In 1973, Carroll and White showed that words acquired earlier in life tend to be processed faster and more accurately than later acquired words in a picture naming task. From that moment on, age of acquisition (AoA) was considered to be a relevant psycholinguistic variable, and several experimental studies have replicated the effect found by Carroll and White in many paradigms (e.g., picture naming-Barry, Holmes & Ellis, 2006;Lambon Ralph & Ehsan, 2006; written-word naming-Brysbaert, Lange, & Van Wijnendaele, 2000;Monaghan & Ellis, 2002a, 2002bMorrison & Ellis, 1995 lexical decision-Gerhand & Barry, 1999;Turner, Valentine, & Ellis, 1998). In fact, AoA seems to be the main variable that influences the speed of lexical processing, and its effect remains even when other variables, such as cumulative frequency, are controlled (Pérez, 2007). Recently, an extensive study conducted by Cortese and Khanna (2007) , and Portuguese (Vicente, 2004;Vicente, Castro, & Walley, 2008). It is also consistent throughout the entire life span, since it has been found in children (D'Amico, Devescovi, & Bates, 2001), young adults, and older subjects (Barry, Johnston, & Wood, 2006). Similarly, early acquired words are the last to deteriorate in neuropsychological disorders such as Alzheimer's disease (Sartori, Lombardi, & Mattiuzzi, 2005) and aphasia (Hirsh & Ellis, 1994). These studies highlight the vital importance of AoA in lexical processing and emphasize the need to control AoA in experimental and neuropsychological studies. For that purpose, researchers need to have access to AoA norms.There are two ways of collecting AoA data: (1) using adult estimates (see, e.g., Bird, Franklin, & Howard, 2001;Marques, Fonseca, Morais, & Pinto, 2007) or (2) analyzing the performance of children in naming tasks (e.g., Chalard, Bonin, Méot, Boyer, & Fayol, 2003). In the first method, adults are asked to estimate the age at which they possibly learned a given word. In the second method, the experimenter presents pictures to children of different age cohorts, and the average AoA is usually calculated when at least 75% of the children can name the picture (Morrison, Chappell, & Ellis, 1997). It has been shown that adult estimates are as valid and reliable as the data collected with children and that the norms provided by the two methods are highly correlated (Ghyselinck, De Moor, & Brysbaert, 2000;Morrison et al., 1997;Pind, Jónsdóttir, Gissurardóttir, & Jónsson, 2000). Moreover, the correlations between the two types of norms are still significant even when other variables, such as familiarity, frequency, and phonological length, are controlled . So, when it is not possible to obtain objective data, adult estimates seem to be a valid and reliable AoA measure.Current lexical databases of AoA norms use both methods. Table 1 summarizes the main AoA norms that have been published since 1997. There is only one published lexical database for Portuguese (Marques et al., 2007). In that database, the authors collected AoA estimates for 834Age-of-acquisit...

No abstract