While we can acknowledge that there is a connection between a word (sign) and its referent, the connection isn't the direct type that most people think it is. This makes sense because, as we realized as we beagan this study of the nature of language, the connection bewteen a word and its referent is really an arbitrary one that exists only as long as the users of a language agree to that connection.

This leads us to consider the role of the users of a language in the assignment of meaning to a word. It turns out that the users of a language are far more important in determining the meaning of a word than the referent of the word. Now we will explore in greater depth how the users create the arbitrary connections that they do.

Activity: Connotation

Click Connotation to open the activity.

Connotation

Connotation helps a great deal when trying to explain how meanings are attached to words. We can define connotation as the attempt to determine meaning as a set of associations mapped onto a word, allowing individual variations in meaning:

Skinny to a fashion model might be an attractive or positive atttribute

Skinny to a weightlifter might be an unattractive or negative attribute.

Winter to a Minnesotan is the season when it snows.

Winter to a Seattlite is the season when it rains.

Winter to a person from Phoenix is the season when temperatures are tolerable.

Now we can begin to see how the relationship of a word to its user is an important piece in understanding meaning in language. While a word's relationship to its referent disregards the ideas of any individual speaker, the relationship of a word to its user adds something to the meanings of words that is completely left out if we only look at the connection of a word to a referent.

While it might seem as if we have completely understood how words get their meanings, we still have one puzzle left to explain, and we will see the puzzle if we return to the Pictionary activity. Connotations can explain why a person from Minnesota might select a picture of a snowflake and a person from Seattle might select a picture of rain when given the word winter, but we still don't have an explanation of why most people select a picture of a robin or sparrow for the word bird and not a picture of a flamingo or penguin.

Generally, when presented with this puzzle, most will say that the meaning depends on the experiences a person has. That is, they point to the associations mapped onto a word. That, is, they use connotation as the entire explanation, arguing in this way:

People in Minnesota experience snow in winter, so they associate snow with winter. That is, Minnesotans perceive winter as a season when snow falls because they experience snow falling in winter. People from Phoenix don't perceive winter as a season of snow falling because they don't experience snow falling in winter.

But, what if we change the example? Can we use the same line of thinking to explain why most people select a picture of a straight-backed chair for chair instead of an armchair? Why is one kind of chair slected more than another kind of chair? If we use the argument just given, we must argue like this:

Most people experience straight-backed chairs when they sit down, so they associate straight-backed chairs with the word chair.

But does this argument really make sense? Think about each of the chairs you have sat on today. Think about students in school and the kinds of chairs they usually sit in. Think about office workers and the kinds of chairs they usually sit in. Can we really say that people have more experience with straight-backed chairs than they do with office chairs? The argument begins to fall apart when we start considering all words and try to explain how all words get their meanings. A more convincing argument is made by a psychologist in what is called the prototype theory.

Prototype Theory

Psychologist Eleanor Rosch conducted a study to learn how people define words by having them make judgments about whether a word would be applied to an image or not. Rosch learned that instead of a simple "Yes this word is used for that thing," or "No, this word isn't used for this thing," that people made judgments on a sliding scale. When presented with the word bird and a picture of a sparrow, people readily judged the word as applicable to the image. When presented with the word bird and a picture of a penguin, people judged the word as applicable, but less readily; they took more time to make the judgment. When presented with chair, the same thing happened, with a straight-backed chair being readily judged and other chairs taking more time. And again with the word furniture, the same thing happened, with certain pieces of furniture being readily judged as furniture and other pieces taking more time.

This study indicates that people's perceptions are important, but so are the essential features of a referent. The thing is, the essential features are parts of mental images in people's minds, not the exact features of the referents themselves. That is, words don't map onto real objects of the world, they map onto the mental images people hold in their minds. The development of meaning really progresses like this:

I experience many objects in the real world, and I develop a mental image encompassing the essential features based on my experience with those objects. A word is then attached to the mental images I hold and subsequently applied to objects in the real world as I encounter them. When the real objects in the world match the mental image I hold in my mind, my prototype, I readily apply a word to that object. When an object is close to, but not exactly like, the mental image I hold in my mind, I must take some time to judge whether it is close enough to use the same word I use for my mental image.

Thus, a robin is a more typical bird than a flamingo, and a flamingo is a more typical bird than a penguin. A straight-backed chair is closer to my mental image of a chair than an office chair, so in Pictionary, I draw a straight-backed chair, not an office chair, even though I spend much more time in real life in an office chair than in a straight-backed chair.

Many readers will still want to hang on to the argument that meaning is tied simply to our real world experiences and not to a mental image, but think about the furniture. Is it really logical to accept the claim, “I have more experience with a high backed chair than with an armchair,” or “I have more experience with sofas than lamps”? No, it seems that we hold a mental image of a word’s meaning in our heads, and that mental image is the referent for words.

Prototype theory links all the three kinds of word connections together. It explains how a word is related to the users of a word (they carry the mental image that is the true referent of a word and they are the judges of whether the word can be applied to something in the real world), and it explains how words are related to the real word referents (they are the basis of the mental image that words map to), and it explains how words are related to each other (the user judges a word's applicability to a real world object and either selects a word or deselects it by selecting another word).

In addition, Prototype Theory can explain how we can have words like unicorn, for which there is no real world object. If words map onto mental images and not onto real world objects, then word meaning comes from mental images, ideas, and abstract thoughts, not from the real world. If we can imagine it (notice the root image in the word imagine?), we can put a word on it.

If we must have real world objects in order to have words, then we can't talk about things that don't exist. And we talk about things that don't exist all the time. We talk about the future, and the past, and other things that aren't around us.

Remember what we said about language being symbolic and not indexical?

An indexical sign is one that points to something else that it coexists with.
If words referred to real world referents, then they would point to them, and the basis of language would be indexical signs.

A symbolic sign, however, is one that is arbitrarily connected to its referent, and it needn't coexist with it.

Because words map to mental images, which lack existence in the real world as physical objects, words are symbolic signs. Words do have an indexical aspect, too, but it is through pointing to the mental images, not through pointing to real world referents.

Although words map onto mental images, some of the mental images we carry around are more clear-cut than others. In addition, meanings can be varied according to individual experience, as connotation suggests. Because concepts are fuzzy, so are words.

This is why word meanings can change over time, creating historical divisions, and from group to group, creating sociolinguistic divisions. The words least likely to change are those that have the least image-based detail attached to them: the functional word categories. Pronunciation of the functional words is the primary basis for changes in them over time, not reconcpetualization of their meanings.

The development of mental images by the users of language also accounts for the linguists' arguments that language isn't passed down from one generation to the next so much as it is recreated by members of the next generation. That is, first language acquisition consists of the development of mental images and the mapping of words onto those images (as Vygotsky suggested in the intertwining of thought and language). The words for a new generation may be heard from the older generation of speakers, but the mental images are all their own, and the mapping is unlikely to take place in exactly the same way from one generation to the next.