[scribus] Sort of OT: Reading and typesetting

John Jason Jordan johnxj at comcast.net
Wed Feb 23 20:09:46 CET 2011

<New thread from "Need Arial, Times New Roman Font "equivalents"
to look/print nice in PDF">

On Wed, 23 Feb 2011 11:25:01 -0500
John Culleton <john at wexfordpress.com> dijo:

>There is however this fact that everyone seems to be skipping over. 
>Times (New) Roman is an extremely narrow font designed for newspaper 
>columns. In a book unless it is used in a fairly large point size the 
>number of characters per line will be excessive. People read not 
>letter by letter or word by word but by line segment. Fluent readers 
>(I used to be one) will have only two or three eye movements per line 
>of text.  But that presumes a number of words small enough to be 
>comprehended at a single glance. If you have a smaller point size (say 
>8 points) and a decent sized measure (say 5 inches) there will be too 
>many characters per line which in turn will upset normal eye

The eye movements that John mentions above are called saccades
(pronounced sǝˈkɑdz, to rhyme with façades). The number of saccades per
line of text varies, not just with the length of the line and the
quality of the typesetting, but due to many other factors as well. There
has been a considerable amount of research on how saccades work,
including the following interesting facts:

1) While reading there are back movements, which are known to be
generated by disfluencies. E.g., the more complex the syntax and the
less frequent the lexemes, the more likely the reader will need to go
backwards over previously read portions of the text. In one study a
Swedish girl was tested while reading English, not her native language.
Her saccades were much shorter with many more back movements due to her
unfamiliarity with the language. The point is that typesetting and font
choices are important for legibility, but they are not the only factors.

2) The saccades do not focus on the entire section that is viewed.
Human vision focuses in different ways, so that (typically) only about
four or five letters of each saccade are actually in sharp focus. The
surrounding letters are seen, but fuzzy. We use top-down processing to
understand the fuzzy parts, e.g., fluent readers of English know that in
English a noun is likely to be preceded by "the," so if "the" is fuzzy
it doesn't matter. In fact, function words are rarely in focus; we skip
over articles, prepositions, conjunctions, complementizers and other
syntactic markers. The words we choose to focus on are the nouns,
verbs, and modifiers.

3) In a study of people who were equally bilingual in Hebrew and
English, their saccades were the same as other native English speakers
when reading English, but much shorter when reading Hebrew. The reason
is because Hebrew words are shorter, as Hebrew is written mostly only
with consonants. In other words, reading speed is determined by how
fast we can comprehend, not how fast the eyes can go.

There are lots of other fascinating facts about how humans actually
read. I can give references to journal articles where the above facts
were reported after controlled scientific experiments, in case someone
is interested in studying the matter in more detail.

More information about the scribus mailing list