Reading upside down

[simont]

I've noticed a couple of times recently that about half my brain is naturally able to read upside down without thinking about it, but not the other half.

Yesterday I happened to see a blog post about image processing, which included a sample image apparently scanned from a map the wrong way up. On this image were the capital letters ‘ATH’ (presumably part of a word), upside down. Now when I see them out of the corner of my eye or while my eye scans past them to look at something else, I read ‘HIV’ (because my pattern-recogniser loses the horizontal of the T and the crossbar of the A, and reads what's left the right way up); but the odd thing is that when I look directly at those letters, I instinctively read ‘ATH’, without having to think about it, and can't see them as anything else.

The first time this happened, the blog post in question was scrolling past at high speed on my LJ friends-of-friends page; I was somewhat confused, and I braked to a halt and backpedalled the scroll bar to find out why I'd just seen the letters ‘HIV’ in a post about image processing. I stared directly at the letters ‘ATH’, saw them as ‘ATH’, and assumed I must just have hallucinated. Then the next time I worked the scrollbar, *blinks* there was ‘HIV’ again. It took me several goes before I figured out what was going on, because when I looked at the letters with the top half of my brain I absolutely could not see anything other than ‘ATH’ there.

In the kitchen at work, there's a big box sitting on the floor with ‘Squashes’ written on it. (Presumably it contains bottles of soft drink, rather than unusual vegetables, but I haven't opened it to check.) The orientation of the box is such that when I walk past it I see the word ‘Squashes’ upside down; and usually at first glance I somehow read it as ‘Squashages’. Then I look at it again and can't see why I would have thought that, because obviously it has no g in it and clearly says ‘Squashes’ – but then it happens again the next time I go to the kitchen.

It's taken me a couple of days to figure this one out, but now I think I've got it. The last few letters of ‘Squashes’, when read upside down, look quite like the first few letters of ‘sausages’. So I think what's happening is that my right-way-up reading instinct sees ‘sausa…S’ and assumes the existence of the ‘ge’ in the middle, and then my upside-down instinct starts to kick in and realises that the word starts ‘Sq’, and some overenthusiastic part of my brain's language module has combined the two to synthesise ‘Squashages’ before I manage to notice that it's being silly.

So the conclusion I draw from all this is that there are two largely independent reading subsystems in my brain. One of them can only read the right way up, apparently does so by whole-word pattern recognition, and produces answers extremely quickly; the other one can read upside down as well, has a slightly higher activation threshold (it won't fire at all if I'm not deliberately trying to read something) and takes a fraction of a second longer to give its answer, but once it's working it overrules the other one to the point where I can no longer perceive what it had been telling me.

It certainly doesn't surprise me to find that my brain supports multiple strategies for extracting meaning from the written word. Having a fast word recogniser and a slower but surer letter recogniser seems like an entirely sensible architecture, since the latter can fill in gaps and correct errors in the output of the former. But it does surprise me that they're so independent of one another that they can fight.

Comments

[(Anonymous)]

The overarching theme is that cognitive psychology, like clinical pyschiatry, rarely presents neatly-defined single-symptom disorders and conditions.

Instead of the convenient categories and labels so beloved of the press and the education authorities, you get broadly-defined syndromes named after one symptom that are actually complex sets of symptoms, some of which will be present to a lesser or greater degree; and some of which will be absent - including the 'headline' symptom.

This is the point with dyslexia: if clinical practice is to be based on rational diagnostic methodology then we have to accept that children and adults who read well (sometimes exceptionally so) but still present an overlapping set of visual, cognitive and spatial deficits that is heavily correlated with 'dyslexic dyslexics' are people with the same condition.

Or a very, very closely related cognitive abnormality that causes dyslexia in others. And your description of left-eye/right eye differences in image-processing - and outright cognitive dissonance between the two inputs - really rings a bell with anyone who's read about remedial reading tutors trying out coloured spectacles and blanking lenses in the belief that it's a visual disorder.

[hairyears.livejournal.com]

sorry, that was me - session expired.