If there are multiple words used the most frequently, the program just chooses the first match.

Perhaps the following doesn't address your general question, but it provides the option of showing all words with the highest frequency:

use strict;
use warnings;
use List::Util qw/max/;

my ( %words, %count );
my $file = $ARGV[0];

while (<>) {
    $words{ lc $_ }++ for split /\W+/;
}

push @{ $count{ $words{$_} } }, $_ for keys %words;
my $max = max keys %count;

print "The most frequent word(s) in $file: @{ $count{$max} }; Times se
+en: $max.\n"
[download]

This uses a hash to store word/count pairs. Next, it creates a hash of arrays (HoA), where the key is the count and the value is a reference to a list of words associated with that count. It then uses List::Util to find the max count, and that count is used to display the word(s) and frequency.

Hope this helps!