Email's filing cabinet metaphor

How it hurts productivity

by Kaitlin Duck Sherwood

It actually isn't at all surprising that email overload is such a problem. Every time there has been a new communications technology, it has taken a very long time to figure out how to best use it.

For example, page numbers did not appear in books until sixty-five years after Gutenberg developed the printing press. The earliest moving pictures didn't have close-ups, pans, or cuts.

Sometimes it just takes a while for the technology to mature to a point where you can do useful things. For example, it was hard to do zooms in early cinema because the focus changed as you zoomed. The first "parafocal" zoom--i.e. one that stayed in focus as you zoomed--didn't go on the market until 1946.

Legacy Systems

But it also takes time for people to try different things, to learn what works and what doesn't. This is hampered by memories of the old way of doing things. For example, in Gutenberg's time, lots of scholarly works were still written on scrolls--which have no pages to number.

Even for books with individual pages, numbering wasn't an obvious concept because books were made one at a time. With different handwriting and different page sizes, what was on the 34th page of one copy was frequently completely different from what was on the 34th page of a different copy.

More fundamentally, if you'd never seen a page number before, how would you know that they were useful?

Email is in the same awkward stage as early books. One symptom of email's immaturity is that different email programs have radically different features and capabilities. Oh, sure, the basics--reply, forward, delete--are the same, but once you get past the basics, the capabilities are very different.

For example,

If we all knew the best way to handle email, all email programs would have -- by and large -- the same features.

Email's Legacy: Filing Cabinets

When programmers created the first email programs, there was no way to easily store messages anywhere but the inbox. This meant that messages that users didn't need to deal with any more were mixed in with messages that users still needed to read, respond to, or act upon. Programmers naturally looked at how they filed paper documents when they developed the metaphor for filing email messages.

However, there are big differences between filing cabinets and email folders. You can't put many paper memos into a file folder, and filing cabinets don't have 'Search' boxes. Not only could you not put many memos into a file folder, but if you weren't extremely careful, it could take hours (days?) to find a memo again. With a filing cabinet, it was thus important to use a filing strategy that minimized how long it took to find something again: using many many folders, each with only a few items in it.

With email, it's more important to use a strategy that minimizes filing time. I suspect users can usually get by with just two folders: the inbox for messages-in-progress and another folder for messages that users are done with.

Underutilization of Filters

But wait, there's worse news.

At some point, people clamored for a way to organize their incoming messages into different groups. Because of the legacy of filing cabinets, people don't make good use of built-in tools (called filters or rules) to organize and prioritize messages.

The canonical email management advice is to use filter to move messages into different folders as the messages arrive. But I found when interviewing email users that most didn't like that technique. They wanted to see all the messages that they had to deal with in one place, not spread across a bunch of different folders.

This makes sense: you don't put your postal mail into your filing cabinet before you read it--you organize it into different piles but leave it on your desk where you can see it. So really what you want is for your active messages to stay in your inbox, but in some sort of organized and prioritized list.

Breaking the Old Patterns

What you need to do instead is use your filters to change something about the message that you can sort by, then sort your inbox by that.

Unfortunately, email programs don't make that technique spectacularly easy to discover or implement.

The good news is that email programs of the future will be much better. Spam filters are going to get a lot better, and it will be much easier to organize messages.


Note: I have a bibliography of the effects of novel communications technology on society that you might also be interested in. Kaitlin Duck Sherwood
Updated 12 June 2002.