In addition to the bug pointed out by toolic, I believe that there is a further bug in as much as it [your code] will merely result in a frequency report for the first line, not the whole file as you expect - the line my @lines = <FILE>; won't slurp in the whole file, it'll merely result in only the first line i.e. all text up to the first end of line char, being read in, you need to undef the line delimiter as in Laurent_R's code i.e. undef $/;.
in reply to Find most frequently used word in text file.
A user level that continues to overstate my experience :-))