Saturday 13 June 2015

The paradox of vision

Here's a list of ten random words.  Please look at them and then answer the following four questions:

heavy
golden
brace
owl
follow
proud
match
island
forget
empire

(1) Are they all legible simultaneously?
(2) What is the top word?
(3) What is the bottom word?
(4) Can you read the top and bottom words simultaneously?

If you're like me, you'll have answered (1) "yes", (2) "heavy", (3) "empire" and (4) "no".

The last one came as something of a surprise the first time I tried it.  They're all legible at once, and I know what the top and bottom words are, so I should be able to read them both at once, shouldn't I?  Apparently not - it's not physically possible.  An anatomical fact about the focal length of the eyeball means that the maximum number of lines it's possible to focus on for the purposes of reading is about five.

Try it for yourself.  Stare fixedly at "heavy" at the top.  You'll probably be able to see down to about "follow", but the rest will be an illegible blur.  And yet, when you look at the list as a whole, all ten words are perfectly legible.  So what's going on?

I don't entirely know, but the explanation appears to be something like this: the eyes automatically scan up and down the list, unconsciously to the viewer.  The impression on the retina persists for something like 1/16 of a second.  Therefore, as long as the eye is capable of repeatedly scanning the list every 1/16 of a second, the entire list will appear completely in focus, even though the eye isn't capable of focusing on the whole list simultaneously.

It seems that some people, when asked question (1) above, will answer "no".  They report that the list is only legible down as far as about "follow", and the rest is an illegible blur.  Presumably these people aren't capable of scanning the list within 1/16 of a second - their eyeballs don't move so easily for some reason.

This seems to be the root cause of a horrendous dispute that I've got into on Usenet.  Some people appear to be saying that the answer to question (1) should be "no" for everyone - in other words, that the entire list shouldn't be simultaneously legible for anyone.  Since the entire list appears simultaneously legible to me and many other people, this is clearly false, but I have had a number of people telling me in no uncertain terms that I'm under an illusion and that I should literally doubt the evidence of my own eyes.

It is clearly absurd to suggest that no one can ever perceive two words more than five lines apart as simultaneously legible.  Many simple, everyday tasks would become impossible or extremely difficult.  As an example, take map-reading.  Hardly any two place-names on a map are less then five lines apart (and if they are, you probably don't need the map!).  If I want to plan a journey from, say, Leicester to Nottingham (which are on the same page in my map), I need to have both "LEICESTER" and "NOTTINGHAM" in view simultaneously so that I can visualize the best route between them.  In this instance I can see a fast route via the M1 motorway, and a slower but more direct route along the A6 to Loughborough and then the A60.  I clearly wouldn't be able to do that if I couldn't, in some sense, see "LEICESTER" and "NOTTINGHAM" at once.

As an experiment, I tried viewing the map in such a way that I had to have the text directly in front of me - in other words, so that I had to keep all text literally in focus.  The task changed from a simple one into a horrendous challenge.  Even locating the correct map square (H5) for Leicester became difficult, because I couldn't simultaneously view the "H" along the bottom and the "5" at the right.  And when "LEICESTER" finally came into view, I had no way of knowing that Nottingham was directly north of me - I'd have had to look it up again separately.  Although I could see a small part of the M1 to my left, I couldn't see the number and it seemed to be going off in the wrong direction.  I lifted my gaze slowly upwards along the A6 (as it turned out to be) until I hit Loughborough, where there was a choice.  But I didn't know which way to go because I still couldn't see my destination!

It's clear to me that I couldn't sensibly read a map like this, and yet if you can only see text in focus that's directly in front of you, there would appear to be no other way of doing it.  Now that I recall, my grandfather used to spend hours peering over maps in a similar fashion when he had to plan a route, a task that he hated.  My grandmother and I, on the other hand, could both simply pick up the map and pick out a route visually.  So perhaps some people literally do not have the ability to see text in focus unless it's straight in front of them.

This would help to explain all sorts of odd behaviour that I've noticed.  For example, I was at Stratford International railway station in London recently with a friend of mine.  Although I could see "Platforms 2 and 3" directly over the doorway leading out to the platforms, I couldn't see the tracks themselves as they were down a long flight of steps.  I confidently set off down the steps but my friend told me I was going the wrong way.  He presumably hadn't seen the sign because it wasn't directly in his field of vision.

If I couldn't make out text that wasn't directly in my field of vision, I'd be hopelessly lost.  Almost all signs are either above or to one side of what they're labelling.  I don't drive, but road signs are a case in point.  They clearly can't be placed directly in front of the driver - they have to be either at the side of the road or above it.  Presumably the intention is that the driver takes in the information from the sign while "quasi-simultaneously" looking at the road ahead - i.e. casts his or her eyes over to the side for just enough time to read the sign while maintaining an impression on the retina of the road, so it looks as though the road and the sign are both in focus even though they physically can't be.  That's certainly what I did when I took driving lessons.

Some people, however, seem to report that they actually have to turn and look at the sign momentarily in order to read it.  This would fit in with the theory that they can't read text unless it's in front of them.  It strikes me as a more dangerous way of reading the sign, as you have to take your eyes off the road for a moment.  But if they have this particular visual arrangement, it's the only way they're able to do it.

So my conclusion is this.  Some people have the ability to keep text apparently in focus that's not in their direct field of vision, while others don't.  Is there general agreement on this point?

Sunday 7 June 2015

Vision experiment (2)

boycott
fireman
crimp
litter
patricide
heavy
pleasantry
easily
perfect
sudden
evil
parachute
sloppy
fling
cleft
stonemason

Saturday 6 June 2015

Vision experiment

Brixton
Stockwell
Vauxhall
Pimlico
Victoria
Green Park
Oxford Circus
Warren Street
Euston
King's Cross St Pancras
Highbury & Islington
Finsbury Park
Seven Sisters
Tottenham Hale
Blackhorse Road
Walthamstow Central