How does typicality of category members affect the deductive reasoning? An ERP study

Abstract

The typicality effect describes a phenomenon whereby a typical item is easier to be judged as a member of a category than are atypical items. This effect has been intensively studied in the context of category verification tasks. The present study further investigated the typicality effect using our newly developed category-based deductive reasoning task. Subjects were required to judge whether an incoming stimulus had the properties described in the premise presented before. The stimuli were either typical or atypical members of four target semantic categories or were non-target stimuli. According to the ERP results, three phases were needed to determine whether the object has the property associated with the category in the premise. First, significant amplitude differences were seen between typical and atypical items at N1, which suggested that attention processing was influenced by the expectation in this up-to-down (deductive) process. The premise automatically induced the expectation of the prototype of one concept, i.e. the expectation for the prototype of birds was induced when the premise was "Birds possess the property C". Typical items (e.g., sparrow) were more similar to the prototype; hence, they were easier to be matched with the prototype induced by the premise than were atypical items (e.g., ostrich). Additionally, there was a dissociation between typical and atypical items at P2, which suggested that the participants' early detection of an item's category membership was influenced by the typicality. Thirdly, N400 effect is related to the process of semantic processes and determining whether the object has the property associated with the category in the premise. N400 mean amplitudes during the 300-500 ms epoch were significantly greater for non-target members than for target members, while words of lower typicality (atypical) evoked greater N400 amplitudes during the 350-450 ms epoch than did words of higher typicality (typical).

How types of premises modulate the typicality effect in category-based induction: diverging evidence from the P2, P3, and LPC effects

“…A similar typicality effect was found in other tasks, such as category-based deduction, semantic categorisation, and lexical decision6121314. For example, when participants were required to decide whether, or not, an incoming stimulus (conclusion) had the property described in the category presented before (premise), reaction time data revealed a significant typicality effect, and ERP data indicated that larger N1, P2, and N400 amplitudes were elicited by atypical words relative to typical words6.…”

“…For example, the verification of a category member was faster if the member was a typical item than if it was an atypical item4. This is known as the typicality effect, which has been examined in many semantic tasks, such as category or picture naming5, category-based verification or reasoning367, and reading and sentence production89. Unfortunately, the neural underpinnings of the typicality effect remain poorly understood.…”

“…For example, when participants were required to decide whether, or not, an incoming stimulus (conclusion) had the property described in the category presented before (premise), reaction time data revealed a significant typicality effect, and ERP data indicated that larger N1, P2, and N400 amplitudes were elicited by atypical words relative to typical words6. Recently, Wang and colleagues15 explored the typicality effect via different a categorisation method: that of either inclusion or exclusion.…”

Event-related potential parameters of category and property violations during semantic category-based induction

“…In a category-based deductive reasoning task using Chinese characters, participants were told that birds had the property X and then were asked to judge whether or not a sparrow or a table had the same property X. Lei et al (2010) found that the non-target category (table) evoked a larger frontal N400 than the target category (sparrow). ERP studies on semantic category-based induction by Liang et al (2010) also found that a noninduction evoked a larger frontal N400 than an induction.…”

On the influence of typicality and age of acquisition on semantic processing: Diverging evidence from behavioural and ERP responses

“…This so called typicality effect has repeatedly been demonstrated in written category-member-verification tasks where a semantic relation, including a superordinate and a subordinate item, is visually presented in form of a sentence (e.g., "A SPARROW is a BIRD"; Mervis and Rosch, 1981;Smith et al, 1974) or as a word pair (e.g., "BIRD -SPARROW"; Hampton, 1997;Kiran et al, 2007;Larochelle and Pineau, 1994). In addition, TYP effects have been found in semantic tasks involving category-based induction and deduction (e.g., Lei et al, 2010;Rein et al, 2010), visual living/non-livingdecisions (Morrison and Gibbons, 2006), category naming (Casey, 1992;Hampton, 1995), and in tasks involving both lexical and semantic processes like picture naming (Dell'Acqua et al, 2000;Holmes and Ellis, 2006), reading (Garrod and Sanford, 1977), sentence production (Kelly et al, 1986) or category-member-generation (e.g., Hernández-Muñoz et al, 2006). Concerning different forms of categories, TYP effects are not restricted to perceptual (e.g., GEOMETRIC FIGURES or COLOURS;Posner and Keele, 1968;Rosch, 1973a) or natural taxonomic categories (e.g., biological: FRUITS, ANIMALS or artefacts: FURNITURE, VEHICLES; Larochelle et al, 2000), but also exist in ad-hoc categories (e.g., "things to buy at the bakery"; Barsalou, 1983;Sandberg et al, 2012) and well- defined categories (e.g.…”