![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
E-mail versus instant messaging
Dec. 14th, 2010 02:26 pmI read, maybe a year back, an article claiming that teens of today don't really use e-mail, since they find it hopelessly slow in comparison to avenues such as Facebook. That made me raise a mental eyebrow: Whaddya mean e-mail is slow? It doesn't take any more time to write an e-mail than it does to send a Facebook private message.
The funny thing is that a lot of the "new" features that show up on the Internet are really psychological and not technological developments. Take instant messaging. I got ICQ back when I was maybe thirteen or so, and it opened to me a whole new world of communication. But when you think about it, instant messaging clients really only provide two services that e-mail doesn't: the ability to know when your buddies are online, and much less centrally and as a later development, the ability to know when the other person is typing.
In principle, nothing would prevent us from using e-mail the way we use instant messaging: for a rapid exchange of short messages a sentence or two long, if that. But most of the time, we don't - probably because our mental model for e-mail is too strongly based on traditional mail and long, well-thought out letters. And while Gmail provides threading for e-mail messages, many popular mail clients don't, making the IM format inconvenient to use with it.
While the ability to know when your buddy is online is convenient and certainly helps, it's not a requirement for IM. I and many of my friends use IRC shells, IRC clients running persistently on reliable servers. I can at any time log on to the server to see who has sent me messages, but since the IRC client running on the server is always logged on, people can't automatically tell when I happen to be reading it or not. This has the consequence that people often leave me messages without knowing whether I happen to be online at the moment, trusting for me to answer when I happen to log on. If I happen to reply when they are also online, this will lead to a "traditional" IM-type conversation, but the discussion can also get rather spread out in time. One friend of mine regularly logs on in the evenings and sends "are you there"-type messages to which I respond when I am. People have also used text messages and phone calls to get me online.
Admittedly, IRC does have two features that make it easier for somebody to figure out whether or not you're online: you can set an away message for when you're gone, and you can often find out the amount of time that someone has been idle by doing a whois on them. But these would be easy to implement in e-mail: simply specify a special e-mail content that you could use to query somebody for their status. Upon receiving this message, a mail client would automatically send back a response, depending on the status its user has set.
(The functionality of IRC channels could also pretty much be replicated via mailing lists.)
There are also some other minor features that IM has, but many of those could be replicated in e-mail with little effort. For instance, on IM, you can block people, but you can also tell your e-mail client to delete any messages from a specific sender. We currently employ spam filtering to great success to eliminate unwanted messages.
I suspect the functionality of buddy lists have some specific psychological purpose that you couldn't get in e-mail. When you request to add someone to your buddy list and they approve that, it gives you an implicit permission to bug them, while one might be uncertain of whether it's okay to e-mail somebody with something minor. On the other hand, AIM used to be one of the most popular IM clients, and you didn't need the other person's permission to add them to your buddy list in that one.
If all of the main features of IM could be implemented via minor additions to e-mail, why did everyone go to the effort of defining their own IM protocols and setting up separate servers people could connect to? Why didn't they just create a custom e-mail client that was built for instant messaging purposes, and take advantage of the existing e-mail infrastructure? I think there are two main reasons:
1. The business reason. If you create your own IM service where the clients rely on your servers, you control the service. If you just wrote a custom e-mail client, you would very quickly lose all control of it.
2. The psychological reason. I suspect that a lot of people would have felt it rather silly if you had told them "hey, by downloading this customized e-mail client, you can talk to your friends in a new way!" People underestimate the effect of trivial inconveniences, and would have felt that a client offering just minor changes to the way e-mail is shown wouldn't have been worth the effort to download. On the other hand, branding the program as an entirely new kind of program makes it reasonable to want to try it out. And once you associate IM-type conversation with IM programs and e-mail-type conversations with e-mail, you don't really question that assumption anymore.
The funny thing is that a lot of the "new" features that show up on the Internet are really psychological and not technological developments. Take instant messaging. I got ICQ back when I was maybe thirteen or so, and it opened to me a whole new world of communication. But when you think about it, instant messaging clients really only provide two services that e-mail doesn't: the ability to know when your buddies are online, and much less centrally and as a later development, the ability to know when the other person is typing.
In principle, nothing would prevent us from using e-mail the way we use instant messaging: for a rapid exchange of short messages a sentence or two long, if that. But most of the time, we don't - probably because our mental model for e-mail is too strongly based on traditional mail and long, well-thought out letters. And while Gmail provides threading for e-mail messages, many popular mail clients don't, making the IM format inconvenient to use with it.
While the ability to know when your buddy is online is convenient and certainly helps, it's not a requirement for IM. I and many of my friends use IRC shells, IRC clients running persistently on reliable servers. I can at any time log on to the server to see who has sent me messages, but since the IRC client running on the server is always logged on, people can't automatically tell when I happen to be reading it or not. This has the consequence that people often leave me messages without knowing whether I happen to be online at the moment, trusting for me to answer when I happen to log on. If I happen to reply when they are also online, this will lead to a "traditional" IM-type conversation, but the discussion can also get rather spread out in time. One friend of mine regularly logs on in the evenings and sends "are you there"-type messages to which I respond when I am. People have also used text messages and phone calls to get me online.
Admittedly, IRC does have two features that make it easier for somebody to figure out whether or not you're online: you can set an away message for when you're gone, and you can often find out the amount of time that someone has been idle by doing a whois on them. But these would be easy to implement in e-mail: simply specify a special e-mail content that you could use to query somebody for their status. Upon receiving this message, a mail client would automatically send back a response, depending on the status its user has set.
(The functionality of IRC channels could also pretty much be replicated via mailing lists.)
There are also some other minor features that IM has, but many of those could be replicated in e-mail with little effort. For instance, on IM, you can block people, but you can also tell your e-mail client to delete any messages from a specific sender. We currently employ spam filtering to great success to eliminate unwanted messages.
I suspect the functionality of buddy lists have some specific psychological purpose that you couldn't get in e-mail. When you request to add someone to your buddy list and they approve that, it gives you an implicit permission to bug them, while one might be uncertain of whether it's okay to e-mail somebody with something minor. On the other hand, AIM used to be one of the most popular IM clients, and you didn't need the other person's permission to add them to your buddy list in that one.
If all of the main features of IM could be implemented via minor additions to e-mail, why did everyone go to the effort of defining their own IM protocols and setting up separate servers people could connect to? Why didn't they just create a custom e-mail client that was built for instant messaging purposes, and take advantage of the existing e-mail infrastructure? I think there are two main reasons:
1. The business reason. If you create your own IM service where the clients rely on your servers, you control the service. If you just wrote a custom e-mail client, you would very quickly lose all control of it.
2. The psychological reason. I suspect that a lot of people would have felt it rather silly if you had told them "hey, by downloading this customized e-mail client, you can talk to your friends in a new way!" People underestimate the effect of trivial inconveniences, and would have felt that a client offering just minor changes to the way e-mail is shown wouldn't have been worth the effort to download. On the other hand, branding the program as an entirely new kind of program makes it reasonable to want to try it out. And once you associate IM-type conversation with IM programs and e-mail-type conversations with e-mail, you don't really question that assumption anymore.
