One day last week, Doug got up at 7:00, in an attempt to be able to fall asleep faster that night. He’d been trying to do it for several days, without success. He would just turn off his alarm without even waking up. I suggested the low-tech solution I’d used in college: Put the alarm clock on the opposite side of the room, so he’d have to get up out of bed to turn it off. And it worked. Now, here he was, up and dressed by 7:30, eating toaster waffles and microwave bacon.
Adam, though, was still asleep at 8:00. I put the remaining two slices of bacon back in the microwave to keep them out of our cats’ reach until Adam could get to them. I had also spooned some yogurt into a bowl, and had a piece of proto-toast in the toaster for him. I wanted Adam’s breakfast to be ready for him when he got up, because I would be running an errand by then. I didn’t want him to just come downstairs and skip breakfast in favor of playing video games.
So where to put the yogurt? Back in the fridge? OK, but the bacon had to stay in the safe. Room-temperature bacon is all right, but not refrigerator-cold bacon. And what about the toast? Darn it, by the time Adam came down, it would probably be stale. All right, I decided. Adam would just have to get up and get his butt downstairs for breakfast before he got dressed or anything else, that was all. I placed all three items on his placemat, and then went up to knock on his door.
“Who is it?” I heard a muffled voice ask.
“It’s me. Hey, I’m going to run an errand. Your breakfast is on the table. You might want to come down and eat it while…
…the toast is still warm, the yogurt’s still cool, and the bacon is still un-nibbled by cats.”
Awright! I was just trying to get my breakfast-making duties out of the way, but in doing it, I had spontaneously created a bracketing paradox!
Here’s the deal. Un-, everyone agrees, is a prefix. It can attach to one adjective to create another adjective. In this case, it’s attaching to the adjective (more specifically, past participle) nibbled to create the adjective un-nibbled, i.e. “not nibbled”. Then the prepositional phrase by cats attaches to that to give us the adjective phrase un-nibbled by cats, as shown in the diagram below:
But wait. Can PPs do that? Can they just attach to an adjective to give you an adjective phrase? Sure, if you have the right kind of adjective. Fond forms an AdjP when it attaches to an of-PP; so do great and with child. But un-nibbled isn’t an adjective that takes a PP, any more than, say, green or scary are. Green by cats? Scary by cats? What would those phrases even mean?
The meaning we’re after is, “It is not the case that the bacon is nibbled by cats,” so why not parse the phrase so that nibbled by cats forms a chunk, and then let the un- attach to that? Something like this:
Great! Now the negation clearly takes scope over the entire part about being nibbled by cats. But now un- isn’t a word prefix anymore. It might as well be the free-standing word not, the way it’s sitting outside the phrase nibbled by cats. Hence, the bracketing paradox.
Now there is one other parse of un-nibbled by cats, one that isn’t a bracketing paradox. It exists because of a peculiarity of the prefix un-. As Ben Zimmer wrote in a 2009 “On Language” column:
Ever since Old English, the un- prefix has come in two basic flavors. It can be used like the word “not” to negate adjectives (unkind, uncertain, unfair) and the occasional noun (unreason, unrest, unemployment). Or it can attach to a verb to indicate the reversal of an action (unbend, unfasten, unmask).
So if we take un- in its guise as a verb prefix, then we can parse un-nibbled by cats this way:
Unfortunately, a completely different meaning comes with this parse. And not only is it not the meaning I want; it’s a meaning that can’t even happen in this world. Living with five cats, I can tell you that they never un-nibble anything!






Doug and Adam like visiting their Aunt Carrie and Uncle Mark, because they have a flat-coated black retriever named Charlie that Doug and Adam like to play with. They’ll usually bring him a new toy, and Charlie is always eager to get it. He comes bounding up to the car, sniffing at us as we get out. My wife will pull the new toy out of the bag it’s in, and throw it into the yard for Charlie. He’s so used to the routine that it caused a problem one time when we didn’t bring a new toy — but did bring one of Doug and Adam’s stuffed animals. Charlie was so excited and so eager to get to work on that stuffed animal that we had to hide it in a bedroom drawer.
I was reminded of this when I listened to Deborah Tannen’s invited talk at the LSA conference. She’s done a study on how family members will use other family members in order to change the tone of a tense interaction. For example, I’ve sometimes asked Doug or Adam, “What would your mother say if she knew you were walking around in the cold house with no socks or slippers on?” It’s kind of a weenie’s way out to fob off the sock requirement on my wife, but hey, it makes me look a little less like the bad guy. And besides, she really would tell them to put on socks or slippers!
It was the point about a voice coming along with certain personality traits that reminded me of the Charlie voice. Trouble comes when we’re back at home, and Doug has the occasion to speak as our cat Nick. When he ventriloquizes Nick, he uses the Charlie voice. My wife can’t abide this. Nick and Charlie have two such different personalities that giving them the same voice is simply unacceptable. It bugged her so much that she even had me create separate voices for Nick and our four other cats. But Doug can’t do the Nick voice, so he’ll still sometimes use the Charlie voice for Nick. “No Charlie voice!” my wife tells him.
Doug’s friend Grant likes petting our cats, and is especially pleased when one of the shy ones lets him pet him. I remember his excitement when 











