Native speaker intuition

For more than fifty years, generative linguistics have posited a notion of native speaker intuition based on Chomsky's (1965) concept of linguistic competence and the idealized native speaker. After fifty years of cognitive linguistic studies, socio- and applied linguistics, and developments in learning theory, most people studying SLA have abandoned the notion of native speaker intuition, although one still sees the term used in textbooks and classrooms to this day.

Instead, most teachers and researchers have developed a nuanced approach to native speaker intuition (and I will continue to use that term in this blurb). With ever growing exposure and experience with the target language, language learns acquire native speaker intuition in a manner similar to learners acquiring their first language — by gaining a sense of grammaticality first, then a growing awarness of appropriateness, and finally a sensitivity to idiomaticity in the target language.

As teachers and students of language, we can use corpora as a way of testing native speaker intuition. Consider the tools available at Mark Davies' corpus.byu.edu.

Figure 1: corpus.byu.edu Corpus tools to explore language quickly and empirically for free.

Choosing the COCA corpus, I can use several different tools to explore the use of a word. For example, assume I want to know which form of an English preposition, toward or towards, is the most common.

Figure 2: corpus.byu.edu The search tool

Figure 3: corpus.byu.edu The frequency tool

Figure 4: corpus.byu.edu The context (KWIC) tool

We can also do more complex, contextualized studies. For example, I would like to know which is the more common, the inflected genitive (such as an (or the) doctor's apointment) or the periphrastic genitive (as in a (or the) appointment with a (or the) doctor)

Figure 5: corpus.byu.edu Using "tags" to find an inflected genitive (COCA)

Figure 6: corpus.byu.edu Using tags to find periphrastic genitives (COCA)

The results shown in the screen captures above are for all of COCA, which includes 520 words of spoken as well as written English. The table below shows the results of the same search as above in written English only and academic English only.

written English only	inflected: 15    periphrastic: 6
academic English only	inflected:  4     periphrastic: 1

Search engines can also be used as tools to study word or phrase frequency by using Boolean operators, either directly or through the advanced search options page. Studying the word or phrase is much more difficult since search engines will only supply links to pages that contain the search query, and they have no ability to show the KWIC content as a corpus tool can.

Figure 7: google.com Using Boolean operators to find an inflected genitive

Figure 8: google.com Boolean operators for the matching periphrastic genetive

Notice that the results in the search engine are the opposite of the results in COCA. This apparent contradiction allow us to end on an important point: the corpus we choose to investigate the language is just as important as the tools we use.