PopSciGuy and Chris M: Perception of an object’s size is pretty much independent of perception of distance. If you get people to estimate the size of objects at various distances, then they give pretty much the right answer consistently (i.e. they don’t think something far away is smaller than the same object nearby). This seems to hold true in both vision and touch – if you extend touch perception by giving people rods with which to feel objects at a greater distance than normal, they still estimate the size accurately. I would personally go so far as to suggest that size and shape of objects is represented in much the same way by both vision and touch, but that’s perhaps getting a bit off-topic…

Anyway, there are some tasks that have been done with things like length estimation which would be a lot like what you’re suggesting, PopSciGuy – I’ll try to dig out some references when I’m not in such a rush!

JMW: people in Haggard’s group have been doing something along those lines (localization of different kinds of stimulus (a simple touch, or a nociceptive stimulus, things like that) with palm up and palm down, and there seem to be slightly different biases – nothing published yet, just seen the poster at a conference last week.

they overestimated the width of their hand by around 67%,

No, from the figure and data that doesn’t seem to be the case. (Stupid paywall, so no original paper handy.) Perhaps the mean overestimate of individual positions is ~ 70 %.

About 20 minutes later, I tried it with my palm up. Again, the base knuckles were pretty good, and this time I placed my finger tips only about a quarter-inch shorter of where they should be.

Now, to be truly indepedent of bias, this experiment should be done with 50 people doing the plam-down first, and 50 doing the palm up first. I could have been influenced by knowing that I had been short the first time.

I think you’re making my point. As size/distance estimation increases, the error would increase (ie. ratio). And yes, to the proprioceptive system distance doesn’t matter so much, whereas it does to the visual system. Hence, our ability to resolve distance with our eyes closed (purely based on touch ) wouldn’t be as good as when we have our eyes open.

As a separate issue, estimating size based on the touch of your tongue would be an altogether different thing. The tongue has a huge representation within the somatosensory system and might overrepresent size – which might explain why my occasional cavity feels so big!

I strongly suspect it has to do with how we actually use our hands. We are consummate tool users, and the majority of our precision work occurs when our hands are closed around a tool, which means it has something much closer to the proportions the article cites. I’d be very interested to see whether we do better at determining our hand position when it’s in use.

@PopSciGuy, I can write with my eyes closed. We’re really just bad at gauging distances, because they aren’t strictly relevant. We’re a lot better at comparing things, eyes open or closed, since the ratio is what often matters most. Anyway! That ratio thing continues up along the scale, because it’s far and away the best way to get a good estimate while covering a good range of data. Can you imagine what it would be like to estimate a mile with the same accuracy as a centimeter? Say you can recognize ten different distances; do you want 1-10cm, or 1cm, 10cm, 100cm, etc.? There’s a trade-off being made between how precise your sensory system is and how many different ranges it can deal with.

Me, I’ve got good visual accuracy for short ranges, but it just gets shot to hell at long ranges, because I’m nearsighted. My eyes don’t cover the most useful range for a human, but I see over a similar log-scaled range that someone with good eyes does.

I digress. Anyway, similar trade-offs are made within our proprioceptive system, and I suspect being tool users has caused us to misconstrue our own hands out flat, in favor of precision with our hands the way they are when we actually use them.

One question that leaps out at me: Are we ANY good at estimating things with our eyes closed? Humans, by nature are visual creatures. It may be that the ‘resolution’ for us to work in touch is very different compared to when we use our vision. One easy demonstration (which I’m sure has been done somewhere by someone) would be to ask someone to estimate 10cm with their eyes open. Then ask them to do it with their eyes closed. And I guess you would be able to see the increase in error as the distance got bigger and bigger.