Likert items

How to represent the data? As one column, with text factors, or as numbers, or aggregate counts, for example:

Don't agree Neutral Agree Question
342244Question 1
45134Question 2

(used by HH Likert function) How to automatically generate these counts (especially when you want to partition by other variables)?

library(HH)
data(ProfChal)
likert(ProfChal)

generates