Consider, if you will, the dance honeybees use to communicate the location and richness of a pollen source. The length of their dance indicates the distance to the source. The direction in which they perform the steps indicates the direction of the source. The speed of their dance indicates the quantity of pollen available. The bee’s dance is certainly a form of communication, but it is not a language.
Now consider American Sign Language (ASL). Until very recently, it was not considered to be a language, either. In the 1960s, Charles Hockett developed a list of criteria which could be used to determine whether a given communication system is a language. It was oral-centric, designed a priori to exclude ASL and other signed languages. If, however, one were to eliminate the mode bias from his list, one would find that ASL meets all of the criteria he set forth for qualification as a language. In this post, I would like to discuss his criteria and to look at ASL as a case study for application of the criteria. Please note that, as I will discuss in the next post, there are actually many different sign languages used around the world. I am limiting my discussion to ASL both because it is the most studied sign language and it is also the one that I know personally. (I used to be pretty fluent in it; unfortunately, I have forgotten a lot since I haven’t had the opportunity to use it much in the past decade or so.)
Non-language communication systems may meet some of the following criteria, but only languages will meet all of them.
Hockett’s first three criteria are mode specific and would a priori exclude ASL as being a language. These criteria are no longer considered valid for determining whether a system is a language. The first criterion claims that languages use the vocal tract to produce sounds to represent meaning. Sign languages, clearly, transmit messages via the visual rather than the aural mode. The next two criteria really are property of all sounds, not just linguistic ones: The signal will be transmitted omnidirectionally but the receiver will be able to determine the direction whence it came, and the signal will fade quickly after it is made.
With languages, but not necessarily with other communication systems, there will be at least one sender and one receiver of the signal (soliloquies and the like excepted, of course), and the individuals involved can exchange roles – senders can become receivers, and vice versa. Furthermore, senders are able to perceive the signal they are sending.
The next set of criteria involves meaning. First, specific signals (words, signs) are associated with specific meanings. For the most part, these signal-meaning pairings are arbitrary. There is nothing inherent in the sounds of the word “dog” to make one think of a canine; and indeed, different spoken languages have different words to refer to canines. Moreover, a given word might be associated with different meanings in different dialects or languages. I’ll take as my example a word which Corrina (Female Hawk) has used in her fanfic Aussie Rules; in some parts of the world, “rubber” means a condom, in other parts, an eraser. Similarly, different sign languages associate different signs with a given meaning, and a given sign might have different meanings in different places. The sign that means “shoes” in most of the United States means “homosexual” in New York City. (At least that was the case several years ago, when I had been an active signer.) This is not to say that *all* words/signs are completely arbitrary; there is onomatopoeia in both signed and spoken languages.
Lexemes (words/signs) are made up of smaller, discrete articulatory units. If one substitutes one unit for another, a change in meaning will result. Replace the first sound of “cat” with the “b” sound, and you have a different word – “bat”. (A side note to Julie and to others who may already be versed in phonetics and phonology: I don’t wish to get side-tracked with a lengthy discussion of phonetics here; therefore, I am purposely glossing over some details which I might discuss in later posts.) Signs, too, are decomposable into discrete units. William Stokoe, the father of sign language linguistics, was the first to recognize this. He described signs as having a few distinct parameters, and a finite number of possible values for each parameter. In his analysis, signs could be decomposed into hand shapes, movements and orientation, and location. Change the value of any of these, and the meaning of the sign changes. For example, the ASL signs UGLY and DRY are articulated identically except for their location. (For ease of notations, signs are often referred to in writing by capitalizing their nearest English counterpart.) The signs PLEASE and SORRY vary only by their hand shape. The signs NICE and SCHOOL vary only in movement. Stokoe’s analysis of ASL has since been determined to be inadequate to describe ASL accurately; for example, it completely ignores facial expressions. Far more sophisticated analyses have since been conducted. Nevertheless, it was groundbreaking at the time and was instrumental to having sign languages recognized as true languages.
Not only are lexemes composed of smaller units, there are rules for such composition. Although “thap” is not an English word, it could be. In contrast, speakers of English recognize that “ngap” is not a possible word in English. (Unless, of course, it happened to be borrowed directly from another language which had different rules for assembling words.) Similarly, ASL signers know that no ASL signs would use a hand shape in which the ring and little fingers are extended and the other three fingers curled into a fist.
In a similar fashion, sentences are also composed of smaller units (lexemes), and each language has rules for such assembly. English speakers realize that although, “Superman caught the fleeing villains” is a sentence, *”Fleeing the caught Superman villains” is not, despite the fact that both strings contain the same words. ASL has its own rules for assembling sentences. Sign order is much freer in ASL than word order is in English; but, like Latin, ASL uses markers to specify information such as the subject vs. the object of the sentence. In the case of ASL, the markers are usually spatial. It also, incidentally, can convey syntactic information non-manually. In ASL, the hands would perform identical actions to convey each of the following ideas:
- Clark loves Lois.
- Clark doesn’t love Lois.
- Does Clark love Lois?
In the first sentence, the face would be neutral with possibly a brief up-and-down nod for minor emphasis. In the second sentence, the head would be shaken side-to-side. (If one wished to be emphatic that Clark REALLY does (or does not) love Lois, then an additional sign would be used.) To indicate a question that can be answered with a simple “yes” or “no,” the signer looks directly at the addressee, tilts his/her head forward, and raises his/her eyebrows. (As an aside, a question that needs more than a simple yes/no answer – a WH-question – has identical non-manual movements except that the brows are lowered. Rhetorical questions are made by with a backward tilt to the head and raised brows.)
Hockett’s displacement criterion specifies that languages can be used to discuss things which are displaced in both time and space. Indeed, they can discuss things that never existed at all. Any language, including ASL, would be capable of communicating the meanings conveyed by the following sentences:
- The Superman character was created many years ago.
- The Lois & Clark TV show will probably be remain popular for many years to come.
- Kal-El’s spaceship was far enough away from Krypton to be safe from the planet’s explosion.
In fact, any language is capable of conveying any message conveyable in another language. Of course, some meanings would require considerably lengthier utterances to convey in some languages than in others. As Ann has pointed out in some of her posts, language and culture are closely tied together. A meaning conveyed in a single word/sign in one language might require many words/signs to convey in another language. For example, some languages don’t have words for a particular technology simply because the technology is not known to the people who speak the language. In order to convey a message involving such technology, one would first have to demonstrate the technology or else provide a potentially lengthy description of it. And nuances inherent to a given word/sign might have to be stated explicitly in another language. But the overall meaning should be conveyable in all languages.
All languages are capable of using a finite number of words/signs to communicate a potentially infinite number of meanings. It is possible in any language to produce a never-before-made utterance and have it be intelligible to others who use the language. I would venture that what I am about to type has never been said before, and yet you who read it will understand it:
When Clark entered his apartment, he was surprised to find a small Kryptonian elephant hovering in front of his television set and using its purple and green striped trunk to eat popcorn from a bowl.
The final four criteria for determining whether a communication system is a language are self-explanatory, and are met by ASL as well as by spoken languages:
Language signals are intended for communication.
Languages can be used to convey false messages (lies) as well as truthful ones.
Languages are culturally transmitted.
Languages can be used to discuss languages.
Let us return to the bees’ dance. We can see that it does not meet the criteria for being a language: The dance could not be used to convey the fact that there had been a wonderful pollen source available last year, for example. In fact, the dance is limited to a single topic (pollen). It cannot be used to lie; bees do not use the dance to trick bees into thinking there is a pollen source where there is, in fact, none. (That, of course, may say more about bees themselves than about their dance.) The dance cannot be used to discuss bees’ dances. It is my understanding that the dance appears to be innate rather than culturally transmitted.
In contrast, ASL meets all of the non-mode-specific criteria for languages.
My next post will discuss sign languages in more detail.
