Typeface selection is subordinate to doc design (posting #2)

[Paul Sholar <pks -at- GENSYM -dot- COM>]

(Many pardons for the non-standard line length of posting #1.  --pks)

Gentle technical communicators,

Talk of duelling typefaces is all well and good, but I find that the
selection of a set of compatible faces for a given doc must be constrained
by the doc's overall design.  When you have more freedom in design rather
than less, decisions about typefaces should be one of the _last_ decisions
that the doc designer makes.  Of course, I don't believe that this decision
is a trivial one, just that, after going through a doc design process, the
alternatives for "appropriate" and "compatible" typefaces become
significantly fewer in number.  This is nothing new to many of you.

I would rather see my professional co-conspirators approach the subject of
typefaces from within the larger (and more difficult) realm of "appropriate
doc design," rather than the other way around.  Of course a doc's overall
design is subordinate to its purpose (such quick tutorial, quick reference,
exhaustive reference, etc.), and so on.  That is another thread . . .

BTW: I have yet to pick up "Designing With Type," and I'm interested in
whether its author reflects my particular "school-of-hard-knocks" bias.

Let's assume that the technical communicator thinks of docs as sets of
information built from:

*   Discursive text:  The "discussion" or "narrative" (depending on the
                      "style" of the doc) that presents information in a
                      series of paragraphs.  Even though the reader
                      should have to "read" as little of this as possible,
                      this text should appear in as "pleasing" a typeface
                      as possible.

*   Headings:         Text that organizes (typically as a hierarchical
                      framework) the discursive text. Headings must be
                      easily distinguished from the discursive text, but
                      must also be "compatible" with other "tags" for
                      information used in the doc.

*   Exhibits:         Figures, tables, etc. that are integral to the doc's
                      "discussion" (i.e., that elaborate a specific concept
                      or reader activity).

*   Parentheticals:   Notes, warnings, cross-references (whether uniquely
                      tagged or not), and even certain exhibits, that are
                      not integral to a particular point in the "discussion."

*   Separators:       Design elements that help the reader distinguish
                      different kinds of information in the doc.

The selection of typefaces has to support the overall doc design, which in turn
has to be appropriate for the "mix" of information elements (for example, the
text-to-exhibit ratio) that the doc must present.  Depending on this "mix,"
for instance, it might be more important (or less important) to distinguish the
text in tables from the text in the discussion itself and from the text that
appears in other text-based exhibits (examples, lists of procedural
instructions, etc.).  The importance of making this kind of differentiation
in a particular doc (or doc set) affects how the designer selects the typefaces.

To differentiate a doc's various kinds of information, the designer can use
primarily _text characteristics_ (typeface, boldface, underline, etc.), or
primarily _other design elements_ (rules, white space, icons, other graphical
techniques).

I'll stop here and watch where the thread leads . . .