Language is perhaps the defining characteristic that makes us humans. It is the vehicle through which we think and we communicate. And while animals are able to think and communicate to some extent — after all, they possess some degree of intelligence — their ability to do so is far less sophisticated than ours. This difference stems almost entirely from the use of language, and in particular, the use of symbolic language. Unlike our bestial brethren, who might rely on chemical signals or mating calls, we are also able to harness the power of symbols, and consequently, elevate our language to new levels of expressivity and complexity.
Symbolic language works by assigning a symbol to certain important concepts. The symbols allow us to effectively manipulate concepts, both internally (thought) and externally (communication). The concepts, on the other hand, are effectively categories, to which certain things belong and certain things do not.
For example, take the symbol [cat]. Here are some things that belong to the associated concept category:
- Your friend’s cat.
- The ragdoll cat you saw in a social media video.
- Tom from Tom & Jerry or Garfield from Garfield.
Likewise, here are some things that do not belong:
- Your neighbor’s dog.
- The spotted genet you saw at the local zoo.
- The catgirl that meowed a lot in that weird anime.
Additionally, for each such category, there are two main things we can do. First, we can pick a thing to represent the category. This representative is not unique; in fact, any two representatives, whether selected by different people or by the same person at different times, might be similar but are almost never identical.
Second, we can identify certain characteristics of the things in the category. These characteristics are expressed in relation to other concepts. For instance, we can say that cats have whiskers, or that cats are harder to train than dogs are. One thing to note is that the characteristics we identify are often statistical, in the sense that they are true for most things in the category, but there are exceptions. Going back to our example, most hairless cats do not have whiskers, and cats can be potty trained to use the toilet (good luck getting a dog to do that).
Regarding the exact biological mechanisms by which our brains go about learning these concepts or the myriad philosophical debates over whether the representatives or the characteristics come first, there is an abundance of existing research, as well as many unanswered questions still left. The main takeaway here is our language is based on mapping symbols to concept categories.
What makes symbolic language so powerful is that we can string together multiple symbols to convey a very specific concept. As languages developed, not only did we expand and refine our symbol-to-concept mappings, but we also created systems by which these symbols can be strung together to create expressive and complex concepts. In linguistics, the study of these systems is syntax, and the study of the corresponding concepts is semantics. In more mundane terms, the system is grammar and the concept is meaning.
Take the example of the symbol string [the cat pushes the bottle with its paw]. The symbol string still maps to a concept category, and there are still multiple things in the category. Here are two examples: an innocent kitten nudges a bottle of milk in hopes of getting fed; a naughty cat pushes a glass bottle off the edge of the table. And while there are many other things that belong to this category, the things that do belong share a much more narrow set of representatives and characteristics.
The potential for expressivity and complexity that stringing together symbols brings us is stunning. Let us put into perspective just how powerful this is. On average, a native English speaker knows around 20,000 words, which we can say translates to around 20,000 concept categories. This is less than, say, the number of scents we can distinguish, which is around 1 trillion, so chemical signals could really communicate quite a lot of information. However, the number of reasonable sentences, say those that are grammatically correct and are no longer than 50 words, is far higher: at 100⁵⁰ or so, it far exceeds the number of atoms in the universe!
Going back to our example, an animal may have a decent idea of what a cat is, of what a bottle is, of what a paw is, and even of what the act of pushing is. However, because they do not use symbolic language, there is no way for them to process this particular conceptual category in a way that is distinct from all the other 100⁵⁰ possible conceptual categories.
This sheer number of conceptual possibilities is how symbols and categories — things that seem second-nature to us — enable our unique ability to think and communicate in an intelligent manner.
