Ch20. Constructional change and motion charts

the summary of:

Levshina,N.(2015). 20.Constructional change and motion charts, How to do Linguistics with R:Data exploration and statistical analysis(pp 387-393), John Benjamins Publishing Company

20.1 The past and present of the future: Diachronic motion charts of be going to and will

20.1.1 Theoretical background and data

Add-on packages

library(Rling);library(googleVis)

Theoretical Background

• The auxiliary shall, which used to be obligatory with the first person subjects, has given way to will
  

• be going to has developed from a purposive motion construction to a full-fledged future marker

• Has be going to taken over some ‘territory’ of will?

Data structure

• the Distribution of the infinitives that follow be going to and will
  • any changes?
    
    
• Data
  • Ngrams from the corpus of Historical American English, or COHA(Davies 2011)
  • The infinitives were extracted from the negram lists of am/is/are going to+Inf and will+Inf
  • The auxilaries be and have were removed, only verbs that occur in all time periods were considered
    
    
• The dataset: fut

data(fut)
str(fut)

## 'data.frame':    6973 obs. of  4 variables:
##  $ Verb   : Factor w/ 367 levels "accept","accompany",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Period : int  1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 ...
##  $ will   : num  2.02 1.52 1.81 2.55 4.38 ...
##  $ goingto: num  0 0 0 0 0 ...

head(fut)

##     Verb Period     will goingto
## 1 accept   1820 2.021075       0
## 2 accept   1830 1.524613       0
## 3 accept   1840 1.807208       0
## 4 accept   1850 2.546407       0
## 5 accept   1860 4.379536       0
## 6 accept   1870 3.385248       0

20.1.2 Motion charts

Motion Charts

• the package googleVis, the function gvisMotionChart()
• needs a browser Flash and an Internet connection
• Three obligatory arguments of the functions
  • dataframe
  • idvar : the column in the data frame that contains the identities of the points
  • timevar: the column that corresponds to the time dimension

mch<-gvisMotionChart(fut,idvar="Verb",timevar="Period")
plot(mch)

the changes in the use of constructions

• the horizontal axis : the normalized frequencies of the verbs after will
• the vertical axis : after be going to
• the bubbles : individual verbs
• scale:linear or logarithmic scale
  
  
• the normalized frequencies of the verbs in the 1820s are much higher in the will construction
• In the 1900s, the frequencies of verbs in be going to future have increased
• In the 2000s, many verbs are now used more frequently with be going to
• Still, it would be wrong to say that be going to is the predominant marker because the weighted frequency of will is still greater than that of be going to

the changes in the use of individual verbs

• click on circles
• check the boxes corresponding to the individual verbs
• Trails
• come and go
• in case of go
  • initial meaning of directed motion is bleached
  • the grammaticalization of be going to was still going on
    
    
• Hilpert(2013:Section 2.4.4)
  • interaction of register and time
  • be going to increased its relative frequency in informal registers, whereas will came to be used more often in formal contexts