[...] we can strip off all grammatical clues to sentence structure, all affixes and prepositions, and yet still achieve communication. Thus restricte… - Colin Cherry

"Information" in most, if not all, of its connotations seems to rest upon the notion of selective power. The Shannon theory regards the information source, in emitting the signals (signs), as exerting a selective power upon the ensemble of messages. In the Carnap-Bar-Hillel semantic theory, the information content of statements relates to the selective power they exert upon ensembles of states. Again, at its pragmatic level, in true communicative situations [...] a source of information has a certain value to a recipient, where "value" may be regarded as a "selective power." Gabor, for example, observes that what people value in a source of information (i.e., what they are prepared to pay for) depends upon its exclusiveness and prediction power; he cites instances of a newspaper editor hoping for a "scoop" and a racegoer receiving information from a tipster. "Exclusiveness" here implies the selecting of that one particular recipient out of the population, while the "prediction" value of information rests upon the power it gives to the recipient to select his future action, out of the whole range of prior uncertainty as to what action to take. Again, signs have the power to select responses in people, such responses depending upon a totality of conditions. Human communication channels consist of individuals in conversation, or in various forms of social intercourse. Each individual and each conversation is unique; different people react to signs in different ways, depending each upon their own past experiences and upon the environment at the time. It is such variations, such differences, which gives rise to the principal problems in the study of human communication. (p. 244-5)

The dictionary definition of communication [...] includes the communication of goods and supplies. [...] But transport of goods is not communication in the sense we are adopting here, and does not raise the same subtle and difficult questions. What "goods" do we exchange when we send messages to one another?

Physically, we transmit signals or signs -- audible, visual, tactual. But the mere transmission and reception of a physical signal does not constitute communication. A sign, if it is perceived by the recipient, has the potential for selecting responses in him. Physically, when we communicate, we make noises with our mouths, or gesticulate, or exhibit some token or icon, and these physical signals set up a response behavior.