Top-down vs. bottom-up analysis? Do both!

[Geoff Hart <Geoff-h -at- MTL -dot- FERIC -dot- CA>]

Anthony Markatos wondered <<Which is the best way to
perform an analysis: top-down or bottom-up? Some experts
say that analysis is all about postponing detail to the
appropriate time, else the analysis will "drown in all the
details". Others advocate bottom-up.  They say top-down is
too complex.  They believe in a bit-by-bit approach.>>

If I've understood your question properly, there is no "best
way". Speaking as an editor, and thus someone who
confronts this issue several times a day (once per manuscript,
in fact), I'd say both your sets of experts are far too deeply in
love with their own pet theories to have a firm grasp of
reality. ("C'mon Geoff... tell us what you really think!" <g>)

When I have a look at a manuscript, I approach it from both
angles: the manuscript must hang together as a whole (on the
macro level: organisation, logic. content, etc.), but it must
also work at the detail level too (on the micro level: grammar,
word choice, clarity, etc.). No manuscript succeeds unless it
succeeds at both levels. Time permitting, I try for a quick
read-through before I ever start editing so that I'll have an
idea how well it hangs together at the macro level; even when
I don't do that, I always evaluate the manuscript at least three
times before it gets published (once for substantive issues,
once for copyediting, and once for proofreading). I have
complete confidence that certain of my authors can produce a
sufficiently well-structured manuscript that the initial read-
through isn't worthwhile; with others, I know that it's pointless
to even start copyediting until I've completely repaired the
overall structure.

--Geoff Hart @8^{)}     Pointe-Claire, Quebec
geoff-h -at- mtl -dot- feric -dot- ca

"If pro is opposite of con, then what is the opposite of progress?"--Anon.

 From ??? -at- ??? Sun Jan 00 00:00:00 0000=