Font size for headings (was:Font size)

Subject: Font size for headings (was:Font size)
From: Geoff hart <ghart -at- attcanada -dot- ca>
To: "Techwr-L (E-mail)" <techwr-l -at- lists -dot- raycomm -dot- com>, "'Michael West'" <mbwest -at- bigpond -dot- com>, "'Tim Altom'" <taltom -at- simplywritten -dot- com>
Date: Mon, 4 Sep 2000 19:24:27 -0400

Michael West, responding to my comments about a "20% difference in font size"
being a good idea when you want to create a visual distinction, wondered: <<the
original question had to do with "odd" font
sizes, and I'm still in the dark as to which of several possible meanings of
"odd" was intended.>>

My fault: should've changed the subject line. I thought the original thread had
been resolved to everyone's satisfaction; I think the original thread was about
whether it was better to use an odd font size (e.g., 11 point) rather than an
even size (e.g., 12 point). My response to that part of the question was that
no, there was no advantage to any particular size, and I don't think I saw any
replies that diverged grossly from that suggestion. The questions then evolved:

<<are we talking about headings and subheadings? If so, I've often read that
three points is the minimum recommended differential for showing subordination
of headings --particularly if size is the *only * marker. (Most designers will
use colour and placement as supplementary markers.) I doubt that a two-point
size difference -- by itself -- would be effective for this purpose. In fact, I
* know * it wouldn't>>

You have to be careful about what you read; a lot of typography books make
recommendations based solely on the author's personal taste rather than on any
consideration of the reader's needs, and those suggestions often don't check
out in the real world. In my response, I suggested that a minimum difference of
20% in font size is necessary (in the absence of other changes such as color,
boldfacing, position, or underlining) to clearly distinguish between body text
and a heading in the same font. I also stated that this is based on personal
experience, supplemented by a bit of theory (and I cautioned that my
extrapolations from Miller's theory might not be valid), and that the
recommendation might not apply equally well for all type sizes. For example, a
20% change at 6 point might not produce a detectable difference, but the
difference between 10 point and 12 point for most fonts is quite clear.
Conversely, a fixed difference of 3 points on a 720-point font used for road
signs would be completely undetectable to the audience of highway drivers.

<<I've often seen (and used) layouts that use a one-point difference in body
text size depending on whether the text is set in a normal column or a narrow
column (e.g., in a table column). The one-point difference is, in this case,
not * meant * to be noticed, but it works just fine.>>

Which is my point: that's less than 20%. I'm not surprised that this works;
making a small change in type size is one of several copyfitting tricks you can
use to fit more text in a given space than the default typographic specs would
otherwise permit.

<<So where are we with regard to this peculiar notion of "using odd sizes"

See above. Typography is too contextual (too many other factors are working
simultaneously) for simplistic generalisations about a single typographic
factor to work well.

Tim Altom continued the thread: <<I don't think this has anything to do with
Miller's cognitive studies, which focused on memory, not on perception. Rather,
it has to do with the Just Noticeable Difference (JND), a common measure in

A little of both, from what I recall. There was certainly a discussion of JND
in "The magic number seven", but if memory serves, Miller also discussed the
number of JNDs that study participants could distinguish at a single time, and
that this, combined with the JND, determined how many items a participant could
remember (in short-term memory). As you said:

<<The JND varies a good deal depending on many factors, including the intensity
of the stimulus. The lower the stimulus, the smaller the JND. The JND also
varies a lot depending on which senses are involved. Your example of gray
shading on graphs is just such an example of the JND. You might find that the

shades needed varied according to the pixel density and how much area the gray
covered on the page.>>

That's more along the lines of my intended argument, though I expressed it

--Geoff Hart ghart -at- netcom -dot- ca
Pointe-Claire, Quebec, Canada
"Most business books are written by consultants and professors who haven't
spent much time in a cubicle. That's like writing a firsthand account of the
Donner party based on the fact that you've eaten beef jerky."--Scott Adams, The
Dilbert Principle

Previous by Author: Font size?
Next by Author: RE: Graphics, Graphics, Graphics
Previous by Thread: RE: New TECHWR-L Poll Question
Next by Thread: Re: Font size for headings (was:Font size)

What this post helpful? Share it with friends and colleagues:

Sponsored Ads

Sponsored Ads