conceptsC.4.a.(1). Literature review and background Studies in normal subjects have shown that concrete words, compared to abstract ones, have anearlier age of acquisition (Brown, 1957; Centner, 1982), are easier to learn and remember (Paivio, 1991), andare processed faster in lexical decision tasks (James, 1975; Kroll & Merves, 1986). This phenomenon, termedthe 'concreteness effect,' has been addressed at length in several cognitively oriented theoretical accounts(Kieras, 1978; Paivio, 1971, 1991; Plaut & Shallice, 1993; Schwanenflugel, 1991). A strong effect ofconcreteness has also been demonstrated in many brain-damaged patients (see Saffran & Shod, 1999, forreview). For example, aphasics show poorer comprehension and production of abstract words, compared toconcrete ones (e.g., Crutch & Warrington, 2005; Franklin et al., 1995; Goodglass et al., 1969; Goodglass &379Project 5: Memory and Language DAMASIO, ANTONIO R.Wingfield, 1997; Marshall & Newcombe, 1966). This same pattern is typical of patients with deep dyslexia anddeep dysphasia (e.g., Coltheart, 1980; Coltheartet al., 1980; Crutch, 2006; Franklin, 1989; Howard & Franklin,1988; Martin, 1996; Morton & Patterson, 1980). Comparable results have been obtained in regard to high-imageable versus low-imageable words, where high-imageable words tend to have an advantage (e.g.,Franklin et al., 1992; Marshall et al., 1996). Crutch and Warrington (2005) argued that abstract and concreteword meanings are based in representational systems that have qualitatively different properties. Given the potency of the concreteness effect, it is intriguing to find instances in which patientsdemonstrate the opposite pattern, namely, superiority in processing abstract words and concepts (e.g.,Warrington, 1975; Warrington & Shallice, 1984). One of our patients defined duck as 'has fur, four legs, and isabout 4 feet tall'; for health, though, he said, 'the way you are feeling...you should have this to live a longtime.' Other cases of superiority for abstract concepts have been reported (Breedin et al., 1994; Carbonnet etal., 1997; Cipolotti & Warrington, 1995; Marshall et al., 1996; Sirigu et al., 1991; Srinivas et al., 1997). Theavailable literature suggests that a relative superiority for concrete concepts is associated with damage to theperisylvian region of the left hemisphere, whereas superiority for abstract concepts is linked to bilateraldamage to the anterior and inferior temporal lobes. This pattern has received some mixed support fromfunctional imaging studies (Beauregard et al., 1997; Chertkow et al., 1998; Friederici et al., 2000; Noppeney &Price, 2004; Wise et al., 2000). It has been shown that simple imagery of the referents of concrete words(compared to abstract ones) activates structures in the left inferior temporal region (D'Esposito et al., 1997).Binder and colleagues (2005) found that abstract words, relative to concrete words, activated left inferior frontalregions; in general, concrete concepts triggered greater involvement of bilateral association areas, andabstract concepts triggered almost exclusively left hemisphere activation. Whatmough et al. (2004) found adifference between abstract and concrete concepts in the lateralization of fusiform activation (abstract-right;concrete-left); however, both types of concepts activated the left temporal cortex. Fiebach and Friederici(2004) found that abstract words activated a subregion of the left inferior frontal gyrus more strongly thanconcrete words, whereas
Showing the most recent 10 out of 353 publications
