Managed to figure something out:
# How to load a text file as a corpus.
import nltk
from nltk.corpus import PlaintextCorpusReader
from nltk.corpus.util import LazyCorpusLoader
afrikaans = LazyCorpusLoader('afrikaans', PlaintextCorpusReader, r'(?!\.).*\.txt')
afrikaans.sents()[1]
af = nltk.Text(afrikaans.words())
af.concordance("mense")
This assumes your corpora text file is in C:\nltk_data\corpora\afrikaans\afrikaans.txt