When I was experimenting with language evolution, my models (and those of others) made some basic assumptions that were so simple they seemed axiomatic: signals specialized into words, and evolved through syntax to still more specific expressions. The experimental representations of these elements varied, but words always preceded syntax; symbols came before structure.
At this time I did not own a cat.
I was also taking a seminar in the philosophy of computing that looked at a lot of alternative concepts of programs, data, and computing. The idea of the situated agent was strong then: a computing entity that was not a mathematical device but a physical one that could sense and act in its environment. It would seem that having to cope with input from around it would confuse a computer rather than make it smarter, but surprisingly this was not always the case. The world has persistence, and hence can store things, and it has spatial relations that create structure. This meant that an agent could use less memory and simpler programs to accomplish the same task, the way we leave a book by the door in order to help accomplish getting it to the library.
It didn't occur to me then that this idea could be extended to language: that the world could hold one's words, and that language could leapfrog the stage of word evolution and advance directly to creating syntactic structure. Then I got a cat.
I have a wordless cat. With the exception of two or three nuances of "meow", she doesn't have sounds for things. But she doesn't have to, because the world is already full of things. Instead, she has paws and sentence structure.
Here are some cat sentences at work.
There is more at work here than cat cuteness; there is the pattern of language. Notice that:
These aren't steps of genius on the part of the cat, and they're trivial for a dog. They aren't complex, but they are the recognizable and unique characteristics of language.
The last point is controversial. It's common to hold that a word is a word precisely because it is disconnected from the object it represents. This is an accurate description of human speech, but a poor one of sign language, and just wrong when it comes to early writing, both of which depend heavily on mimicry and direct representation. And I think it's just this manner of finding words in the world that makes a bridge from words to syntax that I'd been overlooking before. After all, human babies spend a huge portion of their early life immobile, yet making speech. They have little choice but to create words as substitutes for actual things. Cats have no such opportunity. Long before babies, cats walk and paw and get straight to the business of word order, grabbing their words when they need them.
Producing sentences is only half of language; understanding them is the other. My cat seems to have real difficulties with this (we are still working on "No"), but a new study out in Science demonstrates a large vocabulary of English words (up to 200) in a dog, acquired in much the same way that children do. These are genuine, arbitrary sounds that stand for objects, first-class nouns if ever there were any. The lesson for the would-be student of language evolution is, I think, that the strict ladder of words to word sequences to sentence syntax is too idealized. The world that we live and act in breaks that progression and makes possible the parallel development of all the parts of language, objects, patterns, and abstractions together.
Read about Rico, the dog that learns words, and Alex, my other favorite linguistic animal
Consider this trend. The Queen Mary 2 was launched recently, ancestor of the great ocean liners that once crisscrossed the Atlantic. And Space Ship One is soon to lift off on its maiden voyage, carrying a person to the edge of space and back. Two great ships that are writing the future of travel in their respective realms. Two ships that are going nowhere in particular.
And here's a roster of ships now sealed in the past: the Normandie, last of the ocean liners, burned at her dock in 1942. The Concorde, put to pasture from its Atlantic crossings last year. The major airlines, losing money to the last plane in the face of smaller and cheaper discount flyers.
Great ships once went places, and grandly. They turned a necessary trip into a voyage. The idea of voyage seems to have finally expired, though, for if there is a new prevailing idea, it is the idea of the destination: to get there quickly, and to notice the intervening miles as little as possible.
This is, perhaps, part of the great telescoping of things that is one with the Internet and the global village. More and more it's not necessary to travel in order to get places. This makes the idea of travel less and less a part of us. It has been replaced by the concept of the commute, which is travel that takes us from where we'd like to be to where we'd rather not be.
And our two newest ships don't alter this trend. They do not, as did the Concorde or the first Queen Mary, serve the basic purpose of ships, which is to carry things from one place to another. The newest Queen Mary is a cruise ship, whose purpose is to start and leave from the same place, albeit slowly. Space Ship One will go up, but not up to anything. Its great goal is to get back home as well. Neither is a voyage in a real sense, which depends on its destinations to give it shape.
It is pleasing to spend little and arrive soon, or, via the Internet, to spend nothing and publish instantly. But I still miss the rich delay. I may be the only person who sometime reads e-mail with all its transmission headers included, a long and exotic list of worldwide servers that have been the message's millisecond ports-of-call. They are the last remnant of the postmark and the foreign stamp — the telltales of voyage.