For the purpose of creating a variable pool that might allow the application of techniques of numerical taxonomy to a set of communication media, a set of twelve characteristics of communication media were selected. They were not selected on the basis of an exhaustive examination of the literature and there are undoubtedly many other characteristics of media that could be added to the list. The list does reflect many of the themes that are echoed in existing communication research, however. Each characteristic selected can be justified on its own merits.
Three of these measures are concerned with various forms of what we will call, metaphorically, "bandwidth". In information theory, "bandwidth" is a measure of the amount of information that can cross a channel. This is a highly relevant measure when one is examining the information carrying potential of transmission media like telephone lines, radio waves, and light. Because the signal capacity of the modalities of human communication are very difficult to estimate (let alone measure), however, it has not proven highly relevant to human communication media.
A person will typically speak at around 150 words a minute, but a word is not a unit of information. A single word can contain volumes of information, and a complete speech can contain almost no information. Additional layers of meaning may surround those words, and even when subvocal and/or other nonverbal cues are not available, several layers of information can be available in the way things are said.
Hence the measures of bandwidth used here are not pure measures of bandwidth, but rather measures of channel availability. One measure is concerned with the number of channels (presuming the principle human modalities of communication) a medium provides. Two are measures of the extent to which certain kinds of information can be transmitted within the constraints of a particular form of communication.
The first bandwidth measure, Personal Bandwidth, is a measure of the number of channels of communication that are available for use by both the message source (the speaker or transmitter) and the the message target (the reader/listener or receiver). Personal bandwidth is not, therefore, a pure measure of bandwidth. Rather, it is a simple count of communication modalities, incremented by the availability of any of the following transmission channels:
Although it can be argued that some of these channels are more effective, and more important, than others, all will be counted as equals. The minimum value for personal bandwidth is 1. The maximum value is 5.
Does this medium allow for the transmission of the spoken word or its written equivalent. If verbal bandwidth is available for a given form of communication via either the sound or written channel, this characteristic is coded as a yes (1). If not, it is coded as a no (0).
To what extent can analogic information, including, but not restricted to, both images and subvocalizations, be exchanged via this medium. If non-linguistic information can be exchanged via a given medium this characteristic is coded as a yes (1). If not, it is coded as a no (0).
Note that the combination of verbal bandwidth and nonverbal bandwidth provide for three conditions. A medium can transmit one, the other, or both. It is not possible for a medium to transmit neither.
Note also that Personal Bandwidth is, in some sense, an interaction effect of the two. Given the fact that one combination of verbal and nonverbal bandwidth is not possible, it is probably inevitable that there will be a negative correlation both between these two variables and a positive between each of these two variables and personal bandwidth.
These measures of bandwidth belong to a broader class of what might be called transmission characteristics. Other transmission variables include:
How far does a message typically travel in this medium. This measure has been scaled logarithmically using base 12 logarithms (inches are the base measure) as follows:
o Inches -- 1 o Feet -- 1.93 o Tens of Feet -- 2.85 o Hundreds of Feet -- 3.78 o Miles -- 4.45 o Tens of Miles -- 5.38 o Hundreds of Miles -- 6.30 o Thousands of Miles -- 7.23
Note that by scaling this measure logarithmically, estimation errors are minimized. If a medium estimated at hundreds of miles actually reaches tens of miles or thousands of miles, the error inherent to the estimate is not large. Similarly, if a medium estimated in feet actually reaches inches or tens of feet, the error is not large. The possibility of significant errors in estimates is further minimized in the actual analysis when the variable is standardized.
How fast does a message typically travel from its source (at the end of preparation) to its target (at the time of receipt) on this medium. This measure has also been scaled logarithmically, using seconds as the base (base 60 logarithms), as follows:
o Immediate -- 0 o Seconds -- 1 o Minutes -- 2 o Hours -- 2.78 o Days -- 3.25 o Weeks -- 3.61 o Months -- 4.22 o Years -- 5
Note that years have been rounded to the nearest single digit.
What is the shortest period of time within which a message can travel from its source (at the end of preparation) to its target (at the time of receipt) on this medium. This measure has also been scaled logarithmically which the same scale as Speed, immediately above.
Note that typical speeds and theoretical speeds are very different measures. Theoretical speed presumes an ideal situation in which, among other things, the receiver is available to receive the message as soon as the message becomes available. Typical speed assumes the real world, in which this situation rarely obtains.
Nonetheless, typical speeds and theoretical speeds will often be similar, especially when logarithmic scaling is taken into account, and these variables can be expected to correlate strongly.
A second class of characteristics relates to the storage capabilities of a human communications medium. These characteristics include:
How long are messages typically stored in this medium. This measure has also been scaled logarithmically, using seconds as the base (base 60 logarithms), as follows:
o None -- 0 o Seconds -- 1 o Minutes -- 2 o Hours -- 2.78 o Days -- 3.25 o Weeks -- 3.61 o Months -- 4.22 o Years -- 5 o Indefinite -- 6
Note that this measure is related to how long the medium will retain the message and is unrelated to the retention capabilities of the receiver.
How long could information potentially be stored on this medium. This measure has been scaled using the same logarithmic scale used for persistence, above.
Because storage can be a precious commodity, persistence and theoretical persistence need not be strongly related. Hence a medium that has an indefinite theoretical persistence may have a typical persistence of minutes or even seconds. Note, however, that the theoretical and typical persistence of a medium can be identical, especially at the extremes (none and indefinite), and that the theoretical persistence of a medium cannot be less than its typical persistence. As a result, persistence and theoretical persistence will be, by definition, positively correlated.
A third class of characteristics relates to to target audience to which a message is transmitted. These audience characteristics include:
How many people will a message that is transmitted via this medium typically reach? This variable is coded logarithmically as follows:
o Ones -- 1 o Tens -- 2 o Hundreds -- 3 o Thousands -- 4 o Tens of Thousands -- 5 o Hundreds of Thousands -- 6 o Millions -- 7
How many people can a message transmitted via this medium potentially reach, given the limits of the medium? This measure is scaled using the same logarithmic scale that is used for size, shown above. Because size can never be larger than potential size, but the two can match, the two measures can be expected to be positively correlated. The measures need not match, however, and they are ultimately very different measures.
How quickly will replies to messages typically be received? Note that this time estimate, which is measured from the time the first message is sent, will include both transmission times and the preparation time for the second message. This measure is scaled (with base 60 logarithms) as follows:
o Immediate -- 0 o Seconds -- 1 o Minutes -- 2 o Hours -- 2.78 o Days -- 3.25 o Weeks -- 3.61 o Months -- 4.22 o Years -- 5 o Never -- 6
Does the message typically travel to the audience, or the audience to the message. If the message travels to either the audience or its address, this measure is coded as No (0). If the audience travels to the location where the message is delivered, this variable is coded as a Yes (1).
It should be noted, in encoding this variable, that the message always does some of the job of traveling to the target, and target always does some of the job of traveling to the message. The question, however, is which piece of traveling is more important.
Is it, using a movie as an example, more important that the audience traveled to the movie theatre to see the movie or that the movie traveled to the movie theatre. Is it, using a television show as an example, more important that the television show was broadcast to the targets address, or that the target walked to the living room. In the first instance one would say that the target traveled to the message, as the movie theatre is neither the targets address nor outside the transmission process. In the second instance, one would say that the message traveled to the target, as the message is delivered, if not to the target, to the targets address.
It should be fairly easy to make the case that this set of characteristics is not comprehensive. It should, however, allow us to provide a fairly good picture of how various communications media are related to each other, and can be readily expanded to accommodate other measurable characteristics as they emerge.