The psychological validity of non-adjacent collocations: Evidence from event-related brain potentials

The overarching aim of my PhD thesis was to find out whether or not there is a difference in the way that the brain processes pairs of words where the second word is highly likely to follow the first word (e.g. clinical trials), compared to pairs of words where the second word is not likely to follow the first word (e.g. clinical devices). In order to achieve this aim, I carried out a series of four EEG experiments. EEG, short for "electroencephalography", is a method which involves placing electrodes across a participant's scalp. These electrodes detect the participant's brain activity while they are carrying out a particular task. In my experiments, the task was to read sentences presented one word at a time on a computer screen. Within these sentences, I embedded pairs of words where the second word is highly likely to follow the first word (these are known as collocations, or collocational bigrams), as well as pairs of words where the second word is not likely to follow the first word (these are known as non-collocations, or non-collocational bigrams).

The results of my PhD experiments show that there is a difference in the way that the brain processes these two different types of word pairs. When native speakers of English read non-collocational bigrams, their brain seems to work much harder than when they read collocational bigrams. This makes sense, as reading a non-collocational bigram violates the expectation of which word is likely to come next that was set up by the first word in the pair. The results also show that, when non-native speakers of English read the non-collocational bigrams, their brain seems to work even harder than that of the native speakers of English exposed to the same stimuli. This can be explained by considering the amount of exposure that both participant groups have had to the English language. Native speakers are likely to have encountered the non-collocational word pairs before, even though they are highly infrequent, while non-native speakers have probably never encountered these word pairs before. The collocational violation is therefore likely to be more extreme from the perspective of a non-native speaker.

My PhD thesis focused on word pairs rather than longer sequences of words in order to reduce the number of factors that might influence how the word sequences were processed, making it feasible to conduct controlled experiments. However, this is actually a very narrow way of conceptualizing the notion of collocation; in practice, words are considered to form collocations when they occur in one another's vicinity even if there are several intervening words, and even if the words do not always occur in the same order. In the additional research that I intend to undertake during the term of the Fellowship, I will therefore focus on these non-adjacent collocations.

Specifically, in an initial EEG experiment, I aim to find out whether or not there is a difference in the way that the brain processes collocations with one intervening word (e.g. take something seriously), compared to matched non-collocations of the same length (e.g. take something happily). If I find that there is a distinct processing difference, I will then carry out another EEG experiment, this time looking at collocations with two intervening words (e.g. take the matter seriously). I will use the results of these two experiments to inform the design of research questions and methods for future work (after the funded period) engaging with yet more varied types of collocational pattern. My findings, present and to come, will have a number of key implications for both our understanding of how language works in the mind, and for practical aspects of language teaching (since the teaching of collocation is a major part of most current approaches to teaching advanced language learners).