RE: Numbering the headings. What is the point?

["Murrell, Thomas" <TMurrell -at- alldata -dot- net>]

To number heads or not to number heads; that is today's question.

My question is: Has anyone done a real live study to determine under what
conditions, if any, numbered heads make things easier for the reader?

My problem with numbered heads is that the answer to the Why question
usually revolves around either tradition (We've always done it that way) or
cross-referencing (Refer to Section 1.4.25.9.a. for a detailed discussion of
the "ramafratz.")  All that might have been fine in the time of the
typewriter and typesetting of manuals, but I have always felt it was
archaic, a throwback to the days of wall-to-wall text, no graphics, and
passive voice prose where everything was in three- or five- or seven-hole
binders.

As Ms. Backer points out, today we have software that can keep track of
where a reference is and keep the page number updated.  Since more and more
documentation is being delivered electronically (and viewed electronically)
the reader can actually select a link and go right to the reference without
any confusion whatsoever, in a well-designed document.  So, I don't think
the cross-referencing argument applies, unless you want to argue that it is
easier for the writer to remember section 1.4.25.9.a. from the outline than
to do the cross-reference.  And it looks so tacky these days to actually SAY
you're doing something because it's more convenient to the writer.

I guess the short version is that I don't like numbered heads.  In the
absence of clear unambiguous data as to the circumstances under which
numbered heads are superior, I would avoid them.

Tom Murrell
Senior Technical Writer & Young At Heart Curmudgeon