Making a noise

There’s a sense of quiet after a snow fall, which brings to mind the importance of noise in everyday life. We think of noise as random and unattributable sounds. More technically, and as developed by mathematician and information theorist Claude Shannon, noise is any unstructured or random signal. Noisy signals are those with high entropy. Like the molecules of a heated liquid, or a gas, noise contains a lot of energy and random movement.

Shannon identifies noise with distortion and the introduction of errors, but he also indicates that noise introduces ambiguity into a signal, which he calls “equivocation”: “the average ambiguity of the received signal” (p.20). A noisy signal is not only full of errors, but it contains a multiplicity of signals.

Shannon wasn’t concerned with the meanings of messages, but his theory suggests that when a signal is noisy it’s not that we can’t read enough from the signal, but there are too many signals. Think of a mistuned radio, or a poor telephone signal. It is not only that we can’t hear what is being said, but that we hear too much.

We don’t only hear “it’s about to rain,” but “I’m on the train,” “I’m under strain,” and other variations. The listener eventually gets the message thanks to strategies of repetition and the human capacity to derive meaning from context. In fact, in so far as we think of communication as the passage of a signal through a conduit, it invariably involves a multiplicity of filters that eventually transform the noise into something intelligible.

So noise is a highly productive phenomenon, enabling in various ways, a property of ambiguity and of language in general. We gain great utility from language because it is ambiguous, not in spite of ambiguity. We can make use of language because words don’t have precise and fixed meanings. Words slip and slide,  affecting each other with colour and nuance.

If this is so of language then we can celebrate something similar in other areas of creation, such as designing buildings, making websites, and (I assume) composing music.

The economist Jacques Attali’s polemical book Noise: The Political Economy of Music advances provocation along these lines: “work noise, noise of man, and noise of beast. Noise bought, sold or prohibited. Nothing essential happens in the absence of noise” (p.3).


  • Attali, Jacques. 1985. Noise: The Political Economy of Music. Minneapolis: University of Minnesota Press.
  • Shannon, Claude E. 1948. A mathematical theory of communication. The Bell System Technical Journal, (27)379-423, 623-656.


  1. Pier Daniel says:

    I’m a strong believer of the importance of multiple signals in “normal” and mediated speaking communication. In the first case both speakers communicate, not only via the meaning of the message, but also because they are giving additional information by modulating their voice tone, their body language and by interacting with each other. On a phone, on the other hand, one can’t convey additional feelings and information from a visual perspective; but nevertheless, many non-voluntary sound cues are given. Think for example when there is a cell phone communication. We can tell if the other speaker is moving, where he is, if he is relaxed or on a hurry, if he needs to tell a secret, if he wants to be vague etc.
    Also, in mediated communication one easily forgets about the fact that he is projecting himself, therefore giving us a lot of information unconsciously, because he is not seeing the body language and face feedback of the listener. Moreover he can’t be too sure about what the other speaker is hearing and really understanding. Anyhow, he is telling us more than just the meaning of words, even if he doesn’t want too. This also means that he can be more easily misunderstood and that it might be more difficult to say if he is being honest about what he is saying. On the other hand one can’t be sure if he fooled his listener when he is saying a lie. So, mediated communication makes speech more ambiguous, but also more endearing. This is why I think that we are often more interested in this voluntary and involuntary “noise” of a conversation rather than on the actual information the other speaker is giving us via the meaning of the words. In fact, our real human nature is always projected when we talk, but is hidden behind the meaning of words, which might seem more important. However we attribute the importance of word meanings to how much we can glimpse of the speaker’s emotions when he is talking about them. I guess that it is for this reason that randomness and noise are often related to human nature. In conclusion, it would be interesting to investigate how much we consciously know about how we project ourselves when we are on the phone or using other means of communication. Moreover, as listeners, do we perceive phone communication more “noisy”, as in more intriguing, than normal speech?

  2. Xi Ge says:

    If noise are signals, then it is there, with its hidden meaning, no matter whether there is some one trying to figure out its meaning. I think the idea here is to pay attention to things otherwise neglected. However, as long as we begin to dig meaning from the “noise”, can it still be regarded as noise? Donna Haraway uses “noise” parallel with “heat” as one of her modern and postmodern binary concepts. To her, both represent the inefficiencies within a productive process. In industrial ages, heat is the outcome of wasted input, while in informational productive process, noise representative the waste as “whatever isn’t signal”. I find Haraway’s view on noise more convincing to me. What I find tricky here is the rhetoric definition or boundary of “noise”. Sometimes when people try to theorize something, we may make appropriation unconsciously, ending up with signifying other or new concept.

  3. kimo says:

    A friend of mine used to show me a semi-finished painting by herself, and ask me “is it too noisy”? I guess it is one of the way how “noise” should have far more than one definition apart from its HZ properties. The idea that treating noise as a signal is fancy. The noise conveys information through different channel under different background, sometime the noise itself would be of none meaning when we remove its context. I think the definition from this signal angle highlights more on the functional rather than physical propitious of noise, which is why it is special and meaningful.

Leave a Reply