Acquisition of English expletives is probably late by default

This semester, I had to run a modest experiment for my seminar in second language acquisition, an experiment somehow grounded in previous work I’d done in syntactic theory. For me, that meant testing for acquisition of expletives.

Expletives are non-referential pronouns that exist only to satisfy a structural requirement, namely Chomsky’s extended projection principle (EPP). The EPP demands that all clauses have a subject. If no subject moves to Spec-IP in D-structure, then an expletive pronoun is inserted during the application of transformation rules. The expletive pronoun receives no theta role; it simply fulfills the structural requirement that every subject have a clause, as in this example from German:

slabloggermanThe verb regnet (rains) moves from V to T and finally to spec-CP, because the AdvP is fronted and German requires that verbs always appear in second position. However, regnet, despite all its movement in this clause, does not possess a theta grid. In other words, it has no argument structure and simply doesn’t demand an AGENT or THEME or any other kind of argument that might function as the clausal subject.

So, an expletive must be inserted to satisfy the syntactic demand for a subject. All languages want every construal to have a subject and a predicate: a thing and something about the thing. If a lexical entry possesses an argument structure that needn’t provide a thing, or an argument structure that makes the thing optional, expletives are inserted instead.

In terms of language acquisition, I think it’s likely that expletives will be acquired relatively late, for two reasons:

1) There simply aren’t many lexical entries whose theta grids demand or allow expletive insertion. Is likely is one such entry. It assigns only a single theta role to a proposition CP . . .


. . . which can then move to spec-IP to become the clause’s subject . . .

[That Jim likes bourbon [is likely <that Jim likes bourbon>]

. . . OR which can optionally stay in object position, thus requiring expletive insertion:

slablog2Likely is thus one of those lexical items whose argument structure makes expletive insertion possible. Again, these items are rare in the English lexicon. I wouldn’t expect them to be acquired in the early stages of the L2 grammar, especially given that most of them can be construed in other ways. E.g., the example above can be construed using a simple adjunct: Jim probably likes bourbon.

Same meaning. And likes and probably are—err—probably in the L1 input much more often than is likely. No reason to acquire the latter when the former gets the job done.

2) Not all languages handle expletives the same way English handles them, so L2 learners probably receive a lot of negative input early in the acquisition process that inures them to avoiding expletive construction in English. For example, English There is is construed in German as Es gibt, literally It gives. Both There and Es are expletives; both introduce existential clauses; yet both are construed very differently. Through early input, a German L2 speaker would learn that “It gives . . .” doesn’t work in English as an existential. He would then avoid existential constructions entirely until he received a lot of positive input, which would, again, mean a relatively late acquisition of the target expletive.

One exception to late expletive acquisition is the use of weather expletives (e.g., It’s raining). These clauses may be memorized expressions, and may not indicate an underlying competence. Working with my participants, it was therefore important to set up a testing situation in which expletives other than weather expletives would be appropriate and also statistically preferred by natives speakers.

To summarize my hypothesis: expletives are acquired late by default because a) lexical items whose theta grid allows or requires expletive insertion are rare and thus not readily acquired, and b) differing expletive construction between the L1 and English necessitates a certain amount of syntactic reconfiguration, dependent upon a long period of positive input from English.

In a manner of speaking, though, expletives themselves aren’t acquired at all; they’re results of a structural requirement of Universal Grammar. Given (a), what’s acquired late are English lexical items that force the expletives to be used. Given (b), what’s acquired late is enough positive input to make L2 speakers confident in their use of English expletives, which are quite different from the same expletives in the L1.


I presented four pictures showing various scenes to L2 speakers of English. My sample size was absurdly small: 7 L1 Chinese speakers and 5 L1 German speakers. I also showed the same pictures to 7 native speakers of English. I asked the participants to describe each picture using complete sentences. I was testing for the use of two types of expletive pronouns: (a) weather expletive it, and, more importantly, (b) existential there (e.g., There is a snake in my boots.) Being a native speaker, I know that English speakers prefer existential expletives when describing things, so I figured picture-description was a good situation for testing expletive use amongst L2 speakers. All the speakers were high-intermediate or advanced. Had they acquired the preference for expletives? These charts show how many times each speaker used an expletive while describing the four pictures.


Again, my sample size is absurdly small. I’d need to redo the experiment with at least 40-50 L2 speakers of several languages to get a more robust result. However, it’s always fun to see your hunch confirmed after even a small peek into the data. Native speakers had a much higher rate of expletive use than non-native speakers. Non-native speakers chose to describe the pictures using transitive verbs, which didn’t force the speaker’s mental grammar to bother with expletives: e.g., I see a deer, or more directly, A deer is standing next to a stream, rather than, There’s a deer standing next to a stream, or, It looks like the deer’s thirsty or whatever.

There’s an increase in expletive use by the two advanced speakers in each group. The increase is slight for the German speakers; however, the advanced Chinese speakers use expletives at the same rate as the native speakers. Given that I know these participants personally, I’m confident about the reason behind this phenomenon: they’re both Anglophiles who have reached near-native fluency.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s