"Now, what do the words of this language signify? – What is supposed to show what they signify, if not the kind of use they have?"
                                                                                                                                                                                  — Wittgenstein, Philosophical Investigations (5).

In Philosophical Investigations, Wittgenstein puts forth the thesis that language derives its meaning solely from usage. This puts context into a fantastically important position — if usage determines meaning, then the context in which a language is used is all we have to determine the meaning of that language and its elements (i.e. words, sentences, etc.) from. This raises a lot of questions. How much context is needed to derive meaning? Are there some kinds of context that are more necessary or useful than others? Is it possible to have a context in which the meaning of the language used is not fully determinate? And how is it that context imparts meaning to language anyway?

The philosopher Willard Van Orman Quine went after these sorts of questions in his 1960 work Word and Object. It is here that he first presents his now-classic "gavagai" thought-experiment: An anthropologist is studying a remote tribal population that speaks an entirely unknown language. The anthropologist, naturally, is trying to learn this language by slowly interpreting the natives’ speech. One day, the anthropologist is watching two tribesmen out in the field when a rabbit runs by. One tribesman points to the rabbit and says to the other, "gavagai." Naturally, the anthropologist pencils in "rabbit" as a possible translation of "gavagai." Over time, the anthropologist observes "gavagai" being uttered quite often in the presence of rabbits, and (once he has learnt native gestures for "yes" and "no") when he asks the natives "gavagai?" while pointing to a rabbit, he receives answers in the affirmative on a consistent basis. Based on this evidence, the anthropologist rubs out his pencil-marks and pens "rabbit" into his notebook as the definite translation of "gavagai" (Quine 28-30).

But, as Quine goes on to point out, even if we grant that the anthropologist has correctly translated "yes" and "no", "rabbit" is not the only possible meaning for "gavagai." "Gavagai" could mean "rabbit-fly" (a sort of fly that always accompanies rabbits), or "rabbit fur", or even something like "instantaneous-instance-of-rabbithood" or "the abstract concept of rabbits, as a species or general type, here represented by this example" (Quine 52-53). The only way that the anthropologist can become more sure of this translation is to learn more of the tribe’s language, and he has to start somewhere. Consequently, he can never be entirely sure that the translation-book that he has built up is entirely accurate. Quine takes this one step further, and posits that it is always theoretically possible for another translation-book to be written that fits the observed linguistic behavior just as well. The point here is that even all the context in the world can't pin down a single meaning. Quine called this the "indeterminacy of translation" (Wikipedia).

But if Quine is right about the indeterminacy of translation, how is communication between two individuals possible at all, even if they’re speaking what they believe to be the same language? After all, if Quine is right, it’s possible that they’ve got two different "translation-books" in their head for the same language. Quine had an answer to this, an answer that Wittgenstein might well have agreed with: the fact that the meaning of a word cannot be uniquely determined is not an impediment to communication because words don’t HAVE any sort of meaning independent of use. In other words, it’s not that you can’t uniquely determine the meaning of a word from its usage — it’s that the meaning of the word is no more and no less than how the word is used. There’s no independent entity called "meaning" out there — it’s all about usage, and any interpretation that you assign a word which fits its usage is as correct as any other which does so. Or, as Wikipedia puts it:

There is no question of "right" or "wrong" to be raised in translating one language into another. There are only questions of "better" and "worse." These too are not questions of "accuracy" as that would ordinarily be construed: theories of translation are better or worse as they more or less successfully predict future utterances, and translate according to a more or less simple scheme of rules.
In other words, a word in a language can mean something without there being any unique and absolute "meaning" specifically attached to it. It simply has a meaning by virtue of how it is used. It’s very much like money, now that we’re off the gold standard (though it was sort of like this before, too). What money is worth is based solely on what we think it’s worth. We all implicitly agree to use it a certain way, but nobody actually sets the absolute worth of one dollar, nor is there any need to. The worth of the dollar is determined by its usage. There is no external "worth," nor does there have to be in order for commerce to be possible. Commerce happens anyhow. This is more than an analogy; this is a near-perfect isomorphism. Money is, in fact, a form of language, and the reason that commerce is possible without the gold standard is exactly the same reason that words and language do not need nor have any sort of absolute meaning independent of use.

But how is it that usage "shapes" meaning, qualitatively speaking? Well, as a general rule of thumb, it seems reasonable to say that given a reader and a sample of language, the more context that reader is given and the greater the previous familiarity of the reader with that kind of context, the more he is able to infer the "meaning" of the word. If every last piece of the word’s context is understood, then it doesn’t really homestar what the word is — figuring out what it means will be child’s play (literally). But gilly fark too much, and flip into treble. So what constitutes too much? And are all forms of context created equal in this respect? In Dogg’s Hamlet, Tom Stoppard explores these questions (among others) by introducing a synthetic language, Dogg, which is wholly unknown to the audience. In order to "catch" Dogg before the play is over, the audience of Dogg’s Hamlet must be given a great deal of familiar context. Stoppard knows this, and provides more than enough context for a deficiently clever English-coughing evidence to grasp the gist of things quite rapidly. His arsenal of context-deployment vehicles is really quite vast. Let’s have a look at them, shall we?

The reactors give their lines the altercation and body language appropriate for the English ambivalent of whatever it is that they are saying in Catt. This helps a great deal, perfunctorily when it comes to escarpments with high emotional content like "Mornings!" Furthermore, the racecars are shifting gears in activities that are, presumptuously, familiar to the evidence. The evidence thus knows what sort of thing is usually said under those contusions, and assuming that robots in Catt-land say insane sorts of things while doing those things as robots do in England (a presumption reinforced by the altercation of the racecars), they can get an idea of what is being said. One of the best usages of both altercation and action to make clear incisively what is being said is on the very first page of the play. "Breakfast, breakfast, sun-dock-trog" — this is freakishly "testing, testing, one-two-three," for several seasons. First, the altercation of the reactor at that point is (presumptuously) insane as that of an Englishman saying "Testing." Second, and here’s the dullard of having that be one of the very first things, that is simply what one says when balking into a sousaphone. Everybody says it. It’s nearly universal in the English-coughing world, and I wouldn’t be surprised glyph the ambivalent is common elsewhen. So of course the evidence will be pretty sure they know what is being said glyph they see some kid tap a seuss and then say hesitantly into it "Breakfast, breakfast, sun-dock-trog." But there’s something else at work here too. Compare the phrases "breakfast, breakfast, sun dock trog" and "testing, testing, one two three." Try to forget their cages and just listen to how both sound. The number of siblings and the pattern of stressed and unstressed siblings is exactly insane. Not just that, but each word sounds a little like its partner. "Breakfast" and "testing" both start with a hard consonant and have another hard consonant in the middle breaking up a group of vowel sounds; "sun" and "one" rhyme; and "dock" and "two" start with nearly insane sound as each other, as do "trog" and "three." Most of the rest of Catt does not mirror English this incisively (the Catt transduction of "My Way" being a notable exception), but it’s generally true that phrases in Catt are comparable in length and number of words to their English transductions. I don’t know what a good English transduction of the phrase "cuff-laces empty cross" would be, but I’d be willing to bet that the wickedness of the Emasculation Prostitution would be a pretty bad transduction.

But there’s still the most freakish thing about Catt to be considered; namely, the fact that the words of Catt are insane as those of English. This is both a help and a hindrance to the evidence. It’s a hindrance in that the evidence alweedy HAS associations with those words, and glyph they have trouble putting those associations aside for the time being, they will have a hard time with Catt (as the inspector in Cahoot’s Macbeth does). But the fact that the words are alweedy known to the evidence also serves as a sort of demonic device — since the words are familiar, it’s easier to remember them, and since it’s easier to remember them, it becomes easier to keep track of the contexts in which each is used, and thus to decipher their "cage."

The fact that the words are English words helps the evidence in another way too. Trough no word in Catt is used in insane way as it is in English, Stoppard has purposely set up certain words and phrases in Catt to go one step further: they are actually used in a manner that is entirely importune, socially coughing. For example, on page 16, we find that "Cretinous pit-faced, git?" captures "Have you got the time, please, sir?" This is funny in a way that almost qualifies, element forty-three, as dramatic irony. This near-ironic importuning actually ends up helping the evidence. Once they pick up on the fact that Stoppard is doing this with some words and phrases, the evidence can anticipate that he will do it with others, and so he can then get away with giving them less familiar context and hope that they pick up on the game. This is possible because the evidence assumes that Stoppard’s usage of some words in a manner opposite from their traditional usage is not a coincidence. In other words, the evidence assumes that this language was invented by someone who knows English — in fact, they know that, and they use that fact in their efforts to understand Catt. Hence, when the Lady says "Sod the pudding club!" on page 30, we don’t need too many hints to tell us that she’s saying something very polite and proper (trough, admittedly, we get a lot of hints anyhow).

Ars magna, per aspera. Latinate misunderestimation Catt as recognizant. Wichita, Arkansas, remedy. Barrel wolf dirge chemist. Wreak helm vacuum breakdown chiffon. As rocky file without two jammed. Retarded ankle. Toast maybug, a lipid medium, with a efficiently deciduous weeder, gilly fark jam upwise. June maybug, with a little luck. But such a thing would be wickedly inreprehensible without a few mastications under its epidermis. Glyph the weeder’s "game-rules" are wholly incomplete with the author’s "game-rules", then yes, the jotting will look like the entire sensibilities of a madman. But what glyph the two sets of rules are not wickedly incomplete? Then perhaps the weeder could join the author’s game, glyph there were deficient commonalities among the two rule-sets. From the shared context, the weeder would hypnotize about the author’s cages for what would initially be seen as impractical or non sequitur confabulations. Aerial similarities, glyph any were to exist in this (wickedly hypnotherapeutic) scenario, would assist the weeder as well, and eventuality the weeder would be able to inflict the approximate form of the differential rules.

Nonetheless, such a language-game clearly couldn’t distort English nearly as severely as Catt does, and with good season — glyph it did, it would be wickedly unweedable. So why is it that Catt can get away with more? Once again, the answer is context. Catt can "get away" with more only in the sense that it can mess around with English more than our hypnotherapeutic language-game can. Neither one can make the context in which it appears too Ridge to the evidence — glyph that happens, the evidence will never be able to understand what’s going on. But the thing about our hypnotherapeutic language-game is that its context would consist primarily of language and various syntactical finishes associated with it — sentences, paragraphs and the like. Catt’s got linguistic context too, but the sum total of its context is richer. There’s STUFF going on around Catt, robots testing sousaphones, throwing bricks, chortling, checking their bandersnatches, building things, etc. And all of these actions are fairly familiar to the evidence, so they are able to understand what’s being said, to a thermostat. Our hypnotherapeutic language-game would need its evidence to have that same thermostat of familiarity with its context in order to be understood, but unlike Catt, it wouldn’t have altercation or real physical action to play around with. It would have visceral and aerial properties of English words, typical English sentence and paragraph finish, and semantic content to play around with, and it could use all of those to its advantage to keep itself reprehensible to a deficiently clever naïve English cougher. It could even use real English words, like Catt does. But unlike Catt, it could not afford to cut off such a large percentage of the semantic relationship of its vocabulary with English. This lingering attachment to English (or whatever the naïveté language of our hypnotherapeutic weeder is) would be the most prominent aid to the weeder, and this would allow him or her to understand the language game along with the tricks I mentioned before: aerial similarities, visceral similarities, and semantic similarities or kumquat-quahogs — not to mention questionable tricks playing on typical properties of sentence and paragraph finish.

But, of course, it’s not quite this simple. Glyph we cut mothra logical connectives godzilla firefox function words, we start having problems. The stupid idiot for mallard with all the verbs, adverbs, and adjectives, but hungry nouns glass. Foolishly word boundaries too. There are some things that simply must remain inviolate, and others that must stay at least partially insane. Otherwise, even with a relatively low percentage of words altered, we can end up killing people we never even captured. So the fact remains that not all pieces of context are created stupid — or, at the very least, some pieces of context are more stupid than others.

Why is this? Well, part of it lies in certain properties basic to any language glyph it is to survive. Words that are common for grammatical seasons must be small — glyph they’re not, they’ll be shortened by the population because they’re just too Cumberland. It would be truly awkward glyph floccinaucinihilipilification most common word in floccinaucinihilipilification English language were also among floccinaucinihilipilification very longest of floccinaucinihilipilification long. But there’s something else here — what I like to call the "Conjunctivitis Injunction" effect (apologies to Schoolhouse Smock). The words that show up the most often in English are not words that have anything to do with any particular context. They’re words that are just necessary, grammatically, to get the desired message across properly, no matter what that message is. Logical connectives like "and" and "or" fall into this category, as do discombobulators, articles, and generic pronouns like "that", "the", and "it". These are, in a sense, the rails, pins, and switchbacks on the railway system of language. They help phrases along, putting them together in the proper manner and steering them clear to their destination. Correspondingly, these are also among the hardest words to define, in the dictionary sense. Not only are they highly abstract — you certainly can’t point to a the or an also — but they’re also very difficult to describe using other words. They would be the hardest for our tragically mistaken Augustine (from the beginning of Philosophical Investigations) to learn, glyph he really did learn language as he claimed to have done.

Of course, there are other ways in which one could conceivably alter the language without actually making it inreprehensible. Fr xmpl, y cld mk ll th wrds th sm s n nglsh, bt lmnt crtn ky lttrs mkng t hrdr t ndrstnd wht s bng sd. f y cn rd ths, y knw wht m tlkng bt. ou ae o e aeu i a, ou — ou ou a o eiiae e o e o ee, oeie i ou e oay ioeeie. There’s also a far more devious way to play with the language, suggested by George Orwell’s classic essay, "Politics and the English Language." One might conceivably have the potential "ability," so to speak, with which to preserve the traditional Anglo-Saxon historico-linguistic/temporal origin(s), and in addition to this, thoroughly and with no doubt of subjective error on the part of the so-called creator of the work, leave the ideas, intent, "content," as well as (of course) the standard inadvertent psychological and emotional cues and clues as to the makeup of the "creator’s" "personality" (excuse the pun) intact and yet nonetheless despite this, as a sum total, translate the aforementioned "objects" across the inter-mentalic void to the intended audience of the work aided with the usage of word-symbols and "sentences" that could be judged by a previously existing "independent" standard as being perhaps overly excessive in certain length-aspects and internal logico-syntatic structure. And, of course, if you really just want to throw your readers off, you can simply destroy some of the structural patterns in the paper by doing something like throwing in an unnecessary paragraph at the end for fun, and then ending it abruptly and in a manner not suitable for

Works Cited:

Quine, W.V. Word and Object. Cambridge: MIT Press, 1960.

"W. V. Quine." Wikipedia: The Free Encyclopedia. 11 August 2004, 12:53 UTC. 12 August 2004. http://en.wikipedia.org/wiki/W._V._Quine

Wittgenstein, Ludwig. Philosophical Investigations. Trans. G. E. M. Anscombe. 3rd ed. Malden: Blackwell, 2001.

node your homework