Monday, July 4, 2011

Keep talking…

Reed's 2nd Law: on communications
(An Elaboration)

Reed’s 3rd Law on Group Forming Networks is widely known. His 2nd Law was invented to explain a paradox he observed: most communications channels spend most of their 'bits' conveying 'information' already known to the receiver. In information theory these ‘bits’ are redundant. Human factors explain why we need them.

This is David Reed’s 2nd Law: Communications media exist to confirm the prejudices of their audience.

And this is my elaboration of the law: The more open the communications loop the higher the level of information restatement as senders: compete for Focus, need to demonstrate Agreement on meaning, and must establish Trust, with the receiver.


David Reed wrote: “This law was invented to explain a paradox that troubled me for a long time: why do most communications channels spend most of their 'bits' conveying 'information' already known to the receiver.

In information theory, a bit of information is defined to be the distinction between two equally likely possible states, where the actual state is unknown to the receiver, but known to the sender. Sending a bit of information from the sender to the receiver allows the receiver to reach the same state of knowledge as the sender. If the information is actually known to the receiver already, it is, of course, unnecessary to transmit anything to ensure that the receiver knows what the sender knows.

The paradox appears when we look at any of the dominant communications media, such as television broadcast news. Stories are repeated over and over again, commentators make the same, entirely predictable comments, etc. When you think about it, it's quite easy for any member of an audience to simulate or satirize the content on almost any communications medium. If so, why do we need to waste the expensive bandwidth to deliver to the audience what it already knows is coming?

Thus we must look beyond information theory to explain why we need all those television bits, or radio bits, or newspaper bits, etc.

Looking beyond information theory—something I often do—I see there are a few significant factors which come into play when communicating with people. These can be summarised as focus, agreement and trust, and the more remote the communication the more important these factors are.

Broadcasting messages to a human audience, is perhaps the most open form of communications loop.  Focus, agreement and trust are, I believe, are the key factors which promote re-statement in human communications.


Firstly humans are limited communication devices; we can only accept and process a limited number of bits at any one time. Even when we appear to be attending to a message our focus may be elsewhere. Perhaps the most common experience of this is when we see a favorite movie or painting again and notice something we had missed previously. Psychologist George A. Miller demonstrated that, in tests, humans can attend to roughly seven distinct simple stimuli. Most communications are complex and could easily overload us as receiver if we did not filter out irrelevant information. More recently neuroscience suggests that in order for our brains to work quickly and efficiently we also need to forget what our brains judge irrelevant.

Thus many, perhaps the majority, of information ‘bits’ received by humans will be discarded. So for a ‘bit’ to get through it may need to be resent many times.


Secondly humans communicate using tokens which encode the meaning of events and data. Sender and receiver will often not have an agreed shared meaning for a new bit of data. If agreement on meaning could be taken as a given then all phone calls to help desks would be answered simply and easily! In reality human communication is plagued by differences in meaning. We see these communications differences clearly when we look at learning styles. There are many different theories about learning styles. All suggest that to get a group of people to learn a piece of information it may need to be repeated in different ways to get a closer match with each learners preferred style of receiving communications. Some people need the detailed spelled out, some need to see the big picture first.

Repeating known information in slightly different ways can be used to build up shared agreement.


Thirdly, in information theory all bits have equal value, because information theory does not look at quality of information. In human communications quality can be vital. (For example: Can I trust this person to tell me the safest way across the desert?)  The issue of trust is most acute where the data has programmatic effect, in other words accepting the data as valid will cause us to think differently. (For example: They are telling me that white birds landing on that tree mean that it is going to rain. Should I act accordingly in future?).

So how do we know who we can trust? Usually by association—can we rely on what they have told us before, are the contradicting themselves, does someone we respect agree with them.
Information about the identity of the sender, consistency of message and authoritative support can help ensure that new ‘bits’ of information are accepted as trustworthy.


The old saying that ‘half the money spend on advertising is wasted, you just don’t know which half’ may in fact be missing the point. It can be essential to ‘waste’ resources on communicating with people in order to get your message through.

If you need to get your message through then:
  • Keep getting the message out there, so that you catch the audience’s attention.
  • Use different ways of getting the message across. Not everyone will respond to tightly worded text. Some people need to see the picture to understand.
  • Be consistent and reliable if you want people to trust your message. Even better show that the message is supported by someone the audience already sees as an authority.
 If you want to be heard, keep talking..