Linguistic commentary from a guy who takes things too literally

Archive for the ‘Syntax’ Category

Multiple-Level Coordination with And and But

Posted by Neal on December 31, 2014

I’ve continually been hearing this one ad on a podcast I listen to, promoting some kind of investing service. The speaker introduces the subject like this:

Want to save more, invest for the future, but don’t have time to be a full-on investor?

The fact that this sentence is an interrogative gets in the way of what I actually want to talk about, so I’ll just rephrase it as a declarative:

I want to save more, invest for the future, but don’t have time to be a full-on investor.

If you take this sentence to contain a syntactically parallel coordination, then it is saying that these three things are true about me:

  1. I want to save more.
  2. I invest for the future.
  3. I don’t have time to be a full-on investor.

That’s not the intended meaning, though. The intended meaning is that:

  1. I want to save more.
  2. I want to invest for the future.
  3. I don’t have time to be a full-on investor.

In short, this sentence is another example of a multiple-level coordination. Depending on how you parse the coordination, the first conjunct is either the present-tense verb phrase want to save more or the plain-form verb phrase save more; the second is the plain-form verb phrase invest for the future; and the final conjunct is the present-tense verb phrase don’t have time to be a full-on investor. Whichever way you go, the conjuncts are not of the same syntactic category.

In previous posts, I’ve summarized the analysis of Beavers & Sag (2004) for this kind of coordination, and written too-hastily about what seemed like a problem for their analysis. This sentence, though, really is a problem for B&S … that is, if it’s truly grammatical, and not simply a mistake.

As a recap, here is how B&S analyze the MLCs It’s sick, twisted, and smells like old socks and women who are nursing, pregnant, or may become pregnant. For the first one, the and is conjoining these three clauses:

  1. is sick
  2. is twisted
  3. smells like old socks

Because the second conjunct begins with some material that’s identical to the beginning of the first conjunct (namely is), that can be ellipsed (i.e., left unspoken). For the second MLC, or is conjoining these other three clauses:

  1. are pregnant
  2. are nursing
  3. may become pregnant

Applying B&S’s analysis to our investment example, we get something like the diagram below. (Actually, they have it so that it’s binary-branching all the way down, with want to save more coordinated with the entire chunk want to invest for the future but don’t have time…. But my diagram will do for our purposes.)

The second "want to" is not pronounced

B&S argue against interpreting MLCs as having a structure like the following, which the and or or is coordinating two items rather than three, and there is an understood extra conjunction in there, coordinating two smaller items within the first conjunct:

  1. is sick and twisted
  2. smells like old socks
  1. are pregnant or nursing
  2. may become pregnant

Applied to our investment example, the kind of analysis that B&S advise against would give us a structure like the one below. In this diagram, I put an explicit conjunction between the first two conjuncts, to show where the understood conjunction would go. Notice that the main coordination has only two conjuncts, one being want to save more and want to invest for the future, and the other being but don’t have ….

See the extra conjunction I put in there?

One of B&S’s reasons for not favoring such an interpretation is that you can’t just go calling commas or pauses conjunctions. Where would it end? Going that way would license sentences like I saw a dog a cat as grammatical. There other reason is that there is nothing to force that understood conjunction to be the same as the explicit one. In other words, why couldn’t we take these examples to mean “It’s sick or twisted, and smells like old socks” and “woman who are pregnant and nursing, or who may become pregnant”?

With that in mind, notice that the extra conjunction I supplied was not another but, but rather an and, so that the meaning of the whole coordinate structure is “A and B but C.” This makes sense, because the two things I want are similar, and should be joined by and. The but serves to draw a contrast between them on the one hand and the thing I don’t have time for on the other.

So if that second diagram is incorrect, let’s look again at B&S’s analysis, shown in the first one. There’s a problem with this one, too. As I wrote in this Grammar Girl episode, but isn’t like and and or, able to coordinate any number of items. It’s only able to coordinate two:

That’s because and shows addition, and or shows alternatives, and it’s easy to imagine lots of additions to a set, and lots of alternatives. But, on the other hand, shows contrast, and it’s hard to conceive of a contrast between more than two things. We can say slow but steady, but we can’t say slow, careful, but steady.

According to the B&S analysis, our example sentence could be expanded out to have the form “A but B but C,” in violation of the two-conjuncts limit for but.

So what now? We could appeal to semantics. Although but conventionally implicates that there is a contrast between two propositions, there is no truth-conditional difference between and and but. So we could say that’s why the but gets away with this sort of behavior here.

Or there’s the cop-out solution of saying that the sentence is a grammar mistake, something that the writer would correct by themself if they thought about if more carefully. In favor of the cop-out solution is the fact that MLCs depending on and and or are too numerous for me even to bother blogging about anymore, whereas this is the only MLC with but that I’ve ever come across. So if you come across more MLCs with but, definitely leave a comment and tell me about them!

Posted in Multiple-level coordination | 4 Comments »

He Conquers Who Endures

Posted by Neal on November 29, 2014

I saw this on the back of a T-shirt when I was at the grocery store:

He conquers who endures.

Too bad for those people who endure. Even after all their endurance, they get conquered in the end. He, whoever “he” is, is a patient conquerer.

However, I suspect the wearer of the T-shirt probably didn’t realize that this was the meaning it was conveying. He probably thought it meant something like “The person who endures conquers,” or “He who endures conquers.” (Or to put it more gender-neutrally, “They who endure conquer.”) But that would mean that two unusual things were going on in this sentence. Neither of them is unprecedented, but both of them happening in one short sentence is noteworthy.

First, the clause who endures would have to be a relative clause modifying he. This doesn’t happen so much in present-day English. The best-known example in recent years is probably the epithet He Who Must Not Be Named for Voldemort in the Harry Potter novels. And even here, speakers didn’t realize they could change the He to Him when the name was a direct object, as observed by Q. Pheevr here.

Second, this relative clause who endures is separated from he. Now sometimes relative clauses do get separated from their head nouns: a book was published that would be read for centuries by countless generations; a woman appeared who was also carrying her head in her hands; What type of workers were there who participated in building the Pyramids. However, this usually happens when the subject of a clause would be ridiculously long if you refused to break it up. He who endures is just three words.

With my interpretation, though, there’s only one unusual thing going on: who endures isn’t modifying a noun at all, but is acting like a noun phrase all by itself. This is somewhat unusual, but not terribly so. It’s unusual because this kind of clause (known as a fused relative), more typically refers to things than to people. In other words, although sentences like That’s what I want and What you did was inexcusable are common enough, fused relatives like this one and the one in Who told me was my dad are somewhat rare. Exceptions include Can I help who’s next? and To whom it may concern.

Overall, then, my parse is the better choice syntactically. After a bit of internet-searching, though, I found that this is a translation of a Latin quotation from an ancient Roman satirist named Persius, although the opinion seems to be that he wasn’t being satirical when he wrote this:

Vincit qui patitur.

People who explain this quotation talk about the need for persistence in order to achieve victory, which definitely sounds more like the “They who endure conquer” interpretation. OK, so maybe it’s possible that I chose the incorrect interpretation for that guy’s T-shirt. But now I can write about how Latin is more precise than English, and you pick up this ambiguity in translation! Except that the same ambiguity exists in the Latin phrasing. Here’s how…

Vincit means “conquers”. Like its English translation, it can be transitive (as in Omnia vincit amor, “Love conquers all”) or intransitive (as in In hoc signo vinces, “By this sign you will conquer”), so you have to use the context to tell whether a nearby noun phrase is a subject or direct object. Usually in Latin, case endings do this, as illustrated below:

Vincit rex. “The king conquers.”
Vincit regem. “He/she conquers the king.”

Qui patitur means “who suffers (or endures)”, and it’s acting as a fused relative, just like its translation in English. Even in Latin, though, we can’t tell if that fused relative is a subject or an object. It’s the same problem that confuses English speakers about whoever and whomever. So actually, what we have here is a translation that is faithful even in preserving the ambiguity of the original!

Posted in Ambiguity, Fused relatives, Pronouns, Relative clauses | 8 Comments »

Tense Travel

Posted by Neal on October 14, 2014

My wife called me in to her office last night to make me watch this segment of The Big Bang Theory, in which Sheldon, Leonard, Raj, and Howard are watching Back to the Future, Part ii. Harrison Tran has helpfully bootlegged it onto YouTube, complete with a transcript:

Howard: Wait, hold on. Pause.
[music stops]
Howard: Something doesn’t make sense. Look. In 2015 Biff steals the Sports Almanac and takes the time machine back to 1955 to give it to his younger self. But as soon as he does that he changes the future, so the 2015 he returns to would be a different 2015. Not the 2015 that Marty and Doc were in.
Leonard: This is Hot Tub Time Machine all over again. Look. If future Biff goes back to 2015 right after he gives young Biff the Almanac, he could get back to the 2015 with Marty and Doc in it. Because it wasn’t until his 21st birthday that 1955 Biff placed his first bet.
Sheldon: But whoa, whoa. Is placed right?
Leonard: What do you mean?
Sheldon: Is placed the right the right tense for something that would’ve happened in the future of a past that was affected by something from the future?
Leonard: [thinks] Had will have placed?
Sheldon: That’s my boy.
Leonard: Okay. So, it wasn’t until his 21st birthday that Biff had will have placed his first bet and made his millions. That’s when he alters the timeline.
Sheldon: But he had will haven’t placed it.
Howard: What?
Sheldon: Unlike Hot Tub Time Machine, this couldn’t be more simple. [laugh track] When Biff gets the Almanac in 1955, the alternate future he creates isn’t the one in which Marty and Doc Brown ever used the time machine to travel to 2015. Therefore, in the new timeline, Marty and Doc never brought the time machine.
Leonard: Wait, wait, wait. Is brought right?
Sheldon: [thinks] Marty and Doc never had have had brought?
Leonard: I don’t know, you did it to me.
Sheldon: I’m going with it. Marty and Doc never had have had brought the time machine to 2015. That means 2015 Biff could also not had have had brought the Almanac to 1955 Biff. Therefore, the timeline in which 1955 Biff gets the Almanac is also the timeline in which 1955 Biff never gets the Almanac and not just never gets: never have, never hasn’t, never had have hasn’t.
Raj: He’s right.

I love that Harrison included the laugh track in his transcript.

So of course all this is just a good excuse to combine two kinds of geekery: sci-fi and grammar. The main way Sheldon and Leonard twist the verb-tense syntax is to allow auxiliary have to take complements that it doesn’t take in monolinear-time Standard English. First, they put it with the modal will, when modal auxilaries are always the first in a series of auxiliary verbs, even in outlandish strings such as will have been being eaten. Second, they put it with itself, in had have had brought and had have hasn’t.

Another kind of auxiliary combination we don’t get in monolinear-time Standard English is the past tense (or past participle) auxiliary had after a modal, in could not had have had brought. In our English, modals always take a base form, not a past-tense form.

Finally, there’s the combination of the negative contraction haven’t with will in will haven’t placed. In our English, the contraction has to come in the first auxiliary verb: won’t have placed.

But never mind the seeming syntactic violations. In a world with time travel, the language will have to evolve. My question was whether Sheldon and Leonard’s new syntactic rules for these tenses were semantically consistent with each other. In short, they’re not.

The main situation Sheldon and Leonard are discussing is an event time (young Biff placing a bet) that occurs later than a reference time (when young Biff receives the Almanac), but before the time of utterance (2014, in Sheldon and Leonard’s apartment). Ordinarily, this kind of situation would call for the so-called “future in the past“: Biff would place his bet sometime later. The complication is that the event time and the utterance time are now in separate timelines. Even for this situation, though, ordinary English has an appropriate choice: the perfective version of this future in the past: would have placed. But there’s one more complication: We’re talking about an event that not only did not happen in our own timeline, but also did happen in an alternative timeline. So how do Sheldon and Leonard propose to designate such an event? They use the past-tense form had, followed by a normal future perfect tense (will have had), to get had will have had. Let’s call this the alternate future in the past tense. This tense also puts its negative contractions in a different place, as noted earlier: had will haven’t placed.

However, Sheldon and Leonard don’t follow these rules the very next time they discuss an event that occurs later than a reference time but before the time of utterance, in an alternative timeline: Doc and Marty’s trip in the time machine. By the rules created so far, it would be never had will have brought (or if they wanted to use a negative contraction, had will haven’t brought), but instead, they go for the stacking of forms of have instead, leading to Sheldon’s culminating string never have, never hasn’t, never had have hasn’t. These are elliptical forms, which I surmise would all finish with gotten if they were fully spoken.

Still amusing, but it would have been funnier if these tenses had turned out to be consistent with each other, even after my poking at them.

Don't panic.

The grammar problem posed by time travel was also explored in 1980, in Douglas Adams’ The Restaurant at the End of the Universe. I was inspired to look up what he wrote, and was hoping to compare his tenses with Sheldon and Leonard’s. After reading the first paragraph below, I was excited to notice that Adams’ subscribed to the view of time travel in which you cannot alter timelines, unlike the Back to the Future scenario in which you can. Would this differing assumptions be reflected in the hypothetical grammars?

Sadly, no. Although Adams imagines more kinds of twisted temporal situations than Sheldon and Leonard discuss, he never bothers creating new verb tenses. Instead, he chooses to leave them to the readers’ imaginations, while using funny grammar jargon to name them. Here’s the relevant part, from the beginning of Chapter 15:

One of the major problems encountered in time travel is not that of accidentally becoming your own father or mother. There is no problem involved in becoming your own father or mother than a broad-minded and well-adjusted family can’t cope with. There is no problem about changing the course of history–the course of history does not change because it all fits together like a jigsaw. All the important changes have happened before the things they were supposed to change and it all sorts itself out in the end.

The major problem is quite simply one of grammar, and the main work to consult in this matter is Dr. Dan Streetmentioner’s Time Traveler’s Handbook of 1001 Tense Formations. It will tell you, for instance, how to describe something that was about to happen to you in the past before you avoided it by time-jumping forward two days in order to avoid it. The event will be described differently according to whether you are talking about it from the standpoint of your own natural time, from a time further in the future, or a time in the further past and is further complicated by the possibility of conducting conversations while you are actually traveling from one time to another with the intention of becoming your own mother or father.

Most readers get as far as the Future Semiconditionally Modified Subinverted Plagal Past Subjunctive Intentional before giving up; and in fact in later editions of the book all the pages beyond this point have been left blank to save on printing costs.

These two examples are the only pieces of pop culture I know of that specifically deal with new verb tenses to handle time travel, although collisions between time travel and ordinary English verb tenses have their own page on TV Tropes and Idioms.

Posted in Books, TV, Verb tense | 8 Comments »

He’ll Be None the Wiser

Posted by Neal on October 9, 2014

Since I began this blog in 2004, I’ve been vague about where in central Ohio I live, but tonight I’m proud to say that I live in Reynoldsburg, where phenomenal community support for our public school teachers has seen them through a summer of appalling disrespect from the local board of education (except for one notable member) and superintendent, who did everything they could to cause a teacher strike. That strike began on September 19, and might finally be ending today, if a tentative agreement is approved by the teachers.

In other news, Doug and Adam, who have been sick this entire time, might finally be showing signs of recovery. In the mornings, I’ve still been packing a lunch for Adam before I head to work. It’s not that he can’t get himself lunch, but if I don’t make him one, he’ll end up just eating Cheerios for every meal. So a couple of days ago, I opened the bag of bread and pulled out the three slices that were left: two heels and a whatever-you-call-a-slice-that’s-not-a-heel. Dang it, it had happened again!

I remembered a conversation from a month ago, when I had been encouraging Adam to make his own sandwich for a change, and he said he couldn’t, because he’d have to use a heel.

“That’s not a problem,” I told him. “Just do what I do. I put the pieces of bread together, like this, and turn the heel crust-side-in, like this. Then I grab these kitchen shears and cut off the edges of both pieces of bread, like I always do.” (Yes, I cut off the crusts. I don’t have to anymore, because Adam has recently started to eat his sandwiches with the crusts left on.) “Then I spread the peanut butter on the crusty side of the heel, finish making the sandwich, and you’re none the wiser.”

“Oh, I most certainly am the wiser!” Adam said. “Every time you do that, my sandwich tastes funny.”

Almost as interesting as the fact that Adam was, and had been, hip to my trick, was his phrasing I am the wiser. At first I thought he had used a negative polarity item in a positive polarity context, you know, like he did when he was four years old, during another sandwich-related occurrence. But as I thought about it more, I realized that the plus any comparative did happen outside negations and questions, in phrases like all the better and even somewhat the wiser. In any case, you can use NPIs in positive contexts if you’re really emphasizing them, as in Yes, I do give a damn!, and Adam was definitely emphatic. Well, maybe it was because Adam used the wiser without a specifier saying how much the wiser he was: No somewhat, no all, not even very much. But a specifier isn’t necessary, either: I could just as easily have said: You’re never the wiser. So maybe it was the combination of his using the wiser with positive polarity and without a specifier. I don’t know, so that’s one reason I never wrote it up here. Besides, delivering “We Support Reynoldsburg Teachers” yard signs, going to rallies, passing around petitions, writing letters, and picketing the residences of members of the board of education on top of my actual job made it tough to find the time.

Anyway, here I was again, making a sandwich out of a heel. So I put the pieces of bread together, turned the heel crust-side-in, grabbed the kitchen shears and cut off the edges of both pieces of bread, like I used to do. I spread the peanut butter on the crusty side of the heel, sliced the banana on top of it, and laid on the top slice of bread. I put the whole thing in a sandwich container, and stuck a note to the top. The note read, “He’ll be none the wiser.”

When I arrived home that afternoon, Adam had eaten the lunch I’d made. On the table, he’d left my note, with an edit:

Oh, heel naw!

Ha! The linguistic hook of my potential post about the syntax and semantics of none the wiser had just become a post about the homophony of he’ll and heel … or rather, the homophony of he’ll and hill outside of careful speech, and not in a dialect that lowers [i] to [I] before [l] as a matter of course. My guess is that it’s a frequency effect, because he’ll is such a common word. Similarly for we’ll/will and she’ll/shill, but not for Neal/nil.

Posted in Adam, Food-related, Negative polarity items, Ohioana, Syntax | 6 Comments »

Willing and Able

Posted by Neal on August 19, 2014

Nine years ago, I was inspired to write a post after hearing a flight attendant give the pre-flight safety presentation, and say, “Please move from the exit rows if you are unwilling or unable to perform the necessary actions without injury.” On my most recent flight, instead of listening to the attendant, I tried to pick up some Spanish vocabulary by reading the safety-information card from the seatback in front of me. And what do you know, inspiration struck again, when I read this Spanish sentence on the very same topic of sitting in the exit rows:

Toda persona que esté sentada en un asiento de salida debe estar dispuesta y ser capaz de realizar las siguientes funciones.
Every person who is seated in an exit seat must be willing and be able to execute the following functions.


Once again, the interesting part comes from trying to coordinate the adjectives for “willing” and “able”. If you’ve taken even first-year Spanish, you’ve had to learn about the two Spanish verbs that both mean “be”: ser, and estar. The former usually goes with what semanticists call individual-level predicates: properties that are generally true of someone, and less subject to change, such as hair color or nationality. These predicates stand in contrast to stage-level predicates, which are true of someone only for a limited time, such as emotional state or physical location. Now if I had thought of it when I was learning about ser and estar in junior high school, I would have tried to stump the teacher by asking something like, “What if you want to say that someone is Chinese and happy? Do you use ser or estar?”

It looks like the answer is that you don’t choose one or the other; you use both, each laying claim to one of the adjectives: estar dispuesta “be willing”; ser capaz “be able”. What a burn! English speakers don’t have to say be twice, but Spanish speakers do! A similar thing happens in French. If you like your salad with oil, it’s à l’huile (literally “at the oil”). If you like it with with vinegar, it’s au vinaigre (“at the vinegar,” with “at the” collapsed into the single word au). But if you’re like me, and like your salad with oil and vinegar, do you choose à or au? Neither! You have to use them both: à l’huile et au vinaigre.

On the subject of prepositions, though, I realized I needed to take a closer look at the adjectives dispuesto and capaz. The single preposition de goes with both of them in this sentence, but is de actually the preposition that typically goes with dispuesto in Spanish? A bit of Googling indicates that it’s not; the typical collocation seems to be dispuesto a. So why didn’t the translation have to use both a and de, along with both ser and estar, like this?

estar dispuesta a y ser capaz de realizar las siguientes funciones

One possibility is that for adjectives that take complements with mismatching prepositions, you just choose the one that goes with the closest adjective, in the same way in English we say Neither you nor I am the winner, because the closer subject is I. (This kind of solution is known as a resolution rule.) Another possibility is that both prepositions should have been used, and the translator simply made a mistake by using only one.

If we’re considering that possibility, though, maybe Spanish uses a rule of resolution to choose between ser and estar, too, and the translator should have just used the one appropriate for the nearest adjective, dispuesto.

So now I’m curious. I ask my Spanish-speaking readers, which of the following sounds best?

Posted in Other weird coordinations | 3 Comments »

Seldom Is Heard a Discouraging Word

Posted by Neal on August 2, 2014

The wife has spent numerous hours planning the itinerary for a trip we’re going to take out west later this summer. She has tricks up her sleeve that I never would have thought of for finding the best prices for airfare, car rentals, and hotels, so I bow to her travel-savvy. But all our discussions about the American West did get me to do some intense thinking of my own.

So you know the song “Home on the Range”? When I first heard it sung, on Captain Kangaroo when I was probably four or five years old, there were two lines in it that nagged at me. One was Where the skies are not cloudy all day, with its strange scope interaction between the negation of not and the universal of all. (Only cloudy some of the day? Uncloudy all day?) The other was the line that came right before it:

Where seldom is heard a discouraging word

Seldom was not in my vocabulary yet, but whatever it meant, it apparently must be a real bummer of a word to hear. Why might that be? And why would you be so likely to hear it? Were people out west in the habit of just walking up to one another and saying this particular discouraging word? “Hey, guess what!” “What?” “Seldom! Ahahaha!”

Some years later I learned what seldom meant, and eventually plugged the meaning into the “Home on the Range” lyrics. Then I was able to parse the line as a kind of negation inversion along the lines of Never have I ever…, Only rarely is he in his office, or Not a word did she speak. Mystery solved!

Or was it?

Let’s start off by taking out the seldom, an dputting this line into non-inverted, ordinary English word order of subject verb complement:

(no inversion) A discouraging word is heard.

If the line in the song were the same kind of negation inversion as in my examples, I’d expect the negative-like word first, then the auxiliary verb (which in this case is also the main verb, is), and then the subject and any complements, like this:

(Negative inversion) Seldom is a discouraging word heard.

…instead of what we actually have:

Seldom is heard a discouraging word.

Compared to the typical negative inversion, the subject and the complement have swapped places. This seems more like copular inversion, where a subject and a complement connected by a finite form of be switch places. This is the kind of inversion we get in utterances like Of greater concern is the fact that you lied, or Absorbent and yellow and porous is he. But this doesn’t get us our desired output, either. If we take a discouraging word as the subject (as we’ve been doing), and take the phrase seldom heard to be the complement (instead of just heard, as we’ve been doing), we get this:

(Copular inversion) Seldom heard is a discouraging word.

To sum up: two kinds of inversion, and neither of them produces the relevant line in our classic American folk song. At least, not individually…

Together, copular inversion and negative inversion can get us what we want, if we just assume that seldom doesn’t get into the picture until step 3:

  1. (no inversion) A discouraging word is heard.
  2. (copular inversion) Heard is a discouraging word.
  3. (negative inversion) Seldom is heard a discouraging word.

I find it interesting that this negative inversion can do its work on whatever appears in the typical subject position, whether or not it’s a subject. Does this really happen in other examples? Let’s try another one:

  1. (no inversion) The man who finds wisdom is happy.
  2. (copular inversion) Happy is the man who finds wisdom.
  3. (negative inversion) ??Seldom is happy the man who finds wisdom.

That one doesn’t work so well, so now I’m not sure I’ve identified what’s going on in “Home on the Range.” Maybe there’s some locative inversion with the where going on as well, like what Robert Burns has in his poem “Tam O’Shanter”. The poem is about how this guy and his friends like to have fun drinking together, and don’t worry about how late they’re going to arrive home,

Where sits our sulky sullen dame.
Gathering her brows like gathering storm,
Nursing her wrath to keep it warm.

I’ve tried locative inversion, though, and haven’t found a satisfactory combination of moves to get the desired output yet. Now isn’t that a discouraging word?

Posted in Inversion, Music | 8 Comments »

Sleep Like Death, Death Like Sleep

Posted by Neal on May 19, 2014

The boys, the wife and I watched the latest episode of the rebooted Cosmos last night. About 10 minutes in, Neil deGrasse Tyson began talking about the idea that life on Earth may have begun by arriving on meteorites. It’s known that rocks from Mars, for example, have ended up on Earth this way. It’s also known that some bacteria are able to survive in space, as proven by bacteria that survived a stint traveling on the outside of the International Space Station. Finally, it’s known that some bacteria can survive for a long time without a food source. On this point, Tyson talks about some recently revived bacteria found in Antarctic ice:

Even more amazing are these creatures, awakened from a death-like sleep of eight million years…

I was interested to hear Tyson put it that way, because I’ve also been hearing another person talking about death-like sleeps recently, but she phrases it differently:

Did you hear that? She said:

Before the sun sets on her sixteenth birthday, she will fall into a sleep like death!

Both phrases are talking about a sleep, not about death. We know this from the context, and from the fact that the verbs fall and awaken collocate more strongly with sleep than with death. But they’re phrased in completely opposite orders from each other! Furthermore, it’s syntactically possible for each phrase to be referring to death, not to a sleep. No, I haven’t actually found any examples of this, but it could happen, OK?

Here are the structural differences all sorted out. The diagrams on the left refer first to a death that is like sleep, and then to a sleep that is like death. In these parses, the adjective like is looking for a noun-phrase complement on its right to form an adjective phrase. The diagrams on the right refer to a sleep that is death-like, then to a death that is sleep-like. Here, the adjective like forms a compound adjective with the noun phrase on its left.

Dead, or Just Resting?

The situation reminds me of Shel Silverstein’s “snake eating cake”.

Posted in Attachment ambiguity, Compound words, Diagramming, Movies, TV | Leave a Comment »

Chemicals, Castro, and Last Year’s Jeans

Posted by Neal on May 8, 2014

Here are a few items I recorded some time ago, and just found lying in my stash of draft posts.

Doug and Adam each spontaneously uttered a double passive within a day or two of each other sometime in the past year. Here’s Doug’s:

… standing in front of a chemical plant that‘s threatened to be bombed.

I don’t even remember what the context was for that. It’s a good thing I wrote this down, or it’d be a memory lost forever, like all that thing he did that one time that I didn’t bother putting in his baby book, or that other thing he did that other time that I never put in a diary or anything. Anyway, if the clause about the chemical plant were in the active voice, it would be something like “a chemical plant that someone threatens to bomb.” But if you don’t know who made the threat, how do you say this? You do like Doug did, and use the passive voice. The trouble is, you have two verbs to deal with: threatened, and bomb. Which one do you make passive? Option 1 below

just doesn’t make any sense. Option 2 is grammatical, but it removes the human agency from threaten. It sounds like conditions are such that the chemical plant is likely to be bombed, in the same way that It’s threatening to rain means, “Conditions are such that it’s likely to rain.” So he went with option 3, making both verbs passive.

  1. *a chemical plant that is threatened to bomb
  2. a chemical plant that threatens to be bombed
  3. a chemical plant that is threatened to be bombed

Now, on to Adam’s double passive:

Fidel Castro has been attempted to be assassinated over 600 times.

I think he got this off a history website or something, like maybe this Mental Floss article.

Robin Dodsworth sent me the weirdest case of possible right-node wrapping that I’ve seen. I’ve come to believe that nonparallel coordinate structures such as

wash and put the dishes away

are actually part of many people’s English grammar. Usually they consist of an ordinary transitive verb (e.g. wash) and a phrasal transitive verb (put away) taking a single direct object, with the preposition from the phrasal verb coming after the direct object. Phrased as a parallel coordinate structure, this would be “Wash the dishes and put them away.” Robin’s example, though, is different. An old high school friend wrote on Facebook:

Scariest moment of the year — the first cool day of Fall when you have to put on (and find out) if last year’s jeans still fit.

Put into a parallel structure, this would be “Put on last year’s jeans and find out if they still fit.” So instead of being a direct object with respect to the parts I’ve colored red and green in the quotation, it’s a direct object for put on, and an embedded subject in find out if ___ still fit. This is so unlike other RNW examples that I suspect it was just a mistake, but I don’t know. Does it sound comparable to wash and put the dishes away to you?

Posted in Adam, Double passives, Doug, Right-node wrapping ("Friends in Low Places" coordinations) | 6 Comments »

Thoughts on Just Because X Doesn’t Mean Y

Posted by Neal on April 4, 2014

On her Grammar Girl podcast this week, Mignon Fogarty is running a guest script that I wrote on the just because X doesn’t mean Y construction, a thriving piece of English syntax that has come into its own in the last 50 years or so. My favorite example of JBX-DMY is

Just because you’re paranoid doesn’t mean they’re not out to get you.

In this post, I go into some of the details that didn’t make it into the script. First off, quite a bit has been written about this construction; the sources I read while writing the Grammar Girl script are:

Weird syntax
JBX-DMY is so common that it’s easy to overlook that it doesn’t follow the regular syntactic or semantic rules. On the syntactic side, what’s the subject of doesn’t mean? Is it the just because clause? Or is it an understood subject that sometimes gets pronounced, as in Just because it’s easier to raise VC money, that doesn’t mean you should, or Just because it’s wrong, it doesn’t mean it’s not funny?

Weird semanticsFor comparison, let’s look at more regular sentences that contain just because. For example, think about the scope of the “cause” meaning and the negation meaning in Bill isn’t speaking to me, just because I served him a cat-food sandwich as a joke. Depending on the intonation, we can get two readings. I’ll illustrate them by using the notation CAUSE(A)(B) to mean that A caused B. With the intonation suggested by the comma I put in the sentence, we get this reading, with CAUSE taking scope over the negation:

CAUSE(serve cat-food sandwich)(NOT(speaking to me))
(i.e. my serving Bill the CFS caused him to not speak to me)

Without a pause before the just because, and with a a rise-fall intonation at the end of the sentence, we get a reading with a wide-scoping negation:

NOT(CAUSE(serve cat-food sandwich)(speaking to me))
(i.e. Bill is speaking to me, but for reasons other than my serving him a CFS)

However, when we put the just because clause in front, as in Just because I served him a cat-food sandwich as a joke, Bill isn’t speaking to me, the only reading we get is the wide-scoping CAUSE.

Now consider just because X doesn’t mean Y. In passing, let’s note that just because is now not talking strictly about cause and effect, but about inference. This is not unusual in and of itself, because “inferential because” is a well-known phenomenon. For example, you can say, “Classes just let out, because the hallways are full of students,” but not mean that the crowded hallways caused classes to let out. What you really mean is that the fact that the hallways are full of students allows you to infer that classes have just let out. But what’s interesting about just because X doesn’t mean Y requires that NEG scope wider than (inferential) CAUSE, just the opposite of our cat-food sandwich example:


Variations on the just because and the doesn’t mean
Another point made in several of the articles is that JBS-DMY requires neither just because nor doesn’t mean to work. You can do it with simply because, or in the right context, plain old because. And instead of doesn’t mean, you’ll also hear other negations, such as is no reason to, or doesn’t make, or even rhetorical questions that imply a negative answer.

Just because with of complements, and “because X”
By analogy with ordinary because and because of, Kanetani observes, the just because construction now has a variant with just because of, as in

Just because of his dominance doesn’t mean they’re going to win games of footy or win or the clearances. (link)

The much-talked-about because X construction (in which X is something other than a full clause, such as a noun phrase, an adjective, a participle, or an interjection) has also now been folded into the analogy. Taking the top result from Tyler Schnoebelen’s listing of the most common words to appear in tweets after because, I did a search for “just because YOLO doesn’t mean” and found numerous examples like Just because YOLO doesn’t mean you can act like a moron.

Hilpert believes that the origin of the unusual semantics of JBX-DMY followed a sequence like this:

  1. Unremarkable sentences beginning with just because were already in the grammar, and some of them happened to have negated main clauses; for example, Just because I served him a cat-food sandwich as a joke, Bill isn’t speaking to me.
  2. Such constructions eventually came to be used with the inferential CAUSE meaning rather than a pure causation meaning.
  3. Finally, the inference-denying meaning of today’s JBX-DMY.

There’s a problem, though. In step 1, the semantics has CAUSEcausation>NOT, as discussed earlier. In step 3, the semantics has NOT>CAUSEinference. So there are two changes that have to happen in between: the scope change, and the change from causal to inferential just because. Whichever of those changes happens first, we end up with something that doesn’t seem to have the appropriate meaning, as illustrated below:

Just because he’s my nephew I didn’t hire him!
(Desired meaning impossible: that I hired my nephew for non-nepotistic reasons.)
Just because the streets are wet, it didn’t rain.
(Desired meaning impossible: that the sole fact of wet streets allows us to infer that it rained.)

I think a more likely progression is suggested by Bender and Kathol’s paper. Noting that there is flexibility over how the DMY part gets negated, imagine a sentence like this, in which the speaker seems to question the validity of an inference:

Just because I let him borrow my computer once, he seems to think he is allowed to use it any time he wants to.

The semantics here is the straightforward one you’d expect: For the sole reason that I let him borrow my computer once, he thinks he can use it whenever he wants. The negation, which is unspoken, permeates the whole utterance: This guy is wrong to think he can borrow my computer any old time now. From here, it’s a short step to turn that implied negative he seems to think into an actual one:

Just because I let him borrow my computer once, he shouldn’t think it’s his to borrow whenever.

Now we have NOT>CAUSEcausation, which we couldn’t get in our nepotism example. From here to that doesn’t mean doesn’t seem quite such a jump now. Subsequent elimination of the overt subject that or it, Hilpert argues, was due to the mostly empty meaning of the pronouns, plus the phonetic similarity of their final /t/ and the beginning /d/ of doesn’t. This part of his argument I’m inclined to believe.

Update, April 14, 2015: Doug was re-watching Iron Man this afternoon, and I heard this line of dialogue, uttered by bad guy Obadiah Stane to Tony Stark:

Do you really think that just because you have an idea, it belongs to you?

Without the framing rhetorical question of Do you really think…?, this is a straightforward instance of causal just because: if you have an idea, that makes it yours. But inside the rhetorical question, the clear meaning is that simply having an idea doesn’t make it yours. It’s an example, in the wild, of the type I was describing in the original post.

Posted in Scope ambiguity, Syntax | 2 Comments »

What She Cooks Like

Posted by Neal on March 9, 2014

One day last month, Doug and his classmates watched part of a Disney movie during one of the many wasted class periods he’s had this year (thanks to the busiest, most pointless, and most disruptive standardized-test schedule I’ve ever seen). He liked it, he said, and he’d figured out that the person who voiced a dragon in the movie was that guy who had done Donkey in Shrek.

“Oh! Mulan!” I said. “That’s one of the last movies your mom and I saw before you were born.” I also clued him in on the name of “that guy who did the voice of Donkey,” as he and his classmates think of Eddie Murphy. He wanted to put it on our Netflix queue so he could see the rest, so we did.


I didn’t tell Doug my secret reason for putting Mulan in the queue: a line in one of the songs that I’ve occasionally considered blogging about, but hadn’t wanted to go to the trouble of watching the movie again so I could get the exact wording. But if Doug wanted to watch the movie anyway, I could conveniently accomplish the goal.

So last weekend, I saw Mulan for the second time. Adam pointed out that the voice of Mulan herself was done by one of the stars of Agents of SHIELD. Doug noticed that the enemies that were clearly supposed to be Mongols were actually referred to as Huns, probably because Huns was easier to rhyme in a song than Mongols. (They rhymed it with sons.) The wife noticed an “American Gothic” reference she hadn’t remembered. And I got to hear the line in the song I’d been trying to remember. I had to pause and rewind a couple of times before I could write it all down, but luckily, nobody minded.

It comes in the song “A Girl Worth Fighting For,” as the members of the Chinese army who are prominent enough to have names sing about their dream women. About 40 seconds into the song, the baby-faced big fat one, named Chien Po, sings

I couldn’t care less what she’ll wear or what she looks like.
It all depends on what she cooks like.

Hah! Looks like … cooks like! Oh, and notice the standardly negated couldn’t care less, too. But still: You can say someone looks like a god, or cooks like a chef, so why is looks like … cooks like so funny?

In the question what she looks like, the what corresponds to the missing object of like. But the key, I think, is that what she looks like has essentially the same meaning as how she looks, where how could be standing in for an adjective (she looks good) or a prepositional phrase (she looks like a statue). With these two equivalent sentences available, we can set up an analogy:

how she looks : what she looks like :: how she cooks : X

What does X equal? what she cooks like, naturally! But why is it so funny?

How is the question word we use in order to ask about a predicate adjective. Questions like How do you feel?, How does it taste?, How did they sound?, and How does she look? are typically answered with adjectives: great, good, bad, swell, or maybe prepositional phrases such as like a million bucks. But how is also the question word we use to ask about the manner in which something was done. Questions like How did he do it? and How does she cook? are typically answered with an adverb, like well or poorly, or some other kind of phrase that tells how something was done: with a ball-peen hammer, for example. Only the how corresponding to an adjective means the same thing as what … like, and the analogy that gets us what she cooks like totally ignores this fact.

Posted in Kids' entertainment, Movies, Syntax | Leave a Comment »