#!/bin/bash

cd /net/corpora/nlnieuws

select i in `find . -name '*data.dz' | sort`
do
    alto "$i" fp:'//node[@cat="mwu" and node[@pt="spec"] and not(@his="normal") and not(@his_1="decap")]' tt:%w | \
        sort | uniq -c | sort -nr | head -n 40
done