Variation and Bulgarian Inner Aspect Introduction: Slavic aspect often focuses on the effects that prefixes have on the verb (see Filip 1999, Schoorlemmer, 1995, Svenonius 2004 etc.). Nevertheless, there is a subset of verbs whose aspect is not dependent on prefixes: biaspectual verbs. In this talk we outline systematic aspectual differences between Bulgarian standard verbs and biaspectual verbs, and note the consequences for the "size" of impact of a parameter (Baker 2008), a notion which has been questioned since at least Kayne (2005), since it may not be very informative. Biaspectual Bulgarian verbs: There are three systematic properties that Bulgarian biaspectual verbs show that standard verbs do not: 1. The internal argument of the verb affects the aspectual interpretation of the predicate (Slabakova 1997), shown in (1). While both the time span and durational adverbial are compatible with a count noun internal argument, as in (1a), when the internal argument is a mass noun, only the durational adverbial is compatible, as in (1b). 2. Bare plurals (BPs) elicit a multiple events interpretation, indicated by the compatibility of both the time span and durational adverbial at the same time, shown in (2). 3. Goal prepositions can turn an atelic predicate into a telic predicate, illustrated in (3). Standard Bulgarian verbs: Standard (non-biaspectual) verbs in Bulgarian pattern with Slavic verbs quite generally (see, for example, Svenonius 2004). They have simplex imperfective forms, which (ignoring habitual interpretations) are atelic, behaving like activities, as in (4a). With the addition of a lexical prefix, the predicate becomes telic (Markova 2007), as in (4b). Furthermore, they systematically lack the three properties that biaspectuals show: 1. The internal argument does not affect the telicity of the predicate, shown in (5). Note that bare nouns are incompatible with prefixed verbs in Bulgarian, shown in (5a); yet, it is not clear that the definite article here determines the aspect of the predicate (see Jackendoff 1996 for English, Nishihda 1994 for Spanish); minimally a count noun does not affect the atelicity of a simplex imperfective verb, shown in (5b). 2. BPs do not elicit a multiple events interpretation, illustrated in (6). 3. A goal P does not turn an imperfective (atelic) predicate into a telic predicate, shown in (7a); a prefix is necessary, shown in (7b). The account: English eventives show the same three properties that Bulgarian biaspectuals show. MacDonald (2008, in press) accounts for the English properties by claiming that they are all dependent on an aspectual projection, AspP, between vP and VP; each establishes distinct relations with AspP. He also observes that English statives systematically lack these three properties; he claims English statives lack AspP. Along these lines, we claim that Bulgarian biaspectuals possess AspP and Bulgarian standard verbs lack AspP. AspP shows a clustering of properties tied to an abstract element, which is something we expect for an abstract parameter (Kayne 2000). Moreover, language variation reduces to the presence vs. absence of an element, along the lines of Chomsky (2000) regarding the EPP feature. The scope of AspP: Both English and Bulgarian have AspP, and when AspP projects the same properties are found. When it does not project, we find different properties: In English, stative predicates result; In Bulgarian, the standard Slavic paradigm results. This difference can be conceived of in terms of the scope of AspP. AspP in English scopes over all verbs, thus, the impact of AspP is reflected in all verbs; they are either stative or eventive. In contrast, in Bulgarian, the scope of AspP is restricted to a well-defined subclass of verbs: biaspectuals, which are borrowings (Slabakova 1997, Stambolieva 2008). In these terms, the "size" of impact of AspP is larger in English than in Bulgarian, since it scopes over all verbs; AspP is more macro-like in English and more micro-like in Bulgarian. In a sense, AspP is both a macro- and a microparamter. This conclusion only makes sense if the macro/micro distinction is not absolute, but relative (perhaps to the degree of (historical) relatedness between languages, in the sense of Kayne (2005)). Moreover, in order for the scope of AspP to be distinct, it must be activated differently in English and Bulgarian. One possibility, if activation of parameters is a part of the acquisition process (Roberts & Roussou 2003), is that the time at which a parameter is activated affects the scope it has over the language.