I Typoed a Whole Post for the Language Log

Those darn linguists! Sometimes I think that linguistics degrees are really just licenses to be pedantic. I mean, look what they did to poor William Saffire. And now they've caught me! I'm just hatin', though. I like to be pedantic, but when a cognitive psychologist is being pedantic, he's just being pedantic, not doing his job. Also, I wish I had caught my own error, not to correct it, but because it's pretty damn funny. "Square quotes" in place of "scare quotes?" I think it's a typed spoonerism of sorts, in that I got to the q a little too soon, and then ran with it, but it could even be a Freudian typo (I do tend to think both Leiter and analytic philosophy are a bit square). Either way, I hope I get a nice thank you note from the linguists, because I typoed a whole post for them.

Which brings me to my own version of a linguistics discussion. Notice that in the title of this post, and the verb after "because" in the final sentence of the last paragraph, the verb ("typoed") is a bit odd. It's a novel denominal verb used in a strange context. While you may have been confused by it in the post title, hopefully by the time you had all of the context, it started to make sense. By typing the typo, I gave the folks at language log material for a whole post. OK, so my use of a novel denominal verb is a bit of a stretch, but I use it to set the stage for a discussion of how people might understand some similar (and more plausible) denominal verbs. For instance, at Language Log, there's a post entitled, "What can you Bret Easton Ellis to that?" which highlights the following sentence from Ellis' Glamorama:

"As if', I Alicia-Silverstone-in-Clueless back at him.

Even though both examples (the title and the Ellis example) from Language Log use proper nouns, the same sort of thing is going on. A noun ("typo," "Bret Easton Ellis") is used as a verb, and, at least in context, people tend to have little trouble understanding it. This is actually very common in ordinary language, and serves as the impetus for a great deal of linguistic innovation. Think of how many verbs are formed from nouns. How does this work? That's a good question for empirical investigation, don't you think? Well, I do, and so did Michael Kaschack and Arthur Glenberg, who actually went out and conducted a set of experiments on the understanding of novel denominal verbs¹. Since I think it's interesting, it may be that one of you will also, so I'm going to tell you what they found.



Table 1 from Kaschack & Glenberg (2000) presents examples of form-meaning pairs. Click for a larger view.

Before I get to their findings, though, I need to set things up properly, using the discussion in the Kaschack and Glenberg paper (K&G). Let's start with syntax. Children learn and adults understand the meanings of many verbs by recognizing certain syntax-meaning pairings. Kaschack and Glenberg give the following examples (see also Table 1 above):

The transitive sentence structure (N–V– OBJ) is linked with a putative meaning of “X acts on Y,” whereas the double-object sentence structure (N–V–OBJ1–OBJ2) is linked with a meaning of “X transfers Y to Z." (p. 509)

One theory (the theory from which K&G draw) used to explain these form-meaning pairings is the constructional theory², in which the syntactic forms and their paired meanings, or "argument structure constructions," constitute independent language units. K& G write:

Argument structure constructions are believed to be lexically un- filled syntactic forms that specify both syntactic (e.g., nouns, verbs, and objects) and thematic (e.g., agents, patients, and recipients) information. For example, the transitive construction specifies reference to the “agent” and “patient” thematic roles and links these roles to specific syntactic units (the “agent” is the subject, the “patient” is the direct object). (p. 510)

K&G note that this theory provides an explanation for certain novel uses of verbs, such as "sneezed" in the following sentence:

Art sneezed the foam of his beer. (p. 510)

In this sentence, the normally intransitive verb "to sneeze" is used in a transitive construction (N–V– OBJ), and thus the meaning associated with the transitive construction is carried over to the interpretation of the verb. This theory alone, however, cannot explain novel uses of denominal verbs like "typoed" and "Bret Easton Ellis." We need something else to explain how the context in which these denominal verbs are used helps determine their meaning. For this purpose, K&G invoke the Indexical Hypothesis (IH). The IH involves three processes (quoted from K&G):

1. The first of these is indexing (e.g., mapping) the words and phrases in the sentence to (a) referents in the environment or to (b) analog mental representations.
2. The second process... is the derivation of affordances (after Gibson, 1979) from the referents that were indexed. The term affordances refers to the ways in which individuals can interact with things in their environment.
3. The third process... is that the affordances are meshed under the guidance of (a) intrinsic constraints and (b) constraints provided by the syntax of the sentence. Meshing is a process that combines affordances into coherent patterns of actions (Glenberg, 1997); that is, actions that can actually be completed to accomplish a goal. (p. 510-511)

To see how this works, consider the following sentence with a novel use of a denominal verb:

Lyn crutched Tom her apple so he wouldn’t starve. (p. 512)

First, notice that this sentence uses the double-object argument construction (N–V–OBJ1–OBJ2), which is associated with the meaning "X transfers Y to Z." According to IH, the first stage in interpreting this sentence involves mapping Lyn, Tom, the apple, and so on, onto objects in the world or to mental representations. Thus, Lyn is mapped to Lyn (or a representation of Lyn), Tom to Tom, the apple to an apple that Lyn possesses, and so on. The second process involves deriving affordances from the mapped objects. In this case, the key affordance is that of the "crutch," which is used as a verb, but mapped onto an object (a crutch). During this second process, we retrieve the affordances of crutches. Finally, these affordances are meshed with the double-object argument construction, so that the affordance(s) used to interpret the sentence are those that make sense for the meaning "X transfers Y to Z." For instance, we may interpret this sentence as Lyn pushing the apple to Tom with the foot of her crutch.

To test whether IH, combined with the constructional theory can explain peoples' understanding of sentences, K&G conducted four experiments using novel denominal verbs like "crutched" or "typoed." The first two experiments were designed to show that a.) people do in fact use the form-meaning pairs to interpret sentences, and b.) this use cannot be explained by the presence of previously learned words (hence, the use of the denominal verbs). I'll just describe experiment 1, as the results from experiment 1 and 2 are pretty much the same. In experiment 1, K&G presented participants with either a single or pair of sentences like the following (1 & 2 and 3 & 4 represent two pairs):

(1) Lyn crutched Tom her apple so he wouldn’t starve. (double-object form)
(2) Lyn crutched her apple so Tom wouldn’t starve. (transitive form)
(3) Lindsay bought Sam a sweater to please him. (double-object form)
(4) Lindsay bought a sweater to please Sam. (transitive form) (p.512)

Participants who received pairs of sentences subsequently received another sentence (e.g., "Tom got the apple") that was more consistent with either the double-object or transitive form, and were asked which sentence in the pair "most strongly implied" the third sentence. As in the examples, half of the pairs contained novel denominal verbs, and half contained normal verbs. If participants are sensitive to the form-meaning construction, then they should perform as well when they receive sentences with novel denominal verbs as they do when they receive ordinary verbs. Participants who received only one sentence were then given two meanings, one associated with the double-object form, and one with the transitive form, and asked which meaning best fit the sentence. Once again, if participants are using the form-meaning constructions to interpret the sentences, they should be able to map the correct meaning onto both the sentences using the normal verbs and those using the denominal ones. In both experiments, K&G found that participants were, in fact, able to use the form-meaning constructions to interpret the uses of denominal verbs. Participants who received the sentence pairs and the single sentences assigned the correct meanings to the sentences using denominal verbs as often as they did the sentences with normal verbs.

Experiment 3 and 4 were designed to show that participants not only needed to use the form-meaning pairs, but also had to mesh the appropriate affordances with those form-meaning constructions in order to interpret sentences using novel denominal verbs. Again, the results from experiment 3 and 4 are consistent, so I'll just describe experiment 3. In this experiment, participants received a short scenario, like the this one:

Rachel worked for a scientist in a research firm. As part of her duties, she was required to bring the scientist’s mail to his office so he could open it after lunch. On this particular day, Rachel encountered three large boxes among the mail addressed to the scientist. The boxes were way too big for her to carry.

After which, they received an "affordance manipulating sentence" like the following:

In the corner of the room, though, Rachel noticed an office chair with four good/missing wheels.

Finally, they received a critical sentence:

Rachel brought/chaired the scientist his mail.

In one condition ("four missing wheels"), the affordance manipulating sentence created a situation in which the task could not be easily carried out using the actions afforded by the object. When the verb in the critical sentence is a conventional one, this should not be a problem, because people can interpret the sentence using the normal meaning of the verb. However, according to the IH, when the verb is novel, and the affordance is blocked (e.g., by the missing wheels), people should have a difficult time interpreting the sentence, because they cannot mesh the affordance with the form-meaning pairing. This should cause participants to take longer when reading the critical sentences containing novel verbs when affordances are blocked. This is, in fact, what K&G found. When verbs were conventional, participants' reading times were near identical for both the blocked and non-blocked affordance conditions. However, when the verbs were novel, participants' reading times were significantly longer when the affordances were blocked than when they were not.

So, after all that, we get to the punch line. It appears that people are able to interpret these sorts of novel denominal verbs through a simulation process that meshes the affordances of the objects involved with the syntactic constructions of the sentences in which they appear. I must admit that I'm not exactly sure what sort of actions Bret Easton Ellis affords, but if I were guessing, I'd say that instead of affordances, we use our Bret Easton Ellis schema (formed, in the context of the Language Log post, by reading the Alicia-Silverstone-in-Clueless sentence). "Alicia-Silverstone-in-Clueless" also serves to activate a schema, in this case, Silverstone's character's manner of speaking in the movie Clueless. Still, the principle is the same. We "mesh" the activated schemas with the syntactic cues from the sentence, and voila! We've got an interpretation of the sentence.

I hope that my next major typing error yields a new denominal verb for the linguists at Language Log. "I mixing memoried all the way to the bank."

¹ Kaschack, M. P., & Glenberg, A. M. (2000). Constructing meaning: The role of affordances and grammatical constructions in sentence comprehension. Journal of Memory and Language, 42, 508-529.
² Goldberg, A. E. (1995). Constructions: A construction grammar approach to argument structure. Chicago: University of Chicago Press.