Tag Archives: linguistics

Connotations are indelible

Once a connection exists between two concepts, there is no easy way to remove it. For instance if it was publicly decided tomorrow that ‘f***’ should no longer carry connotations of anything other than sweet lovemaking, it would be virtually impossible to remove the other meanings even if everyone wanted to. The connotation will always be a Schelling point for what the word might imply. Whenever the banned connotation made the most sense, people would understand it as that. Listeners know that the speaker knows they will be reminded of the connotation, and since the speaker used the word anyway, they intentionally sent the message.

This is part of why polite terms are constantly changed for concepts which are followed by unwanted negative connotations, such as terms for physically and mentally disabled people and ethnic and racial minorities. As Steven Pinker probably pointed out, the negative connotations people attach to the subject matter get attached to the word, so the word becomes derogatory and we have to get another one for when offense isn’t meant. So these words cycle much faster than other words.

You can’t even refuse to use or accept a connotation yourself. Some people insist that gender stereotypes don’t apply, are offensive, and should never be used. But if someone says to them ‘David was being a bit of a girl’, they can’t help but receive the message. They might refuse to respond, but they have no defenses to receiving. They would like to remove the association of wimpiness from the public understanding of femalehood, but they can’t even opt out themselves.

This is similar to the game mentioned in The Strategy of Conflict where two people are to privately pick a letter from several which have different payoffs to both of them without communicating. If a single suggestion is accidentally uttered, they must pick the letter spoken even if neither of them prefer it to others. It’s the only way to coordinate. If one of them managed to speak another letter, that would weaken the original Schelling point, but not destroy it. Similarly, if you make it clear that femininity suggests strength to you, you can confuse the communication somewhat by making it difficult for either the speaker or listener to guess which of the prominent possible meanings is being communicated, but you can’t destroy the existing meaning. At best the interpretation will depend on the situation, just like in the game it will depend on other cues both parties can use interpret one of the letters as more obvious.

Besides this, the more you point out things shouldn’t be associated, the more you associate them in the public mind. And that’s before taking into account that your bothering to draw attention to the issue advertises to your audience that in the common understanding the concepts are associated, so they should understand them as so in normal discussion. For instance if I argued to you that ‘capitalistic’ shouldn’t be associated with ‘immoral’, you might be persuaded, but you would also get the impression that everyone else thinks they are related. Since the latter matters and the former doesn’t, the net effect on you would be to make you understand ‘immoral’ more strongly the next time you heard someone say ‘capitalistic’.

So I would expect campaigns to attach new associations to things (such as tiny penises to speeding) are likely to be more effective than campaigns to remove associations from things (such as inferiority from femininity). Any evidence on this?