The acquisition of wh-questions: Beyond structural economy and input frequency

ABSTRACT

We present in this article corpus analyses, two experiments, and a preliminary English-French comparison on children’s acquisition of wh-in-situ. Our examination of 10,000 wh-questions from CHILDES reveals that the reported empirical picture of wh-question acquisition in English is incomplete: A type of wh-in-situ, probe questions (PQs), has been left out from most discussions despite its presence in child-directed speech. Unlike wh-in-situ echo questions (EQs), PQs are used to request new information, and parents frequently use PQs and fronted information-seeking questions in alternation. The fact that PQs share the pragmatic space with fronted wh-questions while involving fewer syntactic operations and exhibiting lower input frequency allows us to test both structure-based and frequency-based theories of syntax acquisition. Our comprehension task with 3;06–5;06-year-olds confirms that children accept and understand PQs as information seeking. On the other hand, results from a production task show a strong avoidance of wh-in-situ, which is in line with reported elicited data from French-speaking children. We reason that a structural economy-based approach alone is not sufficient to account for children’s disfavor of wh-in-situ. Depending on the input frequency and consistency, as well as the number of variants licensed by the grammar of a given language, children may treat part of the input as uninformative and initially only learn from higher-frequent, more regularized input. Their intake is thus selective.

Notes

¹ We do not consider structural accounts that rely on intervention effects such as Featural Relativized Minimality (Friedmann, Belletti & Rizzi 2009) because the present study is limited to simple object (what, who) and adjunct questions (where). See Table A1 for details.

² We will systematically represent EQs with a wh-phrase in caps for ease of identification.

³ The Corpus of Contemporary American English (COCA) is a widely used corpus of American English containing 1.1 billion words.

⁴ Since the wh-words in almost all of the samples were in the final position of the sentence, it is impossible to tell whether the differences between EQs and PQs arise from the stress on the wh-word or from the sentence-final prosody. Having PQs or EQs with sentence-medial wh-word or multiple wh-words would address this problem; however, such sentences are rare in corpora. As most studies on EQs have analyzed their unique intonation in terms of their stressed wh-word, we suggest that the differences between PQs and EQs emerge from the wh-word instead of sentence-final prosody.

⁵ The “cousin” is an older child only producing adult-like utterances throughout the corpus.

⁶ Adam frequently produced a PQ immediately after a fronted wh-question, most likely mirroring the adult behavior mentioned in (6). However, adults typically rephrase the original question into an in-situ PQ only if the child fails to answer the fronted one. Adam, on the other hand, did not wait for a response. This suggests that Adam may have used PQs in a different way compared to adults, as he asked these questions without knowing the answer.

(i) Child: What is that?

 Child: It’s a what?

 Researcher: I don’t know what it is, do you?

⁷ In the experiment, we used animation effects to display the building one by one so that it was clear to the child which building was which. Assuming that 4-year-olds cannot read yet, a “bookstore” drawing was chosen to represent the “library” to maximize illustration as “library” drawings are typically a generic building with no books shown.

⁸ The lack of reported data on child-directed speech precludes a fuller comparison with Brazilian Portuguese, but the elicitation results are informative on their own. In general, the large set of available wh-strategies is similar (but not identical) to French, and in-situ wh can be used for information-seeking questions. Children ages 4;06–5;06 are shown to prefer fronted wh (~80%) to in-situ wh in two experimental conditions, one establishing a Common Ground (hence an enriched presuppositional context) and the other not. Adult controls show no preference in the prominent Common Ground condition (50.5% fronted wh vs. 49.5% in-situ wh).

⁹ We thank an anonymous reviewer for bringing this to our attention.

¹⁰ French has verb raising all the way to C in questions; hence fronted questions in the absence of an auxiliary involve longer head movement, which translates into additional instances of Merge under the Derivational Complexity Metric of Jakubowicz (2011). The economy of movement hypothesis predicts more fronting with auxiliaries (shorter chains) than with lexical verbs in French, which has yet to be tested.

¹¹ Zuckerman & Hulk (2001) report a lower production rate of French in situ (6%, after omission of outliers [n = 5], the rate drops to 3%). The very low level of wh-in-situ may at least partly be the consequence of the method of elicitation in which an indirect question with a clause initial wh-phrase and no inversion was used as a prompt (Je veux savoir où il est allé ‘I want to know where he went’) despite the fact that one possible answer is Où il est allé? ‘Where did he go?,’ which is likely to have inflated the proportion of fronted wh without inversion (89%).

¹² The elicited production result (~20% wh-in-situ) is culled from several sources (Strik 2007, 2008) and is an estimate based on limited text descriptions of the results. A pilot study (as reported in Strik, 2008) is not reported because the elicitation method was similar to that used by Zuckerman & Hulk (2001) and discussed in fn. 11.

¹³ It is worth noting that the Leveillé corpus analyzed in Crisma (1992) represents French speech from the early 1970s, which is likely to be more formal than more recent spoken speech. Interestingly, Larrivée (2019) provides adult data from the Orleans corpus at two different time points on comment ‘how’ questions, documenting a change in use of in situ from the early 1970s to 2014. Limiting the analysis to information seeing vs. echo use, he demonstrates that they are increasingly used as requests for new information (as opposed to requests for repetitions/clarifications). This is a clear pragmatic change within a short time interval of 40 years, and it suggests that other pragmatic changes may have taken or are taking place.