This two-part paper is concerned with the processing of two types of compound future in Polish, with infinitival and participial complements. In the first part we present a design and predictions of an ERP study whose goal was to monitor the EEG correlates of two types of temporal mismatches: i) tense mismatches between the future auxiliary and the past tense modifier wczoraj (‘yesterday’) relative to the jutro (‘tomorrow’) baseline and ii) aspect mismatches between the future auxiliary and the perfective aspect of the lexical complement relative to the imperfective baseline. In addition, we wanted to assess whether matching tense specifications in different words of a sentence can cause grammatical illusions. To this aim, we tested whether the presence of the adverb wczoraj (‘yesterday’) (specified for [past]) could give rise to an illusion of grammaticality for perfectives as  l-participles (allegedly [past] marked), but not as infinitives (not having any [past] specification). The study and its results as well as a general discussion of the findings will be presented in Part II of the paper. 

While electronic corpora may not seem adequate sources for anglicisms retrieval, since despite promising attempts they still lack readily available and efficient tools for foreign loans identification, they are indispensable in a systematic verification of the use of preidentified loans. The article offers an assessment of an electronic corpus of Polish in reference to its usefulness for the study of English loans. Though we test a selected corpus and its tools, and use Polish anglicisms as exemplifications, the findings presented in the article pertain to other large corpora and anglicisms in other languages. Corpus tools allow for a multidimensional analysis of loans, yet they fail to meet the requirements of more in-depth analyses of anglicisms, related to their semantics and structure. The limitations of corpora tools will be illustrated with authentic attempted-but-failed corpus searches.

In classic generative phonology (The Sound Pattern of English, Lexical Phonology) underlying representations and associated rules account for generalizations of two types: alternation-based generalizations and distribution-based generalizations. This article addresses the issue of how distribution-based generalizations are handled in Standard Optimality Theory and in Derivational Optimality Theory. The former uses the principle of the Richness of the Base, the latter relies on underspecification. It is argued that the Richness of the Base and the associated principle of Lexicon Optimization are unable to provide an adequate analysis of three types of generalizations: Nasal Assimilation in English, Vowel Retraction in the process of assimilating borrowings into Polish, and a presonorant voicing process called Cracow Voicing.