The Language of Machines :: Technology Communication Essays

The Language of Machines Computers are language machines. By saying this I mean both that language processing is a valuable metaphor for understanding computer computation and that, in a fundamental way, computer computation is language processing; no more, no less. The language understood by a modern computer when it first comes off the assembly line is quite simple. The alphabet of this language consists of two letters, 0 and 1 (or a and b or any other two characters, it doesn't matter), which is stored internally as two intensities of an electrical signal (either high or low). The grammar of this language has two rules: (1) Sentences consist of one word and (2) Words are all of a single specified length (probably either 16 or 32 characters). This computer knows in two ways. It knows what every word in the language means (i.e., what action to perform upon reading that word, information which is stored in the design of the processor), and it knows all of the words it has stored in memory. Each time a comput er reads a sentence (executes a command), a change results in memory, dependent on what the sentence says and what is already in memory. Modern computers are Turing machines (named after the British mathematician Alan Turing), which means that they are language machines which can simulate other language machines. In other words, given a special type of text to read (a program), a Turing machine that understands the simple language described above (for example) can act as if it understands a much more complicated language. This is why modern computer keyboards have more than just 0's and 1's on them. A modern computer comes complete with many virtual computers built on top of it, so to speak, enabling the computer to understand much more complex (although mathematically equivalent) higher-level languages. These are mathematical languages, of course; they have much more rigid structure and precise meaning than natural languages. They lack in many ways what Derrida calls "play." But must they? Is there an intrinsic fundamental difference between mathematical and natural languages, or is the difference instead that we hav e more control over mathematical languages because we know their rules and can understand the system in which they work, while with natural languages we know neither, because we are not in conscious control of their creation and we can not fully grasp how they operate in society and in our heads?