so, even though i haven't done any research yet, i feel pretty comfortable hypothesizing that the layout of the keyboard is one factor in determining how language (text, in this case) is produced. it not only structures how we're able to create text on a computer, it has also irrevocably changed how language is changing. that is, in some circles 'teh' is a word with its own meaning. '!!!11oneeleven!!' is an expression of excitement. 'book' can mean 'cool' because they're frequently confused words in the text messaging system T-9.
the keyboard is a technology. it is a tool that helps humans accomplish a task; it extends the body in a certain way. it intervenes from without and acts upon language. but i would argue that it is also already a part of language. a historical linguist reflecting on this period of development 1000 years from now could not understand the process of language change that English (and probably all languages, to greater or lesser degrees) is undergoing wihout understanding the technology of the keyboard.
linguists have to study the configuration of the human mouth, lungs, and vocal cords in order to understand the way languages change over time and why people make the mistakes they do in processing it. is the mouth also an apparatus of language? does it intervene from without? is it the original extension of the body? is it an extension of the mind (from which language flows directly)?
is the brain an apparatus? is the way it's configured responsible for how language is configured and used? i think most linguists would say yes. how could we ever study language separately from the apparatuses used to produce it? but the brain is viewed as something integral to and inseparable from language use.
whereas the keyboard is not.
we could switch keyboards any time.
but we don't. we can't, actually. there's a better keyboard design ( the Dvorak simplified keyboard) out there. ndividual typists are perfectly capable of learning a new system quickly and effectively. the traditional QWERTY keyboard layout was intentionally designed to slow typists down back when hitting keys too fast would cause typewriters to jam. if the purpose of language were fast communication, and everyone worked toward the end of making communication efficient, we would all go buy Dvorak keyboards tomorrow and start producing different typos.
but everyone doesn't work toward the end of making communication efficient. everyone works toward the most conventional end. Convention (as defined by David Lewis in his book of the same name) is a continuation of the same behavior on the expectation that everyone else will continue the same behavior (or a complementary behavior like understanding you when you speak). how the behavior gets started in the first place doesn't actually matter. it can be established by precendent, by analogy to another situation, by fictional or second-hand analogy to a similar situation, or so on. we try to make language as useful as we can without rocking the boat too much. each individual could switch keyboards or create a private language or make up slang, but for a change to be recognized as part of the language it has to occur on a much wider scale. in fact, it's probably most likely to occur on a wide scale if it's caused by the apparatuses we all use (more or less the same way) to create language.
have i gone in a circle yet with this? just because we CHOOSE a technology for some arbitrary reason doesn't make it any less integral to language than the structures of the mouth or the brain. how a convention begins doesn't matter. if we all keep using it, it will continue to work, and if it keeps working we will continue to use it. there. there's the circle.
i think what i'd like to get at is that all parts of language (phonology, syntax, semantics, etc.) are fundamentally shaped and changed by these processes which are not logical or rational but conventional. what are the odds that structures that formed conventionally over thousands of years and accidents can be re-written in binary or reduced to logic? slim to none, i'd think. why do we formulate 'if then' statements for phonological rules and binary trees to explain syntactic structures?
maybe those are just the conventions we're born into.