Try:
tr -s ' ' <text.txt | cut -d ' ' -f4
From the tr man page:
-s, --squeeze-repeats replace each input sequence of a repeated character
that is listed in SET1 with a single occurrence
of that character