When tokenizing a column in R how do I get the max number of words in each row using NGramTokenizer?

Question

I am trying to tokenize a column of position descriptions to show the most common 4 to six word phrases.
I have been using the NGramTokenizer function with a Weka_cont