>>> from nltk import*
>>> def hedge(text):
	count = 0
	for w in range(len(text)):
		if count == 3:
			text[w] = "like"
			count = 0
		count= count + 1
	return text

>>> import nltk
>>> f=open("donquijote.txt", "rU")
>>> quijote=nltk.wordpunct_tokenize(f.read())
>>> f.close()
>>> quijotito=quijote[:20]
>>> hedge(quijotito)
['El', 'ingenioso', 'hidalgo', 'like', 'Quijote', 'de', 'like', 'Mancha', 'TASA', 'like', ',', 'Juan', 'like', 'de', 'Andrada', 'like', 'escribano', 'de', 'like', 'del']
>>>