The Japanese Psychonomic Society
NII-Electronic Library Service
TheJapanesePsychonomic Society
TheJapaneseJournalofPsychonomi[S[ience
2013,NbL 32,No,1,61-64
Lecture
Vection
strength
is
determined
by
the
plausibility
of
a
stimulus
as
a
representation
ofthe
world
Tbkeharu
SENo
kyushu
Vitiversity
The
previeus
vectien(illusory
self motionperception)
studies canbe
explainedby
a one simple rulethat
is
vection strength
is
determined
by
the
plausibility
of astimulus
as
arepresentation ofthe
world,When
the
worldap-pears
to
be
moving, we must make sense ofthe
situation,but
we refuseto
acceptthe
solutionthat
the
worldis
mov-ing,
Instead,
we
favor
the
solutionthat
we ourselves are moving,because
weknow
that
our
bodies
are movable, whereasthe
worldis
static,I
propose
a
hypothesis
that
ve[tion willbe
stronger when aperceived
visual stimulus rnakes us stronglyinfer
amoreplausible
representation ofthe
world.Key
words: vection,the
plausible
representationof
the
world'
Introduction
Exposure
to
avisual motionfield
that
simulatesthe
retinal opticalflow
generated
by
selfmovement commonly causesthe
'
perception
ofthe
subjective rnovement of ones
own
body,
This
phenomenen
is
known
as
`vection'(Fischer
&
Korn-muller,
1930),
FDr
exarnple, whena
stationary
person
observes atrain
beginning
to
move,they
arelikely
to
perceive
that
they
are moving
in
the
oppositedirectien
to
the
motion ofthe
train,
This
phenomenon
is
known
asthe
`trainillusion',
andprovides
agood
example of vection,This
article reviews awide
range of vection studies conducted overthe
last
several
decades.
The
findings
of recent vection studieshave
elucidated which stimulus attributes are effectivefor
vectioninduction,
providing
new
insight
into
the
phenomenon.
In
this
article,I
propose
anewtheoretical
framework
that
canbe
usedto
un-derstand
the
findings
efprevious
vectien research.According
to
this
novel
perspective,
wegenerally
operate underthe
cog-nitive assumptionthat
the
worldis
static.In
real,the
worldhas
some active and movablefactors,
i.e.
Ieaves
stirredby
the
wind.
Hewever,
when we assumethe
world, almost allparts
are static.Thus
in
this
article,the
word, worldis
indicating
the
staticpart
of
the
reaL
werld,
i.e.
the
mountains andground.
Stimuli
appearingin
awide visualfield
makes usinfer
arepre-Corresponding
address:Institute
for
Advanced
StudF
Kyushu
Universitv,
Faculty
ofDesign,
ICyushu
University,
Research
Center
for
Applied
Perceptual
Science,
Kyushu
University
E-mail:seno@design.k)rushu-u,ac.jp
copY
sentation
of
the
world.When
the
world appearsto
be
moving, we must makesense
ofthe
situatien,but
we refuseto
acceptthe
solution that the worldis
movingbecause
almestal1
parts
ofthe
world always are static,Instead,
wefavor
the
solutionthat
we ourselves are moving,because
weknow
that
our
bodies
are movable,whereas
the
world
is
static,I
propose
ahypothesis
that
vection willbe
stronger when aperceived
visual sttmulus makes us strongly
infer
a rnoreplausible
representationoftheworld,
Historyofvection
In
1875,
Mach
described
vectionthat
was experienced whenhe
observedthe
flow
of a riverfrom
abridge
(Mach,
1875).
While
this
is
the
first
published
report,it
is
likely
that
people
were aware ofthe existence ofvection even
before
this
descrip-tion,
The
history
ofthe
scientific measurement of vection,however,
is
relativelyshort,
The
first
scientific experiment examining vection wasperformed
in
1973
by
Brandt
and colleagues(Brandt,
Dichgan$.
&
Koenig,
1973).
Thus,
vection researchhas
ahistory
of only40
years.
In
their
first
experi-ment,
Brandt
et
al.(1973)
used a mechanically-drivenro-tating
drum
that
subjectsentered
and observed a square-wave modulatedluminance
pattern
inside,
The
researchersmea-sured
the
su.bjective strength of vection,its
duration
andthe
subjective speed of selfimotion,After
Brandt
et aL's(1973)
initial
studp a number of scientistsin
variousfields
attemptedto
research vection.Tbmp
oral
properties
of vectionA
certainduration
of exposureto
a visuaright
2013.
The
Ja
I
rnotion stirnulusThe Japanese Psychonomic Society
NII-Electronic Library Service
TheJapanesePsychonomic Society
62
The
Japanese
Journa1
ofPsychonomic
Science
VbL32,
No,
1
is
requiredto
induce
vection.This
period
is
referredto
asthe
`latency'
of vection, and
has
been
examinedin
sorne studies(e.g.
Kennedx
Hettinger,
Harm,
Ordp
&
Dunlap,
1996,
etc.),The
induction
of vection requiresa
latency
of
at
least
one second,When
subjectively stronger vectionis
obtained,the
la-tency
tends
to
be
shorter.Many
studieshave
reported alatency
of4to
12
seconds(e.g.
Bubka,
Bonato,
&
Palmisano,
2Q08),
which
is
consideredto
be
atypical
latency
range.The
latency
of vection
can
be
measuredby
instructing
subjectsto
press
abutton
andhold
it
down
whilethey
areperceiving
vection.This
procedure
has
remainedlargely
the
samethrough
the
short
history
of vection research.Vlection
candisappear
and reappearduring
long
periods
ef stimuluspresentation.
Thus,
the
cumulativeperied
ofvectionis
also animportant
measure ofthe
phenemenon,
referredto
asthe
`duration' ofvection,Stronger
vectionis
associated with alonger
duration.
Latency
and
duration
are almost always negatively correlated(e.g.
Sene,
Sunaga,
&
Ito,
2010),
Thus,
the
measurement oflatency
andduration
togetheri$
one ofthe
most valid methodsfor
confirming
the
phenomenon,
and canbe
used as anindex
of
the
strength ofvectien,Visual
field
and
vection
The
visual
field
andits
effect on vectioninductien
has
been
a major
topic
of vection research sincethe
initial
studyby
Brandt
et aL(1973),
A
number of studieshave
robustlyre-ported
that
a
wider visualfield
induces
stronger vection(e.g.
Brandt
et al.,1973;
Held.
Dichigans
&
Bauer,
1975;
Lestienne,
Soeching,
&
Berthoz,
1977).
In
addition, many reportshave
indicated
that
the
peripheral
visual
field
is
more effectivefor
vectioninduction
thanthe
central visualfield
(Brandt
et al.,1973;
Held
etal.,1975;
Johansson,
1977;
Dichgans
&
Brandt,
1978).
Brandt
etal.(1973)
reportedthat
stimulipresented
to
the
central30
deg
ofthe
visualfield
could notinduce
vection,However,
stimulipresented
in
the
peripheral
120
ofthe
visualfield
were ableto
induce
strong vection.In
contrast, seme re-searchershave
proposed
that
the
important
factor
is
the
area ofpresentation,
notthe
position
ofthe
stimulusin
the
yisualfield,
andthat
there
is
nodifference
in
the
abilityto
induce
vection
between
the
peripheral
and central visualfield
(e.g.
Post,
1988).
Post
(1988)
reportedthat
nopart
ofvisual
field
was more
effective
for
vectioninduction
than
others.Naka-mura
(2006)
suggeststhat
the
discrepancy
between
peripheral
and central
visual
fields
reportedin
earlier studies canbe
ex-plained
by
the
perceived
depth
effbct.That
is,
stimuliin
the
peripheral
visualfield
maybe
perceived
asfurther
awaythan
those
in
the
central visualfield,
and vectionis
more efficientlyinduced
by
stimulithat
arefurther
awaMPerceived
depth
is
animportant
factor
for
vection, asdescribed
in
the
nextsectien.
Stimulus
depth
and
vection
Stimulus
depth
is
one ofthe
mostimportant
factors
in-velved
in
vection.As
described
above, a number of studieshave
reportedthat
stimulifurther
awayinduce
strongervec-tien
(Delmore
&
Martin,
1986;
Ohmi
&
Howard,
I988;
Howard
&
Heckman,
1989;
Ito
&
Shibata,
2005),
Ito
andShi-bata
(2005)
presented
stimulithat
simultaneouslyinvolved
contraction and expansion, superimposed so
that
onedirec-tion
of movement appeared closeg whilethe
other appearedfurther
awayUnder
these
conditions,the
further
plane
wasfound
to
dominate
the
directien
of vectioninduced;
Le.
if
the
expansion appeared
further
awaythan
the contraction,vec-tion
wasinduced
in
the
forward
direction,
However,
Naka-mura
and
colleagues reported an exceptionto
this
rule (Naka-mura&
Shimojo,
2000,
2003;
Nakamura,
2004),
reportinginyerted
vectionthat
wasdominated
by
the
foreground
stimu-lus.
When
foreground
rnotien
was slow and orthogonalto
the
farther
motion,it
wasfound
to
induce
vectionin
the
same di-rection asthe
foreground
metion.Nevertheless,
apartfrom
this
exception, vectienis
censistently reportedto
be
deter-mined
by
the
direction
ofthe
farthest
motionin
avisualstim-ulus.
The
propesed
plausibility
framework
is
alsoable
to
ex-plain
this
feature
of
vectien, sincethe
mostplausible
frame
of referencefor
interpreting
the
worldis
the
furthest
away.Color
and
vectionBonato
andBubka
(2006)
reportedthat
agrating
[onsisting of sixdifferent
colors
is
mere effectivefor
vectioninduction
than
ablack
and whitegrating,
In
addition,Bubka
andBonato
(2elO)
reportedthat
a natural scene with color was more effectivefor
inducing
vectionthan
a natural scene without color,indicating
that
the
addition of color canfacilitate
vec-tion,
Visual
scenesin
the
real world aregenerally
fi11ed
with color.Thus,
a colorfu1stimulus
is
likely
to
be
interpreted
as a moreplausible
representation ofthe
world,and
thus
induce
strongervection,
Seno,
Sunaga
andIto
(2010)
comparedthe
effectiveness ofred and
green
opticflows.
The
results revealedthat
red was a relativelyineffective
colorfor
inducing
vection,
whilegreen
was aseffectivefor
inducing
vection asagrey
stimulus.Naka-The Japanese Psychonomic Society
NII-Electronic Library Service
TheJapanesePsychonomic Society
SENo
:
Vection
strengthis
determined
by
the
plausibility
ofa stimulus asarepresentation ofthe werld63
mura,
Seno,
Ito,
&
Suanga
(2010)
examinedthe
effect of stimuliwith
dynamic
color changes onthe
induction
ofvec-tion,
We
u$ed an opticfiow
withdots
whose color was changedbetween
red andgrey
coherently at1
Hz
and used an opticflow
withdots
whose colors were changedbetween
red andgrey
randomlMThe
vection magnitude waslarger
in
the
random condition
than
in
the
ceherent condition.Dynamic
changes of color arethought
to
represent color changes ofob-jects
themselves.
However,
the
coherent color change wasper-ceived as changes
in
illumination
of
the
world,
which maydrastical!y
decrease
the
perception
ofthe
stabilityofthe repre-sentationof
the
world.The
dynamic
color changedrastically
reducedthe
stability andthe
plausibility
ofthe
static not moving world,Therefore,
dynamic
color changes appearto
in-hibitvection,
The
object andbackground
hypothesis
of
vection
Overall,
the
evidencediscussed
aboveindicates
that
vectionis
efficientlyinduced
by
astimulus component thatis
inter-preted
asthe
background,
gnd
notby
stimulus componentsthat
areinterpreted
as objects(Sato,
Seno,
Kanaya,
&
Fuka-zawa,2007).
This
distinction
between
object andbackground
appearsto
be
the
critical stimulusproperty
determining
the
effecttveness of avectioninduction
stirnulus.The
properties
reportedto
be
associated withbackground
arefurther
dis-tance,
peripheral
position,
lower
spatialfrequencF
larger
size,and
green
color,Closer
distance,
centralposition,
higher
spa-tial
frequencM
smaller sizeand
red
coler
areproperties
associ-ated with components of astimulus
interpreted
as objects,I
propose
that
the
perception
of vectioni$
driven
by
representa-tions
of the world.If
arepresentation ofthe
worldhas
rnoreplausibi!ity
as
a
frame
of
reference,it
wouldbe
expectedto
in-duce
stronger vection.Plausibility
is
influenced
by
the
proper-ties
ofthe
background,
This
hypothesis
wasdirectly
investi-gated
andprooyed
in
aprevious
study(Seno,
Ito
andSunaga,
2e09).
Applying
this
hypothesis
allows many aspects ofvec-tion
to
be
understoodin
anintuitive
manner.A
novel
perspective
of vectionstimulus
attributes
The
present
article summarized vectien studiesthat
have
elucidated
the
stimulus attributesthat
determine
the effective-ness of vectioninduction
stimuli,Overall,
the
findings
dis-cussed suggest
that
the
critical attributes canbe
explainedby
the
object andbackground
hypothesis
and
the worldplausi-bility
hypothesis.
That
is,
vection appearsto
be
strengthenedwhen a
visual
frame
ofreference makes usinfer
a moreplausi-ble
representation efthe $taticworld.Visual
representations efthe
world are moreplausible
whenthey
appearto
be
bigger,
further
awaFperipherally
positioned
and colored.Vectien
ts
induced
by
attemptsto
make sense of amoving world,In
this
process,
interpretation$
of visual stimuli changefrom
repre-sentatiens of a moving world to representations of a static werld and amoving selfI
propose
that
this
framework
allows vectionto
be
more easily understeod.Acknowledgements
This
workis
supportedby
Program
to
Disseminate
Tlenure
Tracking
Systems,
MEXT;
Iapan.
References
Brandt,
Tl,
Dichgans,
J,,
&
Koenig,
E.
(l973)
Diffbrential
effects of central versus
peripheral
vision en egocentric andexocentric motion
perception,
Experimental
Brain
Research,
5,476-491,
Bonato,
F,
&
Bubka,
A.
(2006)
Chromaticity;
spatialplexits
and selfmotionperception,
Perception,
35,
53-64.
Bubka,
A.,
&
Bonato,
F.
(2010)
Natural
visual-fieldfeatures
enhancevection.Perception,39,627-635,
Bubka,
A.,
Bonato,
E,
&
Palmisano,
S.
(2008)
Expanding
andcontracting optic-fiow
patterns
and vection.Perception,
37,
704-711.
Dichgans,
J.,
&
Brandt,
T.
(1978)
Visual-vestibular
tion:
Effect
on selfmetionperceptien
andpostural
control,In
Held,
R.,
Leibowitz,
H.
W,
&
Tbeber,
H,
L,
(Eds.),
book
ofsensory
pilysiology.
Berlin.
Springer-Vlerlag,
pp.
230.Fische4
M.
H.,
&
Kernmuller,
A.
E.
(1930)
Optekinetic
geloste
Bewegungs-wahrnehmungen
under
Nystagmus.
J
Psychol,
IVburoL,
41,
273-308.
Held,
R.,
Dichigans,
J.,
&
Bauer,
J.
(1975)
Characteristics
ofmoving visual scenes
influencing
spatial
orientation.Wsion
Res.,15,357-365,
Howard,
I,
R,
&
Heckman,
T,
(1989)
Circular
vection as afunction
ofthe
relative sizes,distances,
andpositions
oftwo
competing visual
displays,
Perception,
18,
657-66S.
Ito,
H,,
&
Shibata,
I.
(2005)
Selfmotion
perceptien
frem
panding
and contracting opticalflows
overlapped withbinocular
disparity
Vi'sion
Res,,
45,
397-402.
Johansson,
G,
(1977)
Studies
on visualperception
oftion,
Perception,
6,
365-376.
Kennedv
R.
S,,
Heuingeg
L.
J.,
Harrn,
D.
L.,
OrdM
J.
M,,
&
Dunlap,
Mn
R
(1996)
Psychophysical
scaling of circularvection
(CV)
produced
by
eptokinetic(OKN)
motien:dividual
differences
and effects ofpractice.
J
Vlestibul
Res.,
6,
331-341.
Lestienne,
E,
Soeching,
J.,
&
Berthoz,
A.
(1977)
Postural
The Japanese Psychonomic Society
NII-Electronic Library Service
The JapanesePsychonomic Society
64
The
Japanese
Journal
ofPsychonomic
Science
Vbl.32,
No.
1
Brain
Res.,
28,
363-384.
Mach,
E.
(1875)
Grundlinien
der
Lehre
vonden
gungsempfindungen.Engelman.
Nakamura,
S,
(2006)
Effects
ofdepth,
eccentricity and size ofadditional static stimulus on visually
induced
self-motionperception.
Vision
Res.,
46,
2344-2353.
Nakarnura,
S.,
&
Shimojo,
S,
(2000)
A
slowly movinggreund
can capture an observer's selfmotion a repert ofnew motion
illusion:
Inverted
vection.Vision
Res.,
40,
2915-2923.
Nakarnura,
S.,
&
Shimojo,
S,
(2003)
Sustained
deviation
ofgaze
direction
can affect "inverted vection"induced
by
the
foreground
motion.11ision
Res.,
43,
745-749.
Nakamura.
S.,
Seno,
r,
Ito,
H.,
&
Sunaga,
S.
(2010)
Coherent
Modulation
of
Stimulus
Colour
Can
Affect
Visually
duced
Selfmotion
Perception,
Perception,
39,
1579-1590.
Ohmi,
M.,
Howard,
I.
R,
&
Eveleigh,
B.
(I986)
Optokinetic
nystagmus:
the
efflects of stationary edges, alone andin
combinationwith
central occlusion.Vision
Res.,
26,