Let's try a thought experiment. Suppose that some malign force knocked all the Internet mail servers permanently offline but left everything else intact. How would we cope? Certain e-mail functions would seem to be easy to replace. For example, as some of us have lately been pointing out, newsletters, mailing lists, and other automatic notifications can easily be converted to RSS feeds and can work more effectively in that mode. Of course, if you're a BlackBerry user, you'll push back and say that RSS feeds are only useful when routed to your BlackBerry as e-mail. Maybe so. But remember, we're talking no e-mail at all. None. Nada. SMTP gone completely AWOL. In that case, you'll cozy right up to WAP and RSS.
Replacing e-mail for interpersonal communication would be the harder and more universal challenge. Instant messaging would take up some of the slack. But the odds of finding every intended message recipient online at the same time diminish as you multiply recipients. Asynchronous messaging is one aspect of e-mail's special power. It's not the core benefit, though. If we peel another layer from the onion we find existing solutions that could be leveraged for store-and-forward messaging. Web-based forums, Wikis, and Weblogs are some of the messaging hubs that can enable groups to communicate independently of time and space. What they can't support as easily or as effectively as e-mail is the dynamic formation (and dissolution) of those groups.
Every interpersonal e-mail message creates, or sustains, or alters the membership of a group. It happens so naturally that we don't even think about it. When you're writing a message to Sally, you cc: Joe and Beth. Joe adds Mark to the cc: list on his reply. You and Sally work for one department of your company, Joe for another, Beth is a customer, and Mark is an outside contractor. These subtle and spontaneous acts of group formation and adjustments of group membership are the source of e-mail's special power. Without any help from an administrator, we transcend the boundaries not only of time and space but also of organizational trust.
An ad-hoc group convened by e-mail dissolves unless membership is reaffirmed by each message. This is a feature, not a bug. Many of the groups that perform work in a modern organization are transient. A hallway conversation is over in minutes; a spontaneous collaboration can last a day; a project may take a week. Software that requires people to explicitly declare the formation of these groups, and to acknowledge their dissolution, is too blunt an instrument for such ephemeral social interaction. Like an operating-system thread, an e-mail thread is a lightweight construct, cheap to set up and tear down.
Could a protocol other than SMTP, and an application other than e-mail, support such interaction? Sure, but any other communication medium that has e-mail's special power to convene groups will suffer the same diseases that afflict e-mail: spam, abuse, infoglut. There's no getting around it. We're going to have to figure out how treat these ills with a mix of technical, social, and legal remedies. As we proceed, let's be clear about what e-mail is uniquely qualified to do. There are other ways to publish newsletters, send automated alerts, transfer files, and hold long-running discussions. If we can relieve e-mail of some of these burdens, it will be easier to heal its wounds.