Response: Shannon, Weaver, and Wiener

This week I am responding to “The Mathematical Theory of Communication” by Claude E Shannon and Warren Weaver and The Human Use of Human Beings, chapters 1 and 11, by Norbert Wiener.

Again, beginning with a couple informal responses.

I hate math. One of math’s favorite hobbies is to tear my brain in pieces and then stomp on them while laughing maliciously at the spurting brain juice. So I was a bit apprehensive about the Shannon and Weaver reading. But I actually found it quite interesting, and not as painful as I was afraid it would be. I was especially intrigued by the idea of “noise” as it applies to the transmission of information, and, more specifically, translation between languages. I am double majoring in English and Spanish, studying Mandarin Chinese, and planning to go onto grad school in translation studies. Languages fascinate me, as does the act of translating something said in one language into another. Noise, in this case, would be whatever discrepancies arise from structural, vocabulary, stylistic, or idiomatic differences between languages. As I believe Wiener pointed out, a translator has, broadly speaking, two options: a more direct, literal translation that maintains denotative ideas from the original language, but perhaps loses something of idiomatic meaning or style, or a broader, more liberal translation that may maintain the original tone, but that also, intentionally or unintentionally, inserts meanings and nuances that were not originally present. I would say that the job of the translator, looking at these two options (and the spectrum of choices in between), is to pick the option producing the least amount of noise.

And now, the prompt:

The Marxists we have previously read seemed concerned with the manipulation of information from a socio-political standpoint—how is information being manipulated for the political ends of the ruling class, and how can the revolutionary class counter it? Shannon, Weaver, and Wiener, seem more concerned with manipulation of information (manipulation as in “technical treatment of a given material with a particular goal in mind,” as Enzensberger defines it) from a pragmatic stance—how can we, whoever we may be, manipulate information most effectively, such that it may arrive at the receiver, be that human or machine, with the least amount of distortion occurring during the transmission process. Both articles seem concerned with the idea of information free from any sort of partisan bias. Their concern is the act of communication, especially the changes it will, should, or should not undergo in the face of the development of communicative technologies.

Shannon and Weaver focus on establishing a framework for thinking about information in quantifiable terms. I think it’s significant that they emphasize “information must not be confused with meaning” (161). This idea shifts the focus from the semantic implications of that which is communicated to the data that is transmitted.

With this framework in mind, Wiener’s writing may be understood to discuss how machines can be designed to transmit, receive, and respond to information, and therefore the role they should play in social communication. His idea of a “message” is similar to Shannon’s idea of a “signal”—a stimulus, be it verbal, visual, etc., transmitted from a transmitter to a receiver, with the end of communicating a meaning. Wiener’s concern is control, “the sending of messages which effectively change the behavior of the recipient” (8). If we put in in Shannon’s terms, control is the process of removing noise from a transmitted message: “a message can lose order spontaneously in the act of transmission, but cannot gain it” (7). In terms of machines, control becomes difficult because machines can only respond to a message in one of a predetermined set of responses. Thus, while Wiener seems enthusiastic about the potential usages of communication machines in society, he warns that it is important to determine their place. Placing machines in roles that belong to humans can be dangerous, because they cannot be appropriately controlled, but placing humans in a role that should belong to a machine, now that such machines exist, is degrading to the human.