Sunday, March 14, 2010

Perception and Reception

In an entry I posted a year ago entitled Information Theory and Communication, I touched on Claude Shannon's mathematical theory of communication, which had a major impact of the field of communication during the 50s and 60s, but was later pushed aside and has been all but abandoned.  And while it did not live up to its promise as providing a scientific basis for all communication study, it ought to be, in my view, an important component of any communication curriculum, along with the related notion of cybernetics that was introduced by Norbert Wiener.

When I used to teach introductory communication courses, I introduced information theory and cybernetics early on, and later I moved on to the topic of perception.  In doing so, I noted that perception was most often studied in the fields of biology and psychology, and that in the field of communication, the subject was typically referred to as information reception.  I didn't take the point any further, but between you and me, the reason for this terminology switch is that communication scholars are typically concerned with the sending and receiving of messages, whereas perception is about something more than receiving messages sent by some source, it's about taking in information from the environment through our sense organs.

Now, I don't think that it takes a brain surgeon or rocket scientist, as they say, to make the connection between the reception of information, as understood by engineers and scientists such as Shannon and Wiener, and the process of perception, as understood by biologists and scientists.  Making connections, that's what we do, or at least, that's what we ought to be doing, in the field of communication, and in media ecology, and general semantics.

So it seems that what was intuitively obvious to us philosophical and poetic types has now been discovered and confirmed by the more mathematical and scientific crowd.  Or so I've learned through the Technology Review website's Physics arXiv Blog (the X in arXiv stand for the Greek letter Chi, in case you didn't know, so that it's pronounced "archive"), in a post dated February 24th, and entitled, An Undiscovered Link Between Sensory Perception and Shannon's Theory of Information.  The post begins by stating that the "mathematics that describe both sensory perception and the transmission of information turn out to have remarkable similarities."  It then goes on to discuss what might be termed a mathematical theory of perception:

In 1834, the German physiologist Ernst Weber ... carried out a series of experiments to determine the limits of sensory perception. He gave a blindfolded man a mass to hold and gradually increased its weight, asking the subject to indicate when he first became aware of the change. 

These experiments showed that the smallest increase in weight that a human can perceive is proportional to the initial weight. The German psychologist Gustav Fechner later interpreted Weber's work as a way of measuring the relationship between the physical magnitude of a stimulus and its perceived intensity.

The resultant mathematical model of this process is called the Weber-Fechner law and shows that the relationship between the stimulus and perception is logarithmic. ... The Weber-Fechner law is important because it established a new field of study called psychophysics.

The logarithmic relationship between a stimulus and its perception crops up in various well known examples such as the logarithmic decibel scale for measuring sound intensity and a similar logarithmic scale for measuring the visible brightness of stars, their magnitude.

Okay, so, so far we have this idea of psychophysics, which is maybe a little interesting, but maybe not, and maybe you're saying, hey man, I'm here for the blog, not the log.  So, the main point here is that this process of quantifying perception opens up the possibility of comparison with Shannon's quantification of information.  And that's what come's up next in this arXiv blog post:

Today, Haengjin Choe at Korea University in South Korea, says there is an interesting connection between the Weber-Fechner Law and the famous mathematical theory of information developed by Claude Shannon at Bell Labs in the 1940s.

Shannon's work is among the most important of the 20th century. It establishes the limits on the amount of information that can be sent from one location in the universe to another. It is no exaggeration to say that the world's entire computing and communications infrastructure is based on Shannon's work.
So yes, Shannon's work has always been important, and has become even more significant, and relevant, now that we are living in a digital, online, new and new new media environment.  Now, what does this have to do with perception?
Choe points out that the law developed by Shannon that links the amount of information that can be transmitted by a single symbol is also logarithmic. In fact, it takes exactly the same form as the Weber-Fechner law. 

What that means is that psychophysical phenomena can be treated mathematically in the same way as any other form of information transmission and so opens up a new and extensive mathematical toolbox that may provide new insights into the nature of perception .

Hey, I'm all for unlocking tool boxes, and I look forward to new insights as well.  Indeed, this may be instrumental in the development of artificial sensory organs and neural technological interfaces.  I have to admit that I once thought that that sort of thing, as posited in the science fiction stories of William Gibson, who coined the term cyberspace (and inspired movies like The Matrix), was not feasible, but now it seems that we're well on the way to developing such technologies.  Whether the benefits of such developments outweigh the losses remains to be seen.

But back to the blog, which ends with the following point:
Of course, the idea that sensory perception is a form of communication and so obeys the same rules, is not entirely surprising. What's astonishing (if true) is that the connection has never been noticed before.

I guess it depends on what you mean by noticing the connection.  But hey, my hats off to Haengjin Choe for pinpointing the mathematical correspondence.  The post also provides a link to the abstract of Choe's "Proposal new area of study by connecting between information theory and Weber-Fechner law."  Here's how it reads:

Modern Information theory is generally considered to have been founded in 1948 by Shannon in his seminal work, "A mathematical theory of communication." Shannon's formulation of information theory was an immediate success with communications engineers. Shannon defined mathematically the amount of information transmitted over a channel. Meanwhile, psychophysics is the study of quantitative relations between psychological events and physical events or, more specifically, between sensations and the stimuli that produce them. It seems that Shannon's information theory bears no relation to psychophysics established by German scientist and philosopher Fechner. To our astonishment, it is possible that we combine two fields. And therefore we come to be capable of measuring mathematically perceptions of the physical stimuli applicable to the Weber-Fechner law.

And there you have it.  Now, let me further note that perception and information reception are aspects of what is known in general semantics as the process of abstracting.  In information theory terms, it means that something is always lost in the process of transmission.  In the field of communication, it's a common place to say that the message received is not the message sent.  In general semantics, the point is that we cannot perceive all that there is to perceive about any event or phenomenon.  We only take part of it in, only attend to a portion of what is out there, what is going on, leaving out some of the detail, filtering the information, selecting and simplifing.

Put another way, perception is more, even, than receiving information about the environment, insofar as reception suggests a passive process of taking in whatever comes our way.  More than mere reception, perception is an active process of meaning-making.  We process the sensory data, put the pieces together, interpret, and construct our reality.  This is how we talked about it in the media ecology classes I took with Neil Postman and Christine Nystrom, where we used the term perception, which is, after all, the clearest term to use in this particular instance.

Human beings are meaning-makers.  We interpret and make meaning out of the messages we receive.  We make meaning out of dots and lines   :)    ;)    :D     We make meaning out of the things we perceive in our environment, out of clouds, tea leaves, ink blots, animal entrails, etc.  We make meaning out of the random firing of our neurons while we are asleep (we call that dreaming).  We cannot not make meaning.  And I don't think that's a matter of logarithms.  But it is a matter for my blogger rhythms (oh, I know that's bad, that's bad, but you get my meaning, don't you?).


Bruce Kodish said...

Glad to see you bringing up this new research and making these connections. In modern computer science, Shannon's work remains basic and is being seen as more and more relevant to understanding how we as brains-cum-nervous-systems process/transact with whatever is going on. If that isn't relevant to communication studies I don't know what is. We do live in a world of relations.

You and your readers might be interested in some new work being done in neuroscientific theory and research, Randy Gallistel's work described in his book, Memory and the Computational Brain: Why Cognitive Science will Transform Neuroscience. He goes into the importance of Shannon and information theory. Ginger Campbell of Brain Science Podcast has a nice interview with him here:

Anonymous said...

Claude Shanno was deeply into juggling. Here is a look at his juggling contributions: