Supporting web materials for 'Dynamic visualizations of language change'

Martin Hilpert, Freiburg Institute for Advanced Studies
martin.hilpertfrias.uni-freiburg.de

DOI: 10.1075/ijcl.16.4.01hil.additional

Chart #1: Ambicategorical elements

The first motion chart corresponds to Figure 5 in the paper. The chart is based on data from Mark Davies' COHA corpus. It shows the normalized text frequencies of ambicategorical elements such as work or love, which can be used both as a noun and a verb. The chart tracks whether any of these elements have become more 'verby' or more 'nouny' over time.

Chart #2: Complement-taking predicates

The second chart corresponds to Figure 8 in the paper. It shows 44 complement-taking predicates of English, verbs such as expect, imagine, try, and remember. These verbs project different syntactic complementation patterns, such as that-clauses, to-infinitives, ing-clauses, or simple noun phrases. Of course, different verbs have different preferences with regard to these structures. In the graph below, the verbs in the upper left show a preference for that-clauses, whereas the verbs in the upper right prefer to-infinitives. Verbs towards the bottom of the graph occur mostly with noun phrases. Between the 1870s and the 2000s, the overall configuration of verbs seems to stay rather stable, but there are several interesting developments going on nonetheless: Some verbs change their preferences - check out confirm or dislike! Also, note that towards the 2000s, the middle area of the graph becomes more and more populated. This happens because more and more verbs show an increasing preference for ing-clauses. This has been discussed in the literature as The Great Complement Shift, and we see it in the COHA data.

Making your own motion charts

You can download an R workspace with the data for the above charts here; the googleVis package comes with a demo that allows you to create these charts with literally two lines of code. If you feel you need help, watch the video clips below.