According to the following quote from Cisco VXML Programmer's Guide (12.4(15)T, page 1-55), the "recordutterance" property is supported within a <field> element:
 
"To enable recording during recognition, set the value of the recordutterance property to true. Recording during recognition is not supported for the <transfer> and <record> elements.

Note:  In VoiceXML 2.1 and later versions, the recordutterance property is not supported for the <transfer> and <record> elements."
 
But when we tested the following simple VXML page (see, in particular, the two lines highlighted in red), it did not work as expected:
 
<?xml version="1.0" encoding="UTF-8"?>
<vxml version="2.1"> 
<property name="recordutterance" value="true"/>
<form id="start">  
    <field name="foundation_fld" modal="false" type="boolean">
     <prompt>Please say or enter your input. </prompt>        
     <nomatch>
          I heard you say <audio expr="application.lastresult$.recording"/>
          Please try again.
     </nomatch>
     <catch event="noinput"> no input </catch>      
     <filled>       
        You entered <value expr="foundation_fld"/>
          <goto next="#start"/>
     </filled>
   </field>
</form>
</vxml>
 
For example, when we said "one two three" to trigger a nomatch, we heard "I heard you say. Please try again.", instead of the expected "I heard you say one two three. Please try again." So, the recording of the actual utterance did not get stored in the expected "application.lastresult$.recording" shadow variable.
 
My questions:
1) Is the "recordutterance" property supposed to work with a <field> element?
2) If so, is the above test result expected of the VXML 2.1 gateway?
 
Thanks,
Rachel

A follow-up question re:  IOS Version 12.4(15) and later releases:
 
Is the "recordutterance" property supposed to work only with VXML 2.1? OR, can it also work with VXML Version 2.0?
 
Thanks,
Rachel