2>

  <p><em>This section describes the status of this document at the
  time of its publication. Other documents may supersede this
  document. A list of current W3C publications and the latest
  revision of this technical report can be found in the <a href=
  "http://www.w3.org/TR/">W3C technical reports index</a> at
  http://www.w3.org/TR/.</em></p>

  <p>
  This is the <a href=
  "http://www.w3.org/2005/10/Process-20051014/tr.html#RecsW3C">Recommendation</a>
  of
  "Pronunciation Lexicon Specification (PLS) Version 1.0".

  It has been produced by the
  <a href="http://www.w3.org/Voice/">Voice Browser Working Group</a>,
  which is part of the
  <a href="http://www.w3.org/Voice/Activity.html">Voice Browser Activity</a>.
  </p>

  <p>Comments are welcome on <a
  href="mailto:www-voice@w3.org">www-voice@w3.org</a> (<a
  href="http://lists.w3.org/Archives/Public/www-voice/">archive</a>).
  See <a href="http://www.w3.org/Mail/">W3C mailing list and archive
  usage guidelines</a>.</p>

  <p>The design of PLS 1.0 has been widely reviewed (see the
  <a href="pls-disp.html">disposition of comments</a>)
  and satisfies the Working Group's technical requirements.

  A list of implementations is included in the
  <a href="http://www.w3.org/Voice/2008/pls-ir/">
  PLS 1.0 Implementation Report</a>,
  along with the associated test suite.

  The Working Group made a few editorial changes to the <a
  href="http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20080818/">
  18 August 2008 Proposed Recommendation</a> in response to comments.
  Changes from the Proposed Recommendation can be found in
  <a href="#AppD">Appendix D</a>.
  </p>

  <p><a href="#S2">Section 2. Pronunciation Alphabets</a> describes
  the legal values of the <code>alphabet</code> attribute for
  specifying a pronunciation alphabet. The Working Group is
  requesting the creation of a Pronunciation Alphabet registry with
  IANA so that pronunciation alphabets other than "ipa" can be also
  used. The location of the registry will be provided at <a href=
  "http://www.w3.org/2001/10/synthesis">http://www.w3.org/2001/10/synthesis</a>
  when the registry becomes available. A future version of the PLS
  specification may permit values from this registry to be used in
  the <code>alphabet</code> attribute.</p>

  <p>This document has been reviewed by W3C Members, by software
  developers, and by other W3C groups and interested parties, and is
  endorsed by the Director as a W3C Recommendation. It is a stable
  document and may be used as reference material or cited from another
  document. W3C's role in making the Recommendation is to draw
  attention to the specification and to promote its widespread
  deployment. This enhances the functionality and interoperability of
  the Web.</p>

<p>This document was produced by a group operating under the <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/" shape="rect">5 February 2004 W3C Patent Policy</a>. W3C maintains a <a rel="disclosure" href="http://www.w3.org/2004/01/pp-impl/34665/status" shape="rect">public list of any patent disclosures</a> made in connection with the deliverables of the group; that page also includes instructions for disclosing a patent. An individual who has actual knowledge of a patent which the individual believes contains <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/#def-essential" shape="rect">Essential Claim(s)</a> must disclose the information in accordance with <a href="http://www.w3.org/Consortium/Patent-Policy-20040205/#sec-Disclosure" shape="rect">section 6 of the W3C Patent Policy</a>. </p>

  <p>The sections in the main body of this document are normative
  unless otherwise specified.  The appendices in this document are
  informative unless otherwise indicated explicitly.</p>

  <hr />

  <h2 id="contents">Table of Contents</h2>

  <ul class='toc'>
    <li class='tocline'>1. <a href="#S1">Introduction to
    Pronunciation Lexicon Specification</a>

      <ul class='toc'>
        <li class='tocline'>1.1 <a href="#S1.1">How TTS Uses
        PLS</a></li>

        <li class='tocline'>1.2 <a href="#S1.2">How ASR Uses
        PLS</a></li>

        <li class='tocline'>1.3 <a href="#S1.3">How VoiceXML
        Applications Use PLS</a></li>

        <li class='tocline'>1.4 <a href="#S1.4">What PLS does not
        Support</a></li>

        <li class='tocline'>1.5 <a href="#S1.5">Glossary of
        Terms</a></li>
      </ul>
    </li>

    <li class='tocline'>2. <a href="#S2">Pronunciation
    Alphabets</a></li>

    <li class='tocline'>3. <a href="#S3">PLS Documents</a>

      <ul class='toc'>
        <li class='tocline'>3.1 <a href="#S3.1">Document
        Form</a></li>

        <li class='tocline'>3.2 <a href="#S3.2">Conformance</a>

          <ul class='toc'>
            <li class='tocline'>3.2.1 <a href="#S3.2.1">Conforming
            Pronunciation Lexicon Specification Documents</a></li>

            <li class='tocline'>3.2.2 <a href="#S3.2.2">Using PLS
            with other Namespaces</a></li>

            <li class='tocline'>3.2.3 <a href="#S3.2.3">Conforming
            Pronunciation Lexicon Specification Processors</a></li>
          </ul>
        </li>
      </ul>
    </li>

    <li class='tocline'>4. <a href="#S4">Pronunciation Lexicon
    Markup Language Definition</a>

      <ul class='toc'>
        <li class='tocline'>4.1 <a href=
        "#S4.1"><code>&lt;lexicon&gt;</code> Element</a></li>

        <li class='tocline'>4.2 <a href=
        "#S4.2"><code>&lt;meta&gt;</code> Element</a></li>

        <li class='tocline'>4.3 <a href=
        "#S4.3"><code>&lt;metadata&gt;</code> Element</a></li>

        <li class='tocline'>4.4 <a href=
        "#S4.4"><code>&lt;lexeme&gt;</code> Element</a></li>

        <li class='tocline'>4.5 <a href=
        "#S4.5"><code>&lt;grapheme&gt;</code> Element</a></li>

        <li class='tocline'>4.6 <a href=
        "#S4.6"><code>&lt;phoneme&gt;</code> Element</a></li>

        <li class='tocline'>4.7 <a href=
        "#S4.7"><code>&lt;alias&gt;</code> Element</a></li>

        <li class='tocline'>4.8 <a href=
        "#S4.8"><code>&lt;example&gt;</code> Element</a></li>

        <li class='tocline'>4.9 <a href="#S4.9">Multiple
        Pronunciations for ASR and TTS</a>

          <ul class='toc'>
            <li class='tocline'>4.9.1 <a href="#S4.9.1">Multiple
            Pronunciations for ASR</a></li>

            <li class='tocline'>4.9.2 <a href="#S4.9.2">Multiple
            Pronunciations for TTS</a></li>

            <li class='tocline'>4.9.3 <a href="#S4.9.3">Examples of
            Multiple Pronunciations</a></li>
          </ul>
        </li>
      </ul>
    </li>

    <li class='tocline'>5. <a href="#S5">Examples</a>

      <ul class='toc'>
        <li class='tocline'>5.1 <a href="#S5.1">Simple Use
        Case</a></li>

        <li class='tocline'>5.2 <a href="#S5.2">Multiple
        Pronunciations</a></li>

        <li class='tocline'>5.3 <a href="#S5.3">Multiple
        Orthographies</a></li>

        <li class='tocline'>5.4 <a href="#S5.4">Homophones</a></li>

        <li class='tocline'>5.5 <a href="#S5.5">Homographs</a></li>

        <li class='tocline'>5.6 <a href="#S5.6">Pronunciation by
        Orthography (Acronyms, Abbreviations, etc.)</a></li>
      </ul>
    </li>

    <li class='tocline'>6. <a href="#S6">References</a>

      <ul class='toc'>
        <li class='tocline'>6.1 <a href="#S6.1">Normative
        References</a></li>

        <li class='tocline'>6.2 <a href="#S6.2">Informative
        References</a></li>
      </ul>
    </li>

    <li class='tocline'>7. <a href="#S7">Contributors and
    Acknowledgements</a></li>

    <li class='tocline'>Appendix A. <a href="#AppA">Schema for
    Pronunciation Lexicon Specification</a> (normative)</li>

    <li class='tocline'>Appendix B. <a href="#AppB">MIME Type and
    File Suffix</a> (normative)</li>

    <li class='tocline'>Appendix C. <a href="#AppC">Issues in
    Retrieving Lexical Content</a> (informative)</li>

    <li class='tocline'>Appendix D. <a href="#AppD">Changes</a>
    (informative)</li>
  </ul>

  <h2 id="S1">1. Introduction to Pronunciation Lexicon
  Specification</h2>

  <p><i>This section is informative.</i></p>

  <p>The accurate specification of pronunciation is critical to the
  success of speech applications. Most Automatic Speech Recognition
  (<a href="#term-ASR">ASR</a>) and Text-To-Speech (<a href=
  "#term-TTS">TTS</a>) engines internally provide extensive high
  quality <a href="#term-Lexicon">lexicons</a> with pronunciation
  information for many words or phrases. To ensure a maximum
  coverage of the words or phrases used by an application,
  application-specific pronunciations may be required. For example,
  these may be needed for proper nouns such as surnames or business
  names.</p>

  <p>The Pronunciation Lexicon Specification (PLS) is designed to
  enable interoperable specification of pronunciation information
  for both <a href="#term-ASR">ASR</a> and <a href=
  "#term-TTS">TTS</a> engines. The language is intended to be easy
  to use by developers while supporting the accurate specification
  of pronunciation information for international use.</p>

  <p>The language allows one or more pronunciations for a word or
  phrase to be specified using a standard pronunciation alphabet or
  if necessary using vendor specific alphabets. Pronunciations are
  grouped together into a PLS document which may be referenced from
  other markup languages, such as the <a href="#term-SRGS">Speech
  Recognition Grammar Specification</a> [<a href=
  "#ref-SRGS">SRGS</a>] and the <a href="#term-SSML">Speech
  Synthesis Markup Language</a> [<a href="#ref-SSML">SSML</a>].</p>

  <p>In its most general sense, a <a href=
  "#term-Lexicon">lexicon</a> is merely a list of words or phrases,
  possibly containing information associated with and related to
  the items in the list. This document uses the term <a href=
  "#term-Lexicon">"lexicon"</a> in only one specific way, as
  <a href="#term-Pron-Lexicon">"pronunciation lexicon"</a>. In this
  particular document, "lexicon" means a mapping between words (or
  short phrases), their written representations, and their
  pronunciations suitable for use by an <a href="#term-ASR">ASR</a>
  engine or a <a href="#term-TTS">TTS</a> engine. Pronunciation
  lexicons are not only useful for voice browsers; they have also
  proven effective mechanisms to support accessibility for persons
  with disabilities as well as greater usability for all users.
  They are used to good effect in screen readers and user agents
  supporting multimodal interfaces.</p>

  <h2 id="S1.1">1.1. How TTS Uses PLS</h2>

  <p>A <a href="#term-TTS">TTS</a> engine aims to transform input
  content (either text or markup, such as <a href=
  "#term-SSML">SSML</a>) into speech. This activity involves
  several processing steps:</p>

  <ul>
    <li>Text normalization</li>

    <li>Word pronunciation (lexical stress, phonetic
    transcription)</li>

    <li>Sentence structure (intonation, rhythm)</li>

    <li>Sentence level modification in phonetic transcription
    (co-articulation)</li>

    <li>Computation of prosodic parameters</li>

    <li>Generation of the acoustic signal</li>
  </ul>

  <p><a href="#term-SSML">SSML</a> enables a user to control and
  enhance <a href="#term-TTS">TTS</a> activity by acting through
  <a href="#term-SSML">SSML</a> elements on these levels of
  processing (see [<a href="#ref-SSML">SSML</a>] for details).</p>

  <p>PLS is intended to be the standard format of the documents
  referenced by the <a href=
  "#S4.4"><code>&lt;lexicon&gt;</code></a> element of <a href=
  "#term-SSML">SSML</a> (see <a href=
  "http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/#S3.1.4">
  Section 3.1.4</a> of [<a href="#ref-SSML">SSML</a>]).</p>

  <p>The following is a simple example of an <a href=
  "#term-SSML">SSML</a> document. It includes an Italian movie
  title and the name of the director to be read in US English.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
    xmlns="http://www.w3.org/2001/10/synthesis"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
      http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
    xml:lang="en-US"&gt;

    The title of the movie is: "La vita è bella" (Life is beautiful),
    which is directed by Roberto Benigni.
&lt;/speak&gt;
</pre>

  <p>To be pronounced correctly the Italian title and the
  director's name might include the pronunciation inline in the
  <a href="#term-SSML">SSML</a> document.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
    xmlns="http://www.w3.org/2001/10/synthesis"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
      http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
    xml:lang="en-US"&gt;

    The title of the movie is:
    &lt;phoneme alphabet="ipa" ph="ˈlɑ ˈviːɾə ˈʔeɪ ˈbɛlə"&gt;"La vita è bella"&lt;/phoneme&gt;
    &lt;!-- The IPA pronunciation is:
    "&amp;#x02C8;l&amp;#x0251; &amp;#x02C8;vi&amp;#x02D0;&amp;#x027E;&amp;#x0259;
     &amp;#x02C8;&amp;#x0294;e&amp;#x026A; &amp;#x02C8;b&amp;#x025B;l&amp;#x0259;" --&gt;
    (Life is beautiful),
    which is directed by
    &lt;phoneme alphabet="ipa" ph="ɹəˈbɛːɹɾoʊ bɛˈniːnji"&gt;Roberto Benigni.&lt;/phoneme&gt;
    &lt;!-- The IPA pronunciation is:
    "&amp;#x0279;&amp;#x0259;&amp;#x02C8;b&amp;#x025B;&amp;#x02D0;&amp;#x0279;&amp;#x027E;o&amp;#x028A;
     b&amp;#x025B;&amp;#x02C8;ni&amp;#x02D0;nji" --&gt;
&lt;/speak&gt;
</pre>

  <p>Using PLS, all the pronunciations can be factored out into an
  external PLS document which is referenced by the <a href=
  "#S4.4"><code>&lt;lexicon&gt;</code></a> element of <a href=
  "#term-SSML">SSML</a> (see <a href=
  "http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/#S3.1.4">
  Section 3.1.4</a> of [<a href="#ref-SSML">SSML</a>]).</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
    xmlns="http://www.w3.org/2001/10/synthesis"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
      http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
    xml:lang="en-US"&gt;

    &lt;lexicon uri="http://www.example.com/movie_lexicon.pls"/&gt;

    The title of the movie is: "La vita è bella" (Life is beautiful),
    which is directed by Roberto Benigni.
&lt;/speak&gt;
</pre>

  <p>The referenced lexicon might look something like this:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;La vita &amp;#x00E8;&lt;!-- same as: è --&gt; bella&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈlɑ ˈviːɾə ˈʔeɪ ˈbɛlə&lt;/phoneme&gt;
    &lt;!-- IPA string is:
     "&amp;#x02C8;l&amp;#x0251; &amp;#x02C8;vi&amp;#x02D0;&amp;#x027E;&amp;#x0259;
      &amp;#x02C8;&amp;#x0294;e&amp;#x026A; &amp;#x02C8;b&amp;#x025B;l&amp;#x0259;" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Roberto&lt;/grapheme&gt;
    &lt;phoneme&gt;ɹəˈbɛːɹɾoʊ&lt;/phoneme&gt;
    &lt;!-- IPA string is:
     "&amp;#x0279;&amp;#x0259;&amp;#x02C8;b&amp;#x025B;&amp;#x02D0;&amp;#x0279;&amp;#x027E;o&amp;#x028A;" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Benigni&lt;/grapheme&gt;
    &lt;phoneme&gt;bɛˈniːnji&lt;!-- IPA string is:
     "b&amp;#x025B;&amp;#x02C8;ni&amp;#x02D0;nji" --&gt;&lt;/phoneme&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>The PLS engine will load the external PLS document and
  transparently apply the pronunciations during the processing of
  the <a href="#term-SSML">SSML</a> document. An application may
  contain several distinct PLS documents to be used at different
  points within the application. <a href=
  "http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/#S3.1.4">
  Section 3.1.4</a> of [<a href="#ref-SSML">SSML</a>] describes how
  to use more than one lexicon document referenced in a <a href=
  "#term-SSML">SSML</a> document.</p>

  <h4 id="S1.1.0">Informative Note:</h4>

  <p>Given that many platform/browser/text editor combinations do
  not correctly cut and paste Unicode text, IPA symbols may be
  entered as numeric character references (see Section 4.1 on
  Character and Entity References of either XML 1.0 [<a href=
  "#ref-xml-10">XML10</a>] or XML 1.1 [<a href=
  "#ref-xml-11">XML11</a>]) in the pronunciation. However, the
  UTF-8 representation of an IPA symbol should always be used in
  preference to its numeric character reference. In order to
  overcome potential problems with viewing the UTF-8 representation
  of IPA symbols in this document, pronunciation examples are also
  shown in a comment using numeric character references.</p>

  <h2 id="S1.2">1.2. How ASR Uses PLS</h2>

  <p>An <a href="#term-ASR">ASR</a> engine transforms an audio
  signal into a recognized sequence of words or a semantic
  representation of the meaning of the utterance (see <a href=
  "#term-SISR">Semantic Interpretation for Speech Recognition</a>
  [<a href="#ref-SISR">SISR</a>] for a standard definition of
  Semantic Interpretation).</p>

  <p>An <a href="#term-ASR">ASR</a> grammar is used to improve
  <a href="#term-ASR">ASR</a> performance by describing the
  possible words and phrases the <a href="#term-ASR">ASR</a> might
  recognize. <a href="#term-SRGS">SRGS</a> is the standard
  definition of <a href="#term-ASR">ASR</a> grammars (see [<a href=
  "#ref-SRGS">SRGS</a>] for details).</p>

  <p>PLS may be used by an <a href="#term-ASR">ASR</a> processor to
  allow multiple pronunciations of words and phrases, and also to
  do limited text normalization, such as <a href=
  "#term-acronym-exp">acronym expansion</a> and abbreviations.</p>

  <p>PLS entries are applied to the graphemes inside <a href=
  "#term-SRGS">SRGS</a> grammar rules to convert them into the
  phonemes to be recognized. See the example below and the example
  in <a href="#S1.3">Section 1.3</a> for a PLS document used for
  both <a href="#term-ASR">ASR</a> and <a href=
  "#term-TTS">TTS</a>.</p>

  <p>There might be other uses of PLS, for instance in a dictation
  system or for unconstrained <a href="#term-ASR">ASR</a>, which
  might be beyond the scope of this specification.</p>

  <p>This is a very simple <a href="#term-SRGS">SRGS</a> grammar
  that allows the recognition of sentences like "Boston
  Massachusetts" or "Miami Florida".</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;grammar version="1.0"
  xmlns="http://www.w3.org/2001/06/grammar"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xsi:schemaLocation="http://www.w3.org/2001/06/grammar
    http://www.w3.org/TR/speech-grammar/grammar.xsd"
  xml:lang="en-US" root="city_state" mode="voice"&gt;

  &lt;rule id="city" scope="public"&gt;
    &lt;one-of&gt; &lt;item&gt;Boston&lt;/item&gt;
             &lt;item&gt;Miami&lt;/item&gt;
             &lt;item&gt;Fargo&lt;/item&gt; &lt;/one-of&gt;
  &lt;/rule&gt;
  &lt;rule id="state" scope="public"&gt;
    &lt;one-of&gt; &lt;item&gt;Florida&lt;/item&gt;
             &lt;item&gt;North Dakota&lt;/item&gt;
             &lt;item&gt;Massachusetts&lt;/item&gt; &lt;/one-of&gt;
  &lt;/rule&gt;

  &lt;rule id="city_state" scope="public"&gt;
     &lt;ruleref uri="#city"/&gt; &lt;ruleref uri="#state"/&gt;
  &lt;/rule&gt;
&lt;/grammar&gt;
</pre>

  <p>If a <a href="#term-Pron-Lexicon">pronunciation lexicon</a> is
  referenced by a <a href="#term-SRGS">SRGS</a> grammar it can
  allow multiple pronunciations of the word in the grammar to
  accommodate different speaking styles. Here is the same grammar
  with a reference to an external PLS document.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;grammar version="1.0"
  xmlns="http://www.w3.org/2001/06/grammar"
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
  xsi:schemaLocation="http://www.w3.org/2001/06/grammar
    http://www.w3.org/TR/speech-grammar/grammar.xsd"
  xml:lang="en-US" root="city_state" mode="voice"&gt;

  &lt;lexicon uri="http://www.example.com/city_lexicon.pls"/&gt;

  &lt;rule id="city" scope="public"&gt;
    &lt;one-of&gt; &lt;item&gt;Boston&lt;/item&gt;
             &lt;item&gt;Miami&lt;/item&gt;
             &lt;item&gt;Fargo&lt;/item&gt; &lt;/one-of&gt;
  &lt;/rule&gt;
  &lt;rule id="state" scope="public"&gt;
    &lt;one-of&gt; &lt;item&gt;Florida&lt;/item&gt;
             &lt;item&gt;North Dakota&lt;/item&gt;
             &lt;item&gt;Massachusetts&lt;/item&gt; &lt;/one-of&gt;
  &lt;/rule&gt;

  &lt;rule id="city_state" scope="public"&gt;
     &lt;ruleref uri="#city"/&gt; &lt;ruleref uri="#state"/&gt;
  &lt;/rule&gt;
&lt;/grammar&gt;
</pre>

  <p>Note also that an <a href="#term-SRGS">SRGS</a> grammar might
  reference multiple PLS documents.</p>

  <h2 id="S1.3">1.3. How VoiceXML Applications Use PLS</h2>

  <p>A VoiceXML 2.0 application ([<a href="#ref-VXML">VXML</a>])
  contains <a href="#term-SRGS">SRGS</a> grammars for <a href=
  "#term-ASR">ASR</a> and <a href="#term-SSML">SSML</a> prompts for
  <a href="#term-TTS">TTS</a>. The introduction of PLS into both
  <a href="#term-SRGS">SRGS</a> and <a href="#term-SSML">SSML</a>
  will directly impact VoiceXML applications.</p>

  <p>The benefits described in <a href="#S1.1">Section 1.1</a> and
  <a href="#S1.2">Section 1.2</a> are also available in VoiceXML
  applications. The application may use several contextual PLS
  documents at different points in the interaction, but may also
  use the same PLS document both in <a href="#term-SRGS">SRGS</a>,
  to improve <a href="#term-ASR">ASR</a>, and in <a href=
  "#term-SSML">SSML</a>, to improve <a href="#term-TTS">TTS</a>.
  Here is an example PLS document:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;judgment&lt;/grapheme&gt;
    &lt;grapheme&gt;judgement&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈdʒʌdʒ.mənt&lt;/phoneme&gt;
    &lt;!-- IPA string is:
    "&amp;#x02C8;d&amp;#x0292;&amp;#x028C;d&amp;#x0292;&amp;#x002E;m&amp;#x0259;nt" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;fiancé&lt;/grapheme&gt;
    &lt;grapheme&gt;fiance&lt;/grapheme&gt;
    &lt;phoneme&gt;fiˈɒns.eɪ&lt;/phoneme&gt;
    &lt;!-- IPA string is:
    "fi&amp;#x02C8;&amp;#x0252;ns&amp;#x002E;e&amp;#x026A;" --&gt;
    &lt;phoneme&gt;ˌfiː.ɑːnˈseɪ&lt;/phoneme&gt;
    &lt;!-- IPA string is:
    "&amp;#x02CC;fi&amp;#x02D0;&amp;#x002E;&amp;#x0251;&amp;#x02D0;n&amp;#x02C8;se&amp;#x026A;" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>which could be used to improve <a href="#term-TTS">TTS</a> as
  shown in the following <a href="#term-SSML">SSML</a>
  document:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
    xmlns="http://www.w3.org/2001/10/synthesis"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
      http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
    xml:lang="en-US"&gt;

    &lt;lexicon uri="http://www.example.com/lexicon_defined_above.xml"/&gt;

    &lt;p&gt; In the judgement of my fiancé, Las Vegas is the best place for a honeymoon.
              I replied that I preferred Venice and didn't think the Venetian casino was an
              acceptable compromise.&lt;\p&gt;
&lt;/speak&gt;
</pre>

  <p>but also to improve <a href="#term-ASR">ASR</a> in the
  following <a href="#term-SRGS">SRGS</a> grammar:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;grammar version="1.0"
    xmlns="http://www.w3.org/2001/06/grammar"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://www.w3.org/2001/06/grammar
      http://www.w3.org/TR/speech-grammar/grammar.xsd"
    xml:lang="en-US" root="movies" mode="voice"&gt;

  &lt;lexicon uri="http://www.example.com/lexicon_defined_above.xml"/&gt;

  &lt;rule id="movies" scope="public"&gt;
     &lt;one-of&gt;
             &lt;item&gt;Terminator 2: Judgment Day&lt;/item&gt;
             &lt;item&gt;My Big Fat Obnoxious Fiance&lt;/item&gt;
             &lt;item&gt;Pluto's Judgement Day&lt;/item&gt;
     &lt;/one-of&gt;
  &lt;/rule&gt;
&lt;/grammar&gt;
</pre>

  <h2 id="S1.4">1.4. What PLS does not Support</h2>

  <p>The current specification is focused on the major features
  described in the requirements document [<a href=
  "#ref-REQS">REQS</a>]. The most complex features have been
  postponed to a future revision of this specification. Some of the
  complex features not included, for instance, are the introduction
  of morphological, syntactic and semantic information associated
  with pronunciations (such as word stems, inter-word semantic
  links, pronunciation statistics, etc.). Many of these features
  can be specified using RDF [<a href=
  "#ref-rdf-xml">RDF-XMLSYNTAX</a>] that reference lexemes within
  one or more <a href="#term-Pron-Lexicon">pronunciation
  lexicons</a>.</p>

  <h2 id="S1.5">1.5. Glossary of Terms</h2>

  <dl>
    <dt><b><a id="term-acronym" name=
    "term-acronym">Acronym</a></b></dt>

    <dd>A word formed from the initial letters of a name, such as
    <b><i>PC</i></b> for <b>P</b>ersonal <b>C</b>omputer, or by
    combining initial letters or parts of a series of words, such
    as <b><i>radar</i></b> for <b>ra</b>dio <b>d</b>etection
    <b>a</b>nd <b>r</b>anging, or variations, such as
    <b><i>W3C</i></b> for <b>W</b>orld <b>W</b>ide <b>W</b>eb
    <b>c</b>onsortium.</dd>

    <dt><b><a id="term-acronym-exp" name="term-acronym-exp">Acronym
    expansion</a></b></dt>

    <dd>The action of replacing an <a href=
    "#term-acronym">acronym</a> by the sequence of words it
    represents. Acronym expansion is typically performed to help
    <a href="#term-TTS">TTS</a> engines read acronyms and <a href=
    "#term-ASR">ASR</a> to recognize them.</dd>

    <dt><b><a id="term-ASR" name="term-ASR">ASR, Automatic Speech
    Recognition, Speech Recognition</a></b></dt>

    <dd>The process of using an automatic computation algorithm to
    analyze spoken utterances to determine what words and phrases
    or semantic information were present.</dd>

    <dt><b><a id="term-grapheme" name=
    "term-grapheme">Grapheme</a></b></dt>

    <dd>One of the set of the smallest units of a written language,
    such as letters, ideograms, or symbols, that distinguish one
    word from another; a representation of a single orthographic
    element.</dd>

    <dt><b><a id="term-Homophone" name=
    "term-Homophone">Homophone</a></b></dt>

    <dd>One of a set of words that are pronounced the same way but
    differ in meaning, origin, and sometimes spelling. E.g.
    <i>night</i> and <i>knight</i> in English. Note that
    <i>color</i> and <i>colour</i> are considered multiple <a href=
    "#term-Orthography">orthographies</a> for the same word and not
    homophones, because they are variations of spelling with the
    same pronunciation and meaning.</dd>

    <dt><b><a id="term-Homograph" name=
    "term-Homograph">Homograph</a></b></dt>

    <dd>'a word of the same written form as another but of
    different meaning and usually origin, whether pronounced the
    same way or not, as <em>bear</em> "to carry; support" and
    <em>bear</em> "animal" or <em>lead</em> "to conduct" and
    <em>lead</em> "metal."' [<a href="#ref-dict">DICT</a>] An
    example from French is <i>fils</i> (son) and <i>fils</i>
    (threads).</dd>

    <dt><b><a id="term-IPA" name="term-IPA">International Phonetic
    Alphabet (IPA)</a></b></dt>

    <dd>The International Phonetic Alphabet [<a href=
    "#ref-ipa">IPA</a>] is a <a href=
    "#term-Phonetic-Alphabet">phonetic alphabet</a> used by
    linguists to accurately and uniquely represent each of the wide
    variety of sounds (phones or <a href=
    "#term-Phoneme">phonemes</a>) the human vocal apparatus can
    produce. It is intended as a notational standard for the
    phonetic representation of all languages.</dd>

    <dt><b><a id="term-Lexeme" name=
    "term-Lexeme">Lexeme</a></b></dt>

    <dd>An atomic unit in a language, like a word or a stem. In
    this specification, "lexeme" designates a collection of
    <a href="#term-grapheme">graphemic</a> and pronunciation
    representations (e.g. <a href="#term-IPA">IPA</a>, <a href=
    "#term-SAMPA">SAMPA</a>, Pinyin, etc.) of words or
    phrases.</dd>

    <dt><b><a id="term-Lexicon" name=
    "term-Lexicon">Lexicon</a></b></dt>

    <dd>In its most general sense, a lexicon is merely a list of
    words or phrases, possibly containing information associated
    with and related to the items in the list. This document uses
    the term "lexicon" in only one specific way, as <a href=
    "#term-Pron-Lexicon">"pronunciation lexicon"</a>, which means a
    mapping between words (or short phrases), their written
    representations, and their pronunciations suitable for use by
    an <a href="#term-ASR">ASR</a> engine or a <a href=
    "#term-TTS">TTS</a> engine. However, the word "lexicon" can
    mean other things in other contexts.</dd>

    <dt><b><a id="term-Orthography" name=
    "term-Orthography">Orthography</a></b></dt>

    <dd>A notation for writing and displaying words. Orthography
    includes character sets, white space, case sensitivity,
    diacritics within languages such as Arabic or Persian, and
    accents within languages such as French.</dd>

    <dt><b><a id="term-Phoneme" name=
    "term-Phoneme">Phoneme</a></b></dt>

    <dd>One of the set of the smallest units of speech that can
    distinguish words: for example, the English language has over
    40 <a href="#term-Phoneme">phonemes</a> (19 vowels and 24
    consonants). In American English, /t/ and /p/ are <a href=
    "#term-Phoneme">phonemes</a> that can distinguish the word
    <i>tin</i> from <i>pin</i>.</dd>

    <dt><b><a id="term-Phonetic-Alphabet" name=
    "term-Phonetic-Alphabet">Phonetic alphabet</a></b></dt>

    <dd>A set of symbols that represent the sounds in spoken
    languages, such as English, Chinese, or German, see also
    <a href="#term-IPA">International Phonetic Alphabet</a>.</dd>

    <dt><b><a id="term-Pron-Lexicon" name=
    "term-Pron-Lexicon">Pronunciation lexicon</a></b></dt>

    <dd>The term pronunciation lexicon means a mapping between
    words (or short phrases), their written representations, and
    their pronunciations suitable for use by an <a href=
    "#term-ASR">ASR</a> engine or a <a href="#term-TTS">TTS</a>
    engine.</dd>

    <dt><b><a id="term-SAMPA" name="term-SAMPA">SAMPA</a></b></dt>

    <dd>The Speech Assessment Methods Phonetic Alphabet [<a href=
    "#ref-sampa">SAMPA</a>]: a <a href=
    "#term-Phonetic-Alphabet">phonetic alphabet</a> using only
    ASCII characters, rather than the extended character set used
    by the <a href="#term-IPA">International Phonetic
    Alphabet</a>.</dd>

    <dt><b><a id="term-SISR" name="term-SISR">Semantic
    Interpretation for Speech Recognition</a> [<a href=
    "#ref-SISR">SISR</a>]</b></dt>

    <dd>A W3C specification defining a process to produce a
    semantic result representing the meaning of a natural language
    utterance.</dd>

    <dt><b><a id="term-SRGS" name="term-SRGS">Speech Recognition
    Grammar Specification</a> [<a href=
    "#ref-SRGS">SRGS</a>]</b></dt>

    <dd>A W3C specification defining a language to describe
    grammars (words or phrases) that an <a href="#term-ASR">ASR</a>
    engine can recognize.</dd>

    <dt><b><a id="term-SSML" name="term-SSML">Speech Synthesis
    Markup Language</a> [<a href="#ref-SSML">SSML</a>]</b></dt>

    <dd>A W3C XML language for specifying the rendering of text by
    a <a href="#term-TTS">TTS</a> engine.</dd>

    <dt><b><a id="term-TTS" name="term-TTS">TTS, Text-To-Speech,
    Speech Synthesis</a></b></dt>

    <dd>Converting text into sounds using speech synthesis
    techniques.</dd>

    <dt><b><em><a id="term-uri" name="term-uri" shape="rect" href=
    "http://www.w3.org/TR/webarch/#def-uri">URI: Uniform Resource
    Identifier</a></em></b></dt>

    <dd>A global identifier in the context of the World Wide Web
    [<a href="#ref-web-arch" shape="rect">WEB-ARCH</a>]. A URI is
    defined as any legal <code>anyURI</code> primitive as defined
    in Section 3.2.17 of XML Schema Part 2: Datatypes [<a href=
    "#ref-xmlschema-2" shape="rect">XML-SCHEMA-2</a>]. For
    informational purposes only, [<a href="#ref-rfc3986" shape=
    "rect">RFC3986</a>] and [<a href="#ref-rfc2732" shape=
    "rect">RFC2732</a>] may be useful in understanding the
    structure, format, and use of URIs. Note that IRIs (see
    [<a href="#ref-rfc3987" shape="rect">RFC3987</a>]) are
    permitted within the above definition of URI.</dd>
  </dl>

  <h2><a id="S2" name="S2"></a>2. Pronunciation Alphabets</h2>

  <p>A phonemic/<a href="#term-Phonetic-Alphabet">phonetic
  alphabet</a> is used to specify a pronunciation. An alphabet in
  this context refers to a collection of symbols to represent the
  sounds of one or more human languages. In the PLS specification
  the pronunciation alphabet is specified by the
  <code>alphabet</code> attribute (see <a href="#S4.1">Section
  4.1</a> and <a href="#S4.6">Section 4.6</a> for details on the
  use of this attribute). The only valid values for the
  <code>alphabet</code> attribute are <code>"ipa"</code> (see the
  next paragraph) and vendor-defined strings of the form
  <code>"x-organization"</code> or
  <code>"x-organization-alphabet"</code>. For example, the Japan
  Electronics and Information Technology Industries Association
  [<a href="#ref-jeita">JEITA</a>] might wish to encourage the use
  of an alphabet such as <code>"x-jeita"</code> or
  <code>"x-jeita-2000"</code> for their <a href=
  "#term-Phoneme">phoneme</a> alphabet [<a href=
  "#ref-jeidaalphabet">JEIDAALPHABET</a>]. Another example might be
  <code>"x-sampa"</code> [<a href="#ref-x-sampa">X-SAMPA</a>], an
  extension of the <a href="#term-SAMPA">SAMPA</a> phonetic
  alphabet [<a href="#ref-sampa">SAMPA</a>] to cover the entire
  range of characters in the <a href="#term-IPA">International
  Phonetic Alphabet</a> [<a href="#ref-ipa">IPA</a>].</p>

  <p>A compliant PLS processor <em title="MUST in RFC 2119 context"
  class="RFC2119">MUST</em> support <code>"ipa"</code> as the value
  of the <code>alphabet</code> attribute. This means that the PLS
  processor <em title="MUST in RFC 2119 context" class=
  "RFC2119">MUST</em> support the Unicode representations of the
  phonetic characters developed by the International Phonetic
  Association [<a href="#ref-ipa">IPA</a>]. In addition to an
  exhaustive set of vowel and consonant symbols, this character set
  supports a syllable delimiter, numerous diacritics, stress
  symbols, lexical tone symbols, intonational markers and more. For
  this alphabet, legal phonetic/phonemic values are strings of the
  values specified in Appendix 2 of [<a href=
  "#ref-ipahndbk">IPAHNDBK</a>]; note that an IPA transcription may
  contain white space characters to assist readability, which have
  no implications for the pronunciation. Informative tables of the
  IPA-to-Unicode mappings can be found at [<a href=
  "#ref-ipaunicode1">IPAUNICODE1</a>] and [<a href=
  "#ref-ipaunicode2">IPAUNICODE2</a>]. Note that not all of the IPA
  characters are available in Unicode. For processors supporting
  this alphabet,</p>

  <ul>
    <li>The processor <em title="MUST in RFC 2119 context" class=
    "RFC2119">MUST</em> syntactically accept all legal values.</li>

    <li>The processor <em title="SHOULD in RFC 2119 context" class=
    "RFC2119">SHOULD</em> handle all Unicode IPA codes that can
    reasonably be considered to belong to the current
    language.</li>
  </ul>

  <h4 id="S2.0">Informative Note:</h4>

  <p>Note that there are peculiarities in the IPA alphabet which
  might have implications for implementers, for instance
  equivalent, withdrawn and superseded IPA symbols; see Appendix 2
  of [<a href="#ref-ipahndbk">IPAHNDBK</a>] for further
  details.</p>

  <h4 id="S2.1">Informative Note:</h4>

  <p>When IPA symbols are used to represent the phonemes of a
  language, there can be an ambiguity concerning which allophonic
  symbol to select to represent a phoneme. Note that this may
  result in inconsistencies between lexicons which were composed
  for the identical language.</p>

  <h4 id="S2.2">Informative Note:</h4>

  <p>Currently there is no ready way for a blind or partially
  sighted person to read or interact with a lexicon containing IPA
  symbols. It is hoped that implementers will provide tools which
  will enable such an interaction.</p>

  <div>
    <h2 id="S3">3. PLS Documents</h2>

    <h3 id="S3.1">3.1 Document Form</h3>

    <p>A legal Pronunciation Lexicon Specification document
    <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
    have a legal XML Prolog from Section 2.8 of either XML 1.0
    [<a href="#ref-xml-10">XML10</a>] or XML 1.1 [<a href=
    "#ref-xml-11">XML11</a>]. The XML prolog is followed by the
    root <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element.
    See <a href="#S4.1">Section 4.1</a> for details on this
    element.</p>

    <p>The <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element
    <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
    designate the PLS namespace. This can be achieved by declaring
    an <code class="att">xmlns</code> attribute or an attribute
    with an "xmlns" prefix. See Section 2 of Namespaces in XML
    (Namespaces in XML 1.0 [<a href="#ref-xmlns-10">XML-NS10</a>]
    or Namespaces in XML 1.1 [<a href=
    "#ref-xmlns-11">XML-NS11</a>]) for details. Note that when the
    <code class="att">xmlns</code> attribute is used alone, it sets
    the default namespace for the element on which it appears and
    for any child elements. The namespace for PLS is defined to be
    <code>"http://www.w3.org/2005/01/pronunciation-lexicon"</code>.</p>

    <p>It is <em title="RECOMMENDED in RFC 2119 context" class=
    "RFC2119">RECOMMENDED</em> that the <a href=
    "#S4.1"><code>&lt;lexicon&gt;</code></a> element also indicate
    the location of the PLS schema (see <a href="#AppA">Appendix
    A</a>) via the <code class="att"><a href=
    "http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/#xsi_schemaLocation">
    xsi:schemaLocation</a></code> attribute from <a href=
    "http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/#Instance_Document_Constructions">
    Section 2.6.3</a> of XML Schema Part 1: Structures Second
    Edition [<a href="#ref-xmlschema-1">XML-SCHEMA-1</a>].</p>

    <p>The following is an example of a legal PLS header:</p>
    <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
</pre>
  </div>

  <h3 id="S3.2">3.2 Conformance</h3>

  <p>This section enumerates the conformance rules of this
  specification.</p>

  <p>All sections in this specification are normative, unless
  otherwise indicated. The informative parts of this specification
  are identified by "Informative" labels within sections.</p>

  <p>Individual conformance requirements or testable statements are
  identifiable in the PLS specification through imperative voice
  statements. The key words "<em title="MUST in RFC 2119 context"
  class="RFC2119">MUST</em>", "<em title=
  "MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>",
  "<em title="REQUIRED in RFC 2119 context" class=
  "RFC2119">REQUIRED</em>", "<em title="SHALL in RFC 2119 context"
  class="RFC2119">SHALL</em>", "<em title=
  "SHALL NOT in RFC 2119 context" class="RFC2119">SHALL NOT</em>",
  "<em title="SHOULD in RFC 2119 context" class=
  "RFC2119">SHOULD</em>", "<em title=
  "SHOULD NOT in RFC 2119 context" class="RFC2119">SHOULD
  NOT</em>", "<em title="RECOMMENDED in RFC 2119 context" class=
  "RFC2119">RECOMMENDED</em>", "<em title="MAY in RFC 2119 context"
  class="RFC2119">MAY</em>", and "<em title=
  "OPTIONAL in RFC 2119 context" class="RFC2119">OPTIONAL</em>" in
  this document are to be interpreted as described in [<a href=
  "#ref-rfc2119">RFC2119</a>]. However, for readability, these
  words do not appear in all uppercase letters in this
  specification.</p>

  <div>
    <h3 id="S3.2.1">3.2.1 Conforming Pronunciation Lexicon
    Specification Documents</h3>

    <p>A document is a <em>Conforming Pronunciation Lexicon
    Specification Document</em> if it meets both the following
    conditions:</p>

    <ul>
      <li>It is well-formed according to the version of XML used
      (either XML 1.0 [<a href="#ref-xml-10">XML10</a>] or XML 1.1
      [<a href="#ref-xml-11">XML11</a>]) and conforms to the
      corresponding Namespaces in XML specification (Namespaces in
      XML 1.0 [<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces
      in XML 1.1 [<a href="#ref-xmlns-11">XML-NS11</a>]).</li>

      <li>It adheres to the specification described in this
      document (<a href="#S1" shape="rect">Pronunciation Lexicon
      Specification</a>) including the constraints expressed in the
      Schema (see <a href="#AppA" shape="rect">Appendix A</a>) and
      having an XML Prolog and <a href=
      "#S4.1"><code>&lt;lexicon&gt;</code></a> root element as
      specified in <a href="#S3.1">Section 3.1</a>.</li>
    </ul>

    <p>This specification and these conformance criteria provide no
    designated size limits on any aspect of PLS documents. There
    are no maximum values on the number of elements, the amount of
    character data, or the number of characters in attribute
    values.</p>

    <h3 id="S3.2.2">3.2.2 Using PLS with other Namespaces</h3>

    <p>The PLS namespace <em title="MAY in RFC 2119 context" class=
    "RFC2119">MAY</em> be used with other XML namespaces as per the
    Namespaces in XML Recommendations (Namespaces in XML 1.0
    [<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces in XML 1.1
    [<a href="#ref-xmlns-11">XML-NS11</a>]). Future work by W3C is
    expected to address ways to specify conformance for documents
    involving multiple namespaces.</p>

    <h3 id="S3.2.3">3.2.3 Conforming Pronunciation Lexicon
    Specification Processors</h3>

    <p>A <em>Conforming Pronunciation Lexicon Specification
    Processor</em> <em title="MUST in RFC 2119 context" class=
    "RFC2119">MUST</em> be able to parse and process <a href=
    "#S3.2.1">Conforming Pronunciation Lexicon Specification
    documents</a>.</p>

    <p>In a <em>Conforming Pronunciation Lexicon Specification
    Processor</em>, the XML parser <em title=
    "MUST in RFC 2119 context" class="RFC2119">MUST</em> be able to
    parse and process all XML constructs defined by either XML 1.0
    [<a href="#ref-xml-10">XML10</a>] or XML 1.1 [<a href=
    "#ref-xml-11">XML11</a>] and conforms to the corresponding
    Namespaces in XML specification (Namespaces in XML 1.0
    [<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces in XML 1.1
    [<a href="#ref-xmlns-11">XML-NS11</a>]).</p>

    <p>A Conforming Pronunciation Lexicon Specification Processor
    <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
    conform to the XML 1.0 or XML 1.1 requirements for conformant
    non validating processors.</p>

    <p>A Conforming Pronunciation Lexicon Specification Processor
    <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
    correctly understand and apply the semantics of each markup
    element as described by this document.</p>

    <p>A Conforming Pronunciation Lexicon Specification Processor
    <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
    meet the following requirements for handling of natural (human)
    languages:</p>

    <ul>
      <li>A Conforming Pronunciation Lexicon Specification
      Processor is <em title="REQUIRED in RFC 2119 context" class=
      "RFC2119">REQUIRED</em> to parse all legal natural (human)
      language declarations successfully.</li>

      <li>For any Conforming Pronunciation Lexicon Specification
      Document whose <code>xml:lang</code> attribute (on the
      <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element) has
      a value representing a natural (human) language that the
      Pronunciation Lexicon Specification Processor claims to
      support, the Processor is <em title=
      "REQUIRED in RFC 2119 context" class="RFC2119">REQUIRED</em>
      to successfully parse and treat all text encountered as if in
      that language in order to be a Conforming Processor.</li>

      <li>A Conforming Pronunciation Lexicon Specification
      Processor <em title="SHOULD in RFC 2119 context" class=
      "RFC2119">SHOULD</em> inform the hosting environment when it
      parses a Conforming Pronunciation Lexicon Specification
      Document whose <code>xml:lang</code> attribute (on the
      <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element) has
      a value representing a natural (human) language that the
      Processor does not support.</li>
    </ul>

    <p>When a Conforming Pronunciation Lexicon Specification
    Processor encounters elements or attributes that are not
    declared in this specification and such elements or attributes
    occur where it is not forbidden in this specification, the
    processor <em title="MAY in RFC 2119 context" class=
    "RFC2119">MAY</em> choose to:</p>

    <ul>
      <li>ignore the non-standard elements and/or attributes</li>

      <li>or, process the non-standard elements and/or
      attributes</li>

      <li>or, reject the document containing those elements and/or
      attributes</li>
    </ul>

    <p>Except where stated in this document, there is no
    conformance requirement with respect to performance of
    rendering pronunciations as acoustic structures (models,
    waveforms, etc.) for <a href="#term-ASR">ASR</a> and <a href=
    "#term-TTS">TTS</a>.</p>
  </div>

  <h2 id="S4">4. Pronunciation Lexicon Markup Language
  Definition</h2>

  <p>The <a href="#term-Pron-Lexicon">Pronunciation Lexicon</a>
  markup language consists of the following elements and
  attributes:</p>

  <table border="1" summary=
  "Elements and Attributes for Pronunciation Lexicon">
    <thead>
      <tr>
        <th width="25%" scope="col">Elements</th>

        <th width="25%" scope="col">Attributes</th>

        <th width="50%" scope="col">Description</th>
      </tr>
    </thead>

    <tbody>
      <tr>
        <td><a href="#S4.1"><code>&lt;lexicon&gt;</code></a></td>

        <td><code>version</code><br />
        <code>xml:base</code><br />
        <code>xmlns</code><br />
        <code>xml:lang</code><br />
        <code>alphabet</code></td>

        <td>root element for PLS</td>
      </tr>

      <tr>
        <td><a href="#S4.2"><code>&lt;meta&gt;</code></a></td>

        <td><code>name</code><br />
        <code>http-equiv</code><br />
        <code>content</code></td>

        <td>element containing meta data</td>
      </tr>

      <tr>
        <td><a href="#S4.3"><code>&lt;metadata&gt;</code></a></td>

        <td>&nbsp;</td>

        <td>element containing meta data</td>
      </tr>

      <tr>
        <td><a href="#S4.4"><code>&lt;lexeme&gt;</code></a></td>

        <td><code>xml:id</code><br />
        <code>role</code></td>

        <td>the container element for a single lexical entry</td>
      </tr>

      <tr>
        <td><a href="#S4.5"><code>&lt;grapheme&gt;</code></a></td>

        <td>&nbsp;</td>

        <td>contains <a href="#term-Orthography">orthographic
        information</a> for a <a href=
        "#term-Lexeme">lexeme</a></td>
      </tr>

      <tr>
        <td><a href="#S4.6"><code>&lt;phoneme&gt;</code></a></td>

        <td><code>prefer</code><br />
        <code>alphabet</code></td>

        <td>contains pronunciation information for a <a href=
        "#term-Lexeme">lexeme</a></td>
      </tr>

      <tr>
        <td><a href="#S4.7"><code>&lt;alias&gt;</code></a></td>

        <td><code>prefer</code></td>

        <td>contains <a href="#term-acronym-exp">acronym
        expansions</a> and orthographic substitutions</td>
      </tr>

      <tr>
        <td><a href="#S4.8"><code>&lt;example&gt;</code></a></td>

        <td>&nbsp;</td>

        <td>contains an example of the usage for a <a href=
        "#term-Lexeme">lexeme</a></td>
      </tr>
    </tbody>
  </table>

  <h3 id="S4.1">4.1 <code>&lt;lexicon&gt;</code> Element</h3>

  <p>The root element of the Pronunciation Lexicon markup language
  is the <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element.
  This element is the container for all other elements of the PLS
  language. A <a href="#S4.1"><code>&lt;lexicon&gt;</code></a>
  element <em title="MUST in RFC 2119 context" class=
  "RFC2119">MUST</em> contain zero or more <a href=
  "#S4.2"><code>&lt;meta&gt;</code></a> elements, followed by an
  <em title="OPTIONAL in RFC 2119 context" class=
  "RFC2119">OPTIONAL</em> <a href=
  "#S4.3"><code>&lt;metadata&gt;</code></a> element, followed by
  zero or more <a href="#S4.4"><code>&lt;lexeme&gt;</code></a>
  elements. Note that a PLS document without any <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> elements may be useful as
  a placeholder for future lexical entries.</p>

  <p>The <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  specify an <code>alphabet</code> attribute which indicates the
  default pronunciation alphabet to be used within the PLS
  document. The values of the <code>alphabet</code> attribute are
  described in <a href="#S2">Section 2</a>. The default
  pronunciation alphabet <em title="MAY in RFC 2119 context" class=
  "RFC2119">MAY</em> be overridden for a given lexeme using the
  <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element.</p>

  <p>The <em title="REQUIRED in RFC 2119 context" class=
  "RFC2119">REQUIRED</em> <code>version</code> attribute indicates
  the version of the specification to be used for the document and
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  have the value <code>"1.0"</code>.</p>

  <p>The <em title="REQUIRED in RFC 2119 context" class=
  "RFC2119">REQUIRED</em> <code>xml:lang</code> attribute allows
  identification of the language for which the <a href=
  "#term-Pron-Lexicon">pronunciation lexicon</a> is relevant. IETF
  Best Current Practice 47 [<a href="#ref-bcp47" shape=
  "rect">BCP47</a>] is the normative reference on the values of the
  <code>xml:lang</code> attribute.</p>

  <p>Note that <code>xml:lang</code> specifies a single unique
  language for the entire PLS document. This does not limit the
  ability to create multilingual <a href="#term-SRGS">SRGS</a>
  [<a href="#ref-SRGS">SRGS</a>] and <a href="#term-SSML">SSML</a>
  [<a href="#ref-SSML">SSML</a>] documents. These documents may
  reference multiple <a href="#term-Pron-Lexicon">pronunciation
  lexicons</a>, possibly written for different languages.</p>

  <p>The namespace URI for PLS is
  <code>"http://www.w3.org/2005/01/pronunciation-lexicon"</code>.
  All PLS markup <em title="MUST in RFC 2119 context" class=
  "RFC2119">MUST</em> be associated with the PLS namespace, using a
  Namespace Declaration as described in either Namespaces in XML
  1.0 [<a href="#ref-xmlns-10">XML-NS10</a>] or Namespaces in XML
  1.1 [<a href="#ref-xmlns-11">XML-NS11</a>]. This can for instance
  be achieved by declaring an <code>xmlns</code> attribute on the
  <a href="#S4.1"><code>&lt;lexicon&gt;</code></a> element, as the
  examples in this specification show.</p>

  <p>PLS documents <em title="MAY in RFC 2119 context" class=
  "RFC2119">MAY</em> include the <code>xml:base</code> attribute as
  defined in [<a href="#ref-xml-base" shape="rect">XML-BASE</a>].
  Note that as in the HTML 4.01 specification [<a href="#ref-html"
  shape="rect">HTML</a>], this is a URI which all the relative
  references within the document take as their base.</p>

  <h4 id="S4.1.0.0">Informative Note:</h4>

  <p>Note that in this version of the specification, only the
  contents of metadata can potentially use relative URIs.</p>

  <h4 id="S4.1.1">Example:</h4>

  <p>A simple PLS document for the word "tomato" and its
  pronunciation.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;tomato&lt;/grapheme&gt;
    &lt;phoneme&gt;təmei̥ɾou̥&lt;/phoneme&gt;
    &lt;!-- IPA string is: "t&amp;#x0259;mei&amp;#x325;&amp;#x027E;ou&amp;#x325;" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S4.2">4.2 <code>&lt;meta&gt;</code> Element</h3>

  <p>The <a href="#S4.3"><code>&lt;metadata&gt;</code></a> and
  <a href="#S4.2"><code>&lt;meta&gt;</code></a> elements are
  containers in which information about the document can be placed.
  The <a href="#S4.3"><code>&lt;metadata&gt;</code></a> element
  provides more general and powerful treatment of metadata
  information than <a href="#S4.2"><code>&lt;meta&gt;</code></a> by
  using a metadata schema.</p>

  <p>A <a href="#S4.2"><code>&lt;meta&gt;</code></a> element
  associates a string to a declared meta property or declares
  <code>http-equiv</code> content. Either a <code>name</code> or
  <code>http-equiv</code> attribute is <em title=
  "REQUIRED in RFC 2119 context" class="RFC2119">REQUIRED</em>. It
  is an error to provide both <code>name</code> and
  <code>http-equiv</code> attributes. A <code>content</code>
  attribute is also <em title="REQUIRED in RFC 2119 context" class=
  "RFC2119">REQUIRED</em>. The only <a href=
  "#S4.2"><code>&lt;meta&gt;</code></a> property defined by this
  specification is <code>"seeAlso"</code>. It is used to specify a
  resource that might provide additional metadata information about
  the content. This property is modeled on the
  <code>"seeAlso"</code> property from <a href=
  "http://www.w3.org/TR/2004/REC-rdf-schema-20040210/#ch_seealso">Section
  5.4.1</a> of "RDF Vocabulary Description Language 1.0: RDF
  Schema" [<a href="#ref-rdf-schema">RDF-SCHEMA</a>]. The
  <code>http-equiv</code> attribute has a special significance when
  documents are retrieved via HTTP. Although the preferred method
  of providing HTTP header information is to use HTTP header
  fields, the <code>http-equiv</code> content <em title=
  "MAY in RFC 2119 context" class="RFC2119">MAY</em> be used in
  situations where the PLS document author is unable to configure
  HTTP header fields associated with their document on the origin
  server, for example, cache control information. Note that HTTP
  servers and caches are not required to inspect the contents of
  <a href="#S4.2"><code>&lt;meta&gt;</code></a> in PLS documents
  and thereby override the header values they would send
  otherwise.</p>

  <p>The <a href="#S4.2"><code>&lt;meta&gt;</code></a> element is
  an empty element.</p>

  <h4 id="S4.2.0">Informative Note:</h4>

  <p>This section is modeled after the <code>&lt;meta&gt;</code>
  description in the HTML 4.01 Specification [<a href=
  "#ref-html">HTML</a>]. Despite the fact that the name/content
  model is now being replaced by better ways to include metadata,
  see for instance <a href=
  "http://www.w3.org/TR/2004/WD-xhtml2-20040722/mod-meta.html#sec_20.6.">
  Section 20.6</a> of XHTML 2.0 [<a href="#ref-XHTML2">XHTML2</a>],
  and the fact that the <code>http-equiv</code> directive is no
  longer recommended in <a href=
  "http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html#application-xml">
  Section 3.3</a> of XHTML Media Types [<a href=
  "#ref-XHTML-MTYPES">XHTML-MTYPES</a>], the Working Group has
  decided to retain this for compatibility with the other
  specifications of the first version of the Speech Interface
  Framework (VoiceXML, SSML, SRGS, CCXML). Future versions of the
  framework will align with more modern metadata schemes.</p>

  <h4 id="S4.2.1">Example:</h4>

  <p>This is an example of how <a href=
  "#S4.2"><code>&lt;meta&gt;</code></a> elements can be included in
  a PLS document to specify a resource that provides additional
  metadata information.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
    &lt;meta http-equiv="Cache-Control" content="no-cache"/&gt;
    &lt;meta name="seeAlso" content="http://example.com/my-pls-metadata.xml"/&gt;
    &lt;!--  If lexemes are to be added to this lexicon, they start below --&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S4.3">4.3 <code>&lt;metadata&gt;</code> Element</h3>

  <p>The <a href="#S4.3"><code>&lt;metadata&gt;</code></a> element
  is a container in which information about the document can be
  placed using metadata markup. The behavior of software processing
  the content of a <a href=
  "#S4.3"><code>&lt;metadata&gt;</code></a> element is not
  described in this specification. Therefore, software implementing
  this specification is free to ignore that content.</p>

  <p>Although any metadata markup can be used within <a href=
  "#S4.3"><code>&lt;metadata&gt;</code></a>, it is <em title=
  "RECOMMENDED in RFC 2119 context" class=
  "RFC2119">RECOMMENDED</em> that the RDF/XML Syntax [<a href=
  "#ref-rdf-xml">RDF-XMLSYNTAX</a>] be used, in conjunction with
  the general metadata properties defined by the Dublin Core
  Metadata Initiative [<a href="#ref-dc">DC</a>] (e.g., Title,
  Creator, Subject, Description, Rights, etc.)</p>

  <h4 id="S4.3.1">Example:</h4>

  <p>This is an example of how metadata can be included in a PLS
  document using the "Dublin Core Metadata Element Set, Version
  1.1" [<a href="#ref-dc-es">DC-ES</a>] describing general document
  information such as title, description, date, and so on:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;metadata&gt;
    &lt;rdf:RDF
       xmlns:rdf = "http://www.w3.org/1999/02/22-rdf-syntax-ns#"
       xmlns:dc  = "http://purl.org/dc/elements/1.1/"&gt;

     &lt;!-- Metadata about the PLS document --&gt;
     &lt;rdf:Description rdf:about=""
       dc:title="Pronunciation lexicon for W3C terms"
       dc:description="Common pronunciations for many W3C acronyms and abbreviations, i.e. I18N or WAI"
       dc:publisher="W3C"
       dc:date="2005-11-29"
       dc:rights="Copyright 2002 W3C"
       dc:format="application/pls+xml"&gt;
       &lt;dc:creator&gt;The W3C Voice Browser Working Group&lt;/dc:creator&gt;
     &lt;/rdf:Description&gt;
    &lt;/rdf:RDF&gt;
  &lt;/metadata&gt;
  &lt;!--  If lexemes are to be added to this lexicon, they start below --&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S4.4">4.4 <code>&lt;lexeme&gt;</code> Element</h3>

  <p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element is
  a container for a lexical entry which <em title=
  "MAY in RFC 2119 context" class="RFC2119">MAY</em> include
  multiple <a href="#term-Orthography">orthographies</a> and
  multiple pronunciation information.</p>

  <p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
  contains one or more <a href=
  "#S4.5"><code>&lt;grapheme&gt;</code></a> elements, one or more
  pronunciations (either by <a href=
  "#S4.5"><code>&lt;phoneme&gt;</code></a> or <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> elements or a combination
  of both), and zero or more <a href=
  "#S4.8"><code>&lt;example&gt;</code></a> elements. The children
  of the <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
  <em title="MAY in RFC 2119 context" class="RFC2119">MAY</em>
  appear in any order, but note that the order will have an impact
  on the treatment of multiple pronunciations (see <a href=
  "#S4.9">Section 4.9</a>).</p>

  <p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
  has an <em title="OPTIONAL in RFC 2119 context" class=
  "RFC2119">OPTIONAL</em> <code>xml:id</code> [<a href=
  "#ref-xml-id">XML-ID</a>] attribute, allowing the element to be
  referenced from other documents (through fragment identifiers or
  XPointer [<a href="#ref-xpointer">XPOINTER</a>], for instance).
  For example, developers may use external RDF statements [<a href=
  "#ref-rdf-conc">RDF-CONC</a>] to associate metadata (such as part
  of speech or word relationships) with a lexeme.</p>

  <p>The <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
  has an <em title="OPTIONAL in RFC 2119 context" class=
  "RFC2119">OPTIONAL</em> <code>role</code> attribute which takes
  as its value one or more white space separated QNames as defined
  in Section 4 of Namespaces in XML (1.0 [<a href=
  "#ref-xmlns-10">XML-NS10</a>] or 1.1 [<a href=
  "#ref-xmlns-11">XML-NS11</a>], depending on the version of XML
  being used).</p>

  <p>The <code>role</code> attribute describes additional
  information to help the selection of the most relevant
  pronunciation for a given <a href=
  "#term-Orthography">orthography</a>. The main use is to
  differentiate words that have the same spelling but are
  pronounced in different ways (cf. <a href=
  "#term-Homograph">homographs</a> and see also <a href=
  "#S5.5">Section 5.5</a>). A QName in the attribute content of the
  <code>role</code> attribute is expanded into an expanded-name
  using the namespace declarations in scope for the containing
  <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element. Thus,
  each QName provides a reference to a specific item in the
  designated namespace. In the second example below, the QName
  <code>"claws:VVI"</code> within the role attribute expands to the
  <code>"VVI"</code> item in the
  <code>"http://www.example.com/claws7tags"</code> namespace. This
  mechanism allows for referencing defined taxonomies of word
  classes, with the expectation that they are documented at the
  specified namespace URI.</p>

  <h4 id="S4.4.1">Example:</h4>

  <p>A <a href="#term-Pron-Lexicon">pronunciation lexicon</a> for
  the Italian language with two <a href="#term-Lexeme">lexemes</a>.
  One of them is for the loan word "file" which is often used in
  technical discussions to have the same meaning and pronunciation
  as in English. This is distinct from the homograph noun "file"
  which is the plural form of "fila" meaning "queue". Note that
  this user-specified pronunciation for "file" takes precedence
  over any system-defined pronunciation.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="it"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;file&lt;/grapheme&gt;
    &lt;phoneme&gt;faɪl&lt;/phoneme&gt;
    &lt;!-- This is the pronunciation
      of the loan word "file" in Italian.
      IPA string is: "fa&amp;#x026A;l" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;EU&lt;/grapheme&gt;
    &lt;alias&gt;Unione Europea
      &lt;!-- This is a substitution of the European
      Union acronym in Italian language.  --&gt;&lt;/alias&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>The following is an example of a pronunciation lexicon for the
  word "read":</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      xmlns:claws="http://www.example.com/claws7tags" alphabet="ipa"
      xml:lang="en"&gt;
  &lt;lexeme role="claws:VVI claws:VV0 claws:NN1"&gt;
    &lt;!-- verb infinitive, verb present tense, singular noun --&gt;
    &lt;grapheme&gt;read&lt;/grapheme&gt;
    &lt;phoneme&gt;ri&amp;#x02D0;d&lt;!-- same as riːd --&gt;&lt;/phoneme&gt;
  &lt;/lexeme&gt;
  &lt;lexeme role="claws:VVN claws:VVD"&gt;
    &lt;!-- verb past participle, verb past tense --&gt;
    &lt;grapheme&gt;read&lt;/grapheme&gt;
    &lt;phoneme&gt;red&lt;/phoneme&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>Note that the <code>role</code> attribute is based on
  qualified values (in this example from the <a href=
  "http://www.comp.lancs.ac.uk/ucrel/claws7tags.html">UCREL CLAWS7
  tagset</a> of part-of-speech) to distinguish the verb infinitive,
  present tense and singular noun from the verb past tense and past
  participle pronunciation of the word "read".</p>

  <p>The following is an example document which references the
  above lexicon and includes an extension element to show how the
  <code>role</code> attribute may be used to select the relevant
  pronunciation of the word "read" in the dialog.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.0"
      xmlns="http://www.w3.org/2001/10/synthesis"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
        http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
      xmlns:myssml="http://www.example.com/ssml_extensions"
      xmlns:claws="http://www.example.com/claws7tags"
      xml:lang="en"&gt;
  &lt;lexicon uri="http://www.example.com/lexicon.pls"
      type="application/pls+xml"/&gt;
  &lt;voice gender="female" age="3"&gt;
      Can you &lt;myssml:token role="claws:VVI"&gt;read&lt;/myssml:token&gt; this book
      to me?
  &lt;/voice&gt;
  &lt;voice gender="male" age="43"&gt;
      I've already &lt;myssml:token role="claws:VVN"&gt;read&lt;/myssml:token&gt; it
      three times!
  &lt;/voice&gt;
&lt;/speak&gt;
</pre>

  <p>Here is another example in Chinese that uses SSML 1.1
  [<a href="#ref-SSML-11">SSML-11</a>].</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      xmlns:claws="http://www.example.com/claws7tags"
      alphabet="x-myorganization-pinyin"
      xml:lang="zh-CN"&gt;
  &lt;lexeme role="claws:VV0"&gt;
    &lt;!-- base form of lexical verb --&gt;
    &lt;grapheme&gt;处&lt;/grapheme&gt;
    &lt;phoneme&gt;chu3&lt;/phoneme&gt;
    &lt;!-- pinyin string is: "chǔ" in 处罚 处置 --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme role="claws:NN"&gt;
    &lt;!-- common noun, neutral for number --&gt;
    &lt;grapheme&gt;处&lt;/grapheme&gt;
    &lt;phoneme&gt;chu4&lt;/phoneme&gt;
    &lt;!-- pinyin string is: "chù" in 处所 妙处 --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>This is a sample document which references the above lexicon
  and shows how the <code>role</code> attribute may be used to
  select the relevant pronunciation of the Chinese word "处" in the
  dialog.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;speak version="1.1"
      xmlns="http://www.w3.org/2001/10/synthesis"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2001/10/synthesis
        http://www.w3.org/TR/speech-synthesis/synthesis.xsd"
      xmlns:claws="http://www.example.com/claws7tags"
      xml:lang="zh-CN"&gt;
  &lt;lexicon uri="http://www.example.com/lexicon.pls"
      type="application/pls+xml"
      xml:id="mylex"/&gt;
  &lt;lookup ref="mylex"&gt;
    他这个人很不好相&lt;w role="claws:VV0"&gt;处&lt;/w&gt;。
    此&lt;w role="claws:NN"&gt;处&lt;/w&gt;不准照相。
  &lt;/lookup&gt;
&lt;/speak&gt;
</pre>

  <h4 id="S4.4.2">Informative Note:</h4>

  <p>The SRGS 1.0 [<a href="#ref-SRGS">SRGS</a>] and SSML 1.0
  [<a href="#ref-SSML">SSML</a>] specifications do not currently
  support a selection mechanism based on the <code>role</code>
  attribute. Future versions of these specifications are expected
  to allow the selection of relevant pronunciations on the basis of
  the <code>role</code> attribute.</p>

  <h3 id="S4.5">4.5 <code>&lt;grapheme&gt;</code> Element</h3>

  <p>A <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> contains at
  least one <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
  element. The <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
  element contains text describing the <a href=
  "#term-Orthography">orthography</a> of the <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a>.</p>

  <p>The <a href="#S4.5"><code>&lt;grapheme&gt;</code></a> element
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  contain 'character' child information items. The <a href=
  "#S4.5"><code>&lt;grapheme&gt;</code></a> element <em title=
  "MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
  contain 'element' child information items from any namespace,
  i.e. PLS or foreign namespace.</p>

  <p>In more complex situations there may be alternative textual
  representations for the same word or phrase; this can arise due
  to a number of reasons, for example:</p>

  <ul class="noindent">
    <li>Regional spelling variations e.g. "colour" and
    "color";</li>

    <li>Free spelling variations e.g. "judgment" and
    "judgement"</li>

    <li>Alternate writing systems, e.g. Japanese uses a mixture of
    Han ideographs (Kanji), and phonemic spelling systems (Katakana
    or Hiragana) for representing the <a href=
    "#term-Orthography">orthography</a> of a word or phrase, and
    such mixture sometimes has several variations as in kana
    suffixes following kanji stems (Okurigana) for example "okonau"
    (行なう vs. 行う);</li>

    <li>Reformed spellings e.g. in German some of the words which
    used to have "ß" before are now to be written with "ss".</li>
  </ul>

  <p>In order to remove the need for duplication of pronunciation
  information to cope with the above variations, the <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> element <em title=
  "MAY in RFC 2119 context" class="RFC2119">MAY</em> contain more
  than one <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
  element to define the base <a href=
  "#term-Orthography">orthography</a> and any variants. Note that
  all the pronunciations given within the <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> apply to each and every
  <a href="#S4.5"><code>&lt;grapheme&gt;</code></a> within the
  <a href="#S4.4"><code>&lt;lexeme&gt;</code></a>.</p>

  <h4 id="S4.5.1">Examples:</h4>

  <p>An example of a single <a href="#term-grapheme">grapheme</a>
  and a single pronunciation.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Sepulveda&lt;/grapheme&gt;
    &lt;phoneme&gt;səˈpʌlvɪdə&lt;/phoneme&gt;
    &lt;!-- IPA string is: "s&amp;#x0259;&amp;#x02C8;p&amp;#x028C;lv&amp;#x026A;d&amp;#x0259;" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>Another example with more than one written form for a lexical
  entry, where the first <a href=
  "#term-Orthography">orthography</a> uses Latin characters for
  "Romaji" <a href="#term-Orthography">orthography</a>, the second
  one uses "Kanji" <a href="#term-Orthography">orthography</a> and
  the third one uses the "Hiragana" <a href=
  "#term-Orthography">orthography</a>:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="jp"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;nihongo&lt;!-- "Romaji" --&gt;&lt;/grapheme&gt;
    &lt;grapheme&gt;日本語&lt;!-- "Kanji" --&gt;&lt;/grapheme&gt;
    &lt;grapheme&gt;にほんご&lt;!-- "Hiragana" --&gt;&lt;/grapheme&gt;
    &lt;phoneme&gt;ɲihoŋo
      &lt;!-- IPA string is: "&amp;#x0272;iho&amp;#x014B;o" --&gt;&lt;/phoneme&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S4.6">4.6 <code>&lt;phoneme&gt;</code> Element</h3>

  <p>A <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> <em title=
  "MAY in RFC 2119 context" class="RFC2119">MAY</em> contain one or
  more <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements.
  The <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element
  contains text describing how the <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> is pronounced.</p>

  <p>The <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  contain 'character' child information items. The <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> element <em title=
  "MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
  contain 'element' child information items from any namespace,
  i.e. PLS or foreign namespace.</p>

  <p>A <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element
  <em title="MAY in RFC 2119 context" class="RFC2119">MAY</em> have
  an <code>alphabet</code> attribute, which indicates the
  pronunciation alphabet that is used for this <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> element only. See
  <a href="#S4.1">Section 4.1</a> for the default pronunciation
  alphabet. The legal values for the <code>alphabet</code>
  attribute are described in <a href="#S2">Section 2</a>.</p>

  <p>The <code>prefer</code> is an <em title=
  "OPTIONAL in RFC 2119 context" class="RFC2119">OPTIONAL</em>
  attribute, which indicates the pronunciation that <em title=
  "MUST in RFC 2119 context" class="RFC2119">MUST</em> be used by a
  <a href="#term-TTS">speech synthesis engine</a> when it is set to
  <code>"true"</code>. See <a href="#S4.9">Section 4.9</a> for
  required behavior when multiple pronunciations have
  <code>prefer</code> set to <code>"true"</code>. The possible
  values are: <code>"true"</code> or <code>"false"</code>. The
  default value is <code>"false"</code>.</p>

  <p>The prefer mechanism spans both the <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> and <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> elements. <a href=
  "#S4.9">Section 4.9</a> describes how multiple pronunciations are
  specified in PLS for <a href="#term-ASR">ASR</a> and <a href=
  "#term-TTS">TTS</a>, and gives many examples in <a href=
  "#S4.9.3">Section 4.9.3</a>.</p>

  <h4 id="S4.6.1">Examples:</h4>

  <p>More than one pronunciation per lexical entry:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;huge&lt;/grapheme&gt;
    &lt;phoneme prefer="true"&gt;hjuːdʒ&lt;/phoneme&gt;
    &lt;!-- IPA string is: "hju&amp;#x02D0;d&amp;#x0292;" --&gt;
    &lt;phoneme&gt;juːdʒ&lt;/phoneme&gt;
    &lt;!-- IPA string is: "ju&amp;#x02D0;d&amp;#x0292;" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>More than one written form and more than one
  pronunciation:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;theater&lt;/grapheme&gt;
    &lt;grapheme&gt;theatre&lt;/grapheme&gt;
    &lt;phoneme prefer="true"&gt;ˈθɪətər&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;&amp;#x03B8;&amp;#x026A;&amp;#x0259;t&amp;#x0259;r" --&gt;
    &lt;phoneme&gt;ˈθiːjətər&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;&amp;#x03B8;i&amp;#x02D0;j&amp;#x0259;t&amp;#x0259;r" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>An example of a <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> that changes the
  pronunciation alphabet to a proprietary one.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;color&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈkʌlər&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;k&amp;#x028C;l&amp;#x0259;r" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;XYZ&lt;/grapheme&gt;
    &lt;phoneme alphabet="x-example-alphabet"&gt;XYZ&lt;/phoneme&gt;
    &lt;!-- The above pronunciation is given in a proprietary alphabet
      called: "x-example-alphabet" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S4.7">4.7 <code>&lt;alias&gt;</code> Element</h3>

  <p>A <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element
  <em title="MAY in RFC 2119 context" class="RFC2119">MAY</em>
  contain one or more <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> elements which are used to
  indicate the pronunciation of an <a href=
  "#term-acronym">acronym</a> or an abbreviated term, in terms of
  other <a href="#term-Orthography">orthographies</a>, or other
  substitutions as necessary; see examples below and in <a href=
  "#S4.9.3">Section 4.9.3</a>.</p>

  <p>The <a href="#S4.7"><code>&lt;alias&gt;</code></a> element
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  contain 'character' child information items. The <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> element <em title=
  "MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
  contain 'element' child information items from any namespace,
  i.e. PLS or foreign namespace.</p>

  <p>In a <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element,
  both <a href="#S4.7"><code>&lt;alias&gt;</code></a> elements and
  <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements
  <em title="MAY in RFC 2119 context" class="RFC2119">MAY</em> be
  present. If authors want explicit control over the pronunciation,
  they can use the <a href="#S4.6"><code>&lt;phoneme&gt;</code></a>
  element instead of the <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> element.</p>

  <p>The <a href="#S4.7"><code>&lt;alias&gt;</code></a> element has
  an <em title="OPTIONAL in RFC 2119 context" class=
  "RFC2119">OPTIONAL</em> <code>prefer</code> attribute analogous
  to the <code>prefer</code> attribute for the <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> element; see <a href=
  "#S4.6">Section 4.6</a> for a normative description of the
  <code>prefer</code> attribute.</p>

  <p>Pronunciations of <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> element contents
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em> be
  generated by the processor using pronunciations described by the
  <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element of any
  constituent graphemes in the PLS document and without invoking
  recursive access to the PLS document on the <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> elements of any
  constituent graphemes. The processor <em title=
  "SHOULD in RFC 2119 context" class="RFC2119">SHOULD</em>
  determine the pronunciations of the remaining <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> element contents by the
  same process that it determines the pronunciation of
  out-of-lexicon graphemes.</p>

  <h4 id="S4.7.1">Examples:</h4>

  <p><a href="#term-acronym-exp">Acronym expansion</a> using the
  <a href="#S4.7"><code>&lt;alias&gt;</code></a> element:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;W3C&lt;/grapheme&gt;
    &lt;alias&gt;World Wide Web Consortium&lt;/alias&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>The following example illustrates a combination of <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> and <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> elements. The indicated
  acronym, "GNU", has only two pronunciations. Note that the
  pronunciation described by the <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> element of "Unix" is not
  used as part of the pronunciation of the <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> element contents of "GNU"
  as recursion of <a href="#S4.7"><code>&lt;alias&gt;</code></a> is
  not permissible. The pronunciations described by the <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> elements of "GNU" and
  "Unix" are used by the processor to generate the pronunciation of
  "GNU is Not Unix".</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;GNU&lt;/grapheme&gt;
    &lt;alias&gt;&lt;!-- be careful about recursion here --&gt;GNU is Not Unix&lt;/alias&gt;
    &lt;phoneme&gt;gəˈnuː&lt;/phoneme&gt;
    &lt;!-- IPA string is: "g&amp;#x0259;&amp;#x02C8;nu&amp;#x02D0;" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Unix&lt;/grapheme&gt;
    &lt;grapheme&gt;UNIX&lt;/grapheme&gt;
    &lt;alias&gt;a multiplexed information and computing service&lt;/alias&gt;
    &lt;phoneme&gt;ˈjuːnɪks&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;ju&amp;#x02D0;n&amp;#x026A;ks" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S4.8">4.8 <code>&lt;example&gt;</code> Element</h3>

  <p>The <a href="#S4.8"><code>&lt;example&gt;</code></a> element
  includes an example sentence that illustrates an occurrence of
  this <a href="#term-Lexeme">lexeme</a>. Because the examples are
  explicitly marked, automated tools can be used for regression
  testing and for generation of <a href=
  "#term-Pron-Lexicon">pronunciation lexicon</a> documentation.</p>

  <p>The <a href="#S4.8"><code>&lt;example&gt;</code></a> element
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  contain 'character' child information items. The <a href=
  "#S4.8"><code>&lt;example&gt;</code></a> element <em title=
  "MUST NOT in RFC 2119 context" class="RFC2119">MUST NOT</em>
  contain 'element' child information items from any namespace,
  i.e. PLS or foreign namespace.</p>

  <p>Zero, one or many <a href=
  "#S4.8"><code>&lt;example&gt;</code></a> elements <em title=
  "MAY in RFC 2119 context" class="RFC2119">MAY</em> be provided
  for a single <a href="#S4.4"><code>&lt;lexeme&gt;</code></a>
  element.</p>

  <h4 id="S4.8.1">Example:</h4>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;phoneme&gt;led&lt;/phoneme&gt;
    &lt;example&gt;My feet were as heavy as lead.&lt;!-- possible comment --&gt;&lt;/example&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;phoneme&gt;liːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
    &lt;example&gt;The guide once again took the lead.&lt;/example&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S4.9">4.9 Multiple Pronunciations for ASR and TTS</h3>

  <p>This section describes the treatment of multiple
  pronunciations specified in a PLS document for <a href=
  "#term-ASR">ASR</a> and <a href="#term-TTS">TTS</a>.</p>

  <h3 id="S4.9.1">4.9.1 Multiple Pronunciations for ASR</h3>

  <p>If more than one pronunciation for a given <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> is specified (either by
  <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements or
  <a href="#S4.7"><code>&lt;alias&gt;</code></a> elements or a
  combination of both), an <a href="#term-ASR">ASR</a> processor
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  consider each of them as valid pronunciations for the <a href=
  "#S4.5"><code>&lt;grapheme&gt;</code></a>. See <a href=
  "#S4.9.3ex2">Example 2</a> and following examples in <a href=
  "#S4.9.3">Section 4.9.3</a>.</p>

  <p>If more than one <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> contains the same
  <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>, all relevant
  pronunciations (see discussion in <a href="#S4.4">Section 4.4</a>
  regarding the selection of relevant pronunciations using the
  <code>role</code> attribute) will be collected in document order
  and an <a href="#term-ASR">ASR</a> processor <em title=
  "MUST in RFC 2119 context" class="RFC2119">MUST</em> consider all
  of them as valid pronunciations for the <a href=
  "#S4.5"><code>&lt;grapheme&gt;</code></a>. See <a href=
  "#S4.9.3ex7">Example 7</a> and <a href="#S4.9.3ex8">Example 8</a>
  in <a href="#S4.9.3">Section 4.9.3</a>.</p>

  <h3 id="S4.9.2">4.9.2 Multiple Pronunciations for TTS</h3>

  <p>If more than one pronunciation for a given <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> is specified (either by
  <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> elements or
  <a href="#S4.7"><code>&lt;alias&gt;</code></a> elements or a
  combination of both), a <a href="#term-TTS">TTS</a> processor
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  use the first one in document order that has the
  <code>prefer</code> attribute set to <code>"true"</code>. If none
  of the pronunciations has <code>prefer</code> set to
  <code>"true"</code>, the <a href="#term-TTS">TTS</a> processor
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  use the first one in document order unless the <a href=
  "#term-TTS">TTS</a> processor is documented as having a method of
  selecting pronunciations, in which case the processor <em title=
  "MUST in RFC 2119 context" class="RFC2119">MUST</em> use any one
  of the pronunciations. See <a href="#S4.9.3ex2">Example 2</a> and
  following examples in <a href="#S4.9.3">Section 4.9.3</a>.</p>

  <p>If more than one <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> contains the same
  <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>, all relevant
  pronunciations (see discussion in <a href="#S4.4">Section 4.4</a>
  regarding the selection of relevant pronunciations using the
  <code>role</code> attribute) will be collected in document order
  and a <a href="#term-TTS">TTS</a> processor <em title=
  "MUST in RFC 2119 context" class="RFC2119">MUST</em> use the
  first one in document order that has the <code>prefer</code>
  attribute set to <code>"true"</code>. If none of the relevant
  pronunciations has <code>prefer</code> set to
  <code>"true"</code>, the <a href="#term-TTS">TTS</a> processor
  <em title="MUST in RFC 2119 context" class="RFC2119">MUST</em>
  use the first one in document order unless the <a href=
  "#term-TTS">TTS</a> processor is documented as having a method of
  selecting pronunciations, in which case the processor <em title=
  "MUST in RFC 2119 context" class="RFC2119">MUST</em> use any one
  of the relevant pronunciations. See <a href="#S4.9.3ex7">Example
  7</a> and <a href="#S4.9.3ex8">Example 8</a> in <a href=
  "#S4.9.3">Section 4.9.3</a>.</p>

  <p>Note that a <a href="#term-TTS">TTS</a> processor may have
  language-dependent internal mechanisms enabling it to
  automatically choose between multiple pronunciations. See
  <a href="#S4.9.3ex9">Example 9</a> in <a href="#S4.9.3">Section
  4.9.3</a>.</p>

  <h3 id="S4.9.3">4.9.3 Examples of Multiple Pronunciations</h3>

  <p><i>This section is informative.</i></p>

  <p>The following examples are designed to describe and illustrate
  the most common examples of multiple pronunciations. Both
  <a href="#term-ASR">ASR</a> and <a href="#term-TTS">TTS</a>
  behavior is described.</p>

  <h4 id="S4.9.3ex1">Example 1:</h4>

  <p>In the following example, there is only one pronunciation. It
  will be used by both <a href="#term-ASR">ASR</a> and <a href=
  "#term-TTS">TTS</a> processors.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;bead&lt;/grapheme&gt;
    &lt;phoneme&gt;biːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "bi&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex2">Example 2:</h4>

  <p>In the following example, there are two pronunciations. An
  <a href="#term-ASR">ASR</a> processor will recognize both
  pronunciations, whereas a <a href="#term-TTS">TTS</a> processor
  will only use one. Since none of the pronunciations has
  <code>prefer</code> set to <code>"true"</code>, unless the
  processor is documented to have a different strategy, it will use
  the first of the pronunciations because it is first in document
  order.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;read&lt;/grapheme&gt;
    &lt;phoneme&gt;red&lt;/phoneme&gt;
    &lt;phoneme&gt;riːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "ri&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex3">Example 3:</h4>

  <p>In the following example, there are two pronunciations. An
  <a href="#term-ASR">ASR</a> processor will recognize both
  pronunciations, whereas a <a href="#term-TTS">TTS</a> processor
  will only use the second one (because it has <code>prefer</code>
  set to <code>"true"</code>).</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;phoneme&gt;led&lt;/phoneme&gt;
    &lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex4">Example 4:</h4>

  <p>In the following example, "read" has two pronunciations. The
  first one is specified by means of an alias to "red", which is
  defined just below it. An <a href="#term-ASR">ASR</a> processor
  will recognize both pronunciations, whereas a <a href=
  "#term-TTS">TTS</a> processor will only use one. Since none of
  the pronunciations has <code>prefer</code> set to
  <code>"true"</code>, unless the processor is documented to have a
  different strategy, it will use the first of the pronunciations
  because it is first in document order. In this example, the alias
  refers to a lexeme later in the lexicon, but in general, this
  order is not relevant.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;read&lt;/grapheme&gt;
    &lt;alias&gt;red&lt;/alias&gt;
    &lt;phoneme&gt;riːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "ri&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;red&lt;/grapheme&gt;
    &lt;phoneme&gt;red&lt;/phoneme&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex5">Example 5:</h4>

  <p>In the following example, there are two pronunciations for
  "lead". Both are given with <code>prefer</code> set to
  <code>"true"</code>. An <a href="#term-ASR">ASR</a> processor
  will recognize both pronunciations, whereas a <a href=
  "#term-TTS">TTS</a> processor will only use the first one
  (because it is first in document order that has
  <code>prefer</code> set to <code>"true"</code>).</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;alias prefer="true"&gt;led&lt;/alias&gt;
    &lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;led&lt;/grapheme&gt;
    &lt;phoneme&gt;led&lt;/phoneme&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex6">Example 6:</h4>

  <p>In the following example, there are two pronunciations for
  "lead". <a href="#term-ASR">ASR</a> processor will recognize both
  pronunciations, whereas a <a href="#term-TTS">TTS</a> processor
  will only use the second one (because it has <code>prefer</code>
  set to <code>"true"</code>). Note that the alias entry for "lead"
  as "led" does not inherit the preference of the pronunciation of
  the alias.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;alias&gt;led&lt;/alias&gt;
    &lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;led&lt;/grapheme&gt;
    &lt;phoneme prefer="true"&gt;led&lt;/phoneme&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex7">Example 7:</h4>

  <p>In the following example, "lead" has two different entries in
  the lexicon. An <a href="#term-ASR">ASR</a> processor will
  recognize both pronunciations given here, but a <a href=
  "#term-TTS">TTS</a> processor will only recognize one. Since none
  of the pronunciations has <code>prefer</code> set to
  <code>"true"</code>, unless the processor is documented to have a
  different strategy, it will use the "led" pronunciation because
  it is first in document order.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;phoneme&gt;led&lt;/phoneme&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;phoneme&gt;liːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex8">Example 8:</h4>

  <p>In the following example, there are two pronunciations in each
  of two different lexeme entries in the same lexicon document. An
  <a href="#term-ASR">ASR</a> processor will recognize both
  pronunciations given here, but a <a href="#term-TTS">TTS</a>
  processor will only recognize the "liːd" pronunciation, because
  it is the first one in document order that has
  <code>prefer</code> set to <code>"true"</code>.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;alias&gt;led&lt;/alias&gt;
    &lt;phoneme prefer="true"&gt;liːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;lead&lt;/grapheme&gt;
    &lt;phoneme prefer="true"&gt;led&lt;/phoneme&gt;
    &lt;phoneme&gt;liːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "li&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h4 id="S4.9.3ex9">Example 9:</h4>

  <p>In the following example in French, "1" has three
  pronunciations. The latter two pronunciations are specified by
  means of an alias to "une", which is defined just below it. An
  <a href="#term-ASR">ASR</a> processor will recognize all three
  pronunciations given here, but a <a href="#term-TTS">TTS</a>
  processor will only recognize the "un" pronunciation, unless
  otherwise documented by the processor. A <a href=
  "#term-TTS">TTS</a> processor documented capable of automatically
  choosing between multiple pronunciations will select either the
  "un" or "une" alias (given a grammatical context). If it selects
  the "une" alias then the "yn" pronunciation will be used because
  it has <code>prefer</code> set to <code>"true"</code>.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="fr"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;1&lt;/grapheme&gt;
    &lt;alias&gt;un&lt;/alias&gt;
    &lt;alias&gt;une&lt;/alias&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;une&lt;/grapheme&gt;
    &lt;phoneme prefer="true"&gt;yn&lt;/phoneme&gt;
    &lt;phoneme&gt;ynə&lt;/phoneme&gt;
    &lt;!-- IPA string is: "yn&amp;#x0259;" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h2 id="S5">5. Examples</h2>

  <p><i>This section is informative.</i></p>

  <h3 id="S5.1">5.1 Simple Case</h3>

  <p>In its simplest form the <a href=
  "#term-Pron-Lexicon">Pronunciation Lexicon</a> language allows
  <a href="#term-Orthography">orthographies</a> (the textual
  representation) to be associated with pronunciations (the
  <a href="#term-Phoneme">phonetic/phonemic representation</a>). A
  <a href="#term-Pron-Lexicon">Pronunciation Lexicon</a> document
  typically contains multiple entries. So, for example, to specify
  the pronunciation for proper names, such as "Newton" and
  "Scahill", the markup will look like the following.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-GB"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Newton&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈnjuːtən&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;nju&amp;#x02D0;t&amp;#x0259;n" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Scahill&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈskɑhɪl&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;sk&amp;#x0251;h&amp;#x026A;l" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>Here we see the root element <a href=
  "#S4.1"><code>&lt;lexicon&gt;</code></a> which contains the two
  <a href="#term-Lexeme">lexemes</a> for the words "Newton" and
  "Scahill". Each <a href="#S4.4"><code>&lt;lexeme&gt;</code></a>
  is a composite element consisting of the <a href=
  "#term-Orthography">orthographic</a> and pronunciation
  representations for the entry. For each of the two <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> elements there is a
  single <a href="#S4.5"><code>&lt;grapheme&gt;</code></a> element
  which includes the <a href="#term-Orthography">orthographic</a>
  text and a single <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> element which includes
  the pronunciation. In this case the <code>alphabet</code>
  attribute of the <a href="#S4.1"><code>&lt;lexicon&gt;</code></a>
  element is set to <code>"ipa"</code>, so the <a href=
  "#term-IPA">International Phonetic Alphabet</a> [<a href=
  "#ref-ipa">IPA</a>] is being used for all the pronunciations.</p>

  <h3 id="S5.2">5.2 Multiple pronunciations for the same
  orthography</h3>

  <p>For <a href="#term-ASR">ASR</a> systems it is common to rely
  on multiple pronunciations of the same word or phrase in order to
  cope with variations of pronunciation within a language. In the
  <a href="#term-Lexicon">Pronunciation Lexicon</a> language,
  multiple pronunciations are represented by more than one <a href=
  "#S4.6"><code>&lt;phoneme&gt;</code></a> (or <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a>) element within the same
  <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element.</p>

  <p>In the following example the word "Newton" has two possible
  pronunciations.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-GB"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Newton&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈnjuːtən&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;nju&amp;#x02D0;t&amp;#x0259;n" --&gt;
    &lt;phoneme&gt;ˈnuːtən&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;nu&amp;#x02D0;t&amp;#x0259;n" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>In the situation where only a single pronunciation needs to be
  selected among multiple pronunciations that are available (for
  example where a <a href="#term-Pron-Lexicon">pronunciation
  lexicon</a> is being used by a <a href="#term-TTS">speech
  synthesis</a> system), then the <code>prefer</code> attribute on
  the <a href="#S4.6"><code>&lt;phoneme&gt;</code></a> element may
  be used to indicate the preferred pronunciation.</p>

  <h3 id="S5.3">5.3 Multiple orthographies</h3>

  <p>In some situations there are alternative textual
  representations for the same word or phrase. This can arise due
  to a number of reasons. See <a href="#S4.5">Section 4.5</a> for
  details. Because these are representations that have the same
  meaning (as opposed to <a href="#term-Homophone">homophones</a>),
  it is recommended that they be represented using a single
  <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element that
  contains multiple graphemes.</p>

  <p>Here are two simple examples of multiple <a href=
  "#term-Orthography">orthographies</a>: alternative spelling of an
  English word and multiple writings of a Japanese word.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;!-- English entry showing how alternative spellings are handled --&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;colour&lt;/grapheme&gt;
    &lt;grapheme&gt;color&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈkʌlər&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;k&amp;#x028C;l&amp;#x0259;r" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="jp"&gt;
  &lt;!-- Japanese entry showing how multiple writing systems are handled
          romaji, kanji and hiragana orthographies --&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;nihongo&lt;/grapheme&gt;
    &lt;grapheme&gt;日本語&lt;/grapheme&gt;
    &lt;grapheme&gt;にほんご&lt;/grapheme&gt;
    &lt;phoneme&gt;ɲihoŋo&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x0272;iho&amp;#x014B;o" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>In some cases the pronunciations may overlap rather than being
  exactly the same. For example the English names "Smyth" and
  "Smith" share one pronunciation, but "Smyth" has a pronunciation
  that is only relevant to itself. Hence this needs to be
  represented using multiple <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> elements.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Smyth&lt;/grapheme&gt;
    &lt;grapheme&gt;Smith&lt;/grapheme&gt;
    &lt;phoneme&gt;smɪθ/phoneme&gt;
    &lt;!-- IPA string is: "sm&amp;#x026A;&amp;#x03B8;" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Smyth&lt;/grapheme&gt;
    &lt;phoneme&gt;smaɪð&lt;/phoneme&gt;
    &lt;!-- IPA string is: "sma&amp;#x026A;&amp;#x00F0;" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S5.4">5.4 Homophones</h3>

  <p>Most languages have <a href="#term-Homophone">homophones</a>,
  words with the same pronunciation but different meanings (and
  possibly different spellings), for instance "seed" and "cede". It
  is recommended that these be represented as different
  lexemes.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;cede&lt;/grapheme&gt;
    &lt;phoneme&gt;siːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "si&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;seed&lt;/grapheme&gt;
    &lt;phoneme&gt;siːd&lt;/phoneme&gt;
    &lt;!-- IPA string is: "si&amp;#x02D0;d" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S5.5">5.5 Homographs</h3>

  <p>Most languages have words with different meanings but the same
  spelling (and sometimes different pronunciations), called
  <a href="#term-Homograph">homographs</a>. For example, in English
  the word <em>bass</em> (fish) and the word <em>bass</em> (in
  music) have identical spellings but different meanings and
  pronunciations. Although it is recommended that these words be
  represented using separate <a href=
  "#S4.4"><code>&lt;lexeme&gt;</code></a> elements that are
  distinguished by different values of the <code>role</code>
  attribute (see <a href="#S4.4">Section 4.4</a>), if a <a href=
  "#term-Pron-Lexicon">pronunciation lexicon</a> author does not
  want to distinguish between the two words they could simply be
  represented as alternative pronunciations within the same
  <a href="#S4.4"><code>&lt;lexeme&gt;</code></a> element. In the
  latter case the <a href="#term-TTS">TTS</a> processor will not be
  able to distinguish when to apply the first or the second
  transcription.</p>

  <p>In this example the pronunciations of the <a href=
  "#term-Homograph">homograph</a> "bass" are shown.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;bass&lt;/grapheme&gt;
    &lt;phoneme&gt;bæs&lt;/phoneme&gt;
    &lt;!-- IPA string is: b&amp;#x00E6;s --&gt;
    &lt;phoneme&gt;beɪs&lt;/phoneme&gt;
    &lt;!-- IPA string is: be&amp;#x026A;s --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>Note that English contains numerous examples of noun-verb
  pairs that can be treated either as <a href=
  "#term-Homograph">homographs</a> or as alternative
  pronunciations, depending on author preference. Two examples are
  the noun/verb "refuse" and the noun/verb "address".</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      xmlns:mypos="http://www.example.com/my_pos_namespace"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme role="mypos:verb"&gt;
    &lt;grapheme&gt;refuse&lt;/grapheme&gt;
    &lt;phoneme&gt;rɪˈfjuːz&lt;/phoneme&gt;
    &lt;!-- IPA string is: "r&amp;#x026A;&amp;#x02C8;fju&amp;#x02D0;z" --&gt;
  &lt;/lexeme&gt;
  &lt;lexeme role="mypos:noun"&gt;
    &lt;grapheme&gt;refuse&lt;/grapheme&gt;
    &lt;phoneme&gt;ˈrefjuːs&lt;/phoneme&gt;
    &lt;!-- IPA string is: "&amp;#x02C8;refju&amp;#x02D0;s" --&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h3 id="S5.6">5.6 Pronunciation by Orthography (Acronyms,
  Abbreviations, etc.)</h3>

  <p>For some words and phrases pronunciation can be expressed
  quickly and conveniently as a sequence of other <a href=
  "#term-Orthography">orthographies</a>. The developer is not
  required to have linguistic knowledge, but instead makes use of
  the pronunciations that are already expected to be available. To
  express pronunciations using other <a href=
  "#term-Orthography">orthographies</a> the <a href=
  "#S4.7"><code>&lt;alias&gt;</code></a> element may be used.</p>

  <p>This feature may be very useful to deal with <a href=
  "#term-acronym-exp">acronym expansion</a>.</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;!--
        Acronym expansion
  --&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;W3C&lt;/grapheme&gt;
    &lt;alias&gt;World Wide Web Consortium&lt;/alias&gt;
  &lt;/lexeme&gt;
  &lt;!--
        number representation
  --&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;101&lt;/grapheme&gt;
    &lt;alias&gt;one hundred and one&lt;/alias&gt;
  &lt;/lexeme&gt;
  &lt;!--
        crude pronunciation mechanism
  --&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;Thailand&lt;/grapheme&gt;
    &lt;alias&gt;tie land&lt;/alias&gt;
  &lt;/lexeme&gt;
  &lt;!--
        crude pronunciation mechanism and acronym expansion
  --&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;BBC 1&lt;/grapheme&gt;
    &lt;alias&gt;be be sea one&lt;/alias&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <h2 id="S6">6. References</h2>

  <h3 id="S6.1">6.1 Normative References</h3>

  <dl>
    <dt><a id="ref-bcp47" name="ref-bcp47">[BCP47]</a></dt>

    <dd><cite><a href="http://www.rfc-editor.org/rfc/bcp/bcp47.txt"
    shape="rect">Tags for the Identification of
    Languages</a></cite>, A. Phillips and M. Davis, Editors. IETF,
    September 2006. This RFC is available at <a href=
    "http://www.rfc-editor.org/rfc/bcp/bcp47.txt" shape=
    "rect">http://www.rfc-editor.org/rfc/bcp/bcp47.txt</a>.</dd>

    <dt><a id="ref-ipahndbk" name=
    "ref-ipahndbk">[IPAHNDBK]</a></dt>

    <dd><cite><a href=
    "http://www.arts.gla.ac.uk/IPA/handbook.html">Handbook of the
    International Phonetic Association</a></cite>, International
    Phonetic Association, Editors. Cambridge University Press, July
    1999. Information on the Handbook is available at
    <a href="http://www.arts.gla.ac.uk/IPA/handbook.html">
    http://www.arts.gla.ac.uk/IPA/handbook.html</a>.</dd>

    <dt><a id="ref-rfc2119" name="ref-rfc2119" shape=
    "rect">[RFC2119]</a></dt>

    <dd><cite><a href="http://www.ietf.org/rfc/rfc2119.txt" shape=
    "rect">Key words for use in RFCs to Indicate Requirement
    Levels</a></cite>, S. Bradner, Editor. IETF, March 1997. This
    RFC is available at <a href=
    "http://www.ietf.org/rfc/rfc2119.txt" shape=
    "rect">http://www.ietf.org/rfc/rfc2119.txt</a>.</dd>

    <dt><a id="ref-rfc3986" name="ref-rfc3986" shape=
    "rect">[RFC3986]</a></dt>

    <dd><cite><a href="http://www.ietf.org/rfc/rfc3986.txt" shape=
    "rect">Uniform Resource Identifier (URI): Generic
    Syntax</a></cite>, T. Berners-Lee et al., Editors. IETF,
    January 2005. This RFC is available at <a href=
    "http://www.ietf.org/rfc/rfc3986.txt" shape=
    "rect">http://www.ietf.org/rfc/rfc3986.txt</a>.</dd>

    <dt><a id="ref-rfc3987" name="ref-rfc3987" shape=
    "rect">[RFC3987]</a></dt>

    <dd><cite><a href="http://www.ietf.org/rfc/rfc3987.txt" shape=
    "rect">Internationalized Resource Identifiers
    (IRIs)</a></cite>, M. Duerst and M. Suignard, Editors. IETF,
    January 2005. This RFC is available at <a href=
    "http://www.ietf.org/rfc/rfc3987.txt" shape=
    "rect">http://www.ietf.org/rfc/rfc3987.txt</a>.</dd>

    <dt><a id="ref-rfc4267" name="ref-rfc4267" shape=
    "rect">[RFC4267]</a></dt>

    <dd><cite><a href="http://www.ietf.org/rfc/rfc4267.txt" shape=
    "rect">The W3C Speech Interface Framework Media Types:
    application/voicexml+xml, application/ssml+xml,
    application/srgs, application/srgs+xml, application/ccxml+xml,
    and application/pls+xml</a></cite>, Max Froumentin, Editor.
    IETF, November 2005. This RFC is available at <a href=
    "http://www.ietf.org/rfc/rfc4267.txt" shape=
    "rect">http://www.ietf.org/rfc/rfc4267.txt</a>.</dd>

    <dt><a id="ref-SRGS" name="ref-SRGS" shape=
    "rect">[SRGS]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-speech-grammar-20040316/" shape=
    "rect">Speech Recognition Grammar Specification Version
    1.0</a></cite>, Andrew Hunt and Scott McGlashan, Editors. World
    Wide Web Consortium, 16 March 2004. This version of the SRGS
    1.0 Recommendation is <a href=
    "http://www.w3.org/TR/2004/REC-speech-grammar-20040316/" shape=
    "rect">http://www.w3.org/TR/2004/REC-speech-grammar-20040316/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/speech-grammar/" shape=
    "rect">http://www.w3.org/TR/speech-grammar/</a>.</dd>

    <dt><a id="ref-SSML" name="ref-SSML" shape=
    "rect">[SSML]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/"
    shape="rect">Speech Synthesis Markup Language (SSML) Version
    1.0</a></cite>, Daniel C. Burnett, et al., Editors. World Wide
    Web Consortium, 7 September 2004. This version of the SSML 1.0
    Recommendation is <a href=
    "http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/"
    shape=
    "rect">http://www.w3.org/TR/2004/REC-speech-synthesis-20040907/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/speech-synthesis/" shape=
    "rect">http://www.w3.org/TR/speech-synthesis/</a>.</dd>

    <dt><a id="ref-xml-10" name="ref-xml-10" shape=
    "rect">[XML10]</a></dt>

    <dd><cite><a href="http://www.w3.org/TR/2006/REC-xml-20060816/"
    shape="rect">Extensible Markup Language (XML) 1.0 (Fourth
    Edition)</a></cite>, T. Bray et al. World Wide Web Consortium,
    16 August 2006, edited in place 29 September 2006. This version
    of the XML 1.0 Recommendation is <a href=
    "http://www.w3.org/TR/2006/REC-xml-20060816/" shape=
    "rect">http://www.w3.org/TR/2006/REC-xml-20060816/</a>. The
    latest version is available at <a href=
    "http://www.w3.org/TR/REC-xml/" shape=
    "rect">http://www.w3.org/TR/REC-xml/</a>.</dd>

    <dt><a id="ref-xml-11" name="ref-xml-11" shape=
    "rect">[XML11]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2006/REC-xml11-20060816/" shape=
    "rect">Extensible Markup Language (XML) 1.1 (Second
    Edition)</a></cite>, T. Bray et al. World Wide Web Consortium,
    16 August 2006, edited in place 29 September 2006. This version
    of the XML 1.1 Recommendation is <a href=
    "http://www.w3.org/TR/2006/REC-xml11-20060816/" shape=
    "rect">http://www.w3.org/TR/2006/REC-xml11-20060816/</a>. The
    latest version is available at <a href=
    "http://www.w3.org/TR/xml11/" shape=
    "rect">http://www.w3.org/TR/xml11/</a>.</dd>

    <dt><a id="ref-xml-base" name="ref-xml-base" shape=
    "rect">[XML-BASE]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2001/REC-xmlbase-20010627/" shape=
    "rect">XML Base</a></cite>, J. Marsh, editor. World Wide Web
    Consortium, 27 June 2001. This version of the XML Base
    Recommendation is <a href=
    "http://www.w3.org/TR/2001/REC-xmlbase-20010627/" shape=
    "rect">http://www.w3.org/TR/2001/REC-xmlbase-20010627/</a>. The
    latest version is available at <a href=
    "http://www.w3.org/TR/xmlbase/" shape=
    "rect">http://www.w3.org/TR/xmlbase/</a>.</dd>

    <dt><a id="ref-xml-id" name="ref-xml-id" shape=
    "rect">[XML-ID]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2005/REC-xml-id-20050909/" shape=
    "rect">xml:id Version 1.0</a></cite>, J. Marsh, D. Veillard, N.
    Walsh. World Wide Web Consortium, 9 September 2005. This
    version of the xml:id Recommendation is <a href=
    "http://www.w3.org/TR/2005/REC-xml-id-20050909/" shape=
    "rect">http://www.w3.org/TR/2005/REC-xml-id-20050909/</a>. The
    latest version is available at <a href=
    "http://www.w3.org/TR/xml-id/" shape=
    "rect">http://www.w3.org/TR/xml-id/</a>.</dd>

    <dt><a id="ref-xmlns-10" name="ref-xmlns-10" shape=
    "rect">[XML-NS10]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2006/REC-xml-names-20060816/" shape=
    "rect">Namespaces in XML 1.0 (Second Edition)</a></cite>, T.
    Bray et al., Editors. World Wide Web Consortium, 16 August
    2006. This version of the XML Namespaces Recommendation is
    <a href="http://www.w3.org/TR/2006/REC-xml-names-20060816/"
    shape=
    "rect">http://www.w3.org/TR/2006/REC-xml-names-20060816/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/xml-names/" shape=
    "rect">http://www.w3.org/TR/xml-names/</a>.</dd>

    <dt><a id="ref-xmlns-11" name="ref-xmlns-11" shape=
    "rect">[XML-NS11]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-xml-names11-20040204/" shape=
    "rect">Namespaces in XML 1.1</a></cite>, T. Bray et al.,
    Editors. World Wide Web Consortium, 4 February 2004. This
    version of the XML Namespaces Recommendation is <a href=
    "http://www.w3.org/TR/2004/REC-xml-names11-20040204/" shape=
    "rect">http://www.w3.org/TR/2004/REC-xml-names11-20040204/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/xml-names11/" shape=
    "rect">http://www.w3.org/TR/xml-names11/</a>.</dd>

    <dt><a id="ref-xmlschema-1" name="ref-xmlschema-1" shape=
    "rect">[XML-SCHEMA-1]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/" shape=
    "rect">XML Schema Part 1: Structures Second Edition</a></cite>,
    H. S. Thompson, et al., Editors. World Wide Web Consortium, 28
    October 2004. This version of the XML Schema Part 1
    Recommendation is <a href=
    "http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/" shape=
    "rect">http://www.w3.org/TR/2004/REC-xmlschema-1-20041028/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/xmlschema-1/" shape=
    "rect">http://www.w3.org/TR/xmlschema-1/</a>.</dd>

    <dt><a id="ref-xmlschema-2" name="ref-xmlschema-2" shape=
    "rect">[XML-SCHEMA-2]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/" shape=
    "rect">XML Schema Part 2: Datatypes Second Edition</a></cite>,
    Paul V. Biron and Ashok Malhotra, Editors. World Wide Web
    Consortium, 28 October 2004. This version of the XML Schema
    Part 2 Recommendation is <a href=
    "http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/" shape=
    "rect">http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/xmlschema-2/" shape=
    "rect">http://www.w3.org/TR/xmlschema-2/</a>.</dd>
  </dl>

  <h3 id="S6.2">6.2 Informative References</h3>

  <dl>
    <dt><a id="ref-dc" name="ref-dc">[DC]</a></dt>

    <dd><cite><a href="http://dublincore.org/">Dublin Core Metadata
    Initiative</a></cite>.<br />
    See <a href=
    "http://dublincore.org/">http://dublincore.org/</a>.</dd>

    <dt><a id="ref-dc-es" name="ref-dc-es">[DC-ES]</a></dt>

    <dd><cite><a href=
    "http://dublincore.org/documents/dces/">Dublin Core Metadata
    Element Set, Version 1.1: Reference
    Description</a></cite>.<br />
    See <a href=
    "http://dublincore.org/documents/dces/">http://dublincore.org/documents/dces/</a>.</dd>

    <dt><a id="ref-dict" name="ref-dict">[DICT]</a></dt>

    <dd><cite><a href=
    "http://dictionary.reference.com">Dictionary.com Unabridged (v
    1.0.1)</a></cite>, Dictionary.com. Based on the Random House
    Unabridged Dictionary, © Random House, Inc. 2006. Entries in
    the dictionary are available at <a href=
    "http://dictionary.reference.com">http://dictionary.reference.com</a>.</dd>

    <dt><a id="ref-html" name="ref-html">[HTML]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/1999/REC-html401-19991224/">HTML 4.01
    Specification</a></cite>, Dave Raggett, et al., Editors. World
    Wide Web Consortium, 24 December 1999. This version of the HTML
    4.01 Recommendation is <a href=
    "http://www.w3.org/TR/1999/REC-html401-19991224/" shape=
    "rect">http://www.w3.org/TR/1999/REC-html401-19991224/</a>. The
    latest version of HTML is available at <a href=
    "http://www.w3.org/TR/html/" shape=
    "rect">http://www.w3.org/TR/html/</a>.</dd>

    <dt><a id="ref-ipa" name="ref-ipa">[IPA]</a></dt>

    <dd><cite><a href=
    "http://www.arts.gla.ac.uk/IPA/ipa.html">International Phonetic
    Association</a></cite>.<br />
    See <a href=
    "http://www.arts.gla.ac.uk/IPA/ipa.html">http://www.arts.gla.ac.uk/IPA/ipa.html</a>
    for the organization's website.</dd>

    <dt><a id="ref-ipaunicode1" name=
    "ref-ipaunicode1">[IPAUNICODE1]</a></dt>

    <dd><cite><a href=
    "http://web.uvic.ca/ling/resources/ipa/charts/unicode_ipa-chart.htm">
    The International Phonetic Alphabet</a></cite>, J. Esling. This
    table of IPA characters in Unicode is available at <a href=
    "http://web.uvic.ca/ling/resources/ipa/charts/unicode_ipa-chart.htm">
    http://web.uvic.ca/ling/resources/ipa/charts/unicode_ipa-chart.htm</a>.</dd>

    <dt><a id="ref-ipaunicode2" name=
    "ref-ipaunicode2">[IPAUNICODE2]</a></dt>

    <dd><cite><a href=
    "http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm">The
    International Phonetic Alphabet in Unicode</a></cite>, J.
    Wells. This table of Unicode values for IPA characters is
    available at <a href=
    "http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm">http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm</a>.</dd>

    <dt><a id="ref-jeidaalphabet" name=
    "ref-jeidaalphabet">[JEIDAALPHABET]</a></dt>

    <dd><cite><a href=
    "http://it.jeita.or.jp/document/publica/standard/summary/JEIDA-62-2000.pdf">
    JEIDA-62-2000 Phoneme Alphabet</a></cite>. JEITA. An abstract
    of this document (in Japanese) is available at <a href=
    "http://it.jeita.or.jp/document/publica/standard/summary/JEIDA-62-2000.pdf">
    http://it.jeita.or.jp/document/publica/standard/summary/JEIDA-62-2000.pdf</a>.</dd>

    <dt><a id="ref-jeita" name="ref-jeita">[JEITA]</a></dt>

    <dd><cite><a href="http://www.jeita.or.jp">Japan Electronics
    and Information Technology Industries
    Association</a></cite>.<br />
    See <a href=
    "http://www.jeita.or.jp">http://www.jeita.or.jp/</a>.</dd>

    <dt><a id="ref-rdf-conc" name=
    "ref-rdf-conc">[RDF-CONC]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/">Resource
    Description Framework (RDF): Concepts and Abstract
    Syntax</a></cite>, G. Klyne and J.J. Carroll, Editors. World
    Wide Web Consortium, 10 February 2004. This version of the RDF
    Concepts and Abstract Syntax is <a href=
    "http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/">http://www.w3.org/TR/2004/REC-rdf-concepts-20040210/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/rdf-concepts/">http://www.w3.org/TR/rdf-concepts/</a>.</dd>

    <dt><a id="ref-rdf-schema" name=
    "ref-rdf-schema">[RDF-SCHEMA]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-rdf-schema-20040210/">RDF
    Vocabulary Description Language 1.0: RDF Schema</a></cite>, D.
    Brickley and R. Guha, Editors. World Wide Web Consortium, 10
    February 2004. This version of the RDF Schema Recommendation is
    <a href=
    "http://www.w3.org/TR/2004/REC-rdf-schema-20040210/">http://www.w3.org/TR/2004/REC-rdf-schema-20040210/</a>.
    The latest version of RDF Schema is available at <a href=
    "http://www.w3.org/TR/rdf-schema/">http://www.w3.org/TR/rdf-schema/</a>.</dd>

    <dt><a id="ref-rdf-xml" name=
    "ref-rdf-xml">[RDF-XMLSYNTAX]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/">RDF/XML
    Syntax Specification</a></cite>, D. Beckett, Editor. World Wide
    Web Consortium, 10 February 2004. This version of the RDF/XML
    Syntax Recommendation is <a href=
    "http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/">http://www.w3.org/TR/2004/REC-rdf-syntax-grammar-20040210/</a>.
    The latest version of the RDF XML Syntax is available at
    <a href=
    "http://www.w3.org/TR/rdf-syntax-grammar/">http://www.w3.org/TR/rdf-syntax-grammar/</a>.</dd>

    <dt><a id="ref-REQS" name="ref-REQS">[REQS]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/WD-lexicon-reqs-20041029/">Pronunciation
    Lexicon Specification (PLS) Version 1.0
    Requirements</a></cite>, P. Baggia and F. Scahill, Editors.
    World Wide Web Consortium, 29 October 2004. This document is a
    work in progress. This version of the Pronunciation Lexicon
    Requirements is <a href=
    "http://www.w3.org/TR/2004/WD-lexicon-reqs-20041029/">http://www.w3.org/TR/2004/WD-lexicon-reqs-20041029/</a>.
    The latest version of the Pronunciation Lexicon Requirements is
    available at <a href=
    "http://www.w3.org/TR/lexicon-reqs/">http://www.w3.org/TR/lexicon-reqs/</a>.</dd>

    <dt><a id="ref-rfc2732" name="ref-rfc2732" shape=
    "rect">[RFC2732]</a></dt>

    <dd><cite><a href="http://www.ietf.org/rfc/rfc2732.txt" shape=
    "rect">Format for Literal IPv6 Addresses in URL's</a></cite>,
    R. Hinden, et al., Editors. IETF, December 1999. This RFC is
    available at <a href="http://www.ietf.org/rfc/rfc2732.txt"
    shape="rect">http://www.ietf.org/rfc/rfc2732.txt</a>.</dd>

    <dt><a id="ref-sampa" name="ref-sampa">[SAMPA]</a></dt>

    <dd><cite><a href=
    "http://www.phon.ucl.ac.uk/home/sampa/home.htm">SAMPA computer
    readable phonetic alphabet</a></cite>, J.C. Wells.<br />
    See <a href=
    "http://www.phon.ucl.ac.uk/home/sampa/home.htm">http://www.phon.ucl.ac.uk/home/sampa/home.htm</a>
    for information on it.</dd>

    <dt><a id="ref-SISR" name="ref-SISR" shape=
    "rect">[SISR]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2006/CR-semantic-interpretation-20060111/"
    shape="rect">Semantic Interpretation for Speech Recognition
    (SISR) Version 1.0</a></cite>, Luc van Tichelen and Dave Burke,
    Editors. World Wide Web Consortium, 5 April 2007. This version
    of the SISR 1.0 Recommendation is <a href=
    "http://www.w3.org/TR/2007/REC-semantic-interpretation-20070405/"
    shape=
    "rect">http://www.w3.org/TR/2007/REC-semantic-interpretation-20070405/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/semantic-interpretation/" shape=
    "rect">http://www.w3.org/TR/semantic-interpretation/</a>.</dd>

    <dt><a id="ref-SSML-11" name="ref-SSML-11" shape=
    "rect">[SSML-11]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2008/WD-speech-synthesis11-20080620/"
    shape="rect">Speech Synthesis Markup Language (SSML) Version
    1.1</a></cite>, Daniel C. Burnett and 双志伟 (Zhi Wei Shuang),
    Editors. World Wide Web Consortium, 20 June 2008. This version
    of the SSML 1.1 Working Draft is <a href=
    "http://www.w3.org/TR/2008/WD-speech-synthesis11-20080620/"
    shape=
    "rect">http://www.w3.org/TR/2008/WD-speech-synthesis11-20080620/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/speech-synthesis11/" shape=
    "rect">http://www.w3.org/TR/speech-synthesis11/</a>.</dd>

    <dt><a id="ref-VXML" name="ref-VXML" shape=
    "rect">[VXML]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/REC-voicexml20-20040316/" shape=
    "rect">Voice Extensible Markup Language (VoiceXML) Version
    2.0</a></cite>, Scott McGlashan et al., Editors. World Wide Web
    Consortium, 16 March 2004. This version of the VoiceXML 2.0
    Recommendation is <a href=
    "http://www.w3.org/TR/2004/REC-voicexml20-20040316/" shape=
    "rect">http://www.w3.org/TR/2004/REC-voicexml20-20040316/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/voicexml20/" shape=
    "rect">http://www.w3.org/TR/voicexml20/</a>.</dd>

    <dt><a id="ref-XHTML2" name="ref-XHTML2">[XHTML2]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2004/WD-xhtml2-20040722/" shape=
    "rect">XHTML 2.0</a></cite>, J. Axelsson et al., Editors. World
    Wide Web Consortium, 22 July 2004. This version of the XML
    XHTML 2.0 Working Draft is <a href=
    "http://www.w3.org/TR/2004/WD-xhtml2-20040722/" shape=
    "rect">http://www.w3.org/TR/2004/WD-xhtml2-20040722/</a>. The
    latest version is available at <a href=
    "http://www.w3.org/TR/xhtml2/" shape=
    "rect">http://www.w3.org/TR/xhtml2/</a>.</dd>

    <dt><a id="ref-XHTML-MTYPES" name=
    "ref-XHTML-MTYPES">[XHTML-MTYPES]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html"
    shape="rect">XHTML Media Types</a></cite>, Ishikawa Masayasu,
    Editor. World Wide Web Consortium, 1 August 2002. This version
    of the W3C Note is <a href=
    "http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html"
    shape=
    "rect">http://www.w3.org/TR/xhtml-media-types/xhtml-media-types.html</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/xhtml-media-types/" shape=
    "rect">http://www.w3.org/TR/xhtml-media-types/</a>.</dd>

    <dt><a id="ref-xpointer" name="ref-xpointer" shape=
    "rect">[XPOINTER]</a></dt>

    <dd><cite><a href=
    "http://www.w3.org/TR/2003/REC-xptr-framework-20030325/" shape=
    "rect">XPointer Framework</a></cite>, P. Grosso, E. Maler, J.
    Marsh, N. Walsh. World Wide Web Consortium, 25 March 2003. This
    version of the XPointer Framework Recommendation is <a href=
    "http://www.w3.org/TR/2003/REC-xptr-framework-20030325/" shape=
    "rect">http://www.w3.org/TR/2003/REC-xptr-framework-20030325/</a>.
    The latest version is available at <a href=
    "http://www.w3.org/TR/xptr-framework/" shape=
    "rect">http://www.w3.org/TR/xptr-framework/</a>.</dd>

    <dt><a id="ref-x-sampa" name="ref-x-sampa">[X-SAMPA]</a></dt>

    <dd><cite><a href=
    "http://www.phon.ucl.ac.uk/home/sampa/ipasam-x.pdf">Computer-coding
    the IPA: a proposed extension of SAMPA</a></cite>, J.C. Wells,
    University College London, 28 April 1995. This version is
    available at <a href=
    "http://www.phon.ucl.ac.uk/home/sampa/ipasam-x.pdf">http://www.phon.ucl.ac.uk/home/sampa/ipasam-x.pdf</a>.</dd>

    <dt><a id="ref-web-arch" name=
    "ref-web-arch">[WEB-ARCH]</a></dt>

    <dd><cite><a href="http://www.w3.org/TR/webarch/">Architecture
    of the World Wide Web, Volume One</a></cite>, I. Jacobs, N.
    Walsh, World Wide Web Consortium, 15 December 2004. This
    version of the Architecture of World Wide Web Recommendation is
    <a href=
    "http://www.w3.org/TR/webarch/">http://www.w3.org/TR/webarch/</a>.</dd>
  </dl>

  <h2 id="S7">7. Contributors and Acknowledgements</h2>

  <p>The Contributors who provided ideas, comments, feedback and
  implementation experience to improve this specification.
  <i>(listed in alphabetical order)</i>:</p>

  <dl>
    <dd>Jeff Adams, Nuance</dd>

    <dd>Kazuyuki Ashimura, W3C</dd>

    <dd>Patrizio Bergallo, Loquendo</dd>

    <dd>Ellen Eide, IBM</dd>

    <dd>Max Froumentin, W3C</dd>

    <dd>Richard Ishida, W3C</dd>

    <dd>严峻 (Yan Jun), iFLYTEK</dd>

    <dd>Matt Oshry, Microsoft</dd>

    <dd>Dave Pawson, RNIB</dd>

    <dd>Luc Van Tichelen, Nuance</dd>
  </dl>

  <p>The editor wishes to thank the following W3C groups for their
  helpful comments: WAI and WAI/PF, I18N and MMI.</p>

  <p>This specification was written with the help of the following
  people <i>(listed in alphabetical order)</i>:</p>

  <dl>
    <dd>Debbie Dahl, Conversational Technologies</dd>

    <dd>Ken Davies, HeyAnita</dd>

    <dd>Kurt Fuqua, Vail Systems</dd>

    <dd>Will Gardella, SAP</dd>

    <dd>Makoto Hirota, Canon</dd>

    <dd>Jim Larson, Intervoice</dd>

    <dd>Dave Raggett, W3C/Volantis</dd>

    <dd>Matt Womer, W3C</dd>
  </dl>

  <h2 id="AppA">Appendix A - Schema for Pronunciation Lexicon
  Specification</h2>

  <p><b><i>This section is normative.</i></b></p>

  <p>There are two schemas which can be used to validate PLS
  documents.<br />
  The latest version of the schemas are available at:</p>

  <ul>
    <li>The XML schema:<br />
    <code>"http://www.w3.org/TR/pronunciation-lexicon/pls.xsd"</code></li>

    <li>The RELAX NG schema:<br />
    <code>"http://www.w3.org/TR/pronunciation-lexicon/pls.rng"</code></li>
  </ul>

  <p>For stability it is <em title=
  "RECOMMENDED in RFC 2119 context" class=
  "RFC2119">RECOMMENDED</em> that you use the dated URI available
  at:</p>

  <ul>
    <li>The XML schema:<br />
    <code>"http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20081014/pls.xsd"</code></li>

    <li>The RELAX NG schema:<br />
    <code>"http://www.w3.org/TR/2008/PR-pronunciation-lexicon-20081014/pls.rng"</code></li>
  </ul>

  <h2 id="AppB">Appendix B - MIME Type and File Suffix</h2>

  <p><b><i>This section is normative.</i></b></p>

  <p>The media type associated to Pronunciation Lexicon
  Specification documents is <code>"application/pls+xml"</code> and
  the filename suffix is <code>".pls"</code> as defined in
  [<a href="#ref-rfc4267">RFC4267</a>].</p>

  <h2 id="AppC">Appendix C - Issues in Retrieving Lexical
  Content</h2>

  <p><i>This section is informative.</i></p>

  <p>Speech applications that use a PLS document need a mechanism
  enabling them to retrieve appropriate lexical content. In the
  simplest of cases, an application will search the PLS document
  for <a href="#S4.5"><code>&lt;grapheme&gt;</code></a> elements
  with content that exactly matches the input and retrieve all
  corresponding lexemes. In general, however, the retrieval of
  lexical content is not so trivial; it is necessary to define what
  constitutes an exact match and which lexemes are to be retrieved
  when competing matches can apply.</p>

  <p>Here is an example of an approach to retrieve appropriate
  lexical content.</p>

  <ul>
    <li>The text can be tokenized into the smallest orthographic
    units that might have some meaning to the written language
    addressed by the PLS document. By default, for written
    languages with an alphabet, all consecutive members of the
    alphabet can be grouped into a single token and all other
    characters (such as punctuation, hyphen and apostrophe) can be
    considered as individual tokens. Applications that address
    languages using a logographic writing system might consider
    each logogram as a separate token. Languages that are heavily
    agglutinative need a more elaborate default tokenization
    procedure.</li>

    <li>A speech application may provide a mechanism to segment the
    text into explicit tokens. Such a mechanism can be used to
    overcome a limitation of the default tokenization. For example,
    "don't" can be explicitly tokenized as "do" and "n't" in order
    to match a <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
    element with content "n't".</li>

    <li>Tokens should be processed in the direction of normal
    script reading. Thus, English is processed from left to right,
    and languages such as Arabic and Hebrew, from right to
    left.</li>

    <li>Precedence should be given to the retrieval of lexemes
    having a <a href="#S4.5"><code>&lt;grapheme&gt;</code></a>
    element whose content exactly matches the longest possible
    sequence of consecutive tokens. Thus, a lexeme for "they'll"
    should have precedence over a lexeme for "they" given the input
    "they'll'.</li>

    <li>Lexical retrieval should be performed by the bias of tokens
    rather than characters. Thus, a lexeme for "do" should not
    match the beginning of "done".</li>

    <li>A match can occur only when the Unicode characters are
    equal. Thus, matches should be sensitive to case and
    diacritics. "Lima" should not match "lima", "cure" should not
    match "curé", and "vitæ" should not match "vitae". An
    application may choose to expand ligatures (such as "æ") into
    corresponding letter sequences ("ae") and to use a collation
    table to further loosen the match.</li>

    <li>Multiple consecutive white space characters should be
    reduced to a unique single white space character. Thus, "New
    York" should match "New&nbsp;&nbsp;&nbsp;&nbsp; York".</li>
  </ul>

  <p>This outlined approach is designed principally with the needs
  of English in mind and should be modified to accommodate the
  particular requirements of other languages.</p>

  <p>It is recommended for applications that use a PLS document to
  describe the approach they adopt in retrieving lexical
  content.</p>

  <h4 id="AppC.1">Example:</h4>

  <p>An application that uses the following PLS document:</p>
  <pre class="xml">
&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;lexicon version="1.0"
      xmlns="http://www.w3.org/2005/01/pronunciation-lexicon"
      xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
      xsi:schemaLocation="http://www.w3.org/2005/01/pronunciation-lexicon
        http://www.w3.org/TR/2007/CR-pronunciation-lexicon-20071212/pls.xsd"
      alphabet="ipa" xml:lang="en-US"&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;New York&lt;/grapheme&gt;
    &lt;alias&gt;NY&lt;/alias&gt;
  &lt;/lexeme&gt;
  &lt;lexeme&gt;
    &lt;grapheme&gt;York   City&lt;/grapheme&gt;
    &lt;alias&gt;YC&lt;/alias&gt;
  &lt;/lexeme&gt;
&lt;/lexicon&gt;
</pre>

  <p>should process "New&nbsp;&nbsp; York City" as "NY City" rather
  than "New&nbsp;&nbsp; YC" if it uses the above approach.</p>

  <h2 id="AppD">Appendix D - Changes</h2>

  <p><i>This section is informative.</i></p>

  <h3 id="AppD.5">Changes in this document:</h3>
  <ul>

  <li>Modified the third example because there was a missing "uri"
attribute of &lt;lexicon&gt; element (4.4).</li>

  <li>Updated IPA references (6.2).</li>

  <li>The schema files (pls.xsd and pls.rng) have been
  updated to allow for non-PLS attributes on all PLS elements (App. A).</li>

  </ul>

  <h3 id="AppD.4">Changes in the Proposed Recommendation (18
  August 2008):</h3>
  <ul>
    <li>Updated definition of URI (1.5).</li>

    <li>Clarified usage of white space in IPA transcriptions
    (2).</li>

    <li>Clarified definition of xml:base attribute (4.1).</li>

    <li>Updated example of alternative textual representations for
    German (4.5).</li>

    <li>Applied minor editorial changes.</li>
  </ul>

  <h3 id="AppD.3">Changes in the Candidate Recommendation (12
  December 2007):</h3>

  <ul>
    <li>Added support of IRI and XML 1.1.</li>

    <li>Expanded Conformance section (3).</li>

    <li>Clarified recursion language for alias element (4.7).</li>

    <li>Clarified prefer attribute definition (4.6, 4.7, 4.9).</li>

    <li>Updated schemas and added dated URI for stability (App.
    A).</li>

    <li>Added a note on QNames (4.4).</li>

    <li>Added informative notes on IPA (2).</li>

    <li>Updated Authors/Contributors/Acknowledgements (7).</li>

    <li>Added Appendix to list changes since previous drafts (App.
    D).</li>

    <li>Reviewed normative/informative attributions.</li>

    <li>Added Chinese example (4.4).</li>

    <li>Updated examples (i.e. added SchemaLocation, fixes to IPA
    codes, etc).</li>

    <li>Added clarifications and fixed spelling errors.</li>
  </ul>

  <h3 id="AppD.2">Changes in the second Last Call Working Draft (26
  October 2006):</h3>

  <ul>
    <li>Added a new role attribute to the lexeme element which
    helps to disambiguate Homographs (4.4).</li>

    <li>Added an informative appendix on issues in retrieving
    lexical content (App. C).</li>

    <li>Removed orthography attribute in the grapheme element
    (4.4).</li>

    <li>Removed recursion of alias elements (4.7).</li>

    <li>Styled normative language covered by RFC2119.</li>

    <li>Updated schemas (App. A).</li>

    <li>Modified examples.</li>

    <li>Added requested clarifications.</li>
  </ul>

  <h3 id="AppD.1">Changes in the first Last Call Working Draft (31
  January 2006):</h3>

  <ul>
    <li>Support for the IPA phonetic alphabet is mandatory
    (2).</li>

    <li>Added a new section on multiple pronunciations (4.9).</li>

    <li>Clarified the use of the "prefer" attribute (4.6, 4.7,
    4.9).</li>

    <li>Added a glossary of terms (1.5).</li>

    <li>Fixed XML schema and added Relax NG schema (App. A).</li>
  </ul>
  <!-- <p><a href="http://validator.w3.org/check?uri=referer"><img src=
"http://www.w3.org/Icons/valid-xhtml10" alt="Valid XHTML 1.0!" height="31" width="88" /></a></p><p>
 <a href="http://jigsaw.w3.org/css-validator/">
  <img style="border:0;width:88px;height:31px"
       src="http://jigsaw.w3.org/css-validator/images/vcss"
       alt="Valid CSS!" />
 </a>
</p> -->
</body>
</html>