A Midautumn Night's Diversion

hrant's picture

Point out some good evidence that -more than
a century after Javal- the layman still doesn't
know that we read in saccades.

The best/fastest reply wins a fixation.

hhp

dezcom's picture

ChrisL

hrant's picture

:->

THINK, people!

hhp

Nick Shinn's picture

Javal, wasn't he that guy in Les Miz?

hrant's picture

No man, I mean the Abu Ghraib guy.

hhp

hrant's picture

(All this drawing is making your minds lazy...)

The best evidence that the saccadic nature of reading remains
unknown to the layman is... how the eyes of cartoon/animated
characters move when they're [shown] reading.

hhp

Nick Shinn's picture

Well, saccades are kinda old skool, ennit?
What do the reading "experts" have to say about OMT?

hrant's picture

> saccades are kinda old skool

Yeah, as old as mankind (at least). And here to stay.

What's OMT?

hhp

John Hudson's picture

Ocular microtremor?

hrant's picture

I hadn't heard that term in years (in fact it was Bill Troop who brought it up, when I confronted him about his ignorance of reading mechanics). Micromovements of the eye are a mechanism our vision system uses to prevent everything from going gray - apparently the proper propagation of visual signals depends on differences*. But these movements really are micro: not visible even upon direct close examination of a person's eyes. So basically not relevant here (nor to reading).

* A nice extras support for the "information comes from constrast" school.

hhp

Nick Shinn's picture

>So basically not relevant here.

I wouldn't be so sure.
Kainen (Topological Constancy in the Visual Perception of Resonance) has postulated another reason for OMT other than counteracting retinal adaptation.

Beyond the "voluntary" (macro-) saccades that are typically measured in reading experiments there exist three other fixational movements: micro-saccades, drift, and tremor. I would assume they are all inter-related and have some relevance to the process of reading, in that the minutiae of typography has been developed to take advantage of the fact that every micro-detail of text is seen by this array of data-processing techniques.

This is your argument, Hrant: "We see stuff we don't know we're seeing".

hrant's picture

> I would assume they are all inter-related and
> have some relevance to the process of reading

It's certainly not impossible.
Although there's no evidence.

> the minutiae of typography has been developed to take
> advantage of the fact that every micro-detail of text
> is seen by this array of data-processing techniques.

What do type designers do to accomodate micro-movements?
We have trouble just accomodating the macro ones! :-/

But let's say for the sake of argument that we should design with
micro-movements in mind. So specifically what should we do?

--

In any case, how is this relevant to
the layman's ignorance of saccades?

hhp

Dan Weaver's picture

Why would a layman care about saccades? He/she is more interested in putting a meal on the table and paying the bills.

WurdBendur's picture

People sometimes trace over text smoothly with their finger (or whatever) when reading, rather than lifting and tapping, or jerking between fixations. Okay, so that would be awkward, and to be honest, the movement isn't entirely consistent, but rather follows the fixations pretty closely. And it's only useful where line-skip is a significant problem, like when text is ridiculously too small for it's allotted space but still big enough that you won't lose it under your finger.

But hey, it's all I can think of.

hrant's picture

Dan, I don't think laymen should care. It's just interesting
to see such a disconnect between the rendering of reading and
what actually happens - it's a reflection of how our instincts
can iron out complexity for the sake of getting the big picture,
but losing details that can sometimes be important.

Joseph, fingers are also useful for: looking at more
than one part of a text during the same "task"; and
when a reader gets interrupted.

hhp

Kristina Drake's picture

I find this a fascinating concept -- especially in relation to writing and poetry. (Reading back I see that I've managed to ramble on about it, too...)

Our brains do this all the time -- It is apparent in proofreading how quickly our brains fix typos so we don't notice them, which is great if the goal is to understand the message, but not so much when you want to catch the errors...

Patterns -- think houndstooth and how that can look uniformly gray -- or the thread a while back about "uniform colour" and Hrant's position on the necessity of contrast.

Object perception -- we see a dog with specific physical qualities, but we recognize it as "dog" by removing/ignoring the extra detail and fitting it to our generic concept of dog. Or, to relate this to type, we recognize a lowercase letter a regardless of its typeface, based on certain attributes that have nothing to do with the detail of the particular typeface. Yet it is precisely the unnoticed details which make a typeface readable (or not). I would venture that if the average reader spends too much time focusing on the details of the type, it isn't doing a very good job of being type.

I'd also apply it to the way we understand and process other kinds of information, not just visual details. We make stories our own by filling in the gaps with our own detail as well as by "smoothing out" the details into something which resembles ourselves.

The devil *is* in the detail. You can get bogged down in it, lose all perspective, and become entirely unable to function, but it also makes things work and gives a breath of life to what we do. The average person is intent on the message, not the detail -- the writers, editors and type designers take care of the detail so the layman doesn't have to know or care about saccades.

Kristina

WurdBendur's picture

Joseph, fingers are also useful for: looking at more
than one part of a text during the same “task”; and
when a reader gets interrupted.

And marking pages, and holding your reading material, and pointing...
But I was talking about the usefulness of tracing text.

Another example: Subtitles, especially for lyrics, are often outlined and have a fill that sweeps through as the words progress. This sweeping movement is almost invariably smooth and steady, at least within a given word. This would work if we swept our eyes smoothly across each word before moving to the next one, but in reality it just makes them hard to follow. Much more effective are the ones that simply fill an entire word or syllable all at once. An interesting idea would be to fill the bouma-shapes separately where they may be smaller than words and may not correspond exactly to syllables.

Edit: of course, this may aid fast reading, but it might be awkward if you're trying to sing along.

hrant's picture

> a fill that sweeps through as the words progress.

Ah, good example.
And yes, I agree it's a bad idea.

> fill the bouma-shapes

The problem is fixation points are not highly predictable,
and bouma choices even less so (since it depends on so many
things, like reader experience for one). So simply going for
word-at-a-time* is probably best (especially since that sort
of reading is often largely deliberative, not immersive).

* Or maybe syllables for lyrics, noting that
the music tempo is the real driver there.

hhp

John Hudson's picture

The devil is in the detail.

No, God is in the detail. The devil is a generalist. :)

timd's picture

It is apparent in proofreading how quickly our brains fix typos so we don’t notice them, which is great if the goal is to understand the message, but not so much when you want to catch the errors
Even when I am reading for pleasure (rather than concentrating looking for errors) I pick up misspellings (whether misspelt – braek or misuse in context – brake/break) rather than glossing over them.

The best evidence that the saccadic nature of reading remains
unknown to the layman is… how the eyes of cartoon/animated
characters move when they’re [shown] reading.

I suspect if an animator did try to introduce saccadic reading the viewer would see a jerky movement, partly due to the nature of animation (even at 24 drawings per second, rather than the standard 12), but the layman viewer would prefer to believe that reading appears as a smooth motion despite any knowledge they have to the contrary. Maybe the experience (personal or observed) of reading out loud results in the feeling that reading to yourself should be a smooth progressive action.
Tim

dezcom's picture

"The best evidence that the saccadic nature of reading remains
unknown to the layman is…"

that the layman doesn't know the difference between the saccadic motion in reading than he does the motion of the Cicada, the insect pictured in my first post.

ChrisL

hrant's picture

> I pick up misspellings

You're probably not reading fast enough.

> if an animator did try to introduce saccadic reading ...

OK, I agree it could be an intentional simplification.
On the other hand, jerks are in fact easier than smooth
motion at low frame rates.

> reading out loud

Interestingly, as Peter Saenger has explained in his book,
immersive reading began when reading-out-loud (which was
a necessity before word breaks were introduced) got demoted.

> the Cicada

Is that what you meant! At first I thought you were alluding to cricket sounds*; then I thought you meant that I'll need to wait 14 years for a response. Now I see that it was, shockingly, a pun.

* Often used in cartoons to parody silence.

hhp

dezcom's picture

"...Now I see that it was, shockingly, a pun."

ChrisL

timd's picture

>jerks are in fact easier than smooth motion at low frame rates
No – while intuition might tell you that, to get a proper double-take jerk, the more drawings per second the better.

Tim

Syndicate content Syndicate content