A little over three hundred years ago Joseph Marie Jacquard invented a way to weave richly patterned fabric using punch cards.  These patterns could be changed with new punch cards.  This made it different from the high volume fixed pattern looms also being developed at the same time.  Jacquard's loom pointed the way to industrialization based on flexible specialization instead of rigid mechanization.  Over time, the less flexible looms out competed punch card looms on price and volume.  Rigid mechanization became the pattern for the industrial revolution that followed.

But the idea of flexibility on an industrial scale never died out and in the 1970s is started to reemerge.

Today, if you are manufacturing things in America, flexible specialization is likely the only way you are surviving.  You are part of an American manufacturing base that is now producing a third more output than a decade ago. At the same time you are employing about a third fewer people.  In rough terms, your productivity has doubled.  This is a brutal bit of math, especially for the third who no longer work in manufacturing.  The remaining workers are fewer but more skilled and tend to factory floors full of computer controlled machinery.  The flexible specialization embodied by Jacquard's loom has reemerged, driven by the need for precision manufacturing and overnight distribution.

A similar divide occurred for information technology started the 1980s.  The advent of the Fax machine and the emergence of the TCP/IP as a public standard allowed offices and computer systems to take a quantum leap closer together.  The Fax machine allowed the business contract and the company memo to move about in minutes, instead of days.  TCP/IP provided a shared protocol to transmit electronic files between computers regardless of operating system or hardware vendor.  The rigid machinery of information management populated by file clerks, metal cabinets and racks of spooled tapes became less important.  It became more important to be both flexible and specialized.

But electronic productivity seems stalled.  It hasn't halted, but it is struggling to generate more lift against the baggage that has come along with the computer revolution.

E-mail is one culprit.  It favours a conversational style of prose over concise text.  It is simply too easy to type a lot and to conduct in print an exchange that could pass as the transcript from a technical cocktail party.  While the transmission is instantaneous, the time spent writing and reading has gone up.  Plus the exchange is serialized.  Reading always follows writing.  An e-mail conversation takes twice as long as simply talking to each other.  Reading e-mails is now a daily ritual in office life and it takes up more than just a few minutes.

Faster computers are another culprit.  The promise of high powered processors and lots of memory has led to amazing large desktop applications.  Counterintuitively, a large number of desktop applications are not much faster at handling intellectual content than simpler applications running on slower computers.  When memory footprints were small and processors slow, the consequences of poor design and product choices in software were immediately and painfully apparent to both the developer and the user.  Writing a letter or contract in WordPerfect on an IBM XT is not that much slower than doing the same thing in a modern word processor on a quad core.  The software bloat has kept up with computing speed. 

The challenge is always how to remain disciplined in the application of design and content.  A short e-mail has to focus on its message.  A simple application probably has most of the features a user needs and does an excellent job at what it does.   Flexible specialization is the new norm and at its heart is the skilled use of efficient tools.