Content uploaded by Fabian Bross
Author content
All content in this area was uploaded by Fabian Bross on Mar 14, 2018
Content may be subject to copyright.
i
Preface & Acknowledgements
It was a great pleasure for us, hosting P&P13 in Berlin in the fall of 2017 and we thank all 137
attendees for their coming and input. We listened to 2 keynote lectures and 21 talks and had the
opportunity to discuss 59 posters over the span of this two day conference.
We are now ready to publish the proceedings of this conference and are very happy that P&P13 yet
again provided a forum for young and seasoned scientists from phonetics, phonology and various
interrelated disciplines to present their research in progress and to benefit from the ideas and input of
colleagues. While the conference is gaining more and more international visibility, the visionary idea
of P&P is to provide a forum for phoneticians and phonologists of all persuasions to discuss their work
in progress and share ideas, tools and methods while maintaining the highest standards of scientific
research and conduct.
Paper submission for this open access publication hosted by the edoc publication server of the
Humboldt-Universität zu Berlin is not mandatory. Therefore, we are very happy that so many
authors submitted their paper for inclusion of these proceedings. All papers benefitted from
comments received during the presentation of the work during the meeting and from comments
provided by the editorial team: all papers that were received were read by the editors to ensure
adherence to the publication guidelines and scientific standards.
Again, we would like to take the opportunity to thank our sponsors: the Deutsche
Forschungsgemeinschaft (DFG) supported us through a conference grant while the Institut für
deutsche Sprache und Linguistik at the Humboldt-Universität zu Berlin provided the location and
potent liquids during the conference reception. The Leibniz-Centre General Linguistics (ZAS)
generously supported the technical assistance and all catering during this conference. The
publisher Frank & Timme generously supported P&P13 with conference binders and pens. We
also thank the HU Computer and Media Service for their support in making these proceedings
accessible.
We are very lucky to have colleagues, staff, a very supportive administration and diligent student
assistants that went out of their way to make P&P13 possible and a pleasurable experience. The cover
of the proceedings was designed by Egor Savin and Maxim Welsch while Gediminas Schüppenhauer
designed the conference logo and took the conference pictures.
We are now very much looking forward to P&P14 in 2018 which will be hosted by the University of
Vienna! See you there!
Malte Belz, Christine Mooshammer
Institut für deutsche Sprache und Linguistik, Humboldt-Universität zu Berlin
Susanne Fuchs, Stefanie Jannedy, Oksana Rasskazova, Marzena Żygis
Leibniz-Center for General Linguistics (ZAS) Berlin
Berlin, January 2018
Cognitive Associations Between Vowel Length and Object Size: A New
Feature Contributing to a Bouba/Kiki Effect
Fabian Bross
University of Stuttgart
fabian.bross@ling.uni-stuttgart.de
Abstract
Previous studies on sound symbolism have
found evidence for the existence of an
association between certain speech sounds and
physical properties like shape or size: a
phenomenon usually called bouba/kiki effect.
Additionally, more and more experiments have
attempted to find out exactly which phonetic
features contribute to this effect. The present
paper reports two experiments on a previously
uninvestigated feature. The experiments show
that short vowels are associated with
small/short objects and long vowels with
large/long objects for speakers of a language
exhibiting a vowel length contrast (German).
While the association between vowel length
and object length seems to be very strong, the
association between vowel length and object
size seems to be weaker.
Introduction
A large body of evidence suggests that there
exists a cognitive association between sounds
and visual cues. This phenomenon is generally
known as the bouba/kiki effect. Tasks showing
such associations typically include a set of
non-words, like bouba and kiki, and shapes
that are depicted as either round or spiky. The
findings of such studies are that non-words
with rounded vowels and voiced consonants
(bouba) are usually considered to correspond
to round objects, while non-words without
rounded vowels and with voiceless consonants
(kiki) correspond to spiky objects (e.g., Köhler
1929; Fox 1935; Ramachandran & Hubbard
2001, 2005; Maurer, Pathman & Mondloch
2006). It was not until recently that researchers
tried to figure out exactly which phonetic
features contribute to this effect. For example,
D’Onofrio (2014) found that vowel backness
plays a role in the sound symbolic effect (see
also McCormick et al. 2015). The goal of the
present study was to extend these insights to
vowel length. To be more precise, the goal was
to test the hypothesis that long/big objects are
associated with long vowels, and short/small
objects with short vowels. Although it is
known that vowel quality is associated with
object size (e.g. /o/ with bigness and /i/ with
smallness) (e.g., Thompson & Estes 2011), the
claim that vowel quantity could play a similar
role, has, to the best of my knowledge, not
been made and tested.
Sapir (1929) was perhaps the first to show
that back vowels, for example /u/, are usually
judged to refer to big objects and front vowels,
for example /i/, are judged to relate to small
objects. Similar correspondences, namely that
some speech sounds, such as /a, u, o, m, l, v, b,
d, g/, are associated with large objects and
other speech sounds, such as /i, e, t, k/, are
associated with small objects, were also found
in later studies (e.g., Newman 1933; Taylor &
Taylor 1962; Vetter & Tennant 1967; Koriat &
Levy 1977; Mauer, Pathman & Mondloch
2006; Thompson & Estes 2011).1 Interestingly,
it has been argued that the association between
object size and speech sounds does not stem
from the articulatory properties of these
sounds, but from their acoustic properties.
Ohtake & Haryu (2013) found that participants
were faster in categorizing the size of an object
when hearing a congruent vowel (e.g., hearing
an /i/ when presented with a small object),
however this effect disappeared when they
held objects in their mouths that led to a mouth
position similar to the oral cavity shape when
producing vowels. This can be interpreted as
evidence in favor of the idea that the effect is
not caused by mouth position itself.
The present paper reports two experiments.
In Experiment 1, the claim that vowel length is
associated with object length is tested.
Experiment 2 is concerned with the presumed
association between vowel length and object
size. The results suggest that vowel length is
1A related idea, namely that the length of an
expression is reflecting larger quantity, was
famously put forward by Jakobson (1965) who tried
to show that plural morphology iconically
represents quantity.
Proceedings P&P13
17
strongly
and als
o
former
h
corresp
o
Experi
m
The
g
whether
and obj
e
native
exhibiti
n
recruite
d
pure vo
w
tense-la
x
with hi
g
with lo
w
reported
particip
a
the Uni
v
and the
i
5.30).
Six
p
Each pa
long vo
w
vowel (
s
crafted
a
and ph
presente
and g
r
graphe
m
the wor
d
a mono
s
that eac
h
of space
Table 1.
and 2; t
h
marking
s
s
yllables
consona
n
either r
e
marking
(
Pseudo-
w
vowel)
<dehla>
<mahko
>
<mutoh
>
<petoh>
<rohta>
<tehko>
In a
d
picture
p
long an
d
associated
o
with ph
y
h
as been o
b
o
ndence.
m
ent 1
g
oal of Expe
an associat
i
e
ct length e
x
speakers
o
n
g a contra
s
d
(note that
w
el length
c
x
contrast:
l
g
her muscul
a
w
er muscul
a
d
any lan
g
a
ted volunt
a
v
ersity of
S
i
r mean ag
e
p
airs of ps
e
a
ir consisted
w
el and on
e
s
ee Table 1)
a
s to have th
e
onemes, a
n
d visually, t
r
aphemes
(
m
es in each p
d
s were size
-
s
paced Cour
i
h
character
o
.
The written
s
h
e German w
r
s
of vowel l
e
are marke
d
n
tal letter; lo
e
main unma
(e.g., <ah> r
e
w
ord (long
>
>
d
dition to t
p
airs were c
r
d
a short v
e
with physi
c
y
sical size,
b
served to h
a
riment 1 w
a
i
on betwee
n
x
ists. For t
h
o
f German,
s
t in vowe
l
German d
o
c
ontrast, but
l
ong vowels
a
r tension an
d
a
r tension).
N
g
uage diso
r
a
rily. All w
e
S
tuttgart, 28
e
was 21.1
9
e
udo-words
of one wor
d
e
word con
t
. All words
e
same num
b
n
d, as the
he same nu
m
(
five lette
r
seudo-word
)
-
matched. F
o
i
er font was
o
ccupied th
e
s
timuli used i
n
r
iting system
e
ngth;
s
hort
d
by doublin
g
ng vowels i
n
rked or re
c
e
presents a lo
Pseudo
-
vowel)
<della
>
<makk
o
<mutto
>
<petto
>
<rotta>
<tekko
>
he pseudo-
w
r
eated, each
e
rsion of th
c
al elongati
o
however,
t
a
ve a stron
g
a
s to determ
i
n
vowel len
g
is purpose,
a langu
a
l
length, w
e
o
es not hav
e
there is als
o
are produ
c
d
short vow
e
N
one of th
e
ders and
e
re studen
t
s
were fem
a
9
years (SD
were creat
e
d
containin
g
t
aining a sh
o
were carefu
l
b
er of syllab
l
words w
e
m
ber of lett
e
r
s and f
o
)
. Additiona
l
o
r this purpo
s
used to ens
u
e
same amo
u
n
experiment
s
allows differ
e
v
owels in o
p
g
the follow
i
n
open syllab
c
eive a len
g
n
g /aː/)
-
word (short
>
o
>
>
>
>
w
ords, twe
l
consisting o
f
e same obj
e
o
n,
t
he
g
er
i
ne
g
th
35
a
ge
e
re
e
a
o
a
c
ed
e
ls
e
m
all
at
a
le,
=
e
d.
g
a
o
rt
lly
l
es
e
re
e
rs
o
ur
l
ly,
se,
u
re
u
nt
s
1
e
nt
p
en
i
ng
les
g
th
l
ve
f a
e
ct
(
e
1
r
a
ti
m
p
a
s
h
ri
t
o
v
e
t
h
v
e
T
h
o
b
be
c
h
t
h
v
i
s
t
l
o
l
o
o
b
(
8
F
i
c
a
s
i
1
1
e
x
h
y
be
e
.g., a key, a
shows sam
p
Figure 1.
Tw
In the act
u
a
ndomly pre
s
m
e. With e
a
a
ir of object
h
ort object
w
g
ht was als
o
o
ld that the
r
e
ry similar
t
h
at objects
i
e
rsions, na
m
h
eir task w
b
jects woul
d
e
tter. It was
h
oose signi
f
h
e pseudo-
w
i
ce versa.
The results
t
rong prefer
e
o
ng object
w
o
ng vowel (
8
b
ject when
t
8
5% of the
i
gure 2. A
a
lculated an
gnificant
p
1
5.185). Se
e
x
periment c
a
y
pothesis th
a
e
tween vow
e
F
igur
e
coat hanger
,
p
le material.
Tw
o samples f
r
u
al experim
e
s
ented with
o
a
ch word,
p
s
at random
w
as presente
d
o
randomize
r
e exists an
t
o our cultu
r
i
n the alien
m
ely a lon
g
a
s to choo
s
d
fit the p
r
expected th
a
fi
cantly mor
e
o
rd contain
e
indicate th
a
e
nce for part
i
w
hen a pseu
d
8
1% of the
h
e word co
n
cases). Th
i
two-tailed
M
d
revealed
a
p
-value <
e
also Table
a
n be regard
e
a
t there is a
e
l and object
e
2. Results
fr
,
or a hairpi
n
r
om Experim
e
e
nt, particip
a
o
ne pseudo-
w
p
articipants
(whether t
h
d
on the left
e
d). Particip
a
alien cultu
r
r
e except fo
r
world exi
s
g
and short
s
e which o
f
p
resented al
i
at participa
n
e
long obje
e
d a long v
o
a
t there was
i
cipants to c
h
u
do-word co
n
cases) and
n
tained a sh
o
i
s is summ
a
M
cNemar’s
a statistical
l
0.0001 (X
²
2. Thus, th
e
d as eviden
c
cognitive a
s
t
length.
fr
om Experim
n
). Figure
e
nt 1
a
nts were
w
ord at a
saw one
e long or
or on the
a
nts were
r
e that is
r
the fact
st
in two
version.
f
the two
i
en word
n
ts would
c
ts when
o
wel and
indeed a
h
oose the
n
tained a
the short
o
rt vowel
a
rized in
test was
l
y highly
²
(1) =
e
present
c
e fo
r
the
s
sociation
ent 1
Proceedings P&P13
18
Table 2
Long v
o
Short v
o
Tot
a
Experi
m
The
o
out if a
n
object s
i
speaker
s
Experi
m
reported
particip
a
the Uni
v
and the
i
3.31).
The
p
same fo
The sa
m
Again, t
w
consisti
n
same ob
j
or a
m
material
instructi
o
except
t
alien cu
l
large an
d
particip
a
b
ig obje
long vo
w
F
igur
Agai
n
also yi
e
20.119,
Howeve
the asso
c
is weak
e
Particul
a
relation
predicte
d
results
o
weak.
. Review of
t
Lon
g
objec
o
wels 175
o
wels 33
a
l 208
m
ent 2
o
bjective of
e
n
association
i
ze exists.
F
s
of German
m
ent 1 were
d
any lan
g
a
ted volunt
a
v
ersity of
S
i
r mean ag
e
p
resent exp
rced-choice
m
e list of
w
elve pictu
r
n
g of a large
j
ect (e.g., a
n
m
icroscope).
. Participa
n
o
ns as the p
a
t
hat they w
e
lture in wh
i
d
a small v
e
a
nts would
c
cts when th
e
w
el and vice
r
e 3. Two sa
mp
n
, a McNe
m
e
lding a si
g
p < 0.00
r, as the pi
e
c
iation foun
d
e
r than the
o
a
rly short v
o
to small o
b
d
effect was
o
f the first
t
he results o
f
g
ts
Shor
t
objec
t
41
183
224
e
xperiment
2
between vo
w
F
or this pur
p
that did no
t
recruited.
N
g
uage diso
r
a
rily. All w
e
S
tuttgart, 24
e
was 20.5
1
eriment co
n
task as in
pseudo-wo
r
r
e pairs wer
e
and a smal
l
n
ambulance,
Figure 3
s
n
ts receiv
e
a
rticipants i
n
e
re told tha
t
i
ch each obj
e
rsion. It wa
s
c
hoose sig
n
e
pseudo-w
o
versa.
mp
les from Ex
p
m
ar’s test
w
g
nificant res
u
01). See
a
e
charts in
F
d
in the pres
e
o
ne found in
o
wels seem
n
b
jects. Take
n
found, but
w
experiment
)
f
Experiment
t
t
s
Tota
l
216
216
432
2
was to fig
u
w
el length a
p
ose, 35 nat
i
t
participate
N
one of th
e
ders and
e
re students
were fem
a
years (SD
n
sisted of
t
Experiment
r
ds was us
e
e
created, e
a
version of
t
a chili pepp
s
hows sam
p
e
d the sa
m
n
Experimen
t
t
there was
ect exists i
n
s
expected t
h
ificantly m
o
o
rd containe
d
p
eriment 2
w
as calculat
e
u
lt (X² (1)
a
lso Table
F
igure 4 sh
o
e
nt experim
e
Experiment
n
ot to evok
e
n
together,
t
w
as (given
t
unexpecte
d
t
1
l
u
re
a
nd
i
ve
in
e
m
all
at
a
le,
=
t
he
1.
e
d.
a
ch
t
he
p
er,
p
le
m
e
t
1
an
n
a
h
at
o
re
d
a
e
d,
=
3.
o
w,
e
nt
1.
e
a
t
he
t
he
d
ly
T
L
S
C
s
u
a
s
l
e
be
w
o
b
o
b
d
i
e
x
e
v
be
l
e
t
h
f
o
s
e
k
n
s
t
N
c
a
e
x
R
D
F
o
T
able 3. Rev
i
L
ong vowels
S
hort vowels
Total
F
igure 4.
R
C
onclusion
s
The result
s
u
ggest that
s
sociation b
e
e
ngth, but
a
e
tween vo
w
w
as unexpec
b
jects are a
l
b
jects. How
e
i
fference b
e
x
periments,
v
idence that
e
tween the
m
e
ngth and ob
j
What rem
a
h
e same cro
s
o
und for au
d
e
ems likely
n
own to be
t
imuli are pr
e
N
ielsen & Re
a
n be replic
a
x
hibiting a c
o
R
eferences
’Onofrio, A.
(
dimensio
n
correspon
d
paradigm.
393.
o
x, C. (1935).
A
merican
https://doi
i
ew of the re
s
Long
objects
138
100
238
R
esults from
E
s
s
of the p
r
there is
e
tween vow
a
surprising
l
w
el length a
n
t
ed given
t
l
so longer,
c
e
ver, as the
r
e
tween the
this can
there is a
c
m
ental conce
p
j
ect size.
a
ins unclear,
s
s-modal co
r
d
itorily pres
e
as sound-
s
consistent r
e
e
sented visu
a
n
dall 2011)
a
ted for spea
k
o
ntrast in vo
w
(
2014). Phon
e
n
ality in soun
d
d
ences: Refin
i
Language an
An experim
e
J
ournal of Ps
y
.
org/10.2307/
s
ults of Exp
e
Short
objects
72
110
182
E
xperiment
1
r
esented ex
p
a strong
w
el length a
n
l
y small a
s
n
d object s
t
he fact th
a
compared t
o
r
e was a ra
t
results of
be interp
r
c
ategorical
d
e
ptualization
however, i
s
r
respondenc
e
ented stimu
l
s
ymbolic e
f
r
egardless o
f
ally or audi
t
and whethe
r
k
ers of lang
u
o
wel length.
e
tic detail and
d
-shape
i
ng the boub
a
n
d Speech, 53
(
e
ntal study of
s
ychology, 47
,
/
1416003.
e
riment2
Total
210
210
420
1
p
eriments
cognitive
n
d object
s
sociation
ize. This
a
t bigger
o
smaller
t
her large
the two
r
eted as
d
ifference
of object
s
whether
e
s can be
l
i (which
ff
ects are
f
whether
orily; see
r
or not it
u
ages not
a
-kiki
(
3), 367–
n
aming.
545-579.
Proceedings P&P13
19
Jakobson, R. (1965): Quest for the essence of
language. Diogenes, 13(51), 21–37.
Köhler, W. (1929). Gestalt psychology. New York:
Liverlight.
Koriat, A. & Levy, I. (1977). The symbolic
implications of vowels and of their
orthographic representations in two natural
languages. Journal of Psycholinguistic
Research, 6, 353–365.
https://doi.org/10.1007/BF01074374.
Maurer, D., Pathman, T. & Mondloch, C. (2006).
The shape of boubas: Sound-shape
correspondences in toddlers and adults.
Developmental Science, 9(3), 316–322.
https://doi.org/10.1111/j.1467-
7687.2006.00495.x.
McCormick, K., Kim, J. Y., List, S. & Nygaard, L.
C. (2015). Sound to meaning mappings in the
bouba-kiki effect. Proceedings of the 37th
Annual Conference of the Cognitive Science
Society, 1565-1570.
Newman, S. (1933). Further experiments in
phonetic symbolism. The American Journal of
Psychology, 45(1), 53–75.
https://doi.org/10.2307/1414186.
Nielsen, A. & Rendall, D. (2011). The sound of
round: Evaluating the sound-symbolic role of
consonants in the classic Takete-Maluma
phenomenon. Canadian Journal of
Experimental Psychology, 65(2), :115–24.
https://doi.org/10.1037/a0022268 .
Ohtake, Y. & Haryu, E. (2013). Investigation of the
process underpinning vowel-size
correspondence. Japanese Psychological
Research, 55, 390–399.
https://doi.org/10.1111/jpr.12029.
Ramachandran, V. & Hubbard, E. (2001).
Synaesthesia: a window into perception,
thought and language. Journal of
Consciousness Studies, 8(1), 3-34.
Ramachandran, V. & Hubbard, E. (2005).
Synaesthesia: a window into the hard problem
of consciousness. In L. Robertson & N. Sagiv
(Eds.). Synaesthesia: Perspectives from
cognitive neuroscience. Oxford: Oxford
University Press. 127-189.
Sapir, E. (1929). A study in phonetic sound
symbolism. Journal of Experimental
Psychology, 12(3), 255–239.
Taylor, I & Taylor, M. (1962). Phonetic symbolism
in four unrelated languages. Canadian Journal
of Psychology, 16, 344–356.
Thompson, P. D. & Estes, Z. (2011). Sound
symbolic naming of novel objects is a graded
function. The Quaterly Journal of
Experimental Psychology, 64(12), 2392–2404.
https://doi.org/10.1080/17470218.2011.60589
8.
Vetter, H. & Tennant, J. (1967). Oral-gesture cues
in sound symbolism. Perceptual and Motor
Skills, 24, 54.
Proceedings P&P13
20
