Language Log: Grilling, staging, and landing

A couple of days ago ("On not allowing Bin Laden to back-burner", 5/3/2011), I noted that English (like other languages)  often turns a noun denoting a place into a verb meaning "cause something to come to be in/on/at that place".  I also noted that other causative change-of-state verbs generally have intransitive/inchoative uses as well (The sun melted the snow versus The snow melted), but denominal locative verbs typically don't. Thus we have transitive causatives like She floored the accelerator and We tabled the motion, but not the corresponding intransitive/inchoative versions *The accelerator floored and *The motion tabled. I also observed that Hale and Keyser's 1993 paper ("On argument structure and the lexical expression of syntactic relations") has spawned a large literature discussing ways to explain these and related facts. And it occurs to me that there have been quite a few corpus studies of causative/inchoative alternations and related things, but I don't know of any that have specifically focused on the case of denominal locative verbs. Correspondingly, the work stemming from Hale and Keyser 1993 seems to be entirely or almost entirely based on lexical and grammatical intuitions. (This is not my field, so I may be wrong: if so, some readers will no doubt set me straight.) So for this morning's Breakfast Experiment™, I decided to take a quick look at the usage in COCA of three of the verbs identified as denominal locatives in the literature: grill, stage, and land. What we'll find is three quite different patterns: grill is mostly causative but has inchoative uses at a low rate, in the sense of cooking on a grill but not in the metaphorical sense of being subjected to intensive questioning; stage is invariably causative in the observed sample; and land is inchoative more often than not, except that the fishing use is always causative, as is the extended sense that means "to catch or get hold of a person, a job, an opportunity, etc.". The goal — aside from satisfying my own curiosity — is to make a plausible case that a more serious empirical study would yield evidence relevant to the theoretical debate. I'll leave for another time the question of what implications such explorations might have.  (And you are hereby warned that relatively few readers are likely to be interested in wandering as far into the lexico-syntactic weeds as this post is likely to go.) The verb grill, as the OED observes, has an transitive/causative use "to broil on a gridiron or similar apparatus over or before a fire" (with the figurative extension "To subject to severe questioning"); and an intransitive/inchoative use "To undergo broiling, to frizzle. Chiefly fig.". The transitive uses are fairly common. Thus the literal causative/transitive: One of the couple's passions is grilling things never cooked that way before. As the sun went down, we grilled burgers and let the kids run around on the grass. Then we grilled the corn, onions and tomatoes over charcoal, with mesquite chips added for smoke. And its metaphorical extensions: She learned all these Sex Pistols songs on guitar and is grilling us for information about what the punk era was like. Senator Carl Levin, a Democrat from Michigan, chaired the panel that grilled the Goldman executives. We grilled Max about his homework and the latest gossip in his class. The intransitive inchoative is quite a bit less common, and (contrary to what the OED says) is always literal rather than figurative in the examples that I found: While escarole is grilling, whisk together lemon zest, lemon juice, garlic, parsley and olive oil. And, kangaroo, water buffalo, yabbie, emu sausage, barramundi as well as crocodile (tastes like chicken) were grilling at an all-you-can-eat picnic. To those of us who buy and eat chicken, it's a nice meat that makes good salads, grills nicely and tastes far better than it ought to when fried to golden crispness. In order to estimate the relative frequencies, I didn't count things like grilled as a modifier ("we liked the grilled fish") or predicative ("we like our fish grilled"). And I also ignored the (small number of) cases of null-complement anaphora, meaning "to grill unspecified stuff", though in fact this is a normal extension of transitive verbs: Above all, we grilled year-round while our Eastern cousins hibernated in the winter. They went down in the first round to Mancelona, and that night we grilled out at Jake's. In order to keep the problem simple, I only checked out the patterns "have|has grilled", "is|are|was|were grilling", and "I|you|he|she|it|we|they grilled". In those sets, I found 136 transitive/causative uses and 8 intransitive/inchoative uses, for a rate of about 5.6% inchoative. Of those 8 inchoative examples, 6 were in recipes. Furthermore, all 8 of the inchoative uses were about literal instances of cooking on a grill, whereas 62/136 = 46% of the causative uses were about intensive questioning. The "figurative" instances of intransitive/inchoative grill in the OED are all about literal heat, or at least an active metaphor of cooking: 1849 R. Curzon Visits Monasteries 2   Malta‥was cool in comparison to the fiery furnace in which we were at present grilling. 1883 J. Hawthorne Dust I. 277   The spleen which was doubtless grilling within him. 1883 R. L. Stevenson Treasure Island I. v. xxii. 177   Walking in the cool shadow of the woods,‥while I sat grilling. So in the COCA sample, the causative uses of the verb grill are dominant — leaving out recipe-language, the inchoative uses are only about 1.4% of the total. And the inchoative uses in this sample are always about being subjected to literal heat, whereas the causative uses are about half heat-related and half about intensive questioning. If the inchoative usage had the same expected proportion of intensive-questioning uses as the causative usage does, this distribution of results would be quite unlikely: it's like flipping a fair coin eight times and getting eight heads. My intuition, for what it's worth, agrees that the intensive-questioning sense doesn't work these days in an intransitive/inchoative frame: "While the expert witnesses were (*grilling) (being grilled), we decided to go out for lunch". At about 5% in this sample, intransitive/inchoative grill is relatively rare. But still, that's a higher proportion than I found  for inchoative uses of the verb stage. Here we're interested the literal locative verb, derived from the noun stage, that the OED glosses as "to put (a play, etc.) upon the stage", and its figurative extensions "To mount or put on (a spectacle). Also, to effect (a recovery); to stage a comeback". (In addition to comeback, I included examples of staging a protest, coup, rally,  walkout, boycott, etc.) Again, I ignored modifier or predicative uses of staged, and  one apparent example of null complement anaphora: Sometimes I think the White House press corps features vaudevillians, journalistic vaudevillians, who are staging and performing for the cameras. I also ignored a few examples of another verb, used in military contexts, which seems to be related to the word that the OED glosses as "intr. To travel by stage or stage-coach; to travel by stages; to journey over by stages", e.g. It was not until 3:45 P.M. that Lieut. Col. Bill David, the commander of the battalion that was serving as the ground element of the reaction force, was ordered to send an infantry company to the airport, where the Rangers were staging. For the analogous set of patterns ("have|has staged", "is|are|was|were staging", and "I|you|he|she|it|we|they staged"), I counted 453 examples of the transitive/causative use, and 0 of the intransitive/inchoative use. In the earlier post, I observed that If the local theater staged Three Sisters in 2009, we wouldn't say that Three Sisters staged in 2009. But in the comments, Duncan countered that I've had some amateur acting experience (mostly in high school and college, US mid- and mountain-west, '80s) and wouldn't find it unusual at all to say that a particular production staged in some year (or season). I don't doubt his testimony, but that usage is rare or specialized enough not to leave a trace in COCA, or at least in the parts that my patterns searched. In the case of land, there are various older and newer literal transitive/causative meanings, among them those glossed by the OED as "To bring to land; to set on shore; to disembark"; "To bring (an aircraft) to earth from the air; to place (an aircraft or spacecraft, or its contents) on the ground or some other surface after a flight"; "To bring (a fish) to land, esp. by means of a gaff, hook, or net". Some of these have intransitive/inchoative counterparts, "To come to land; to go ashore from a ship or boat; to disembark. Of a ship, etc.: To touch at a place in order to set down passengers"; "To alight upon the ground, e.g. from a vehicle, after a leap, etc. Esp. of an aircraft or spacecraft, or a person in one: to alight upon or reach the ground, or some other surface, after a flight". Notably missing from the set of intransitive/inchoative senses is the one that would correspond to the angling expression "bring (a fish) to land". And there are several extended or figurative uses as well, including "To bring into a specified place, e.g. as a stage in or termination of a journey; to bring into a certain position … Also fig. to bring into a certain position or to a particular point in a course or process."; "slang. To get (a blow) home"; and "fig. To catch or ‘get hold of’ (a person); to secure or win (a sum of money, esp. in betting or horse-racing). Also, to obtain (employment)."  The first two of these have inchoative as well as causative uses, but the last one seems to be lack them, in the OED's examples as well as in my COCA sample. Thus COCA has Actor Charlie Sheen has landed in new trouble. … the same disorder that has relegated his father to life in a lift chair and has landed his brother in a group home. Powlus also has landed in an anonymous spot on the Oilers depth chart … This isn't the first time Crowe's temper has landed him in the news. My punch landed on his face and he fell. The fat guy landed a punch to his stomach … But I landed a job as an apprentice fashion copywriter in the Garment District … … he landed a contract to supply the armed forces with silk for parachutes … Now it appeared that Christina had landed a client who could keep the firm busy well into the future. For the last sense, I neither expected nor encountered any examples in which the job, contract, client, etc., was in the subject position. Because land is substantially more common than the other two words I looked at, I limited my tally to the pattern "is|are|was|were landing", for which I counted 26 transitive/causative uses, and 164 intransitive/inchoative ones, or 87% inchoative. The analogous search pattern for grill yielded 30 transitive/causative cases and 2 non-recipe intransitive/inchoative cases, or about 7% inchoative; and the analogous search for stage yielded 126 transitive/causative cases and 0 intransitive/inchoative cases, or 0% inchoative. FWIW, the analogous COCA search patterns for boil and melt are 97% and 96.7% inchoative respectively. So to sum up, grill is mostly causative but has inchoative uses at a low rate, in the sense of cooking on a grill but not in the metaphorical sense of being subjected to intensive questioning; stage is invariably causative in the observed sample; and land is inchoative more often than not, except that the fishing use is always causative, as is the extension that means "to catch or get hold of a person, a job, an opportunity, etc.". Why are grill, stage, and land so different? Why are there also such striking differences among their senses? To a first approximation, this favors Paul Kiparsky's approach over Hale & Keyser's, since (as Kiparsky puts it) The relevant  constraints […] could not be derived from purely combinatorial properties of the primitives of compositional semantics — for H&K,  the lexicosyntactic categories from which word-internal propositional structure  is built. […]  H&K will not be able to  avoid bringing extralinguistic conceptual knowledge into their theory. However, I don't (yet?) see how Kiparsky's specific proposal deals with the differences among the various senses of the three words examined here. More on this at some later time. It's worth noting that none of the corpus-search results discussed here are particularly divergent from native-speaker intuitions, at least in retrospect. However, I don't think that I would have arrived at the same set of observations without looking at the patterns of usage. This reminds me of a point made in Charles Fillmore's 1992 chapter "'Corpus Linguistics' or "Computer-aided armchair linguistics'", which begins Armchair linguistics does not have a good name in some linguistics circles. […] Corpus linguistics does not have a good name in some linguistics circles. […] These two don't speak to each other very often, but when they do, the corpus linguist says to the armchair linguist, "Why should I think that what you tell me is true?", and the armchair linguist says to the corpus linguist, "Why should I think that what you tell me is interesting?" This paper is a report of an armchair linguist who refuses to give up his old ways but who finds profit in being a consumer of some of the resources that corpus linguists have created. I have two main observations to make. The first is that I don't think there can be any corpora, however large, that contain information about all of the areas of English lexicon and grammar that I want to explore […] The second observation is that every corpus that I've had a chance to examine, however, small, has taught me facts that I couldn't imagine finding out about in any other other way. Academics being as endearingly conservative as they are, things have not changed in the intervening 20 years as much as you might expect, although the existence of convenient web-accessible corpus searching has certainly shifted the culture a bit.