Are Faces “Special” Objects? Associative and Sem antic Priming of Face and Object Recognition and Naming

Abstract

Faces represent a “special” class of physically similar stimuli but it remains uncertain whether they are processed by cognitive systems that are functionally separate from those used for objects. This paper reports two experiments, which examine whether there exist qualitative differences in the semantic and associative priming of faces, “structurally similar” objects (living things), and “structurally distinct” objects (artefacts). Recognition was examined in Experiment 1 using the familiarity judgement task for faces and the object decision task for objects, and naming was examined in Experiment 2. Both experiments compared, within subjects, priming by associates (e.g. Eric Morecambe → Ernie Wise, lion → tiger and lock key) and priming by non-associates from the same semantic category (e.g. Keith Richards → Paul McCartney, bee → spider and nail-file → comb) against both “neutral” and unrelated prime conditions. Both experiments produced a remarkably similar pattern of results. For faces, there was a substantial priming effect from associates but no reliable priming from non-associates of the same semantic category. In contrast, both structurally similar and distinct objects were primed reliably by both associates and semantically (i.e. categorically) related non-associates. The results are interpreted within a model that proposes that the semantic representations of objects are inter-connected by abstracted superordinate categories, but that the representations of people (the elements of which, we propose, are specific biographical descriptive information units) are inter-connected by networks of inter-personal relatedness rather than by “categories” of celebrity.

Categorical priming of famous person recognition: A hitherto overlooked methodological factor can resolve a long-standing debate

“…This compares with SOA of 1000ms in Carson and Burton (2001), 1100ms in Barry et al (1998), and 500ms in Young et al (1994). Note that Bruce and Valentine (1986) reported very little difference in associate priming between 250ms and 1000ms SOA.…”

“…They reported significant priming only from close associates, with non-significant categorical priming, in a face familiarity decision (Experiment 2 and 3) and a face naming response (Experiment 3). Barry et al (1998) used a similar design, and also reported significant associative priming and nonsignificant categorical priming, in a face familiarity decision (Experiment 1) and face naming response (Experiment 2).…”

When false memories do not occur: Not thinking of the lure or remembering that it was not heard?

“…frequent co-occurrence) has been demonstrated in the field of familiar people recognition. Indeed, it has been shown that associative relatedness rather than category membership per se is responsible for the so-called "semantic priming" of face recognition (Barry, Johnston, & Scanlan, 1998;Young, Flude, Hellawell, & Ellis, 1994). Investigating the influence of such factors in the future should 12 contribute to a better understanding of why some lists are more effective in inducing the recall of critical lures when these lures are people's names.…”

Getting connected: Both associative and semantic links structure semantic memory for newly learned persons

“…Thus, while Angelina Jolie and Brad Pitt are both highly associated (because they co-occur regularly) and categorically related (since both are Hollywood actors), it is possible to combine two actors who are not associated because they never appeared in the same movie (such as Hugh Grant and Brad Pitt). While initial studies did not observe priming for categorically related pairs (Barry et al, 1998;Young, Flude, Hellawell, & Ellis, 1994), a number of more recent experiments demonstrated this effect (Carson & Burton, 2001;Stone, 2008;Stone & Valentine, 2007;Wiese & Schweinberger, 2008, 2011. The typically observed "associative boost", which reflects stronger priming for associated (e.g., Jolie !…”

“…In the example above, the presentation of Angelina Jolie's name or face will pre-activate the PIN of Brad Pitt, because the two are interconnected via shared semantic units containing categorical information (e.g., both are 'actors'). By contrast, if semantic information is not shared but stored in separate units for each individual, such categorical priming should not exist (Barry et al, 1998;Ellis, 1992). Instead, only people who regularly co-occur, and are therefore connected via direct links, should prime each other.…”