For most of the 20th century our means of communication were as unchanging as Latin grammar. We sent letters or made phone calls, and that was that, with little variation from one generation to the next. Nobody but science fiction writers sought or even imagined any improvements; certainly the organizations providing those services did not offer any.Then, during the 1980s, new forms of communications?voice mail, fax, beepers, cell phones and e-mail?suddenly began to appear, penetrating quickly in the commercial world. By the early \u201990s, the background assumptions about the nature of communications appeared to have reversed themselves: now it was change that seemed normal. The marketplace was packed with vendors promoting their candidates. The only questions were which changes would come next, and how quickly?Perhaps the leading candidate for the next big thing in communications was called unified messaging, also known as integrated or universal messaging. Exactly what the idea meant in practice depended on the constituency. To users, unified messaging meant retrieving and processing every sort of message from one address and over one connection. To engineers, it meant a world in which the technical meaning of every message-type lived within a single family of standards. To a CIO, it offered all messaging functions with a single procurement cycle and vendor. To the CEO, it promised responsibility for all communications in a single department. And to investors, it foreshadowed a massive consolidation across industry sectors as wireless, landline, ISP and cable companies all began offering one another\u2019s services.In April 1994, we wrote a wide-ranging article ("E Pluribus Unum") covering all of those perspectives and endorsing the vision of unified messaging. The basic argument of our piece was that the bandwidth tide was relentless and without limit, and that quite soon?because radical change was just the normal state of affairs?that tide would naturally gather all forms of communication to itself. CIOs were like white-water rafters trying to stay upright in a torrent plunging irresistibly and inevitably toward convergence. But could these CIOs stay afloat?Part of the vision came true?the rise in LAN capacities and Internet connection speeds during the \u201990s was certainly dramatic. However, the big change in communications this decade turned out to be not unification but huge increases in the adoption rates of the component technologies: faxes, e-mail and wireless, which all grew by order of magnitude. By the late 1990s, 900 million voice mail messages were being exchanged and 300 million fax minutes consumed every year, says Tom Harper, executive director of Telemessaging Industry Association.Ian Andrew Bell, a telecommunications consultant working in Vancouver, British Columbia, argues that CIOs understood that when they embraced unified messaging as the components were evolving rapidly they risked losing the gains that might come from their individual development. "If you\u2019re going to be a jack-of-all-trades, it\u2019s best not to become the master of none," he says. Would we have instant messaging today if unified messaging had triumphed five years ago? Perhaps not. In addition, huge user bases built around the individual technologies impose their own inertia.Finally, the transition to unified messaging was not a small step to make. As a practical matter, integrating telephony with e-mail usually means replacing a perfectly functional and probably amortized PBX with a digital or "soft" PBX that can run off the local LAN. If an enterprise had several PBXes, thorny interoperability questions were bound to come up. Programmers would need to wire Web-based interfaces. Someone would need to persuade people to give up their fax machines. Training costs would be involved. The proliferation of input devices and new message types (such as video mail) kept raising the complexity bar.As logical a step as unified messaging might seem, it wasn\u2019t easy to find a business case for it that was detailed, quantitative and convincing, not to mention sufficient enough to pay those costs. What is the value of time wasted collecting phone calls and e-mail over two connections? Printing out and faxing text from a dedicated machine? Or retrieving voice mail and responding to it (store-and-forward telephony), as opposed to having the voice mail processed by sophisticated notification algorithms with access to beepers or cell phones? The savings might be there, but they were hard to pin down.However, the qualitative case for UM (which has been renamed "unified communications" by the marketeers to distance itself from the failed product lines of yesteryear) remains interesting. The steadily rising security standards and the huge increase in the number of messages both seem to cry out for a centralized management point. The rise of the application service provider model, which permits companies to buy just as much unification as they need for the staff that needs it the most, should make it easier to think about cost justification. The tide of bandwidth continues to flood in; it seems hard to doubt that sooner or later every type of message is going to be some twist of IP over Ethernet.And on a deep level, the technology seems to want to be unified. While the infrastructures have stayed distinct, the functions themselves have begun to migrate across boundaries. E-mail, a store-and-forward technology, sprouted instant messaging and voice synthesis. Telephony, a real-time technology, developed voice mail. Voice mail has begun to acquire speech recognition, turning it back into another real-time medium. E-mail started carrying images. Wireless is borrowing features from every direction. Each of the communications technologies seems to be developing a unified messaging solution on its own terms.This may well be the best possible outcome. Perhaps the fastest progress occurs not when companies marketing the same technologies compete but when different technologies do?when CRTs compete with flat LCD screens or tape drives with hard disks. As paradoxical as this might sound, it may well be that all the bandwidth and computational muscle flooding in on us will be best used for carrying not just one unified messaging solution but several.