18 Apr Irrational Language Rules
Now let’s narrow the focus from communication to language. To avoid confusion, let it be clear from the outset that this discussion is not concerned with computer-programming languages, but “natural languages.” Computer languages have built-in or “prescriptive” regularity called a formal syntax. They are easy to describe, and they are obedient to the prescribed sets of rules that govern their functionality. Thus, if we want to learn all about a computer language, we simply buy the book. So many of our words can have different meanings in different contexts, that human languages can seem downright irrational. Ambiguity rules.
|Understanding Context Cross-Reference|
|Click on these Links to other posts and glossary/bibliography references|
|Prior Post||Next Post|
|Learning by Repetition||Learning by Doing or Active Osmosis|
|communication language||Bhatt 2011 – Linguistic Phenomena|
|syntax rules||References on Rules Pinker 1999|
|ambiguity interpretation||Hirst 1987|
There is no similar book on any human language that will tell us, or a computer, everything we need to know to interpret it fully and accurately. We can buy a book on chess, encode the knowledge, and build a persuasively intelligent chess-playing computer. What is it about language that makes it so unfathomable? Sure, we can buy dictionaries and grammar books, but they are not enough. There is more to language than words and syntax.
Perhaps the toughest thing about encoding knowledge about language is the intuitive part buried deep in the psyche of all speakers: common sense and world knowledge. Each game of chess consists of a finite number of pieces, each with a finite range of legal moves. Since the problem space, rules, and objectives are known, chess expertise can be encoded and the game played automatically. For language, the objectives are to produce and comprehend messages. The problem space consists of all the words in the lexicon and all the possible variations of intent they can express. What are the standard rules, and what irrational language rules make automating language understanding so complex?
Section1 described how separate areas of the brain handle different language phenomena, including the expressive (Broca’s) and receptive (Wernicke’s) areas which specialize in speech and hearing. In fact, language phenomena can be subdivided much further. When we break it down below the most basic levels, however, we extend beyond the areas of physiological specialization and into the areas of abstract cognitive behaviors. Consider the art that can go into an act of speech. Say you are in a country where it is common to bargain over price, and your objective is to persuade a vendor to sell you his wares at a low one. To make it interesting, assume that you are a multilingual tourist in a country that is less economically established than your own (no one said this would be a noble speech act). What would you do? Would you select a language, choose some words, and spill your guts? Would you hide your ability to understand the vendor’s tongue, or use his language to your advantage? Would you use more than one language in a sentence? How would you portray yourself? Would you empathize? Language can be used in complex ways, almost as a game. To make it more interesting, assume that you and your husband don’t agree on which item to buy. Now we have multiple opportunities to work out agreements. Negotiation is a prime example of a situation in which multiple language, and behavioral (the interconnections cannot be ignored), phenomena occur.
|interpretation||translation||many many more…|
The many phenomena above give a flavor of some of the constraints with which we deal in the course of our everyday communications. These same constraints will need to be part of the language understanding system that is able to consistently and correctly complete your sentences for you.
|Click below to look in each Understanding Context section|
|4||Perception and Cognition||5||Fuzzy Logic||6||Language and Dialog||7||Cybernetic Models|
|8||Apps and Processes||9||The End of Code||Glossary||Bibliography|