24

When I dust my dresser, I am removing something from the dresser.

When I dust a cake, I am adding something to the cake.

How did this happen? I've heard of words taking different meanings during the transition between British English and American English. I understand how sanction came to mean multiple things. Some words that are auto-antonyms are clear through context, but consider these:

I dusted the dresser with my Swiffer.

I dusted the dresser with a bag of cocaine confetti.

USER_8675309
  • 1,625
  • 3
    It's absolutely normal, in fact ubiquitous, in English that words have many meanings. There are any number of examples of the type you give; it's so common it's not worth mentioning. – Fattie Jul 11 '16 at 17:34
  • 4
    In case anyway wants to get more confused, to dust can also mean to defeat or to kill as in turn [the enemy] to dust, and it can also mean to leave behind or to outrun/outrace as in leave [the other person] in the dust [that is kicked up and left behind as you speed away from them]. Hooray, English! – Todd Wilcox Jul 11 '16 at 19:20
  • 7
    Sometimes when we verb nouns, we employ handy prefixes. Sometimes we don't.

    Consider the pair "encrypt" and "decrypt" for adding and removing encryption. One might imagine the pair of verbs "endust" and "dedust". Ironically, the commercial product "Endust" is intended to dedust things.

    – Monty Harder Jul 11 '16 at 19:41
  • You forget the 3rd usage: When you dust a bad guy, you neither add nor remove but change his/its nature -- from (un)living to (eventual) dust/compost. – Brock Adams Jul 12 '16 at 04:49
  • 1
    @BrockAdams Sure, I mean we could keep going with the slight euphemisms. When I can dust someone while driving by leaving them behind when I speed off. – USER_8675309 Jul 12 '16 at 11:33
  • @Mari-LouA Vaguely. I read this one before I asked the question. To me, this question wasn't so much about dust specifically, as the verbification of nouns which then come to have two opposite meanings. – USER_8675309 Jul 12 '16 at 12:19
  • Just going to leave this here: https://smile.amazon.com/Amelia-Bedelia-Chapter-Book-Box/dp/0062334204/ref=sr_1_1?ie=UTF8&qid=1468777839&sr=8-1&keywords=amelia+bedelia – Jonathan Cast Jul 17 '16 at 17:52

4 Answers4

56

The reason it's even possible for to dust to have two contradictory meanings is simply that there's no single fixed relationship between the underlying noun and one particular activity closely associated with it (logically, one might be just as likely to apply dust as to remove it).

Much the same applies to some other nouns used as verbs...

seed (seed a lawn = apply lawnseed, seed a tomato = remove the seeds)
stone (stone a heathen = throw stones at him, stone a cherry = remove the stone)
screen (screen a movie = present on a screen, screen from view = hide behind a screen)
etc., etc. (more examples on mentalfloss.com)

Thus, to the extent that there's a reason for the dual use, it's because in the real world there are multiple relatively common actions associated with the base noun, and in most cases context makes the relevant meaning obvious, so this doesn't cause problems for native speakers.

FumbleFingers
  • 140,184
  • 45
  • 294
  • 517
  • 8
    @ThunderGuppy: I must be honest. I thought of *seed* all by myself, but it was googling *dust seed verb opposite* that led me to a page titled 14 Words That Are Their Own Opposites where I found the others. Bizarrely though, since the relevant search is still there on my browser "Back" mutton, I rechecked and now notice *25 Words That Are Their Own Opposites* on the same site. I can't quite see how I missed that first time around! :) – FumbleFingers Jul 11 '16 at 17:13
  • 9
    Here's a freshman linguistics puzzle dealing with verbing nouns. There are a lot of different kinds of verbing, but the puzzle deals with the two mentioned here. The kind that refers to depriving something of something else is called the Privative sense; the kind that refers to providing something with something else is called the Provisional sense. – John Lawler Jul 11 '16 at 18:21
  • 7
    @John: Oh, that's very interesting. I never thought it through fully when I was posting the answer, but now you've flagged it up I suppose it's only natural that the two main actions one might associate with a "noun turned into verb" would largely be identifiable as "[de]privation" and "provision" (remove or add "noun"). Perhaps such usages aren't as complicated or arbitrary as they seem at first. – FumbleFingers Jul 11 '16 at 18:29
  • 5
    @FumbleFingers I just +1'd your comment for '"Back" mutton'. – Dewi Morgan Jul 11 '16 at 18:44
  • @FF: There's also Instrumental (horsewhip somebody), Manner (whipsaw somebody), Means (railroad somebody), and plenty of metaphoric and idiomatic uses that deal with all kinds of relations (ladder one's stockings, nose someone out for a position, foot the bill). It's like you said -- the context chooses the interpretation. This is part of what the slogan "Semantics Underspecifies" means. – John Lawler Jul 11 '16 at 18:59
  • @John: Presumably that's part of what makes it so difficult to bring our massively enhanced computing power to bear on the field of linguistics. Especially semantics, since most of the "meaning" of most utterances can't really be gleaned from the text itself - it's "out there", in the form of pragmatics & context. You linguists may no longer be ploughing such a lonely furrow, but you sure picked a tough and tortuous one! – FumbleFingers Jul 11 '16 at 21:49
  • 3
    @FF: Oh, but we can bring it to bear. There's been a lot of work done on contextual analysis; for instance FrameNet, which was developed by Chuck Fillmore, who dreamed up the concept of conceptual frames and hooked them up with metaphor theory. Framenet and Wordnet are both available for anybody, having been developed by US government grants. And they're very widely used. – John Lawler Jul 11 '16 at 21:58
  • @John: I suspect it'll be some time before initiatives like that lead to chatbots that never make daft mistakes. A couple of weeks ago I watched a talk on AI (by Nick Bostrom, I think - love that guy!) where pragmatics would imply his example chatbot simultaneously lived in America and the US, and was simultaneously a happily-married man and a woman looking for romance. I know that one was just based on statistically likely responses at the utterance level, but machine understanding of language still seems some way off even for smarter ("true AI") systems. – FumbleFingers Jul 11 '16 at 22:38
  • Well, there's a certain self-reinforcing regression involved. "True AI" is always on the horizon; if we can already do it, that's not really AI. Whenever AI advances -- by centimeters, like all sciences -- it gets heralded and blown up, and then you never hear about that particular thing again. Meanwhile, downstream, these are making a whole lot of things work better, but it isn't AI, it's "updates". There will always be errors in transcription or translation; there always have been. Why worry about labels? – John Lawler Jul 11 '16 at 22:52
  • 1
    @John: Some labels are good. I really feel I've learned something today by becoming consciously aware of Privative, Provisional, Instrumental, Manner, Means as major subtypes within "verbification" (ta muchly! :). And I'd love to imagine that "big data analysis" will soon allow us to autogenerate easily-absorbed summaries of relative prevalence, preferred context, etc. for all those types and more. I'm no "ontologist", but I feel that knowing more about how language works could teach all of us more about both what the world is, and what it "means". – FumbleFingers Jul 11 '16 at 23:11
  • 1
    Haha, "the ratings board must screen the movie before the cinemas can screen the movie." – detly Jul 12 '16 at 06:36
  • 1
    @JohnLawler Oh my gosh why has no one ever told me of FrameNet or WodNet o.0 – USER_8675309 Jul 12 '16 at 18:06
13

It happened the same way all language happens: somebody needed a way to say something, and found a way. In this case, different people (well, maybe the same person, but probably different people) needed ways to say "put dust on" and "remove dust" in different contexts, and both used the obvious (in English) method of verbing a noun.

It probably never occurred to either of them that it might be ambiguous, because in context, it wasn't. In most contexts, it still isn't.

Colin Fine
  • 77,173
  • 3
    "...the obvious method of verbing a noun." Ah, using the method while describing the method. Well done. – CivFan Jul 12 '16 at 16:54
5

You can tell the meaning of to dust only by the context where it is used. What do you do with Swiffer? You clean the dust. What do you do to a dresser with a bag of confetti? You can't remove dust with it. That's the way English developed over the years with contronyms because you don't need to coin or borrow other verbs as they can never cause any confusion.

There are many contronyms in English as the link indicates. To dust is one of them as it means:

To add fine particles, or to remove them

The linked Wikipedia article calls it auto-antonym:

The terms "autantonym" and "contronym" were coined by Joseph T. Shipley in 1960 and Jack Herring in 1962, respectively. Some pairs of contronyms are true homographs, i.e., distinct words with different etymology which happen to have the same form. For instance cleave "separate" is from Old English clēofan, while cleave "adhere" is from Old English clifian, which was pronounced differently. This is related to false friends, but false friends do not necessarily contradict.

Online Etymology Dictionary explains that "to rid of dust" comes form the noun dust which meant:

Old English dust, from Proto-Germanic *dunstaz (source also of Old High German tunst "storm, breath," German Dunst "mist, vapor," Danish dyst "milldust," Dutch duist), from PIE *dheu- (1) "dust, smoke, vapor" (source also of Sanskrit dhu- "shake," Latin fumus "smoke").

2

The two different connotations developed in the second half of the 16th century, probably by similarity to its original meaning, "to rise as dust".

Dust (v.)

  • c. 1200, "to rise as dust;" later "to sprinkle with dust" (1590s) and "to rid of dust" (1560s); from dust (n).

(Etymonline)