The development of abstract schemas and productive rules in language is affected by both token and type frequencies. High token frequencies and surface similarities help to discover formal and functional commonalities between utterances and categorize them as instances of the same schema. High type frequencies and diversity help to develop slots in these schemas, which allow the production and comprehension of novel utterances. In the current study we looked at both token and type frequencies in two related constructions in German (...) child-directed speech: simple transitive and complement-clause constructions. Both constructions contain high frequency verbs, which potentially support the development of verb-specific schemas. However, only the frequent verbs in the transitive constructions occur with a variety of subject types, which also supports the development of a slot in the subject position. We then used an elicited production task to compare 4- and 5-year-old German-speaking children's productivity with simple transitive constructions and complement-clause constructions. The children were prompted to change the subjects of high and low frequency simple transitive verbs, such as essen ‘eat’ and naschen ‘nibble’, mental-state complement-taking verbs, such as denken ‘think’ and vermuten ‘presume’, and communication complement-taking verbs, such as sagen ‘say’ and berichten ‘report’. In accordance with earlier findings, children had less difficulty producing new utterances with high frequency transitive verbs than with low frequency transitive verbs. For the other verb classes, however, we found either reverse frequency effects or no frequency effects. For these verb classes, children's productivity can be determined by diversity rather than simple token frequency. We discuss how token frequency interacts with diversity, discourse function, semantic complexity, and syntactic complexity. (shrink)

While many linguists view language as either a cognitive or a social phenomenon, it is clearly both: a language can live only in individual minds, but it is learned from examples of utterances produced by speakers engaged in communicative interaction. In other words, language is what (Keller 1994. On language change: The invisible hand in language. London: Taylor & Francis) calls a “phenomenon of the third kind”, emerging from the interaction of a micro-level and a macro-level. Such a dual perspective (...) helps us understand some otherwise puzzling phenomena, including “non-psychological” generalizations, or situations where a pattern which is arguably present in a language is not explicitly represented in most speakers’ minds. This paper discusses two very different examples of such generalizations, genitive marking on masculine nouns in Polish and some restrictions on questions with long-distance dependencies in English. It is argued that such situations are possible because speakers may represent “the same” knowledge at different levels of abstraction: while a few may have extracted an abstract generalization, others approximate their behaviour by relying on memorised exemplars or lexically specific patterns. Thus, a cognitively realistic usage-based construction grammar needs to distinguish between patterns in the usage of a particular speech community (a social phenomenon) and patterns in speakers’ minds (a cognitive phenomenon). (shrink)

A tight connection between competence and performance is a central tenet of the usage-based model. Methodologically, however, corpus frequency is a poor predictor of acceptability – a phenomenon known as the “frequency/acceptability mismatch”. This article argues that the mismatch arises from a “methodological mismatch”, when simple frequency measures are mapped onto complex grammatical units. To illustrate, we discuss the results of acceptability judgments of go/come-v. The construction is subject to a formal constraint (Go see the doctor! vs. *He goes sees (...) the doctor), which results from its mandative semantics (directives, commissives). While a formal model makes no prediction with regard to gradient acceptability of bare (“grammatical”) go/come-v, the usage-based view assumes that acceptability is a function of compatibility with an abstract schema. The experimental ratings are compared with a number of corpus-derived measures: while acceptability is largely independent of (raw) frequency, it is not independent of frequency-related usage distribution. The results add to recent suggestions that the frequency/acceptability mismatch is substantially reduced if the syntactic complexity of a unit is appropriately captured in usage data. (shrink)