VoiceXML and speech have been percieved as tightly linked because that's how VoiceXML was marketed initially, and first impressions tend to stick around for a long time. Going back to the initial 1999 press release that alerted the world of its existence, VoiceXML was marketed as a standard that would help drive the market for voice-enabled, internet-based services. This was a mistake, and it still haunts us nearly eight years later because the speech angle deflected attention from some of VoiceXML's most powerful benefits (described very well by Dan in his posting above.) Here's a link to the original press release, which triggered many stories and actions that further reinforced that critical first impression: http://www.voicexml.org/pr19990302-1.html I was one of the folks who reviewed this press release before it went out, so yes, this is partly my bad. At the time I was marketing speech technology, so let's just say I had a bias. I thought speech would be a great reason for customers to adopt the standard, but looking back I think I got it backwards. The main reasons to adopt the VoiceXML standard have nothing to do with speech... they have to do with it being a standard that's based on internet infrastructure. It just so happens that IF you adopt VoiceXML, THEN it becomes much easier to also adopt speech. Happily, markets are smart. Customers eventually figured this out, and many started treating standards adoption and speech adoption as separate considerations. As early as 2001 we began to see a surge in DTMF-based VoiceXML deployments. I still believe that if we hadn't made the mistake of initially positioning VoiceXML as a speech standard, customers would have adopted it at an even faster rate. But enough speculation. Now that you have the inside scoop, you too can help set the record straight. It's about standards, and speech-enablement is a very attractive but optional by-product.