1 On Tense, Aspect and Genericity in English and Romanian: Insights from Language Acquisition Alexandra Cornilescu 1 & Alexandru Nicolae 1,2 1 University of Bucharest 2 “Iorgu Iordan – Al. Rosetti” Institute of Linguistics Abstract This paper offers a contrastive analysis of English and Romanian generic sentences, arguing that generics originate from a simple conceptual structure. We argue that the comprehension and production of generics require no functional structure, but rather merely the comprehension of lexical aspect in its simplest form, namely states. While English distinguishes generics from episodics via the progressive, Romanian’s imperfective lacks this distinction. Additionally, cross-linguistic variation is evident in the semantics of definite articles. English definite articles encode familiarity and maximality, aligning with constraints on kind reference, while Romanian articles, marked with phi-features, encode maximality and treat familiarity as an optional pragmatic feature. The paper also argues that there are good grounds to give up the covert Gen(eric) operator. Keywords: generic, episodic, aspect, progressive, imperfective, kinds 1. Preliminaries: proposal and aims 1 1.1. The cognitive problem Generic sentences express regularities about entities (objects and kinds) and about situations. Our study of genericity starts from the observation that the current linguistic scene shows a paradox in the description of generic sentences. Acquisition studies and cognitive studies more generally have proved that, despite their semantic complexity, generic sentences are acquired very early, at an age when it is not the case that children have developed all the functional categories of the languages, more specifically they have not put in place the quantificational system yet. Thus two-year-olds understand generic sentences with more ease and earlier than they can correctly interpret quantifiers and even cardinals. On the other hand, linguists assign complex semantic representations to generic sentences, starting from the obvious fact that generic sentences make intensional statements, which are not descriptions of the world. The truth of Mary is smoking a cigarette is not sufficient to establish the truth of Mary smokes. Some generics (1), of special interest in this paper, are more complex in that they involve reference to unobservable, but intuitive entities, known as kinds. Still other generics (2) also show an overt quantificational structure, containing a generic adverbial operator, i.e. a functional element with sentence scope. (1) a. Birds fly. [kind generic] b. Mosquitoes are irritating. (2) a. Typically, Germans drink beer. [overt quantificational generics] b. Generally, Mary prefers apples. In an attempt to give generics a uniform analysis, generic sentences have standardly been analyzed as quantificational structures, with the typical tripartite structure of generalized quantifiers: a sentence operator, a restriction and a nucleus, as in (3). It was and is generally still admitted (cf. Driemel et al. 2025 and references therein) that, when no quantifier is overt, as in sentences (1) above, the logical form (LF) of a generic sentence contains a silent quantificational adverb, notated Gen. Gen is a sentential operator, and it is the silent equivalent of quantificational adverbs like usually, typically, generally, often, many times, as a rule, etc. (3) a. Typically x [Germans (x)] (drink-beer (x)) b. Gen x [dog (x)] (bark (x)) 1 Abbreviations: BP = bare plural; CL = clitic; DP = determiner phrase; DKP = Derived Kind Predication; DOM = differential object marker; Engl. = English; ILP = individual level predicate; LF = logical form; MTS = Model Theoretic Semantics; PN = proper name; PT = present tense; Ro. = Romanian; SLP = stage level predicate; ET = event time; ST = speech time; RT = reference time (cf. Reichenbach 1947).