Flat Orgs & Learning LO4654

GaltJohn22@aol.com
Sat, 6 Jan 1996 12:59:58 -0500

Replying to LO4588 --

In a message dated 96-01-06 01:55:02 EST, you write:

>However, if the flatter organization is
>combined with a communication structure which minimizes status and
>maximizes encouragement to share information, the opposite is true."

Somehow it would appear that this writer assumes that communication is
always good, not just flatness.

Communication is NOT *inherently* good. Hitler, as an extreme example,
was an excellent communicator.

A more contemporary and perhaps more acceptable example is communication
within a software organization. Up to a point, communication is VITAL,
but beyond that point, it is a drag on productivity. The why is simple:

Let us consider an arbitrary person A in a flat software engineering team.
Person A, to communicate an idea to Person B will require X minutes of a
finite day of 48X (I arbitrarily set X=10 so that we have a 480 minute
day, 10 minutes to communicate idea). B will want to communicate an idea
to A also, even as a critique of A's idea. This will also take X so 2X of
48X was spent communicating between two team members. As the number of
persons on the team rises one can quickly find 48X used in communication
and NONE used doing the work.

This is the most important reason why software teams larger than some
arguable number cannot produce software. In fact, IMHO, tho only reason
the number is arguable is that larger teams with a SHARED VISION *can*
produce viable software systems - precisely *because* they can spend
*less* time communicating.

This function of Diminishing Return on Communication (Let's call it DRC
and look highly official) is true of any organization.

In software, its just easier to see the failure and impossible to cover it
up. The software works or it doesn't.

--
Hal Popplewell
GaltJohn22@aol.com