how we read[long]

["Robert H. Bohle" <rbohle -at- CABELL -dot- VCU -dot- EDU>]

A week or so ago I joined in a thread on online typos and such with
an offer to send along some results of research I was doing for an
article. Here it is, and I apologize for its length. Please D now, if
you have no interest. The full article will be in DESIGN: THe Journal
of the Society of Newspaper Design. I have made a decision to edit my
listservs down to at least single digits, and I must now signoff this
one, though you have some int. discussions. So please e-mail me
privately if you have comments or questions. I will also try to find
out if you can get a copy of DESIGN, which normally goes only to
members. This entire issue will be about typography, and many may
find it interesting.

Cheers,
Bob Bohle
Design&
DesignAND -at- aol -dot- com or netdzine -at- richmond -dot- infi -dot- net
URL=http://www.infi.net/~netdzine/
(watch for name change . . . .)

        After going through research on reading (little on monitors specifically), some
 random thoughts on reading on paper vs. on screen:
        - the flicker of fluorescent light affects reading. I wonder if the flicker of
 the monitor also affects how easy it is to look at and focus on.
        - as readers reach *bifocal age,* more light is necessary for good reading.
 Maybe we old duffers (oops, I mean: speaking for myself) need to make sure our
 monitors have good, high contrast. Overall, brightness and contrast seems to be
 the most important factor in easy reading.
        - reading distance may be important. If our monitor is large enough and/or
 close enough, we may have to move our head (instead of just our eyes) to scan
 each wide, word-processed line. The normal 3-degree cone of attention means
 that anything beyond about 14 or so picas (on paper) gets more difficult. On
 screen word processing makes for very wide line measures.
        - the angle of the paper (e.g., on table vs. held in hand) turned out to be a
 factor. A 45-degree angle for what we read turned out to be optimum. My screen,
 for instance, is always much flatter to my line of sight than is my paper
 reading. Maybe thatUs a part, too. My head is also much more erect than it
 likely should be.
        - lights off to the side of the visual field tend to distract. This could be
 distractions around the monitor, as well as the fact that we have more light in
 our peripheral vision by looking ahead at a monitor, rather than by looking
 down at paper on a desktop.
        - readers are very much creatures of habit, and our habit is black on white,
 and also on paper, I suppose. Reverse video reading may be slowed because of
 this. Color seemed to make little difference here, although a background of a
 light blue or cyan seemed to have a slight edge.


        FROMJARTICLE: When we read, we donUt stop and wade through every syllable of a
 word, much less every letter. We make a stop, or fixation, at various points
 along a line of type, gathering meaning as we do so. Sophisticated readers tend
 to identify words by their shape, not by looking at each letter in a word or by
 sounding each syllable, as we did when we were learning to read. So the stops
 can be very quick P between 100 to 250 milliseconds.
        Then we jump to the next stop on the line and look again. These jumps, or
 saccades, vary in length depending on an assortment of typographic variables
 and the skill of the reader. When we reach the end of a line, we must make the
 empty sweep back to find the beginning of the following line so we can
 continue.
        If, during either the momentary fixation on a word (which is referred to as the
 foveal word), on the saccade, or during the empty sweep to the next line,  the
 reader has to struggle, both speed and comprehension suffer. When a reader has
 to struggle he or she also is likely to give up and move on to something else.
 Many studies have shown there is a strong relationship between the visual
 characteristics of the text and the pattern of fixations.
        The initial fixation on a word tends to be halfway between the beginning and
 the middle of a word. Interestingly, not only is the fixated word processed,
 but usually the word immediately to the right is as well. This is called the
 RparafovealS word in the research.
        This is rather like peripheral vision. When we are looking at one thing, we can
 also see other objects surrounding our object of focus. We can move our eyes to
 focus on that other object when we need to gather information from it. Reading
 is similar. When we are focusing on one word, we are also peripherally studying
 the next word in line.
        This parafoveal word is often skipped over at the next fixation or is fixated
 very briefly if our RpreviewS processing comes up with a reasonable meaning,
 either from the shape of the word or from the context, or both. The length of
 the parafoveal word tends to determine where the next fixation will occur, if
 it occurs in that word at all. Studies have shown, for instance, that
 three-letter parafoveal words are often skipped over.
        The importance of the parafoveal word in reading is that some researchers
 believe it has great influence on the length of each fixation and the size of
 the subsequent saccade. If the preview, or peripheral vision glance, of that
 word is good, then reading ease and speed is enhanced.
        The Rpreview benefitS helps to speed the reading process by allowing longer
 saccades and fewer fixations along the line of type. The preview benefit seems
 to come not from perceived structure or meaning of the word, but from low-level
 visual cues, such as the length of the word.           This is why inter-word spacing
 is a critical variable in easy-to-read text: if the reader canUt quickly tell
 where the parafoveal word ends, confusion arises. The word spaces must be large
 enough to be seen peripherally. On the other hand, studies have shown that
 large and irregular word spacing (such as we sometimes get with weak H&J) is a
 major culprit is hampering reading speed.
        Also, all caps are very difficult to process quickly because the shape of the
 word doesnUt have as much variance as does upper and lower case. Also, when we
 read we tend to look only at the top of each letter, and lower case again,
 especially the readability faces with large x-heights, is better suited for
 this. Because the parafoveal word is seen only peripherally, but is very
 important to text processing, making sure the visual characteristics of text
 lend themselves to quick processing is critical.
        Other researchers believe, however, that it is the foveal word processing that
 influences the length of the fixation, and the processing of the following
 parafoveal word that influences the saccade, or the next stop down the line.
 Obviously, the studies continue.
        Good readers also have been found to do a number of other things not directly
 related to typography.
        Readers have been found to apply four levels of attention as they approach a
 printed page: page, line, word, and letter. First, they make a quick assessment
 of the page, noting the straightforward organization of the grid-like line-by
 line structure. Then they enter the presumed first line and begin fixating on
 the foveal words and moving along line by line of type. If they encounter
 difficulty with the word, they break down and look at individual letters.
 Presumably, readers go only as far down the hierarchy as they need to get the
 meaning.
        As they gather word meanings, readers begin grouping them into larger and
 larger meaning units to finally complete the task at the sentence level.
 Clearly, for communication purposes, the more quickly and the higher up the
 hierarchy access to meaning is gained the better.
        The familiar gestalt principles of visual perception come into play as readers
 scan the information to gain access to meaning. Size and shape proximity go
 hand in hand with appropriate letter and word spacing, and to a lesser extent,
 linespacing. Controlling these parameters allows for quicker reading and better
 comprehension.
        Other non-typographic factors that affect reading are familiarity or habit (one
 of the reasons romans are, in general, preferred over sans serif body types P
 at least in the U.S.), interest, redundancy, meaningful grouping of content,
 and context. Because of these and other influences, and because better readers
 use certain schema or approaches to different kinds of reading moments (e.g., a
 textbook vs. a Dave Barry column) and they constantly monitor their
 comprehension, typographic fixes can only go so far in assisting readers.