It is now time to put together the results of the empirical work that lies behind us, to produce an overview. Overall, 45 major analyses have been performed (21 synchronic, 21 diachronic, and 3 poetic). These major analyses may be divided into 83 subanalyses (41 synchronic, 42 diachronic). (Subdivisions of the poetic sections were not made.) None of the major analyses failed completely. The psycholinguistic predictions were borne out either in part or totally. Of the subanalyses, 79 (92 per cent) matched the predictions while 7 (8 per cent) did not. Even without a precise null hypothesis, there is no question that the obtained results cannot be attributed to chance.
The overwhelming majority of analyses strongly support the following main conclusion: The structure of language is shaped by the properties of the mechanism which puts it to use. This link between the processor and the final product allows us to take the former as an explanation of the latter. Language is the way it is because it satisfies processing constraints in probabilistic fashion. Psycholinguistics thus qualifies as one likely1 approach to the explanation of linguistic structure. The explanatory value of the psycholinguistic approach has to be seen at two levels. At the global level, the processing account gains in strength as a general explanatory principle for language structure. That is, whenever an explanation for a given linguistic phenomenon is sought, psycholinguistics is one promising area to turn to. Of course, this is not to say that psycholinguistics is the best of all explanatory approaches, but it is one that has proven quite successful over a wide range of data. At the local level, the uncertainty surrounding a processing account of an individual phenomenon is somewhat reduced. Alternative explanations are always conceivable; however, the success of the psycholinguistic approach at large increases its plausibility as an account of individual patterns. Again, there is no guarantee that it will provide the best explanation in every single case, but at least there is good reason to assume that it provides a probable one. With respect to future analyses, linguists might feel encouraged to look to psycholinguistics for explanations of language structure and change.____________________