S1 (FriAM 1-65) - The Psychonomic Society

Papers 131–137 Saturday Morning
Lexical Representations
Regency ABC, Saturday Morning, 8:00–9:20
Chaired by Stavroula-Thaleia Kousta, University College London
8:00–8:15 (131)
Believe It or Not, Abstract Words Have a Processing Advantage
Over Concrete Words. STAVROULA-THALEIA KOUSTA, DAVID P.
VINSON, & GABRIELLA VIGLIOCCO, University College London—
It is currently taken for granted that concrete words have a processing
advantage over abstract words. This advantage is assumed to arise because
concrete words are more imageable (Paivio, 1986) or have higher
context availability (Schwanenflugel & Shoben, 1983). In three lexical
decision experiments and an analysis of lexical decision and word naming
reaction times for approximately 2,000 words from the English
Lexicon Project (Balota et al., in press), we show that once imageability
and context availability (along with 15 other lexical and sublexical
variables) are controlled for, it is abstract words that have an
advantage over concrete words. We discuss how this advantage arises
as a result of differences in content and organization of semantic representations
for abstract and concrete words. These results force a radical
revision of our theories of the way in which abstract and concrete
lexical knowledge is acquired, represented, and processed.
8:20–8:35 (132)
Psycholinguistic and Corpus Investigations of Verbal Event Structures.
GAIL MCKOON & ROGER RATCLIFF, Ohio State University—Lexical
semantic representations for verbs of two classes are investigated:
“manner of motion” verbs (e.g., run) and “change of
location” verbs (e.g., arrive). The lexical structures of change of location
verbs are complex in that they include a location and a change
to that location by the entity engaging in the verbal event. The structures
of manner of motion verbs are simpler because they include only
an entity engaging in an activity. This differential complexity is
demonstrated empirically in three ways: Lexical access is significantly
faster for manner of motion verbs than for change of location
verbs, as shown by lexical decision response times; short-term memory
is significantly better for manner of motion verbs, as shown by response
times for yes/no recognition; and sentence comprehension is
significantly faster for manner of motion verbs, as shown by sentence
reading times.
8:40–8:55 (133)
The Advantage of Arbitrariness Over Systematicity in Language
Learning. PADRAIC MONAGHAN, University of York, & MORTEN
H. CHRISTIANSEN, Cornell University (read by Morten H. Christiansen)—Why
aren’t systematic sound–meaning mappings (i.e.,
sound symbolism) more prevalent in language? We propose that
sound symbolism is avoided because it is redundant alongside more
reliable contextual information. Neural network simulations corroborated
this suggestion: The model learned arbitrary mappings better
than systematic mappings when contextual information was also present,
as arbitrariness maximized the information present in the environment.
The patterns of errors for systematic mappings indicated
that individual form–meaning mappings were less well differentiated,
resulting in within-category confusion between items. We tested this
effect in an artificial language learning study, where adult participants
learned to match sentences to pictures of either objects or actions. In
the systematic condition, words referring to the same category (objects
or actions) sounded similar. In the arbitrary condition, there was
no within-category similarity. The human data confirmed the modeling
results, suggesting that arbitrary sound–meaning mappings facilitate
language learning in a contextually rich language environment.
9:00–9:15 (134)
Roses Are Red and White Boys Are Stuffy: Automatic Stereotype
Priming Reflects General Prejudice in the Culture. PAUL VERHAE-
GHEN, Georgia Institute of Technology, SHELLEY N. AIKMAN,
21
Gainesville State College, & ANA B. VAN GULICK, Brown University—It
has been argued that stereotype priming (response times are
faster for stereotypical word pairs, such as black–poor, than for nonstereotypical
word pairs, such as black–balmy) reflects an individual’s
level of prejudice. We show, in three experiments, that stereotype
priming effects associated with race, gender, and age can be explained
through semantic co-occurrence in the culture: (1) Once objective associative
strength between word pairs is taken into account, stereotype
priming effects become nonsignificant; (2) the relationship between
response time and associative strength is identical for social primes
and nonsocial primes; and (3) the correlation between associativevalue-controlled
stereotype priming and self-report measures of
racism, sexism, or ageism is near zero. The racist/sexist/ageist in all
of us appears to be a mere reflection of the surrounding culture.
Associative Learning
Regency DEFH, Saturday Morning, 8:00–10:20
Chaired by Dennis J. Delprato, Eastern Michigan University
8:00–8:15 (135)
Failures to Detect Cue-Independent Retroactive Interference.
DENNIS J. DELPRATO, Eastern Michigan University—Nearly a
century after the discovery of what is called retroactive inhibition (or
interference-RI), M. C. Anderson and collaborators offered the first
viable conceptual and operational framework for determining if inhibitory
processes indeed participate in this and related classes of forgetting.
The main theoretical hypothesis is that inhibition acts directly
on representations of target items; therefore, inhibition functions as a
cue-independent source of forgetting that is separate from associative
or cue-dependent effects such as competition/blocking and associative
weakening. Four experiments used paired words across two successive
lists in within-list manipulations of baseline and interference conditions.
Cue-independent RI could have occurred at 11 comparisons
over the experiments when cues not episodically associated with targets
under interference conditions were used on retention tests. Despite
(1) using a variety of cues and testing conditions (e.g., paced,
unpaced, forced) and (2) obtaining cue-dependent RI in all experiments,
cue-independent RI was never found.
8:20–8:35 (136)
A Unified Theory of All-or-None (A-O-N) and Incremental Learning
Processes. CHIZUKO IZAWA, Stanford University—The A-O-N
versus incremental learning debate was examined under study–test–rest
(S–T–R) Presentation-Programs SSSSSSST, SSST, ST, S, and
STTTTTTT in a relatively simple paired-associate learning (PAL) situation.
Learning performances jumped from the unlearned to the
learned state in the A-O-N fashion, notwithstanding small nonsignificant
incremental trends during some precriterion trials. HR and GSR
data noted relaxation occurring abruptly a few trials before actual
learning. Thus, relaxation might facilitate mastery. Unlearned items
lacked such sudden jumps for all three response measures. An inclusive,
third perspective—that is, a unified theory—is advocated: A-O-N
learning can be regarded as a special case of incremental learning (cf.
Estes’ mathematical derivations, 1959). The former is associated with
simple/easy learning situations, whereas the latter involves complex/
difficult learning situations. The two views need not continue as a dichotomy
since they derive from two extremes of the same dimension—
namely, the simple/easy and complex/difficult learning situations.
8:40–8:55 (137)
A Couple of Weird Things About Feedback Effects in Paired-
Associate Learning. REED HUNT, University of Texas, San Antonio,
MELISSA D. MCCONNELL, University of North Carolina, Greensboro,
& REBEKAH E. SMITH, University of Texas, San Antonio—
Pashler, Cepeda, Wixted, and Rohrer (2005) reported that in pairedassociate
learning, immediate feedback did not affect performance on
initially correct items when a subsequent test occurred a week later.