Reading Images

Allan Hanson (hanson@KUHUB.CC.UKANS.EDU)
Thu, 4 Jan 1996 10:50:44 -0600

A fascinating thread, this, and one that is stirring a lot of interest.
It seems to me that information theory is relevant here. According to
that, the intelligibility of any message varies directly with the amount
of redundancy it contains, redundancy being defined roughly as the part
of the message that is already familiar to the receiver. If a message
contains absolutely no redundancy for a given receiver, it will be
completely unintelligible to him or her.

So in Martin Cohen's contribution to the thread, the images seen in
standing water would be the source of redundancy enabling someone who had
never seen a photograph before to understand it.

Here's another example that has always been fascinating to me. European
explorers encounter a group that has absolutely no experience of
firearms. A confrontation develops, the Europeans feel threatened and
fire a gun, wounding or killing someone. Do the people fired upon
understand what has happened? My anticipation is no, at least not right
then. It would require at least two experiences to establish sufficient
redundancy for them to understand (learn) the connection between pointing
and firing a gun and the person pointed at being wounded or killed. (An
exception might be Australian aboriginal groups who have the notion that
a sorcerer can injure someone by pointing a killing stick at them. That
might provide some redundancy to enable understanding of the first
encounter with a firearm, in a way similar to reflections in water
providing redundancy for understanding one's first encounter with a

I'd be very interested in people's reactions to the situation of first
encounters with firearms, and if anyone has some empirical data on it.

Allan Hanson