> From: Ron Jeffries [mailto:[EMAIL PROTECTED]
>
> On Friday, December 17, 2004, at 9:00:22 AM, Steve Bate wrote:
>
> >> Or what? My basic reaction to dividing people is "Don't". Let's hear
> >> some ideas that push in the other direction, just for fun.
>
> > When you react to the notion of dividing people, is the dividing
> > defined as categorizing by knowledge, skills, aptitude, or some
> > other criteria? If so, my reaction would be that although there
> > are risks, this type of dividing is often useful. In the best case,
> > it allows people to do what they do well with support of others
> > doing what they do well. Categorization and it's close relative,
> > association, are two of the key functions of the higher parts
> > of the human brain. It gives humans many unique and valuable
> > capabilities compared to other animals. Whether it's sport
> > teams, Sun Tzu's soldier types, XP roles, MBTI, or the myriad
> > other categorizations, this dividing is as pervasive as metaphor
> > in human experience.
>
> The human mind categorizes and associates all the time. It does it
> rather poorly, which is why all Jews are penurious, all Christians
> are fanatics, all Arabs are terrorists, all Irishmen are drunks, all
> managers are stupid, and so on. Fact is, categories are very
> approximate and highly inaccurate, and too stable over time.
True. Those are some specific examples of the risks I acknowledged.
Most, if not all, of our cognitive abilities can be used well or
poorly. Your examples don't negate the benefits of categorization
done well. You've hinted at how it could be done better: recognition
of the fuzzy boundaries of categories and the willingness to adjust
categories based on new information.
Your last sentence is an example of cognitive ability gone awry.
It is not a fact. Categories vary in precision and accuracy and
not all categorizers insist on overly stable categories.
> The individual knows better than her manager or some other "up
> front" divider what she can and can't work on, what she is
> interested in today, what she wants to learn about today, and so on.
> Therefore I prefer to mix the people together and let them figure
> out how to divide the work.
OK. Would this preference be satisfied if the people decide to
divide the work based on areas of specialization or roles or
whatever? If so, is the issue not categorization but who does
the categorization, when, why, and how often (as Keith Ray
hinted)?
> Furthermore, this approach is mathematically preferable to dividing
> the people ahead of time /in every case/. As soon as we divide the
> resources applied to a single problem (writing this program), we
> create dependencies. One group needs the work of another in order to
> proceed, and so on. These dependencies cannot be balanced perfectly,
> since people do not divide infinitesmally and vary in performance.
> Therefore there will /in every case/ be queing at one place in the
> network, with idleness at one station, and unused inventory building
> up at another.
I'm not understanding your association of categorization and
dependency. Categories don't necessarily create workflow
dependencies and workflow dependencies can exist among members
of a category. Most of what you said appears to be more related to
the side effects of dependencies than to categorization.
In an XP context, do you think it's useful to know who is a
customer and who is a developer (and to understand the implications
of each category/role even when realized by the same person)?
Steve
To Post a message, send it to: [EMAIL PROTECTED]
To Unsubscribe, send a blank message to: [EMAIL PROTECTED]
ad-free courtesy of objectmentor.com
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/extremeprogramming/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
