The IPA transcription is shown beside each word. The middle column shows the known words (in bold). The column to the left shows the minimal pairs with these known words (in bold) that would be used in the correlated learning condition. The column to the right shows a similar list that would be used for the uncorrelated condition. In the subsequent MMN sessions, we would only use the minimal triplets (the top two lines). However, to make the learning task sufficiently challenging, we used extra known words that had a minimal pair (which we could allocate to either learning condition) and novel words that were minimal pairs with each other (these are shown in italics).