Our project is concerned with the notion of the morphome (a term introduced into morphological theory in the 1990s (Aronoff 1994) and not to be confused with the morpheme). 'Morphomic' phenomena are such that their distribution cannot be attributed, synchronically, either to phonological or to functional conditioning. A classic example is the English 'perfect participle', a part of the verb whose form varies unpredictably from one verb to another, but which is consistently and recurrently used in at least two disparate functions: passive and perfect. Thus, for see, write, seek, read, tear, bring, take, sell, shear (etc.), each with different past participle forms, the same form always appears equally in the passive (It is seen/written/sought/read/torn/brought/taken/sold/shorn) and the perfect (I have seen/written/sought/read/torn/brought/taken/sold/shorn it). Another example, involving lexical roots rather than full word-forms, is Spanish decir 'to say', where the variant root forms dig-/dig-/ and dic-/di?-/ have an arbitrary distribution, the so-called L/U-pattern (Maiden 2018), shared with a large number of lexical verbs, that is 'incoherent' in terms of the set of features expressed (dig-/dig-/ is found in first-person singular of present indicative, imperative and throughout the present subjunctive; dic-/di?-/ everywhere else) or phonological conditioning. Such cases remain the centre of a debate in linguistic morphology which has major theoretical ramifications (cf. Luís & Bermúdez-Otero 2016; Herce 2023). An unresolved question in this debate is whether morphomic structures are cognitively real (i.e. whether they exist in speakers' grammar rather than being merely detectable by linguists) and, if so, what kind of mental representation is involved. Until very recently, the evidence for the psychological reality of morphomes had come principally from data derived from the historical morphology of Romance languages (Maiden 2018) and is therefore a matter of indirect comparative-historical inference. For example, the L/U-pattern is well-documented in the history of Romance languages where we find that, repeatedly, if any morphological modification affected one of the paradigm cells 'involved', it affected them all. These data imply speakers' capability to abstract distributional patterns of allomorphy and use them as models for the paradigmatic distribution of completely new types of alternant, so the pattern must have been internalized in their mental grammar. However, there cannot be any significant advance in our understanding of morphomic structures and their place in the architecture of grammar until we can complement indirect diachronic evidence with direct, synchronic, experimental evidence. This is what this project seeks to provide. Having already successfully completed a pilot experiment focusing on the L/U-pattern in Italian, we now propose to develop and carry out the very first comprehensive programme of research into the psychological reality of morphomes which involves new behavioural, eyetracking, and EEG experiments, and aims at testing a wider range of languages. This highly innovative project will make a substantial empirical contribution to our understanding of a major and controversial area of morphological and general linguistic theory and will open unexplored avenues of enquiry in the cognitive sciences and applied linguistics, where the significance of morphomes has still to be fully appreciated.