You've probably heard that in early April a judge enjoined Vonage, a big voice-over-IP (VoIP) company, from taking on any new customers. Vonage's ads imply that only a dope wouldn't buy their service. The thing is, I've always wondered why anyone becomes a VoIP customer.

That's not entirely correct. I know that people sign up for VoIP because the price is right. But that's only because IP-based telephone providers don't have many costs. They don't need to run cables under the sea or across continents, nor do they have to shoot satellites into geostationary orbit. They need only sell the customer a VoIP adapter (i.e., a box that lets you attach a standard phone to the Internet, converting your voice into a series of IP packets) and maintain a set of servers that help IP packets function with the old-fashioned telephone service switches and wires. And often, VoIP customers don't even need a VoIP adapter and can instead talk into an Internet-connected PC through a standard headset. Either way, my objection is the same: The Internet just isn't made to carry audio voice conversations.

Why do I think so? In the late 70s and early 80s, big brains thought a lot about how to extend the existing voice networks to accommodate the growing desire for world-wide voice communications and create the data networks, on which today's private and public organizations now depend. In no time, technology made it cheap and simple to convert analog signals, such as our voices, into digital data streams. By the mid-80s, the majority of US phone calls were digitized at some point during their travels, and more and more data-communications customers were demanding cheap bandwidth for data networks to connect their burgeoning computer systems.

"Why not merge these two fast-growing networks?" these late-20th century techie archons thought. If most long-distance phone calls were mostly digital anyway, why create two separate networks to do what was essentially the same thing? So the big brains crunched the numbers and found that while it was possible to put data and voice on the same network, doing so wasn't efficient. It worked--but to the detriment of both networks.

Internet telephony is possible, but not very efficient--sort of like a railroad transportation provider cutting costs by running his trains on the highways. With some modifications, mile-long trains of freight cars might function on the highways, but the side effects both to the highway infrastructure (trains are a lot heavier than semis) and highway intersections (or switching points) would be significant.

So why does VoIP still exist? Two reasons: First, Internet telephony runs over the Internet, and as we all know, nobody's in charge of what does or doesn't run on the Internet. Second, and most important, I suspect that one of the main drivers of VoIP usage is the fact that many people live in countries where their local phone service is a government-owned monopoly, and their governments exploit that monopoly to impose a hidden tax on its citizens in the form of excessively expensive phone service. But Internet service in many of those countries is delivered by private ISPs.

The result? Longer wait times to download movie rentals--curse it all. But at least I have the comfort of knowing it's all for good causes: freedom, democracy, and electronic capitalism. Meanwhile, I'm praying that the VoIP vendors don't decide to go into the freight business.