An activation-verification model for letter and word recognition: The word-superiority effect

Abstract

Developed an activation–verification model for letter and word recognition that yields predictions of 2-alternative forced-choice performance for 864 individual stimuli that are either words, orthographically regular nonwords, or orthographically irregular nonwords. The model explains why letters embedded in words are recognized more accurately than letters embedded in nonwords. The encoding algorithm uses empirically determined confusion matrices to activate units in both an alphabetum and a lexicon. Predicted performance is enhanced when decisions are based on lexical information, because activity in the lexicon tends to constrain the identity of test letters more than the activity in the alphabetum. Thus, the model predicts large advantages of words over irregular nonwords, and smaller advantages of words over regular nonwords. The predicted differences demonstrate that the effects of manipulating lexicality and orthography can be predicted on the basis of lexical constraint alone. Within each class (word, regular nonword, irregular nonword) there are significant correlations between the simulated and obtained responses on individual items. This model is contrasted with the interactive activation model of J. L. McClelland et al . (29 ref)

Full text

Psychological Review
1982,
VoT 89, No. 5,
573-594
Copyright
1982
by the
American Psychological
Association,
Inc.
0033-295X/82/8905-0573$00.75
An
Activation-Verification
Model
for
Letter
and
Word
Recognition:
The
Word-Superiority
Effect
Kenneth
R.
Paap,
Sandra
L.
Newsome,
James
E.
McDonald,
and
Roger
W.
Schvaneveldt
New
Mexico
State
University
An
activation-verification
model
for
letter
and
word recognition yielded
predic-
tions
of
two-alternative forced-choice performance
for 864
individual stimuli that
were
either words,
orthographically
regular nonwords,
or
orthographically irreg-
ular nonwords.
The
encoding algorithm (programmed
in
APL) uses empirically
determined confusion matrices
to
activate units
in
both
an
alphabetum
and a
lexicon.
In
general, predicted performance
is
enhanced when
decisions
are
based
on
lexical information, because activity
in the
lexicon tends
to
constrain
the
identity
of
test letters more than
the
activity
in the
alphabetum. Thus,
the
model
predicts large advantages
of
words over irregular nonwords,
and
smaller advan-
tages
of
words over regular nonwords.
The
predicted
differences
are
close
to
those
obtained
in a
number
of
experiments
and
clearly demonstrate
that
the
effects
of
manipulating lexicality
and
orthography
can be
predicted
on the
basis
of
lexical
constraint alone. Furthermore, within each class (word, regular nonword, irreg-
ular nonword) there
are
significant
correlations between
the
simulated
and ob-
tained performance
on
individual
items.
Our
activation-verification model
is
contrasted with McClelland
and
Rumelhart's
(1981)
interactive activation model.
The
goal
of the
activation-verification
model
is to
account
for the
effects
of
prior
and
concurrent
context
on
word
and
letter
recognition
in a
variety
of
experimental
par-
adigms
(McDonald,
1980;
Paap
&
Newsome,
Note
1,
Note
2;
Paap,
Newsome,
& Mc-
Donald,
Note
3;
Schvaneveldt
&
McDonald,
Note
4). An
interactive activation
model,
in-
spired
by the
same
set of
sweeping
goals,
has
recently
been
described
by
McClelland
and
Portions
of
this
research
were
presented
at the
meet-
ings
of the
Psychonomic
Society,
St.
Louis, November
1980;
the
Southwestern
Psychological
Association,
Houston,
April 1981;
and the
Psychonomic
Society,
Philadelphia, November
1981.
The
project
was
partially
supported
by
Milligram
Award
1
-2-02190
from
the
Arts
and
Sciences Research Center
at New
Mexico
State
University.
We
would like
to
thank
Ron
Noel, Jerry
Sue
Thompson,
and
Wayne
Whitemore
for
their
contribu-
tions
to
various stages
of
this research. Also,
we
appre-
ciate
the
thoughtful reviews
of a first
draft
of
this
paper
provided
by Jay
McClelland,
Dom
Massaro,
and
Garvin
Chastain.
Sandra
Newsome
is now at
Rensselaer
Polytechnic
Institute
in
Troy,
New
York. James McDonald
is now
at
IBM in
Boulder,
Colorado.
Requests
for
reprints should
be
sent
to
Kerineth
R.
Paap, Department
of
Psychology,
Box
3452,
New
Mex-
ico
State
University,
Las
Cruces,
New
Mexico,
88003.
Rumelhart
(1981).
Although
the
models
complement
one
another
nicely
with
regard
to
some
aspects,
we
will
contrast
the two ap-
proaches
in our final
discussion
and
highlight
the
very
important
differences
between
them.
The
verification
model
was
originally
de-
veloped
to
account
for
reaction
time
data
from
lexical-decision
and
naming
tasks
(Becker,
1976,
1980;
Becker
&Killion,
1977;
McDonald,
1980;
Schvaneveldt,
& Mc-
Donald,
1981;
Schvaneveldt,
Meyer,
&
Becker,
1976;
Becker,
Schvaneveldt,
&
Gomez,
Note
5).
Although
the
various
dis-
cussions
of the
verification
model
differ
about
certain
details,
there
has
been
general
agreement
about
the
basic structure
of the
model.
The
basic
operations
involved
in
word
and
letter
recognition
are
encoding,
verification,
and
decision.
We
refer
to the
model
described
in the
present paper
as the
activation-verification
model
to
emphasize
the
extensive
treatment
given
to
encoding
processes
that
are
based
on
activation
of
let-
ter and
word
detectors.
The
activation
pro-
cess
shares
many
features
with
the
logogen
model
proposed
by
Morton
(1969).
In the
activation-verification
model,
we
have
at-
tempted
to
formalize
earlier
verbal
state-
573

574
PAAP,
NEWSOME,
MCDONALD,
AND
SCHVANEVELDT
ments about
the
verification model.
As we
will
show, this
formalization
permits
a
quan-
titative evaluation
of
aspects
of the
model
with
data
from
the
word-superiority para-
digm.
The
activation-verification model consists
of
encoding, verification,
and
decision
op-
erations. Encoding
is
used
to
describe
the
early
operations that lead
to the
unconscious
activation
of
learned units
in
memory.
In the
case
of
words,
the
most highly activated lex-
ical
entries
are
referred
to as the set of
can-
didate words.
Verification
follows
encoding
and
usually
leads
to the
conscious
recognition
of a
single
lexical
entry
from
the set of
candidates. Ver-
ification
should
be
viewed
as an
independent,
top-down
analysis
of the
stimulus that
is
guided
by a
stored representation
of a
word.
Verification
determines whether
a
refined
perceptual representation
of the
stimulus
word
is
sufficiently
similar
to a
particular
word,
supported
by the
evidence
of an
earlier,
less
refined
analysis
of the
stimulus. This gen-
eral definition
of
verification
is
sufficient
for
the
current tests
of the
activation-verifica-
tion model,
but
more specific assumptions
have
been suggested (e.g., Becker, 1980;
McDonald, 1980; Schvaneveldt
& Mc-
Donald, 1981)
and
could
be the
focus
of
fu-
ture work.
For
example, verification
has
been
described
as a
comparison between
a
pro-
totypical representation
of a
candidate
word
and a
holistic representation
of the
test stim-
ulus. However, within
the
framework
of our
model,
we
could just
as
easily suggest that
verification
involves
a
comparison between
the
letter information available
in an
acti-
vated word unit
and the
updated activity
of
the
letter units
in the
alphabetum.
The
verification process
has
been instan-
tiated
in a
computer simulation that mimics
the
real-time processing involved
in
verifi-
cation (McDonald, 1980).
The
simulated
verification
process
is a
serial-comparison
operation
on the set of
candidate words gen-
erated during encoding. Thus, verification
results
in a
match
or
mismatch.
If the
degree
of
fit
between
the
visual evidence
and the
candidate word exceeds
a
decision criterion,
then
the
word
is
consciously recognized.
If
the
match does
not
exceed
the
criterion, then
the
candidate
is
rejected
and the
next can-
didate
is
verified. Semantic context
affects
the
definition
of the
candidate set, whereas word
frequency
affects
the
order
of
verification
for
words
in the
candidate set. Those words
in
the
candidate
set
that
are
related
to the
con-
text
will
be
verified
before
those that
are
not.
If
the
verification process
finds no
match
among
the set of
related words,
it
proceeds
to
check
the
remaining candidates
in a de-
creasing
order
of
word
frequency.
These pro-
visions produce semantic-priming
and
word-
frequency
effects
in a
simulated lexical-de-
cision task.
The
upper panel
of
Figure
1
depicts
the
important structures
and
pro-
cesses that
are
simulated
for a
typical lexical-
decision task that involves normal stimulus
durations
of 250
msec
or
more.
The
factors
affecting
the
speed
and
accu-
racy
ofperformance
in a
particular paradigm
depend
on
whether decisions
are
based pri-
marily
on
information
from
encoding
or
from
verification. Because verification relies
on
a
comparison that involves continuing
perceptual
analysis
of the
stimulus,
the po-
tential
contribution
of
verification should
be
severely
attenuated whenever
a
backward
mask
overwrites
or
erases
the
sensory
buffer.
Thus,
paradigms that present masked letter
strings
offer
a
potential
showcase
for the
pre-
dictive
power
of our
simulated encoding pro-
cess.
The
bottom
panel
of
Figure
1
shows
the
reduced
model that
is
appropriate
for
very
short stimulus durations
or
stimuli that
are
masked.
Of
primary importance
is the
model's
abil-
ity
to
explain
why
letters embedded
in
words
are
recognized more accurately than
letters
embedded
in
nonwords.
The
current version
of
the
model predicts
not
only this word-su-
periority
effect
(WSE)
as a
general phenom-
enon
but
also
the
relative performance
for
any
given letter string.
The
predictions
are
derived
from
the
following
descriptions
of
the
encoding process
and the
decision rule.
Encoding
Feature
Matching
Like
many others,
we
view
encoding
as a
process that involves matching features
to
various
types
of
units.
The
model assumes
two
types
of
units: whole words stored
in a
lexicon
and
individual letters stored
in an

ACTIVATION-VERIFICATION
MODEL
NORMAL
STIMULUS
DURATIONS
AND NO
MASKING
575
VERY
BRIEF
STIMULUS
DURATIONS AND/OR
MASKING
Figure
1. The
upper
panel
shows
the
important
structures
that
the
model
simulates
for a
typical
lexical-
decision
task
that
involves
normal
stimulus
durations
of
250
msec
or
more;
the
lower
panel
shows
the
reduced
model
that
is
appropriate
for
very
short
stimulus
durations
and/or
stimuli
that
are
masked.
alphabetum.
Each letter
of the
alphabet
is
represented
by a
feature
list, with
the
relative
level
of
activation
for
each
letter
unit deter-
mined
by the
number
of
matching
and
mis-
matching features that have been detected.
Word
units
are
activated
to the
extent that
their constituent letters
are
activated
in the
alphabetum.
The
model also allows
for the
possibility that
the
detection
of
supraletter
features
(e.g., word shape
or
word length)
may
directly contribute
to the
activation
level
of
the
word units. However, because
the
present evaluation
of the
encoding process
consists
entirely
of
four-letter
uppercase
strings,
we
have assumed that there
are no
distinctive supraletter features.
It
is a
straightforward
matter
to
implement
a
simulation based
on
feature
matching.
However,
this strategy
is not
likely
to
succeed
because
the
selection
of the
appropriate
set
of
features relies heavily
on
guesswork.
If in-
appropriate
features
are
used,
a
bogus
set of
candidate words will
be
generated.
Confusion
Probabilities
as
Activation
To
avoid
the
problem
of
selecting
the
cor-
rect
set of
features,
the
activation-verifica-
tion model uses empirically determined con-
fusion
matrices
to
generate activation levels
in the
alphabetum
and
lexicon. Table
1
shows
the
obtained
confusion
matrix
for the
uppercase characters
we
used. Entries
are the
percentage
of
responses (columns)
for
each
letter
as a
stimulus (rows).
The
specific
pro-
cedure
used
to
obtain this matrix
has
been
reported elsewhere (Paap, Newsome,
&
McDonald, Note
3).
We
assume that
confusability
reflects
the
degree
of
feature matching
and the
appro-
priate rules
for
combining matching
and
mismatching information. This definition
of
activation emphasizes
the
role
of
psycho-
physical distinctiveness because
an
identity
match does
not
always lead
to the
same level
of
activation.
For
example, because
the
prob-
abilities
of a
correct response given
K,
S, and
Fas
stimuli
(K/K,
S/S,
&
VIV)
are
.748,
.541,
and
.397, respectively,
the
model assumes
that
S, a
letter
of
average confusability,
re-
ceives less activation than
the
more
distinc-
tive
letter
K,
but
more activation than
the
less distinctive letter
V.
All
of the
matrices used
to
generate pre-
dictions
are
transformations
of the
matrix
shown
in
Table
1.
Transformations
are ap-

Table
1
Confusion
Matrix
for the
Terak
Uppercase
Letters
ON
Stimulus
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
V
w
X
Y
Z
A
45
3
1
1
1
1
2
2
1
1
1
1
1
0
1
3
4
1
1
0
1
1
1
2
1
2
B
6
61
1
0
4
2
3
2
1
0
0
2
0
0
1
3
2
2
3
1
1
0
2
1
1
2
C
0
0
54
0
0
0
2
0
0
1
1
1
0
0
2
0
1
0
1
0
1
1
1
0
0
1
D
1
2
5
66
1
1
4
1
1
4
1
2
1
2
10
0
6
2
2
2
2
1
1
2
1
1
E
2
4
3
1
65
11
1
0
1
1
1
1
2
3
2
3
3
2
4
3
1
2
2
2
1
3
F
2
2
1
0
6
64
2
1
4
1
1
0
1
0
0
2
1
1
4
2
0
0
1
0
1
2
G
2
3
3
1
2
1
61
2
2
3
3
0
2
1
3
4
8
2
5
1
2
1
2
1
1
3
H
8
2
1
2
3
2
1
73
5
2
2
2
6
3
2
2
3
2
3
4
1
1
8
3
6
3
I
0
1
0
2
0
1
1
0
53
6
0
2
2
1
1
1
1
1
0
13
1
1
1
1
0
2
J
1
1
1
0
1
1
0
1
.
2
41
0
1
2
1
0
1
0
2
2
1
1
1
2
0
1
3
K
2
2
1
3
2
1
2
2
2
2
75
2
3
2
1
2
1
2
2
3
1
1
2
9
2
3
L
2
3
2
3
3
1
2
1
6
4
2
64
2
1
1
2
1
2
3
3
1
2
2
1
2
5
M
0
0
0
1
0
0
0
1
0
0
1
1
56
1
0
0
0
1
0
1
0
0
1
2
1
0
N
2
1
2
2
1
0
1
1
3
2
3
1
10
76
1
1
3
3
1
1
1
0
8
4
3
3
O
1
2
9
8
0
1
4
1
1
4
0
2
0
1
58
1
13
0
1
0
4
3
0
0
2
1
P
1
2
1
1
1
2
0
1
1
0
0
0
1
1
1
60
1
1
1
0
1
0
1
0
0
1
Q
1
1
2
1
0
0
3
0
1
0
0
0
0
0
6
0
36
0
0
0
1
0
0
0
0
0
R
16
4
3
0
3
3
1
2
2
2
1
2
2
1
1
9
6
69
5
2
2
1
1
2
2
2
S
2
2
3
3
4
2
1
1
1
1
0
1
1
0
1
1
2
1
54
2
0
1
1
1
1
3
T
1
0
1
0
0
1
2
1
6
4
1
2
1
0
1
1
1
1
1
56
2
0
1
1
2
10
U
1
2
2
2
0
0
1
1
2
11
1
5
0
2
2
1
0
4
1
1
64
35
2
1
5
3
V
0
1
0
0
0
0
0
0
1
2
0
1
0
0
0
0
1
0
0
0
5
40
1
0
1
1
W
0
1
0
1
1
1
0
1
1
2
1
2
2
1
0
0
2
0
2
1
3
3
53
2
3
1
X
1
1
1
0
0
0
0
1
1
2
2
1
2
1
1
1
0
0
1
0
0
0
1
61
1
1
Y
1
0
0
0
0
0
1
1
1
0
0
0
1
0
0
1
0
0
57
2
Z
1
0
1
1
1
0
1
1
1
1
0
1
0
0
0
1
1
0
0
1
1
1
1
2
1
39
>
Z>
S
08
i
-m
^
§
e
5
>
§
o
<
<
w
r
D
Note.
Entries
are the
percentages
of
responses (columns)
for
each letter
as a
stimulus (rows).

ACTIVATION-VERIFICATION
MODEL
577
plied
to
model
any
variable that
is
assumed
to
affect
stimulus quality.
For
example,
if the
onset
asynchrony
between stimulus
and
mask
is
greater than
the
17
msec used
to
generate
the
percentages shown
in
Table
1,
then
the
values
on the
main diagonal (for correct
re-
sponses) should
be
increased, whereas
the
off-
diagonal
values (for incorrect responses)
are
decreased.
The
particular adjustment used
increases each correct response percentage
by
a
percentage
of
the
distance
to the
ceiling
and
decreases each incorrect response percentage
by
a
percentage
of the
distance
to the floor.
The
increments
and
decrements
are
such
that
the
rows
always
sum to
100%.
The
procedure
is
reversed when stimulus quality
is
degraded
rather than enhanced.
Another
effect
that
the
model
can
capture
by
appropriate transformations
of the
basic
matrix
is
loss
of
acuity
for
letters
at
greater
distances
from
the
average
fixation
point.
All
of
the
predictions reported later access
sep-
arate matrices
for
each
of the
four
spatial
positions.
The
extent
to
which separate
ma-
trices improve
the
model's predictions
de-
pends
on
whether correlations between
ob-
tained
and
predicted data
are
based
on all
stimulus items
or
only
those that test
the
same
target position.
To
demonstrate this
we
derived
a
single matrix
in
which each cell
entry
was the
mean
of the
four
confusion
probabilities
found
in the
separate matrices.
When
the
single matrix
is
used,
correlations
between
predicted
and
obtained
perfor-
mance
are
significantly
higher
for the
subsets
of
stimuli that
all
share
the
same target
po-
sition than across
the
entire
set of
stimuli.
When
separate
confusion
matrices
are
used,
the
correlation
for the
entire
set of
stimuli
rises to
about
the
same level
as the
separate
correlations
on
each position.
As
an
example
of how the
encoding pro-
cess uses
the
confusion
matrices, consider
the
presentation
of the
input string
PORE.
As in-
dicated
in
Figure
2,
position-specific units
in
the
alphabetum
are
assumed
to be
activated
"PORE"
SENSORY
BUFFER
f
(
MCI
^
J
~N
dinn
1
oing
l
—
r^
\
VISUAL
REPRESENTATION
X
LEXICON
(GEOMETRIC
MEANS)
PORE
.533
PORK
.
276
GORE
.
275
BORE
.
254
LORE
.
245
POKE
.
242
}\\
m
ALPHABETUM
ENTRIES
AND
CONFUSION PROBABILITIES
Pos.
1
Pos.
2
Pos.
3
Pos.
4
P
.54 0 .66 R .58 E .39
R
.09 D .08 N .03 F .07
B
.04 a .04 H .03 s .05
A
.03
G
.03 B .03 B .05
B
.03 . K .03 L .04
H
.02 , E .02
R
.04
L
.02 .
e
.02 H .04
o
.03
K
.03
Figure
2.
Encoding
the
word
PORE.
(Activation strengths
for
letter
units
in the
alphabetum
are
determined
by
letter-confusion
probabilities.
Activation
strengths
for
word
units
in the
lexicon
are
determined
by
taking
the
geometric mean
of the
corresponding
letter-confusion
probabilities.)

578
PAAP,
NEWSOME,
MCDONALD,
AND
SCHVANEVELDT
in
direct proportion
to
their confusability.
In
the
first
position
the
input letter
P
activates
the
corresponding
P
unit
the
most (.538),
the
R
unit more than
any
other remaining unit
(.091),
and
several other units
(G, A, B,
H,
and
L)
to
lesser
extents.
Patterns
of
activation
are
established
in a
similar manner
for the
other three spatial positions.
Activity
in the
alphabetum
continuously
feeds
into
the
lexicon.
The
encoding algo-
rithm
estimates
the
activation strength
for
each
word
in the
lexicon
by
taking
the
geo-
metric mean
of the
activity levels associated
with
the
constituent letters.
One
consequence
of
using
the
geometric mean
is
that
one
very
inactive letter unit
(close
to
zero)
may
pre-
vent
activation
of a
potential word unit that
is
receiving high levels
of
activation
from
three other
letter
units.
This
may
mirror psy-
chological reality because otherwise identical
versions
of the
model
yield
poorer
fits to the
obtained data
if the
geometric mean
is re-
placed
by the
arithmetic mean
or the
square
root
of the sum of
squares (the vector dis-
tance between another word
and the
input
word
in a
space generated
from
the
letter-
confusion
probabilities).
The
Word-Unit
Criterion
The
decision system does
not
monitor
all
of
the
activity
in the
lexicon.
The
model
as-
sumes that
the
activity
in a
word unit
can be
accessed
by the
decision
system only
if the
level
of
activation exceeds
a
preset criterion.
The
predictions reported
in
this paper
are all
based
on a
word-unit
criterion
of
.24. With
this criterion word stimuli generate
an av-
erage
of
about
3.4
words
in the
candidate
set
compared
to
about
2.1
words
for
stimuli that
are
orthographically
regular pseudowords.
If
the
word-unit
criterion
is
raised,
fewer
words
will
be
accessible
to the
decision system.
In
our final
discussion
we
will
suggest that
a
high
criterion
may
offer
an
alternative explana-
tion
for the
pseudoword-expectancy
effect
reported
by
Carr,
Davidson,
and
Hawkins
(1978).
For the
example illustrated
in
Figure
2, six
word
units exceed
the
criterion
for the
input
word
PORE: PORE
(.533),
PORK
(.276),
GORE
(.275),
BORE
(.254),
LORE
(.245),
and
POKE
(.242).
Nonwords
can
also activate
the
lexi-
con
through
the
same mechanism.
For ex-
ample, when
the
pseudoword
DORE
is
input
to the
simulation, three word units exceed
a
geometric mean
of
.240:
DONE
(.268),
LORE
(.265),
and
SORE
(.261).
Nonwords with
lower
levels
of
orthographic structure tend
to
produce less lexical activity.
For
example,
when
EPRO
(an
anagram
of
PORE)
is
pre-
sented
to the
encoding algorithm,
no
word
units exceed
the
.240 criterion.
Decision
Decision
Criterion
If
the
task requires detection
or
recogni-
tion
of a
letter
from
the
stimulus,
the
decision
process
is
assumed
to
have access
to the
rel-
ative
activation
levels
of all
units
in the al-
phabetum
and
those
units
in the
lexicon
that
exceed
the
word-unit criterion.
It is
further
assumed
that when total lexical activity
ex-
ceeds some preset criterion,
the
decision
will
be
based
on
lexical rather than alphabetic
evidence.
This
decision
criterion
is
different
from
the
individual word-unit criterion,
and
the
distinction should
be
kept clearly
in
mind. Exceeding
a
word-unit criterion makes
that particular lexical entry accessible
to the
decision
system. Exceeding
the
decision
cri-
terion
leads
to a
decision based
on
lexical
activity
rather than alphabetic activity.
It
is
advantageous
to
base
a
decision
on
lexical
evidence when there
is
some minimal
amount
of
activation, because many words
can
be
completely
specified
on the
basis
of
fewer
features
than would
be
necessary
to
specify
their
constituent
letters
when pre-
sented
in
isolation. Accordingly, lexical can-
didates
will
tend toward greater veracity than
alphabetic candidates whenever decisions
are
made
on the
basis
of
partial information.
The
specific decision rules used
to
predict
performance
in a
two-alternative,
forced-
choice letter-recognition task
are as
follows:
For any
stimulus,
the
predicted proportion
correct (PPC) depends
on
contributions
from
both
the
lexicon
and
alphabetum. More spe-
cifically,
PPC is the
weighted
sum of the
probability
of a
correct response based
on
lexical evidence
and the
probability
of a
cor-
rect response based
on
alphabetic evidence:
PPC
=
P(L)
X
P(C/L)
+
P(A)
X
P(C/A),
(1)

ACTIVATION-VERIFICATION MODEL
579
where
P(L)
is the
probability
of a
lexically
based
decision,
P(C/L)
is the
conditional
probability
of a
correct response given that
a
decision
is
based
on the
lexicon,
P(A)
is the
probability
of an
alphabetically based deci-
sion,
and
P(C/A)
is the
conditional proba-
bility
of a
correct response based
on
alpha-
betic information. Because
the
decision
for
each
trial
is
made
on the
basis
of
either lexical
or
alphabetic information,
P(A)
is
equal
to
1
-
P(L).
Correct
Responses
From
the
Lexicon
The
probability
of a
correct response given
a
decision based
in the
lexicon
is
P(C/L)
= 1.0 X
(Swc/Sw)
+ .5
X
(Swn/Sw)
+ 0 X
(SWj/Sw),
(2)
where
Swc
is the
activation strength
of
word
units
that
support
the
target letter,
Swn
is the
activation strength
of
word
units that support
neither
the
correct
nor the
incorrect alter-
native,
Sw;
is the
activation strength
of
word
units that support
the
incorrect alternative,
and
Sw
is the
total lexical activity.
The
general expression
for
P(C/L)
shown
in
Equation
2 was
selected
for
reasons
of
parsimony
and
programming
efficiency.
The
equation
can be
viewed
as the
application
of
a
simple high-threshold model (Luce, 1963)
to
each lexical entry. When
a
word unit
ex-
ceeds
the
criterion,
the
decision system
will
(a)
select
the
correct alternative with
a
prob-
ability
of 1.0
whenever
the
letter
in the
crit-
ical position supports
the
correct alternative,
(b)
select
the
correct alternative with
a
prob-
ability
of 0.0
whenever
the
letter
in the
crit-
ical
position supports
the
incorrect alterna-
tive,
and (c)
guess whenever
the
critical
letter
supports
neither alternative.
The
only
addi-
tional assumption required
is
that
the
deci-
sion
system combine
the
probabilities
from
each
lexical entry
by
simply weighting them
in
proportion
to
their activation strengths.
For the
following
examples, words
had to
exceed
a
criterion
of .24 in
order
to be
con-
sidered
by the
decision system.
If
the
decision
for any
single
trial
is
based
on
lexical activity,
our
underlying process
model assumes that something like Equation
2
does apply. That
is, we
have adopted
the
working
hypothesis that
decisions
based
on
unverified
lexical evidence involve
a
weighted
strength
of the
word units supporting each
of
the
two-choice alternatives. Alternatively,
P(C/L)
could
be
viewed
as the
probability
of
certain word units being
the
most
highly
ac-
tivated units
on
individual trials.
We
note
as
an
aside that
our
general approach
has
been
to find a set of
simple algorithms (with plau-
sible
psychological underpinnings) that
do a
good
job of
predicting performance.
An al-
ternative
approach
is to
begin with
very
spe-
cific
ideas about
the
underlying psychological
processes
and
then derive algorithms
to
suit
these particular assumptions.
We
have shied
away
from
this latter strategy
in the
belief
that both
the
tests
and
selection
of
particular
psychological
explanations would
be
easier
once
we had
developed
a
formal model
that
predicts performance
in
several paradigms
with
a
fair
amount
of
success.
The
factors
that determine
the
probability
of
a
correct response
from
the
lexicon
can
be
easily understood
by
examining
specific
examples.
If the
stimulus word
PORE
is
pre-
sented
(see Figure
2) and the
third position
is
probed with
the
alternatives
R and K, we
have
P(C/L)
= 1 X
(1.583/1.825)
+ .5
X
(0/1.825)4-0
=
.867.
(3)
This relatively high probability
of a
correct
response
is
reasonable because
five of the
highly
activated words
(BORE, PORK, GORE,
LORE,
PORE)
support
the
correct alternative,
whereas
only POKE supports
the
incorrect
alternative.
In
general,
P(C/L)
will
be .70 or
greater
for
words;
but
exceptions
do
occur.
For
example, when
the
word
GONE
is
pre-
sented
to the
simulation,
the
following
words,
with
their activation strengths
in
parentheses,
exceed
the
cutoff:
DONE
(.281),
GONE
(.549),
TONE
(.243),
BONE
(.278),
CONE
(.256),
and
LONE
(.251).
If the first
position
is
probed
with
the
alternatives
G and B, we
have
P(C/L)
= 1 X
(.549/1.858)
+ .5
X
(1.031/1.858)+
0
=
.57
(4)
Lower
values
of
P(C/L)
tend
to
occur when
there
is a
highly
activated word that supports
the
incorrect
alternative
and/or
when there
are
several
highly
activated words that
sup-
port neither alternative.

580
PAAP,
NEWSOME,
MCDONALD,
AND
SCHVANEVELDT
Correct
Responses
From
the
Alphabetum
The
probability
of a
correct response given
a
decision based
on the
alphabetum
is
P(C/A)=
1.0X(ac/Sa)
+
.5
X
(San/Sa)
+ 0 X ta/Sa), (5)
where
«c
is the
activation strength
of the
letter
unit corresponding
to the
correct alternative,
San
is the
activation
strength
of the
letter
units
that
are
neither
the
correct
nor the in-
correct alternative,
and
Sa
is the
total
al-
phabetic
activity.
The
only
difference
be-
tween
the
decision
rule
for the
alphabetum
and
that
for the
lexicon
is
that alphabetic
activity
is not filtered by a
criterion.
Assuming
that
the
third position
is
probed
with
the
alternatives
R and K, the
P(C/A)
for
the
stimulus word
PORE
is
P(C/A)
= 1 X
(.585/1.000)
+ .5
X(.390/1.000)
+ 0 =
.780.
(6)
This
value would,
of
course,
be the
same
for
the
pseudoword
DORE,
the
anagram
EPRO,
or
any
other stimulus that contains
R in the
third
position.
Probability
of
a
Decision
Based
on the
Lexicon
For any
given trial,
it is
assumed that
a
decision
will
be
made
on the
basis
of
lexical
information
if
total
lexical activity exceeds
the
decision criterion. Given noise intro-
duced
by
variations
in the
subject's
fixation
or
attention,
and
within
the
visual processing
system
itself,
it is
reasonable
to
assume that
a
specific
stimulus
will
exceed
or
fall
short
of
the
decision criterion
on a
probabilistic,
rather
than
an
all-or-none,
basis. Accord-
ingly,
the
mathematical
instantiation
of our
verbal
model estimates,
for
each stimulus,
the
probability that
its
lexical activity
will
exceed
the
decision
criterion.
This probabil-
ity
will,
of
course, depend
on
both
the av-
erage
amount
of
lexical activity produced
by
the
stimulus
in
question
and the
current
value
of the
decision criterion.
The first
step
in
estimating
P(L)
normal-
izes
the
total lexical activity produced
by
each
individual stimulus
to
that stimulus that
produced
the
greatest amount
of
lexical
ac-
tivity.
Of the 288
words
that
have been used
as
input
to the
encoding algorithm,
the
word
SEAR
has
produced
the
greatest number
of
words above criterion
(9) and the
greatest
amount
of
total lexical activity (2.779). Thus,
normalization
involves dividing
the
total lex-
ical activity
for a
given stimulus
by
2.779.
Normalization
is
simply
a
convenience
to
ensure
that
the
amount
of
lexical activity
generated
by
each stimulus
will
fall
in the
range
of 0 to 1
and, consequently, that
P(L)
will
also
be
bounded
by 0 and
1.
Because
this
transformation
simply involves dividing
by
a
constant,
we are not
altering
the
relative
lexical
strengths that were
initially
obtained
by
summing
the
geometric means
of all
words
above
the
word-unit criterion.
In any
event,
we
certainly
do not
mean
to
infer
that
subjects
must somehow know
in
advance
the
greatest amount
of
lexical activity
that
they
will
experience during
the
course
of the ex-
periment. Rather,
we
simply assume that
to-
tal
lexical activity
is one
important
deter-
miner
of
P(L).
The
contribution
of the
decision rule
to
P(L)
is
reflected
by a
second step that raises
each
of the
normalized activation levels
by
a
constant power between
0 and
1.
This
yields
the
estimated
P(L)
for
each stimulus.
Stringent decision criteria
can be
modeled
by
using high exponents (near
1).
This
proce-
dure generates
a
wide range
of
P(L)
across
items,
and a
decrease
in the
average
P(L).
Lax
decision
criteria
can be
modeled
by
using
low
exponents (near
0). A
very
lax
criterion
compresses
the
range toward
the
upper
boundary
and
thus causes
the
mean
P(L)
to
approach
1.
Consequently, when
a
very
lax
criterion
is
used,
P(L)
tends
to be
quite high
for
any
level
of
lexical activity. Using
an ex-
ponential transformation
is a
convenient
way
to
operationalize
decision
rules
as
diverse
as
"use lexical evidence whenever
it is
avail-
able" (exponents near
0) to
"use lexical
ev-
idence only
for
those stimuli
that
produce
substantial amounts
of
lexical activity" (ex-
ponents near
1). All of the
predictions dis-
cussed later
are
based
on a
constant value
(.5)
for
this parameter.
Because
P(L)
is
derived
from
total lexical
activity,
it
will generally
be
the
case that stim-
uli
like
PORE
that excite
six
word units above
threshold will have higher
probabilities
than

ACTIVATION-VERIFICATION
MODEL
581
stimuli like
RAMP
which produce only
one
suprathreshold
word unit.
In
summary,
the
probability
that
a
decision
will
be
based
on
lexical evidence
is
estimated
for
each stim-
ulus
using
the
following
equation:
P(L)
=
(7)
where
W{
is the
total lexical activity
for
stim-
ulus
i,
Wmax
is the
total
lexical activity
for the
stimulus producing
the
greatest activity,
and
the
exponent
n is a
parameter that
reflects
the
stringency
of the
criterion.
P(L)
for the
stimulus
PORE
would
be
P(L)
=
(1.825/2.779)-5
=
.810.
(8)
When
the
exponent
«
is set to .5,
f\L)
for
word
stimuli will range
from
about
.4 to
1.0,
with
a
mean
of
about
.6.
Finally,
it is
assumed that when
total
lex-
ical activity
is
less than
the
criterion,
the de-
cision will,
by
default,
be
based
on
alphabetic
information.
Accordingly,
the
probability
of
an
alphabetic decision,
P(A\
is
P(A)=l-
P(L).
(9)
Predicted
Probability
Correct
Table
2
uses
Equation
1 to
show
the
der-
ivation
of the
overall probability
of a
correct
response
for
two
sets
of
stimuli. Each
set
con-
sists
of a
word,
a
pseudoword that shares
three letters
in
common with
the
word,
and
an
anagram
of the
word.
The first set was
chosen
because
it
produces predictions that
are
similar
to
most sets
of
words
and
non-
words
and
illustrates
why the
model
will
yield
different
mean PPCs
for
words, pseudo-
words,
and
anagrams.
The
second
set is ab-
normal
and
illustrates some principles that
account
for
variations within stimulus classes.
As
exemplified
by
PORE,
the
probability
of
a
correct response based
on
lexical evi-
dence
is
usually greater than
that
based
on
alphabetic evidence.
The
overall proportion
correct
falls
somewhere between
the
lexical
and
alphabetic
probabilities
and
will
ap-
proach
the
lexical value
as
P(L),
the
prob-
ability
of a
lexical decision, increases.
In
gen-
eral, words should provide
better
context
than nonwords
to the
extent that
(a)
P(C/
L) >
P(C/A)
and (b)
P(L)
is
high. Because
these
conditions
are met for the
stimulus
PORE,
the
model predicts
a
4.2% advantage
over
the
pseudoword
DORE
and a
6.6%
ad-
vantage over
the
anagram
EPRO.
The
model predicts that some words should
actually produce
word-inferiority
effects.
This
can
only occur,
as in the
example
LEAF,
when
lexical
evidence
is
poorer than alphabetic
evidence.
Because
the
probability
of a
lexical
decision
is
estimated
from
total
lexical activ-
ity,
regardless
of the
veridicality
of
that
in-
formation,
the
model predicts that
LEAF
will
be
judged
on the
basis
of the
inferior lexical
evidence about
two
thirds
of the
time. This
leads
to a
predicted 8.4% disadvantage rela-
tive
to the
pseudoword
BEAF
and a
6.1%
dis-
advantage relative
to the
anagram
ELAF.
Table
2
Simulation
of
Word, Pseudoword,
and
Anagram
Differences
for Two
Examples
Simulated values
ClassStimulus Alternatives
WSE SPC =
P(L)
X
P(C/L)
+
P(A)
X
P(C/A)
Typical
Word
Pseudoword
Anagram
Atypical
Word
Pseudoword
Anagram
PORE
DORE
EPRO
LEAF
BEAF
ELAF
R,
K
R,
K
R,K
F.P
F, P
F, P
+.042
+.066
-.084
-.061
.852
.810
.786
.621
.705
.682
=
.810
=
.535
=
.000
=
.591
=
.428
=
.000
X
.867
X.831
X
.000
X.677
X
.736
X.OOO
+
.190
+
.465
+
1.000
+
.323
+
.572
+
1.000
X.786
X.786
X
.786
X
.682
X
.682
X
.682
Note.
WSE =
word-superiority
effect;
SPC is the
simulated
proportion
correct;
P(C/L)
is the
probability
of a
correct
response
from
the
lexicon; P(C/A)
is the
probability
of a
correct
response
from
the
alphabetum;
and
P(L)
is the
probability
of
basing
a
decision
on
lexical information.

582
PAAP,
NEWSOME, MCDONALD,
AND
SCHVANEVELDT
Test
and
Evaluation
of the
Model
The
model
can be
tested
at two
levels.
First,
by
averaging
across
stimuli
in the
same
class,
the
model
can be
used
to
predict
the
magnitude
of the WSE for
words over pseu-
dowords
or
words over anagrams. Second,
the
model should
be
able
to
predict item vari-
ation within
a
stimulus class.
Four experiments provide
the
basis
for the
following
tests (Paap
&
Newsome, Note
1,
Note
2;
Paap,
Newsome,
McDonald,
&
Schvaneveldt,
Note
6). All
experiments used
the
two-alternative, forced-choice letter-rec-
ognition
task.
Each experiment
compared
performance
on a set of 288
four-letter words
to a set of 288
nonwords.
The
nonwords used
in
two of the
experiments were
orthograph-
ically
regular pseudowords.
In the
remaining
two
experiments,
the
nonwords were formed
by
selecting that anagram
for
each word stim-
ulus
that minimized
the
amount
of
ortho-
graphic structure.
The two
alternatives
se-
lected
for
each stimulus both formed words
for
word stimuli
and
nonwords
for the
non-
word
stimuli.
Word
and
Pseudoword
Advantages
Our
first
approach
to
evaluating
the
model
was
to use the
algorithm described
in the in-
troduction
to
predict
the
proportion correct
for
each
of the 288
words, pseudowords,
and
anagrams.
The
mean output
of the
model
for
words, pseudowords,
and
anagrams
is
shown
in
Table
3. The
simulation predicts
a
2.8%
advantage
for
words
(.841)
over pseudowords
(.813),
and an
8.6% advantage
for
words over
anagrams (.755). These
differences
compare
favorably
to the
obtained
WSEs
of
2.6%
and
8.8%,
respectively.
Across
all 288
words,
the
number
of
lexical
entries exceeding
the
cutoff
ranged
from
1
to 9,
with
a
mean
of
3.4. These word units
constrain
the
identity
of the
critical letter
more
effectively
than
it is
constrained
by the
activity within
the
alphabetum.
Thus,
the
word
advantages predicted
by the
model
occur because lexical information
is
used
63%
of the
time
and the
mean probability
of
a
correct
response
from
the
lexicon (.897)
is
greater than that based
on the
alpha-
betum
(.758).
The
major reason
why the
model yields
lower
proportions correct
for
nonwords than
words
is not the
quality
of the
available lex-
ical evidence,
but
rather
its
frequent absence.
That
is, the
probability
of a
correct response
based
on
lexical evidence
for the 253
pseu-
dowords that produce
at
least
one
word
above threshold
is
nearly identical (about
.90)
to
that
for
the 288
words. Similarly,
P(C/
L) for the 44
anagrams that
produce,
at
least
one
word above
the
cutoff
is
.94. Thus,
the
quantity
and not the
quality
of
lexical
infor-
mation
is the
basis
for the
WSE.
Orthograph-
ically
regular pseudowords excite
the
lexicon
almost
as
much
as
words
(2.1
vs. 3.4
entries)
and
lead
to
small word advantages, whereas
orthographically
irregular anagrams generate
much
less lexical activity
(.2 vs. 3.4
entries)
and
show much larger word advantages.
Item-Specific
Effects
The
model's ability
to
predict performance
on
specific
stimuli
is
limited
by the
sensitivity
and
reliability
of the
data.
Our
previous work
provides
two
sets
of
word
data
and one set
Table
3
Simulated
Values
for
Words. Pseudowords,
and
Anagrams
Lexical
class
Words
Pseudowords
Anagrams
PPC
.841
.813
.755
P(C/L)
.897
.791
.144
Simulated values
P(C/A)
.758
.758
.758
P(L)
.634
.415
.073
NW
3.4
2.1
.2
Note.
PPC is the
predicted proportion correct; P(C/L)
is the
probability
of a
correct
response
from
the
lexicon;
P(C/A)
is the
probability
of a
correct response
from
the
alphabetum; P(L)
is the
probability
of
basing
a
decision
on
lexical information;
and NW is the
number
of
words that exceeded
the
criterion.

ACTIVATION-VERIFICATION
MODEL
583
for
each
of the two
types
of
nonwords. Each
of
the 288
items
in a set was
presented
to 24
different
subjects. This means that
the ob-
tained
proportions
correct
for
individual
items
vary
in
steps
of
.04. Given these lim-
itations,
a
correlation
of
data against data
provides
an
index
of the
maximum
amount
of
variation that could
be
accounted
for by
the
model.
The
correlation between
the two
sets
of
word
data
was
.56.
A
similar
deter-
mination
of the
reliability
of the
pseudoword
and
anagram data yielded correlations
of .48
and
.39, respectively. However, because only
24
subjects
saw
each
nonword stimulus, these
lower
correlations
are
due,
in
part,
to the
fact
that each half consisted
of
only
12
observa-
tions compared with
the 24
available
in the
word
analysis.
Table
4
shows
the
correlations between
the
various
sets
of
obtained data
and the
values
generated
by the
model. Because each cor-
relation
is
based
on a
large number (288)
of
pairs,
significant
values
of
r
need only exceed
.12.
For all
three stimulus classes, there
are
significant
correlations between
the
obtained
data
and (a) the
predicted proportion correct,
(b)
the
probability
of a
correct
response
from
the
lexicon,
and
(c)
the
probability
of a
cor-
rect response
from
the
alphabetum.
The
cor-
relations
are
quite high considering
the
lim-
itations discussed above.
For
example,
the
correlation
between
the first set of
word data
and the
predicted
proportion
correct
is .30
compared
to .56
for
data against
data.
Taking
the
ratio
of the
squared values
of
these cor-
relations
(.09
and
.31, respectively)
leads
to
the
conclusion that
the
model
can
account
for
29%
of
the
consistent item variation (both
correlations
are
based
on 24
observations
per
data point,
and no
correction
for n is
needed).
As
a final
check
on the
model's ability
to
predict variation within words,
the 288
words
were
partitioned into thirds
on the
basis
of
their predicted performance,
and
mean
ob-
tained
performance
was
computed
for
each
group.
Obtained proportion correct
for the
upper
third
was .85
compared
to
.82,
and
.78
for the
middle
and
bottom
thirds.
The
source
of the
model's success
in
pre-
dicting
interitem
variation
is
difficult
to
trace.
Because
decisions
about word stimuli
are
made
on the
basis
of
lexical evidence more
often
than
on
alphabetic evidence,
P(L)
=
.63,
it is
clear that both
the
lexicon
and al-
phabetum
contribute substantially
to the
overall
PPC,
and
accordingly, both branches
must
enjoy
some predictive power
in
order
to
avoid diluting
the
overall correlation
be-
tween
obtained
and
predicted correct. Fur-
thermore,
it
should
be
noted that
the
corre-
lation
between
P(C/L)
and the
obtained data
is
quite sensitive
to the
word-unit criterion
(because
this
affects
the
average number
of
candidate words). This
is
consistent with
the
view
that
the
predictive power
of the
lexical
branch primarily depends
on
getting
the
cor-
rect
set of
candidate words
and is not a
simple
transformation
of
alphabetic activity.
The
item-specific predictions
are far
from
exact,
but
they
are
quite encouraging because
our
lexicon
contains
only
the
1,600 four-let-
ter
words listed
in the
Kucera
and
Francis
(1967)
norms. Because
P(C/L)
for any
item
is
determined
by the
activation strengths
of
visually
similar words
in the
lexicon, sub-
stantial variation
for
a
particular item
can be
Table
4
Correlations
Between Obtained Proportion
Correct
and
Simulated
Values
Stimulus
type
Words
Setl
Set 2
Anagrams
Pseudowords
PPC
+.30
+.26
+.37
+.35
P(C/L)
+.28
+.23
+.21
+.17
Simulated
values
P(C/A)
+.29
+.27
+.34
+.38
P(L)
-.05
+.01
+.17
+.15
NW
-.05
.00
+.14
+.16
Note.
PPC is the
predicted proportion correct;
P(C/L)
is the
probability
of a
correct response
from
the
lexicon;
P(C/A)
is the
probability
of a
correct response
from
the
alphabetum; P(L)
is the
probability
of
basing
a
decision
on
lexical information;
and NW is the
number
of
words that exceeded
the
criterion.

584
PAAP, NEWSOME, MCDONALD,
AND
SCHVANEVELDT
introduced
if
just
one
highly similar word
is
either added
of
deleted
from
the
lexicon.
Lexical
Constraint
The
test words consisted
of the 288
words
used
by
Johnston
(1978)
in his
influential
test
of
sophisticated-guessing theory. Half
of the
words
were
defined
by
Johnston
as
high-con-
straint words,
and the
other half
as
low-con-
straint words.
Johnston
assumed
that
lexical
knowledge
will
constrain
the
identity
of the
critical letter
in
inverse proportion
to the
number
of
different
letters that
will
form
words
given
the
remaining context.
For ex-
ample,
the
context
_ATE
supplies much
less
constraint than
the
context
_RIP
because
10
letters
form
words
in the
former
context,
but
only
three
in the
latter. Johnston rejected
the
hypothesis
that lexical constraint contributes
to the WSE
because performance
on the
high-constraint
words (.77)
was
slightly lower
than
performance
on the
low-constraint
words
(.80).
Our
model shows that when
the
same par-
tial information,
in the
form
of
letter-con-
fusion
probabilities,
is
provided
to
both
the
alphabetum
and
lexicon, lexical activity
can
support
the
critical
letter
more
often
than
does
the
alphabetic activity. This
difference
between
P(C/L)
and
P(C/A)
provides
an in-
dex
of the
potential amount
of
lexical
benefit
for
any
word.
We
view
this measure
of
lexical
benefit
as an
alternative definition
for the
global
concept
of
lexical constraint. Thus,
Johnston's
(1978)
conclusion that lexical
constraint does
not
contribute
to the WSE
may
have been premature
and the
product
of
a
less appropriate definition
of
lexical con-
straint.
Concerns that
we
have raised previ-
ously
(Paap
&
Newsome,
1980a)
can now be
extended
in the
context
of our
model
and the
alternative definition
for
lexical constraint.
Johnston (1978) obtained both free-recall
and
forced-choice responses. First, consider
those
trials
on
which
the
three context letters
were
correctly
reported.
The
conditional
probabilities
of a
correct critical-letter report
given
a
correct report
of all
three context let-
ters
were
.90 and .86 for
high-
and
low-con-
straint pairs, respectively. This
is
extremely
high
performance
for
free
recall,
and any
sig-
nificant
differences
due to
lexical constraint
may
be
obscured
by a
ceiling
effect.
More-
over,
if one
assumes that
the
same stimuli
presented
to the
same subjects under
the
same
conditions would yield performance
distributions with some variability, then
it
would
seem quite reasonable
to
characterize
these trials
as
samples that have been drawn
from
the
upper
end of the
distribution
and
that
reflect
trials
on
which
the
level
of
visual
information
was
unusually high.
When
stimulus information
is
high,
the
effects
of
lexical constraint
may be
low.
Our
model
makes exactly this prediction.
If
stim-
ulus
quality
is
enhanced
by
transforming
the
correct responses
in the
confusion matrices
upward,
and the
incorrect
responses down-
ward,
the
difference
between
the
lexical
and
alphabetic branches disappear.
For
example,
if
stimulus quality
is
raised
to the
extent that
the
probability
of
a
correct response based
on
the
alphabetum
is
increased
from
.758
to
.889,
the
advantage
of
lexical over alphabetic
evidence
decreases
from
13.9%
to
-.5%.
When
stimulus information
is low
(when
only
a
few
features
are
detected
in
each letter
location), lexical knowledge should
be
more
beneficial.
However,
when
the
subject
has
only
partial
information
about each letter,
Johnston's
(1978)
procedure
for
computing
lexical
constraint (based
on
complete knowl-
edge
of the
three
context
letters
and no in-
formation
about
the
target)
may no
longer
correlate with
the
lexical constraint provided
by
a
partial
set of
features
at
each letter
lo-
cation.
Our
analysis completely supports this
hypothesis:
Johnston's high-constraint words
yield
a PPC of
.830 compared
to
.852
for the
low-constraint
set. Furthermore,
the
average
number
of
word units exceeding criterion
is
exactly
the
same (3.4)
for
both sets
of
words.
It is
clear that there
is
absolutely
no
relation
between
the
number
of
letters that will
form
a
word
in the
critical position
of a
test word
(Johnston's
definition
of
lexical constraint)
and the
number
of
words that
are
visually
similar
to
that word (the candidate words
in
the
activation-verification model).
In
contrast, when lexical constraint
is de-
fined as the
amount
of
lexical
benefit,
the
effects
of
lexical constraint
are
apparent
in
the
data.
For
each
of the 288
stimuli
of
each
type,
we
subtracted
P(C/A)
from
P(C/L)
and
then
partitioned
the
stimuli into thirds
on

ACTIVATION-VERIFICATION MODEL
585
the
basis
of
these
differences.
For
both sets
of
word data
and the
pseudoword
data,
ob-
tained performance
on the
most highly con-
strained third
is
about
5%
greater than that
on the
bottom third. There
were
no
differ-
ences
for the
anagrams,
but
this
is to be ex-
pected
because
our
anagrams rarely activate
the
lexicon. Although
the
effect
of
lexical
constraint
(denned
as
lexical
benefit)
is
small,
it
appears
in all
three data sets where
it was
predicted
to
occur. Furthermore, this mea-
sure provides
a
pure index
of the
predictive
power
of the
lexical branch
of our
model.
This
is
true because
the
psychophysical
dis-
tinctiveness
of the
target letter
is
removed
by
subtracting
P(C/A).
Differences
in
lexical
constraint
are due
only
to the
mixture
of
can-
didate words that support
the
correct,
incor-
rect,
or
neither alternative.
Another
way of
appreciating
the
role
of
lexical
constraint
in our
data
is to
compare
the
high-constraint (top third)
and
low-con-
straint (bottom third) words
to the
high-
and
low-constraint
anagrams.
The
magnitude
of
the WSE is
about
10%
for the
high-constraint
set
compared
to
only
5% for the
low-con-
straint set.
One
might speculate that
a
com-
parable
effect
of
lexical constraint could
be
found
in
Johnston's
(1978)
data
if
they were
analyzed
on the
basis
of our new
measure
of
lexical
constraint.
Orthography
Massaro
and his
associates
(Massaro,
1973,
1979;
Massaro, Taylor, Venezky,
Jastrzemb-
ski,
&
Lucas, 1980; Massaro, Venezky,
&
Taylor,
1979),
have convincingly advocated
a
model
in
which
letter
recognition
is
guided
by
inferences drawn
from
knowledge
of or-
thographic structure.
Our
model
has no
pro-
vision
for the
dynamic
use of
orthographic
rules,
nor
does
it
assume
a
syllabary
of
com-
monly
occurring letter clusters that could
be
activated
by, or in
parallel with,
the
alpha-
betum.
Although
it is
clear that
the
model
does
not
need
any
orthographic mechanism
in
order
to
predict
the
advantage
of the
reg-
ular
pseudowords over
the
irregular ana-
grams,
the
present experiments
offer
a
large
set
of
stimuli
and
data
to
assess
the
possible
contribution
of
orthography
within
the
word,
pseudoword,
and
anagram classes.
In
accordance with
the
procedure advo-
cated
by
Massaro,
the sum of the
logarithms
of
the
bigram frequencies (SLBF)
was
computed
for
each stimulus.
The
correla-
tions between SLBF
and the two
sets
of
word
data
were
.
11
and
.04. Apparently,
there
is
no
relation between this measure
of
ortho-
graphic structure
and
performance
on
indi-
vidual
items. This
is
also true
for the
correla-
tion between SLBF
and the
pseudoword
data
(r
=
.09).
In
contrast,
the
correlation between
SLBF
and the
anagram
data
is
much higher
(r
=
.30). This pattern
of
correlation
is
similar
to a
previous analysis
of
orthographic struc-
ture (Paap
&
Newsome,
1980b)
and
further
supports
our
conclusion
that
orthographic
structure
will
predict performance only when
very
low
levels
are
compared
to
somewhat
higher
levels
of
structure.
Although
current data
do not
permit
one
to
rule
out the use of
orthographic rules
in
letter
and
word recognition,
our
model shows
that both
the
lexical (advantage
of
words over
well-formed
pseudowords)
and
orthographic
(advantage
of
pseudowords over irregular
strings)
component
of the WSE can be
pre-
dicted
on the
basis
of
lexical constraint alone.
Furthermore, lexical access
may
also account
for
the
apparent
effect
of
orthography
on an-
agram
performance.
In the
activation-veri-
fication
model,
the
contribution
of
lexical
activity
is
determined
by the
probability
of
a
decision based
on the
lexicon,
P(L),
and
the
probability
of a
correct response based
on
lexical activity,
P(C/L).
The
correlation
between
orthography (SLBF)
for
each ana-
gram
and its
corresponding
P(L)
is
.49. Fur-
thermore,
the
correlation between SLBF
and
P(C/L)
is
also .49.
In
terms
of our
model,
there
is no
direct
effect
of
orthographic struc-
ture
on
letter recognition. Rather,
it is
simply
the
case that extremely irregular letter strings
rarely
excite
the
lexicon and, therefore, can-
not
benefit
from
lexical access.
On the
other
hand, less irregular anagrams will occasion-
ally
activate
a
word unit,
and
that unit
is
likely
to
support
the
correct alternative.
Recently,
Massaro (Note
7)
conducted
simulations
of his
fuzzy
logical model that
are
similar
to the
activation-verification
model
in
that top-down evidence (e.g.,
log
bigram
frequencies)
is
combined with
an in-
dex
of
visual
evidence based
on
letter-con-

586
PAAP,
NEWSOME, MCDONALD,
AND
SCHVANEVELDT
fusion
probabilities.
For
six-letter anagrams
visual
evidence alone
is a
poor predictor;
the
correlation between predicted
and
observed
results
for
160
anagrams
is
only
.08. Adding
the
log-bigram
frequency
component
to the
model
raises
the
correlation
to
.59. Orthog-
raphy
does seem
to
have
a
considerable
im-
pact
and
suggests
the
possibility
that
percep-
tion
of
longer strings
may-be
influenced
by
orthographic
regularity
to a
much greater
extent than
is
perception
of
shorter strings.
On
the
other hand,
it is
entirely possible that
the
activation-verification model
may
also
be
able
to
account
for the
orthographic
ef-
fects
in
Massaro's
six-letter anagrams
on the
basis
of
lexical access
and
without recourse
to any
orthographic mechanism.
The
outcome
of
Massaro's simulation
for
the 40
six-letter words
is
less informative.
The
correlation between obtained data
and
that predicted
from
the
visual component
alone
was .48
compared
to
only
.43 for the
model
that combines both
the
visual
and
orthographic components. This suggests that
the
impact
of
orthography
on the
perception
of
six-letter words
may be
quite weak,
but it
may
be
important
to
note that performance
levels
were
not at all
comparable
for the
words
(90% correct)
and
anagrams (75% cor-
rect).
Comparisons
of the
Interactive Activation
and
Activation-Verification
Models
McClelland
and
Rumelhart
(1981;
Ru-
melhart
&
McClelland,
1982)
have proposed
an
interactive activation model that extends
to the
same wide scope
of
letter
and
word
recognition
paradigms that have been
the
tar-
get
of our
activation-verification model.
Both
models share many basic assumptions:
(a)
that stimulus
input
activates spatially
specific
letter
units,
(b)
that
activated
letter
units
modulate
the
activity
of
word units,
and (c)
that letter
and
word recognition
are
frequently
affected
by
important top-down
processes. These generally
stated
assump-
tions permit both models
to
predict
and ex-
plain
the
effects
of
lexicality,
orthography,
word
frequency,
and
priming. However,
the
specific
operations used
to
instantiate these
general
assumptions
in
McClelland
and Ru-
melhart's
computer simulation
and in our
computational algorithms
offer
a
large num-
ber of
provocative
differences
with respect
to
the
specific
mechanisms responsible
for the
various
contextual phenomena. Further-
more,
the two
models
are not
always equally
adept
in
accounting
for the
various context
effects.
The
Word
and
Pseudoword
Advantage
The WSE is
often
characterized
as
con-
sisting
of two
effects.
The
lexical
effect
refers
to the
benefits
that accrue
from
accessing
the
lexicon
and is
estimated
from
the
obtained
advantage
of
words over well-formed pseu-
dowords.
The
orthographic
effect
refers
to the
benefits
derived
from
the
reader's
knowledge
of
orthographic redundancy
and can be es-
timated
from
the
obtained advantage
of
pseudowords
over irregular nonwords. Both
the
activation-verification
and
interactive
activation
models assume that lexical acti-
vation
accounts
for
both lexical
and
ortho-
graphic
effects.
In
the
interactive
activation
model,
lexical
access facilitates letter recognition through
excitatory
feedback
from
activated word
units
to
their constitutent letter units. Word
stimuli
are
very
likely
to
activate word units
that
reinforce
the
letters presented, thereby
increasing
the
perceptibility
of the
letters.
In
contrast,
irregular nonwords
will
rarely
ac-
tivate a
word unit,
and
accordingly,
the
per-
sistence
of
activity
in the
correct
letters
units
will
not be
extended
by
feedback. Because
pseudowords share many
letters
in
common
with
words, they
too
activate word units
that
produce excitatory feedback
and
strengthen
the
letter units that give
rise to
them.
Given
the
detailed
encoding
assumptions
of
the
interactive activation model
and the
particular
set of
parameter values needed
to
predict
the
basic
pseudoword advantage,
McClelland
and
Rumelhart conclude
that
the
amount
of
feedback,
and
hence
the
amount
of
facilitation, depends primarily
on
the
activation
of
word units
that
share three
letters with
the
stimulus. They call
the set of
words
that
share three letters with
the
stim-
ulus
its
neighborhood.
The
amount
of
facil-
itation
for any
particular target letter will
be
primarly
determined
by the
number
of
word
units
in the
neighborhood that support
the

ACTIVATION-VERIFICATION MODEL
587
target
("friends")
and the
number that sup-
port some other letter
("enemies").
This generalization provides
a
good basis
for
comparing
the two
models, because
the
amount
of
facilitation produced
by
lexical
access
in our
model will
be
primarily deter-
mined
by the
number
of
friends
and
enemies
in
the
candidate
set
generated
by our
encod-
ing
algorithm.
The set of
words
in the
neigh-
borhood
of a
particular stimulus
is
likely
to
be
quite
different
from
the set of
candidate
words.
One
major reason
for
this
(as
pointed
out
earlier
in the
discussion
of the
geometric
mean
as a
measure
of
word-unit activation)
is
that
word units
that
share three letters with
the
stimulus will
fail
to
exceed
the
word-unit
criterion
if the
mismatching letter
is not
very
confusable
with
the
letter actually presented.
For
example,
for the
input string
SINK
with
S as the
test letter,
our
encoding algorithm
generates only three
friends
(SING,
SINE,
and
SINK)
and
four
enemies
(LINK, WINK, FINK,
and
RINK).
In
addition
to all of
these words,
the
neighborhood includes
five new
friends
(SICK, SANK, SINS, SILK,
and
SUNK)
and two
new
enemies
(PINK
and
MINK).
Thus,
the
ratio
of
friends
to
enemies
is 3:4 for our
model compared
to 8:6 for
their model.
Using
the
candidate
set
generated
by our
model
and the
neighborhood
denned
by a
search
of our
lexicon (the 1,600 four-letter
words
in the
Kucera
and
Francis, 1967,
norms),
we
computed
the
proportion
of
friends
for
each stimulus according
to
each
of
the two
models.
In
order
to
compare
the
predictive power
of the two
models,
we
then
correlated
the
proportion
of
friends against
the two
sets
of
word data,
the
anagram data,
and
the
pseudoword data.
For all
four
cases
the
proportion
of
friends
in the
candidate
set
yielded
higher correlations than
the
propor-
tion
of
friends
in
the
neighborhood.
The av-
erage
correlation
for our
model
was .24
com-
pared
to .14 for the
interactive activation
model.
In
summary,
our
model seems
to
have
a
slight edge
in its
ability
to
account
for
consistent
interitem
variation that accrues
from
lexical access.
We
were
also curious
as to the
implications
that McClelland
and
Rumelhart's
encoding
assumptions would have
for the
average per-
formance
on our
words,
pseudowords,
and
anagrams.
To
this
end the
alphabetic branch
of
our
model
was
modified
so
that
(a) the
activity
of
each word
was
boosted
by .07 for
each
matching letter
and
reduced
by .04 for
each
mismatching letter
and (b) the
word-
unit criterion would
be
exceeded
by all
those
lexical
entries
that shared
at
least three letters
in
common with
the
stimulus.
The first
mod-
ification
is
based
on the
values
of
letter-to-
word
excitation
and
inhibition used
by
McClelland
and
Rumelhart
and
amounts
to
assigning
a
strength
of .28 to the
word unit
corresponding
to a
word stimulus,
and a
strength
of.
17
to all the
word units that share
three letters with
a
stimulus.
The
probability
of
a
decision based
on the
lexicon,
P(L),
and
the
probability
of a
correct response based
on
lexical access,
P(C/L),
were
then com-
puted
as
usual.
The
decision rule
was
also
the
same,
but
deserves
a
brief comment.
To
extend
Mc-
Clelland
and
Rumelhart's analysis
of the
neighborhood
to
predictions
of
proportion
correct
in a
two-alternative forced-choice
task,
it is
necessary
to
separate nonaligned
neighbors
from
true enemies. That
is,
word
units
in the
neighborhood that support
the
incorrect
alternative (true enemies)
will
have
a
much more disruptive
effect
on
perfor-
mance than words that support neither
al-
ternative (nonaligned neighbors).
This
is
essentially what
is
done
in
Equation
2 for our
model
when
we
assume that
friends
contrib-
ute to a
correct response with
a
probability
of
1,
nonaligned neighbors with
a
probability
of
.5, and
true enemies with
a
probability
ofO.
When
a
neighborhood based
on the
char-
acteristics
of
the
interactive activation model
is
substituted
for the
candidate
set
generated
by
our
encoding algorithm,
and all
other
op-
erations
are
identical,
the
average predicted
performance
is .80 for
words,
.84 for
pseu-
dowords,
and .74 for
anagrams. This will
not
do at
all, because
the
advantage
of
words over
anagrams
is too
small and, more impor-
tantly, words
are
predicted
to be
inferior
to
pseudowords!
McClelland
and
Rumelhart
have
already discussed
why
pseudowords
tend
to
have
a
high
proportion
of
friends.
We
add to
their analysis
a
similar account
of
why
words
tends
to
have
a
lower proportion
of
friends.
Experimenters select stimulus words
in

588
PAAP,
NEWSOME,
MCDONALD,
AND
SCHVANEVELDT
pairs that
differ
by
only
a
single letter. This
ensures
that
the two
alternatives
in the
target
location
will
both
form
words
in the
remain-
ing
context.
For
example,
two of
Johnston's
(1978)
high-constraint words were
SINK
and
WINK,
with
the first
position being probed
with
the
alternatives
S and W. One
conse-
quence
of
this
is
that
every
word stimulus
will
have
at
least
one
friend
(itself)
and one
true enemy (its mate). Experimenters create
pseudowords
by
substituting
one of the
con-
text
letters
from
the
original word pair.
For
example,
we
created
the
pseudowords
SONK
and
WONK
by
replacing
the Is
from
SINK
and
WINK
with
Os.
The
consequence
of
this
is
that every pseudoword
has at
least
one
friend
(SINK
for
SONK
and
WINK
for
WONK)
but no
built-in
enemy
(WONK
is not an
enemy
of
SONK
because
it is not a
word). This system-
atic bias introduced
in the
selection
of the
materials results
in the
words' neighborhood
averaging
only
70%
friends
compared
to 79%
for
the
pseudowords. Thus, models based
directly
on the
composition
of the
neighbor-
hood
will
predict
an
advantage
of
pseudo-
words over words.
In
fairness
to the
interactive activation
model,
it
should
be
clearly
pointed
out
that
when
its
encoding assumptions
are
placed
in
the
context
of
its own
complete model, rather
than
our
complete model,
the
simulation
shows
the
correct ordering
for the
words,
pseudowords,
and
single letters used
by
McClelland
and
Johnston (1977).
We
sus-
pect
that
their
full
simulation
would
also
pro-
duce
the
correct ordering
of our
words, pseu-
dowords,
and
anagrams.
The
reason
for
this
is
that
the
complete interactive activation
model assumes large (parameter value
=
.21)
amounts
of
inhibition between competing
word
units. Thus, when
a
word
is
presented,
the
initial strength
of the
corresponding word
unit
(about .28)
will
quickly dominate
the
initial activity (about
.17)
of any
potential
enemy.
Thus,
the
effects
of
lexical access
for
word
stimuli
are
almost entirely determined
by
feedback
from
the
corresponding word
unit
and no
others.
This
is an
interesting
con-
trast
between
the two
models.
We
assume
that
both
the
word advantage
and the
pseu-
doword
advantage
are
mediated
by
decisions
based
on the
activity
of a
small
set of
can-
didate words. McClelland
and
Rumelhart
assume
that
the
word advantage
is
mediated
by
feedback
from
a
single word unit (the lex-
ical
entry corresponding
to the
word pre-
sented)
but
that
the
pseudoword advantage
is
mediated
by
feedback
from
large neigh-
borhoods.
This inherent
difference
between words
and
pseudowords
in the
interactive activa-
tion
model produces some undesirable
fall-
out.
Specifically,
if
high levels
of
interword
inhibition permit
the
stimulus word
to
dom-
inate
any
potential competition, then
the
stimulus-driven
differences
between various
words
will
be
eliminated.
In
short, high levels
of
interword
inhibition
mean
that
the
func-
tional amount
of
activation produced
by the
presentation
of all
words
will
be
about
the
same. Thus,
the
significant
correlations
be-
tween
obtained performance
and
that pre-
dicted
from
our
model would stand unchal-
lenged
by the
interactive activation model.
It is
true
that
the
interactive activation model
does predict some variation between words
that
is not
stimulus driven, namely,
that
the
resting
levels
of
word units increase with
word frequency,
but we
will show
in a
sub-
sequent
section that this assumption
is not
a
good one.
Throughout
the
preceding section
we
have
compared
the
predictive power
of our
model's
candidate sets
to
that
of
McClelland
and
Rumelhart's
neighborhood.
Our
encoding
algorithm, which
is
highly sensitive
to
visual-
confusability
effects,
seems
to
enjoy
a
con-
sistent
advantage
in the
tests
we
have con-
ducted. However, this should
not be
viewed
as a
permanent
disadvantage
for the
inter-
active activation model because
the
neigh-
borhoods
we
tested
conform
to
those
ob-
tained when their parameter,
p, for
visual-
feature
extraction
is set to
1.0.
If a
value
lower
than
1.0 is
used, their model
will
gen-
erate neighborhoods sensitive
to
visual con-
fusability
in a way
similar
to
that
of our
can-
didate words. However,
one of the
difficulties
in
using
the
interactive activation model
as
a
heuristic device
is its
inherent complexity.
Accordingly,
it is
difficult
to
anticipate
the
results
of
simulations that have
not
been con-
ducted.
It
should
not be
presumed
in
advance
that
the
interactive activation model would
accurately predict
the
relative
differences
be-
tween
words, pseudowords,
and
anagrams

ACTIVATION-VERIFICATION MODEL
589
when
only partial information
is
gained
from
each
letter location. Furthermore, when
the
contribution
of
visual
confusability
is
intro-
duced through
the
partial sampling
of
sub-
jectively
denned
features
it is not as
likely
to
be
as
predictive
as
when confusability
is
based
on an
empirically derived confusion
matrix.
The
Pseudoword
Expectancy
Effect
One
potential problem
for any
model that
eschews
any
direct
contribution
of
ortho-
graphic knowledge
is
that
the
pseudoword
advantage
seems
to be
more susceptible
to
expectancy
effects
than
the
word advantage.
Carr,
Davidson,
and
Hawkins
(1978)
have
shown
that
if
subjects
do not
expect
to see
any
pseudowords, then performance
on an
unexpected
pseudoword will
be no
better
than that obtained with irregular nonwords.
In
contrast, they showed that
the
advantage
of
words over irregular nonwords
was the
same
regardless
of
whether
the
subject
ex-
pected
all
words
or all
nonwords.
McClelland
and
Rumelhart
can
account
for
this pattern
of
expectancy
effects
by as-
suming
that subjects have strategic control
over
the
degree
of
inhibition between
the
alphabetum
and
lexicon. They assume that
if
subjects
expect only words
or
only irregular
nonwords,
they will adopt
a
large value
of
letter-to-word
inhibition.
More specifically,
the
inhibition parameter
in
their simulation
is
set so
that
the
excitation produced
by
three
matching
letters will
be
precisely countered
by
the
inhibition
from
the
remaining mis-
match. Accordingly,
the
only word unit that
will
produce appreciable feedback
to the
let-
ter
units
is the
word presented. This means
that
the
word
advantage will
be
about
the
same
as
always
but
that
the
pseudoword
ad-
vantage
will
be
eliminated.
Our
activation-verification model
can
also
predict
the
pseudoword expectancy results
by
assuming
that
subjects have control over
one
parameter, namely,
the
word-unit criterion.
All
of the
predictions reported earlier used
a
word-unit criterion
of
.24.
The
average
numbers
of
candidate words produced
by the
three classes
of
stimuli were
3.4 for
words,
2.1
for
pseudowords,
and .2 for
anagrams.
By
adopting this
fairly
lax
criterion,
the
sub-
ject
can
take advantage
of
beneficial lexical
evidence
for
both words and, more impor-
tantly, pseudowords. However, because
the
word
unit corresponding
to a
word stimulus
would exceed
a
much
stiffer
criterion, sub-
jects have
no
motivation
to
maintain
a low
criterion and, therefore,
to
consider larger
sets
of
word units unless they expect
to see
some pseudowords.
The
expectancy
effect
was
modeled
by
raising
the
word-unit criterion
from
.24 to
.29. This resulted
in a
reduction
of the
num-
ber of
candidate words
to 1.4 for
word stim-
uli,
.40 for
pseudowords,
and .04 for
ana-
grams.
The
effect
of
this
on the
predicted
proportion correct
is
negligible
for
words
(.841
versus .856)
and
anagrams (.755 versus
.747)
but
results
in a
sizable decrease
in
pseu-
doword
performance (.813
to
.760).
In
sum-
mary,
raising
the
word-unit criterion
can re-
sult
in the
elimination
of the
pseudoword
advantage
while having
very
little
effect
on
the
word advantage. Although
a
higher cri-
terion does lead
to an
increase
in
P(C/L)
for
word
stimuli, this tends
to be
countered
by
a
decrease
in the
total amount
of
lexical
ac-
tivity
and, hence,
a
decrease
in
P(L).
Both
models
can
predict
the
pseudoword
expectancy
effect
reported
by
Carr
et
al.
(1978).
Although introspection
is at
best
a
weak
test
of two
opposing theories,
we
yield
to the
temptation
to
point
out
that
it
seems
to us
more natural that
a
subject-controlled
strategy
might involve
the
adjustment
of a
criterion
for
considering lexical evidence
rather than
the
adjustment
of the
amount
of
inhibition between letter
and
word
detectors.
Word-Frequency
Effects
for
Masked
Stimuli
Under normal
conditions
of
stimulus pre-
sentation, familiar words
can be
processed
more
effectively
than less familiar ones.
For
example, high-frequency words
are
consis-
tently classified
faster
than
low-frequency
words
in
lexical-decision tasks
(Landauer
&
Freedman,
1968;
Rubenstein,
Garfield,
&
Millikan, 1970; Scarborough,
Cortese,
&
Scarborough.,
1977).
Our
complete model
captures this familiarity
effect
by
assuming
that
the
order
of
verification
is
determined,
in
part,
by
word frequency. However,
it was

590
PAAP,
NEWSOME,
MCDONALD,
AND
SCHVANEVELDT
assumed
that
the
brief stimulus
durations
used
in the
present experiments, together
with
the
masking
fields,
would prevent ver-
ification
from
taking place.
Two
studies
have systematically manipu-
lated word
frequency
under conditions
of
backward masking.
In his first
experiment
Manelis
(1977)
selected 32-word sets
from
the
Kucera and
Francis
(1967)
norms with
high
(94-895), medium
(23-74),
and low (2-
10)
frequency counts. Although proportion
of
correct recognitions increased with fre-
quency
from
.775
to
.794
to
.800,
the
differ-
ences were
not
significant.
In the
second
ex-
periment pairs
of
high-
and
low-frequency
words
shared
the
same critical letter
and as
many
context
letters
as
possible.
Again,
there
were
no
differences
between common (.762)
and
rare (.757) words.
In a set of
three
ex-
periments described
by
Paap
and
Newsome
(1980b),
80
words were selected
from
the
Thorndike-Lorge
(1944)
count
so
that there
were
equal numbers
of
words with frequen-
cies
of 1, 2, 5, 14, and 23 per
million. Words
in the five
frequency
classes were matched
in
terms
of the
identity
and
position
of the
target
letter.
The
proportions
of
correct
re-
sponses,
in
increasing order
of
frequency,
were
.67, .62, .65,
.66,
and
.65.
The
results described above support
our
assumption
that
verification does
not
occur
when
stimulus words
are
followed
by a
mask.
We
have also tested
for
word-frequency
ef-
fects
in the
data
we
obtained with Johnston's
(1978) words.
The
Kucera
and
Francis
fre-
quency
counts were determined
for
each
of
the 288
words
and
correlated against both
sets
of
word data. These correlations
are
shown
in
parentheses
in
Table
5.
There
are
no
significant correlations between word fre-
quency
and
proportion correct,
and in
fact,
the
trend
is
toward poorer performance with
higher word
frequency.
However, when
a
log-
arithmic transformation
is
applied
to the
fre-
quency
counts, positive correlations appear
in
each
of the
data sets.
Because
many
of
Johnston's
(1978)
words
are
quite uncommon
and may not be
entered
in
the
subjective lexicon
of our
typical sub-
ject,
it is
possible that this small
word-fre-
quency
effect
reflects
nothing more than
the
probability
of the
word appearing
in the
lex-
icon. This interpretation
was
investigated
by
sequentially
removing
the
words with
the
lowest frequency
from
the
original
set of 288
words.
As
shown
in
Table
5, the
correlation
between
the
logarithm
of
word frequency
and
performance systematically decreases
as
rare words
are
removed
from
the
sample.
When
only words with frequencies greater
than three
are
considered,
there
is no
effect
of
relative
frequency.
In
order
to
further support
our
claim
that
many
of
Johnston's
(1978)
words
are
unfa-
miliar
to our
population
of
undergraduate
subjects,
we had 147
students
classify
each
of
the
words
as
either
(a) a
word
that
I
know
the
meaning
of, (b) a
word that
I
don't
know
the
meaning
of, or (c) a
nonword. Thirteen
words
were
classified
as
nonwords
by a ma-
jority
of the
subjects
(LAVE,
TING,
BOON,
CRAG,
WHET, JILL, BOLL, WILE, HONE, HEWN,
FIFE,
BANS,
VATS).
Furthermore,
for
many
words
the
responses were distributed quite
evenly
across
the
three categories (e.g.,
FIFE,
BANS,
VATS, TEEM, HEMP, PENT, WANE,
NAVE,
SLAT).
When
we
removed
the 35
words
that
are
most
often
classified
as
non-
words
(and
the
meaning
of
which
is
known
by
only
a
minority
of the
subjects), there were
no
significant correlations between
the
data
for
the
individual words
and the
logarithm
of
their
frequency.
This purging
of our
lex-
icon also
led to a
slight improvement
in the
correlation between predicted
and
obtained
performance
for the 288
words,
r =
.32.
These tests lead
us to
conclude that mask-
ing
almost always prevents
verification
and
that there
is no
need
to
build
word-frequency
effects
into
our
encoding algorithm.
In
order
to
make sure that word
frequency
could
not
enhance
the
ability
of our
encoding algo-
rithm to
predict variation between words,
we
tried several
different
ways
of
having
the
log-
arithm
of
word
frequency
modulate
the ac-
tivity
of the
word units.
Our
basic strategy,
like
that
of
McClelland
and
Rumelhart,
was
to
decrease
the
stimulus-driven activity
of
word
units
in
inverse relation
to
their fre-
quency.
Because
the
correlation between
our
obtained word
data
and log
word
frequency
was
.16,
we
searched
for a
frequency
effect
that would produce
a
comparable correlation
between
our
predicted data
and log
word fre-
quency.
The
desired impact
of
word fre-
quency
was
achieved
when
the
amount
of
stimulus-driven
activity
was
reduced
by
about
5%
for
each
half-log
unit drop
in
word fre-

ACTIVATION-VERIFICATION
MODEL
591
quency.
This means that
the
most common
words
in our
lexicon would receive
no re-
duction
in
activity,
and
those with
a
fre-
quency
of
only
one
would
be
reduced
by
40%.
Because
the
word-frequency
effect
leads
to
an
overall reduction
in
lexical activity,
it was
necessary
to
lower
the
word-unit criterion
substantially
(,14)
in
order
to
maintain can-
didate sets
of
about
3.3
words. Under these
conditions
the
predicted performance
for all
words
was
exactly
the
same (PPC
=
.84)
as
that predicted
from
the
original model
that
has no
provision
for
word-frequency
effects.
The
question
of
interest
can now be an-
swered: Does word
frequency
enhance
the
model's
ability
to
account
for
variation
be-
tween
words?
No, the
correlations between
predicted data
and two
sets
of
obtained data
show
that introducing
word-frequency
effects
produces
no
change
for one
data
set and a
decline
of .06 for the
other.
In
summary,
we can find no
evidence
in
our
data
or
elsewhere that two-alternative
forced-choice
performance
on
masked word
displays
shows
a
word-frequency
effect.
This
is
consistent with
the
activation-verification
model, because
we
assume that word fre-
quency
does
not
affect
activation
of the
word
units,
but
will
affect
the
order
of
verification
when
the
stimulus-presentation conditions
permit verification
to
occur.
The
magnitude
of
the
word-frequency
effects
generated
by
the
interactive activation model
is not
known.
Although
their model
specifically
assumes
that
the
resting activity
of
word units
is
de-
termined
by
familiarity, other factors, such
as
the
decision rules adopted
for the
forced-
choice
task,
may
severely attenuate
the
initial
frequency
differences
between word units
and, thereby, permit
the
prediction
of no
word-frequency
effect.
A
fair
conclusion
with
respect
to
word
frequency
is
that
the
acti-
vation-verification
model
can
correctly pre-
dict
the
magnitude
of
familiarity
effects
in
both
tachistoscopic
and
reaction time
studies
and
that
the
interactive activation model
may
be
able
to do so.
Reaction
Time Studies
As
we
mentioned
in the
introduction,
the
concepts
embodied
in our
activation-verifi-
cation model were originally developed
in
the
context
of
reaction time studies using lex-
ical-decision
and
naming tasks. With this
history
it is to be
expected that
the
model
can
handle
a
variety
of
reaction
time
data.
There
are too
many
findings to
cover
in
detail
here,
but it may be
useful
to
review some
of
this earlier
work
to
provide some idea about
the
performance
of the
model. Because
the
interactive activation model
has not
been
specifically
applied
to
lexical-decision data,
we
cannot
draw
specific
comparisons. How-
ever,
the
interactive activation model
has
been used
to
explain
the
effects
of
semantic
context
and
word
frequency
in
other reaction
time
tasks
(e.g.,
naming tasks),
and we
will
comment
on the
applicability
of
analogous
explanations
of findings
from
the
lexical-de-
cision
task.
The
interactive activation model
and our
activation-verification model
differ
about
the
nature
of
effects
of
prior
semantic context
and
word
frequency
when stimuli
are
pre-
Table
5
Correlations
Between
Obtained
Proportion
Correct
and Log
Word
Frequency
Data
set
Word
frequencies
included
All
All>
1
All>2
All>
3
Word
set
1
2
Number
of
words
/•=.05
.16
(-.04)
.14
(-.09)
288
.12
.14
(-.06)
.11
(-.11)
249
.13
.09
(+.04)
.07
(+.01)
228
.13
.04
(-.08)
.04
(-.14)
220
.13
Note.
Correlations
between
proportion
correct
and the
absolute
word-frequency
counts
are
shown
in
parentheses.
"All
> 1"
means
all
stimulus
words
with
a
frequency
greater
than
1.

592
PAAP,
NEWSOME,
MCDONALD,
AND
SCHVANEVELDT
sented
for
normal durations
and
without
masking.
In the
interactive activation model,
these
two
factors both have
the
effect
of in-
creasing
activation levels
in
relevant word
units.
The
base activation level
of the
word
units
increases
as a
function
of
word fre-
quency.
Also, word units that
are
related
to
the
context have increased activity levels rel-
ative
to
word units
for
unrelated words. Per-
haps
word units that
are
inconsistent with
the
context would have depressed activity
levels
as
well.
In
contrast,
our
activation-verification
model
places
the
effects
of
word frequency
subsequent
to the
activation
of
word units.
Word
frequency determines
the
order
in
which
lexical units
are
verified
in the
verifi-
cation
process.
The
activation-verification
model
also assumes that context increases
the
activity level
of
lexical units that
are
related
to the
context,
but
this activity increase
may
be
high enough
to
cause
the
word units
to
exceed
the
criterion
for
inclusion
in the
can-
didate
set.
The
verification process
is
then
responsible
for the
analysis
of
stimulus
in-
formation.
Thus, verification
can
prevent
a
premature
response. There appears
to be no
comparable mechanism
in the
interactive
activation
model.
In
lexical-decision tasks, there
is
evidence
that
context
and
frequency
have
different
effects
on the
time required
to
classify
a
letter
string
as a
word. Becker
and
Killion
(1977)
found
that context interacts with
the
quality
of
the
visual stimulus whereas frequency
and
visual
quality show additive
effects.
These
findings
imply that frequency
and
context
exert
their
influence
on
performance
in
dif-
ferent
ways, contrary
to
expectations, derived
from
the
interactive,
activation model.
McDonald (1980) developed
a
computer
simulation
of the
verification
model (which
was
the
precursor
to our
activation-verifi-
cation
model). McDonald's simulation pro-
duced both
the
additivity
of
frequency
and
visual
quality
and the
interaction
of
context
and
visual quality. Further,
as we
discussed
earlier,
there
are
apparently
no
word-fre-
quency
effects
in the
word-superiority para-
digm.
This result
follows
naturally
from
our
model
because
frequency
does
not
affect
the
activation
process, which
is the
basis
of the
decision
in the
word-superiority paradigm.
The
activation-verification model
is
also
consistent with
findings on
effects
of
context
on the
classification
of
nonwords
in the
lex-
ical-decision task. Several models (including
the
interactive activation model) handle con-
text
effects
by
inducing
a
bias
in
favor
of re-
lated words. This approach leads
to the ex-
pectation that nonwords that
are
very
similar
to
particular words should
be
erroneously
classified
as
words more
often
in a
related
context
than
in an
unrelated context.
For
example,
the
nonword
NERSE
should
be
mis-
classified
more
often
following
a
word related
to
NURSE
(e.g.,
DOCTOR)
than following
an
unrelated word (e.g.,
LAMP).
In
contrast,
our
model assumes that lexical decisions
are
made
on the
basis
of
verification
rather than
activation
and
that
the
quality
of the
verifi-
cation process
is not
affected
by
context.
Context
affects
the
availability
of
lexical units
for
verification,
but not the
quality
of the
verification
process
itself. Thus, context
should
have
no
effect
on the
liklihood
of
clas-
sifying
a
nonword
as a
word.
The
evidence
on the
classification
of
non-
words
supports
the
predictions
of the
acti-
vation-verification model. Schvaneveldt
and
McDonald
(1981)
found
no
effect
of
context
on
classifying
nonwords when stimuli
re-
mained available until
the
response occurred.
Context
did
facilitate response time
to
words
in
their experiments. Other studies have pro-
duced similar results (Antos,
1979;
Lapinski,
1979;
McDonald, 1977, 1980; Lapinski
&
Tweedy, Note
8).
O'Connor
and
Forster
(1981)
concluded that
a
bias explanation
was
ruled
out by
their
findings
even though
one
of
their experiments showed bias
effects.
In
that experiment, however, error rates were
over
35% on the
critical items, which
is un-
usually high.
In the
context
of
the
activation-
verification
model, such error rates suggest
that subjects
are
responding without
verifi-
cation
on a
substantial proportion
of the
trials.
If
verification
is
optional,
speed-ac-
curacy
trade-offs
may be
partly
due to the
probability
of
verification
in a
particular task.
Schvaneveldt
and
McDonald (1981) also
showed
bias
effects
of
context with
a
brief
stimulus
display
followed
by a
masking stim-
ulus.
As we
argued earlier,
we
assume that
'these
stimulus conditions prevent
verifica-
tion.

ACTIVATION-VERIFICATION
MODEL
593
Overall,
the
activation-verification model
appears
to
handle
a
considerable amount
of
data
from
reaction time experiments (see
Becker,
1980,
and
McDonald, 1980,
for
fur-
ther
examples).
We
believe that
one
impor-
tant
characteristic
of the
model lies
in the
independent
top-down
analysis
of the
stim-
ulus
(verification)
that
is
sensitive
to
devia-
tions
from
the
stored representation
of a
word.
These deviations might
be
further
di-
vided
into permissible (identity preserving)
and
illegal (identity
transforming)
distortions
of
the
stored representation. Verification,
then, amounts
to
determining whether
the
stimulus impinging
on the
senses could
be
reasonably
interpreted
as a
particular word
after
context
or the
senses
had
suggested that
the
stimulus might
be
that word.
We
have presented
our
solution
to
what
we
perceive
as an
important theoretical prob-
lem
in
pattern-recognition theory
in
general
and
word recognition
in
particular. That
problem
is to
specify
the
nature
and
inter-
action
of
bottom-up
and
top-down infor-
mation-processng
activities
in
recognition.
There seems
to be
wide acceptance
of the
necessity
for
both
of
these types
of
processes.
There
is
less agreement about just what they
are
and how
they interact.
Our
solution
to
this theoretical problem provides
a
top-down
process
that
involves
comparing stimulus
in-
formation
to
prototypes stored
in
memory.
As
such,
the
top-down process
may
enhance
perception
of
discrepancies rather than
in-
duce
a
perceptual
or
decision
bias
in
favor
of
expected stimuli.
We
believe that
the ev-
idence supports
our
view,
but we are
eager
to
pursue
the
matter
further
with additional
research.
We
hope that
our
theoretical anal-
ysis
and the
contrasts
of two
theoretical
ap-
proaches
will help
to
focus
further
experi-
mentation.
Reference
Notes
1.
Paap,
K.
R., &
Newsome,
S. L. The
role
of
word-
shape
and
lexical constraint
in the
word superiority
effect.
In C.
Cofer
(Chair), Some
new
perspectives
on
word
recognition.
Symposium presented
at the
meet-
ing
of the
Southwestern Psychological Association,
Houston, April
1981.
2.
Paap,
K.
R.,
&
Newsome,
S. L.
Lexical
constraint:
Redefined
and
resurrected.
Paper presented
at the
meeting
of the
Psychonomic Society, Philadelphia,
November
1981.
3.
Paap,
K.
R.,
Newsome,
S.
L.,
&
McDonald,
J. E.
Further
tests
of
the
contribution
of
perceptual
confu-
sions
to the
WSE.
Paper
presented
at the
meeting
of
the
Psychonomic Society,
St.
Louis, November 1980.
4.
Schvaneveldt,
R.
S.,
&
McDonald,
J. E. The
verifi-
cation model
of
word recognition.
In C.
Cofer
(Chair), Some
new
perspectives
on
word
recognition.
Symposium
presented
at the
meeting
of the
South-
western
Psychological Association, Houston, April
1981.
5.
Becker,
C.
A.,
Schvaneveldt,
R.
W.,
&
Gomez,
L.
Semantic,
graphemic,
and
phonetic
factors
in
word
recognition.
Paper presented
at the
meeting
of the
Psychonomic
Society,
St.
Louis,
November 1973.
6.
Paap,
K.
R.,
Newsome,
S.
L.,
McDonald,
J.
E.,
&
Schvaneveldt,
R. W. The
activation
verification
model:
The
effects
of
cuing,
masking,
and
visual
angle.
Manuscript
in
preparation, 1982.
7.
Massaro,
D. W.
Simulating
letter
and
word
recog-
nition:
A
fuzzy
logical
model
of
integrating
visual
in-
formation
and
orthographic
structure
in
reading.
Pa-
per
presented
at the
European Conference
on
Artif-
ical
Intelligence,
Orsay,
France, July 1982.
8.
Lapinsky,
R. H., &
Tweedy,
J. R.
Associate-like
non-
words
in a
lexical-decision
task:
Paradoxical
seman-
tic
context
effects.
Paper presented
at the
Mathemat-
ical Psychology meetings,
New
York
University,
Au-
gust
1976.
References
Antos,
S. J.
Processing facilitation
in a
lexical
decision
task.
Journal
of
Experimental
Psychology:
Human
Perception
and
Performance,
1979,
5,
527-545.
Becker,
C. A.
Allocation
of
attention during visual word
recognition.
Journal
of
Experimental
Psychology:
Human
Perception
and
Performance,
1976,
2,
556-
566.
Becker,
C. A.
Semantic context
effects
in
visual word
recognition:
An
analysis
of
semantic strategies. Mem-
ory
and
Cognition,
1980,
8,
493-512.
Becker,
C.
A.,
&
Killion,
T. H.
Interaction
of
visual
and
cognitive
effects
in
word recognition.
Journal
of
Ex-
perimental
Psychology:
Human
Perception
and
Per-
formance,
1977,
3,
389-401.
Carr,
T.
H.,
Davidson,
B.
J.,
&
Hawkins,
H. L.
Percep-
tual
flexibility in
word recognition. Strategies
affect
orthographic computation
but not
lexical
access.
Journal
of
Experimental
Psychology:
Human
Percep-
tion
and
Performance,
1978,
4,
674-690.
Johnston,
J. C. A
test
of
the
sophisticated
guessing
theory
of
word
perception.
Cognitive
Psychology,
1978,
10,
123-153.
Kucera,
H.,
&
Francis,
W. N.
Computational
analysis
of
present-day
American English. Providence,
R.I.:
Brown
University Press,
1967.
Landauer,
T.,
&
Freedman,
J.
Information retrieval
from
long-term memory: Category size
and
recogni-
tion
time.
Journal
of
Verbal
Learning
and
Verbal
Behavior,
1968,
7,
291-295.
Lapinsky,
R. H.
Sensitivity
and
bias
in the
lexical
de-
cision
task. Unpublished doctoral dissertation, State
University
of New
York
at
Stony Brook, 1979.
Luce,
R. D. A
threshold theory
for
simple detection
ex-
periments.
Psychological
Review, 1963,
70,
61-79.

594
PAAP,
NEWSOME, MCDONALD,
AND
SCHVANEVELDT
Manelis,
J.
Frequency
and
meaningfulness
in
tachisto-
scopic
word perception. American
Journal
of
Psy-
chology,
1977,
99,
269-280.
Massaro,
D. W.
Perception
of
letters, words,
and
non-
words.
Journal
of
Experimental
Psychology,
1973,
100,
349-353.
Massaro,
D. W.
Letter information
and
orthographic
context
in
word
perception.
Journal
of
Experimental
Psychology,:
Human
Perception
and
Performance,
1979,
5,
595-609.
Massaro,
D.
W.,
Venezky,
R. L., &
Taylor,
G. A. Or-
thographic
regularity, positional
frequency,
and
visual
processing
of
letter strings.
Journal
of
Experimental
Psychology:
General,
1979,
108,
107-124.
Massaro,
D.
W.,
Taylor,
G.
A.,
Venezky,
R.
L.,
Jas-
trzembski,
J. E., &
Lucas,
P. A.
Letter
and
word
per-
ception:
Orthographic
structure
and
visual
processing
in
reading.
Amsterdam: North
Holland,
1980.
McClelland,
J.
L.,
&
Johnston,
J. C. The
role
of
familiar
units
in
perception
of
words
and
nonwords.
Percep-
tion
&
Psychophysics,
1977,
22,
249-261.
McClelland,
J.
L.,
&
Rumelhart,
D. E. An
interactive
activation model
of
context
effects
in
letter
percep-
tion:
Part
1.
An
account
of
basic
findings.
Psycholog-
ical
Review, 1981,
88,
375-407.
McDonald,
J. E.
Strategy
in a
lexical
decision
task.
Unpublished
master's
thesis,
New
Mexico State Uni-
versity,
1977.
McDonald,
J. E. An
information
processing
analysis
of
word
recognition.
Unpublished doctoral dissertation,
New
Mexico
State
University,
1980.
Morton,
J.
Interaction
of
information
in
word recog-
nition.
Psychological
Review, 1969,
76,
165-178.
O'Connor,
R.
E.,
&
Forster,
K. I.
Criterion bias
and
search sequence bias
in
word recognition. Memory
&
Cognition,
1981,
9,
78-92.
Paap,
K.
R.,
&
Newsome,
S. L. Do
small visual angles
produce
a
word superiority
effect
or
differential
lateral
masking?
Memory
&
Cognition, 1980,
8,
1-14.
(a)
Paap,
K.
R.,
&
Newsome,
S. L. A
perceptual-confusion
account
of the WSE in the
target search paradigm.
Perception
&
Psychophysics,
1980,
27,
444-456.
(b)
Rubenstein,
H.,
Garfield,
L.,
&
Millikan,
J.
Homo-
graphic entries
in the
internal lexicon.
Journal
of
Ver-
bal
Learning
and
Verbal
Behavior, 1970,
9,
487-494.
Rumelhart,
D.
E.,
&
McClelland,
J. L. An
interactive
activation model
of
context
effects
in
letter
percep-
tion: Part
2. The
contextual enhancement
effect
and
some tests
and
extensions
of the
model.
Psychological
Review,
1982,
89,
60-94.
Scarborough,
D.
L.,
Cortese,
C.,
&
Scarborough,
H. S.
Frequency
and
repetition
effects
in
lexical memory.
Journal
of
Experimental
Psychology:
Human
Percep-
tion
and
Performance,
1977,
3,
1-17.
Schvaneveldt,
R.
W.,
&
McDonald,
J. E.
Semantic con-
text
and the
encoding
of
words: Evidence
for two
modes
of
stimulus analysis.
Journal
of
Experimental
Psychology:
Human
Perception
and
Performance,
1981,
7,
673-687.
Schvaneveldt,
R.
W.,
Meyer,
D.
E.,
&
Becker,
C. A.
Lexical
ambiguity, semantic context,
and
visual word
recognition. Journal
of
Experimental
Psychology:
Human
Perception
and
Performance,
1976,
2,
243-
256.
Thorndike,
E.
L.,
<Sf
Lorge,
I. The
teacher's
word
book
of
30,000
words.
New
York: Teacher's College Press,
1944.
Received
October
21,
1981
Revision
received January
28,
1982