Ou six h sense: Humans loca e he sou ce
o g ound ib a ions wi h hei ee .
Nicolas G imaul 1,4, Eli Kab i i1, Nicolas Ma he on2,3,5 and Fabien
Pe in1,5
1Equipe Cogni ion Audi i e e Psychoacous ique, CRNL, Uni e si y Lyon 1, CNRS,
Inse m, Lyon, F ance.
2ENES Bioacous ics Resea ch Lab, CRNL, Uni e si y o Sain -E ienne, CNRS,
Inse m, Sain -E ienne, F ance.
3Ecole P a ique des Hau es E udes, CHA lab, EPHE - PSL Uni e si y, Pa is,
F ance.
4Lead con ac .
5Co-las au ho s
*E-mail: [email p o ec ed]
***
eTOC blu b
G imaul e al.’s s udy e eals ha humans possess a ib o ac ile “six h
sense”. S anding indi iduals can loca e he sou ce o g ound ib a ions
using hei ee , elying on iming and/o in ensi y di e ences be ween
hem. This abili y, analogous o binau al hea ing, ep esen s a p e iously
unknown channel o human spa ial pe cep ion.
***
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Vision, hea ing and ol ac ion a e conside ed he p ima y senso y channels
o loca ing dis an objec s o li ing beings beyond ou immedia e each1.
While some animals also use g ound-bo ne ib a ions o sense hei
en i onmen , whe he humans could exploi his channel o loca e dis an
ib a ion sou ces emained unexplo ed. He e, we show ha humans
s anding ba e oo can pinpoin he sou ce o mechanical wa es
p opaga ing h ough he g ound. Ou esul s sugges his abili y elies on
compa ing ib a ion in ensi y and a i al ime be ween he wo ee , a
p ocess s ikingly analogous o binau al sound localiza ion by he audi o y
sys em. Ac ing as a di ec ional an enna, he ee p o ide a hi he o
o e looked senso y channel o accessing po en ially i al spa ial
in o ma ion. This disco e y no only e eals a hidden ace o human
pe cep ion bu also aises impo an ques ions abou he impac o
pe asi e an h opogenic g ound ib a ions on how we na iga e ou
senso y wo ld.
Loca ing he sou ces o senso y s imuli is a undamen al p e equisi e
o animals, including humans, o in e p e hei en i onmen and
espond app op ia ely. A a ie y o species a e sensi i e o ib a ional
signals and may use hem o communica e, escape p eda o s o ind
hei p ey
2,3
(Figu e 1A). Some species o in e eb a es and snakes
ha e been shown o loca e he sou ce o g ound ib a ion wi h g ea
accu acy (e.g., sand sco pion, Saha an sand ipe s, sand ish liza d
2,4
).
In mammals howe e , his localizing abili y has been expe imen ally
demons a ed only in mole a s
4,5
. While Humans eadily pe cei e oo
ib a ions
6
, his senso y channel is o en conside ed unc ionally
limi ed, and i is unknown whe he Humans can localize a dis an
sou ce o ib a ion. Ye , his abili y could ha e been o p ima y
impo ance in he con ex o ou ances al hun e -ga he e ’s way o
li e, whe e o ien ed displacemen s including an i-p eda o y beha io s
and o aging decisions depended on a mix u e o senso y cues
p o ided by he en i onmen .
2
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
In he p esen s udy, we es ed human adul s (N= 6 men and 6
women, aged 23-50 yea s) on hei abili y o epo he di ec ion o a
g ound-bo ne ib a ion sou ce unde condi ions ha elimina ed all
isual and audi o y cues. Du ing each ial (n = 80 / pa icipan ),
pa icipan s s ood ba e oo on a me al pla e inside a semi-anechoic
boo h, blind olded and wea ing a
noise-cancelling headphone
playing
low-pass il e ed noise o mask any ai bo ne sound (Figu e 1B). A
ib a o y de ice placed one me e away gene a ed sinusoidal
ib a ions a ei he 30 Hz o 125 Hz. These equencies we e chosen o
di e en ially ac i a e cu aneous mechano ecep o s: 30 Hz s imula es
bo h as -adap ing (FA) and slow adap ing (SA) ecep o s, whe eas 125
Hz p ima ily ac i a es FA ecep o s
7
. Vib a ions we e deli e ed om
one o ou di ec ions ela i e o he pa icipan (0° = on , 180° =
back, + 90° = igh , - 90° = le ). To es ou co e hypo hesis ha
localiza ion depends on compa ing inpu s be ween he ee ,
pa icipan s pe o med he ask while s anding on ei he a single oo
o bo h ee . S imulus o de and condi ions we e coun e balanced
(Supplemen al In o ma ion, Supplemen al Table S2).
Ou esul s p o ide he i s di ec e idence ha humans can
de e mine he di ec ion o g ound-bo ne ib a ions. As illus a ed in
Figu e 1C, pe o mance was signi ican ly abo e he 25% chance le el
when pa icipan s used bo h ee o de ec he 125 Hz s imulus,
eaching an a e age o 50% co ec decisions (Bayesian Fac o > 100,
indica ing e y s ong e idence; Supplemen al Table S1). Unde his
op imal condi ion, accu acy was obus o sou ces loca ed o he on
and sides, hough we obse ed a no able inc ease in con usion o
sou ces o igina ing om he back o pa icipan s (Figu e 1D). This
pa e n is a classic signa u e o spa ial hea ing ha elies on binau al
cues. The d ama ic imp o emen in pe o mance when using wo ee
compa ed o one s ongly suppo s he hypo hesis ha he ee
unc ion as a “ ib o ac ile an enna”, wi h he b ain in eg a ing
in o ma ion om bo h o compu e di ec ion.
3
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
The compu a ional mechanisms appea o be analogous o ha o
audi o y localiza ion, which elies mainly on In e au al Time Di e ence
(delay o sound a i al be ween bo h ea s) and In e au al Le el
Di e ence (di e ence in sound ampli ude). To in es iga e whe he
analogous cues ope a e o ib o ac ile localiza ion ia he ee , we
que ied pa icipan s abou hei s a egy. Ele en ou o wel e epo ed
ac i ely compa ing sensa ions be ween hei ee . This subjec i e
epo aligns wi h he physical eali y o he s imulus. Gi en he
a e age in e - ee dis ance o 26.4 ± 3.3 cm, we calcula ed he
es ima ed In e -Fee Time Di e ence (IFTD) o a la e al sou ce (± 90°)
o be app oxima ely 9 ms (See Supplemen al In o ma ion). This
empo al delay is well wi hin he disc imina i e capaci y o he human
soma osenso y sys em, which can esol e empo al di e ences on he
o de o milliseconds
7,8
. Fu he mo e, we measu ed In e -Fee Le el
Di e ences (IFLD, ampli ude di e ence in e ms o displacemen ) o 7
mm o he 30 Hz s imulus and 0.15 mm o he 125 Hz s imulus
(Supplemen al In o ma ion). Conside ing he known sensi i i y o
cu aneous mechano ecep o s
9
, hese sub le ampli ude di e ences
should also be pe cep ible. Since bo h IFTD and IFLD a y p edic ably
wi h sou ce di ec ion, hey ep esen plausible and su icien cues o
localiza ion. The supe io pe o mance a 125 Hz likely s ems om he
p ope ies o FA ecep o s, whose apid esponse cha ac e is ics a e
ideally sui ed o encoding he p ecise onse iming and ansien
ampli ude in o ma ion c i ical o esol ing hese sub le in e - ee
di e ences
8
.
In summa y, his s udy demons a es ha humans possess a la en
“six h sense”, a capaci y o u ilize ib o ac ile pe cep ion h ough hei
ee o gain spa ial in o ma ion abou dis an ib a ion sou ces. The
unde lying senso y p ocessing appea s o sha e undamen al p inciples
wi h binau al hea ing, sugges ing ha he compa ison o signals om
pai ed senso s is a conse ed, c oss-modal s a egy o spa ial
compu a ion in he b ain. This disco e y opens se e al exci ing
a enues o u u e esea ch. Neu oimaging s udies could explo e
4
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
whe he ib o ac ile localiza ion ec ui s co ical a eas ha o e lap
wi h hose in ol ed in audi o y spa ial p ocessing. Fu he wo k could
also in es iga e whe he his abili y can be enhanced h ough aining,
po en ially o e ing new senso y subs i u ion pa hways o indi iduals
wi h isual impai men s. Finally, ou indings ha e p o ound
implica ions o unde s anding ou in e ac ion wi h mode n
en i onmen s. We a e cons an ly exposed o a cacophony o low-
equency, g ound-bo ne ib a ions om a ic, machine y, and
in as uc u e. This “ ib a ional noise pollu ion” may no only mask ou
abili y o pe cei e meaning ul en i onmen al signals bu could also
ha e unsuspec ed, subliminal e ec s on ou na iga ion, s ess, and
cogni i e load. Unco e ing his do man senso y skill is no jus a
ma e o scien i ic cu iosi y; i p omp s a e-e alua ion o ou
ela ionship wi h he e y g ound benea h ou ee .
Supplemen al in o ma ion
Supplemen al in o ma ion including expe imen al p ocedu es, 2
Supplemen al Tables, and supplemen al e e ences can be ound wi h his
a icle online a XXX.
Acknowledgemen s
We hank E ienne Pa ize and E ienne Gaud ain o help ul ad ice on he
design o he expe imen al and he s a is ical analyses. Funding was
p o ided by CeLyA (ANR 10-LABX-0060), he Uni e si y Lyon 1 (UCBL,
F ance), he Uni e si y o Sain -E ienne and he Cen e Na ional de la
Reche che Scien i ique (CNRS, F ance).
Re e ences
1. Wu Y., Chen K., Ye Y., Zhang T., and Zhou W. (2020) Humans na iga e wi h
s e eo ol ac ion, PNAS 117 (27) 16065-16071. DOIS:10.1073/pnas.2004642117
5
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
2. Coc o , R. B., Gogala, M., Hill, P. S. and Wessel, A. (eds) S udying Vib a ional
Communica ion Vol. 3 o Animal Signals and Communica ion (Sp inge Be lin
Heidelbe g, Be lin, Heidelbe g, 2014). DOIS: 10.1007/978-3-662-43607-3.
3. Hill, P. S. and Wessel, A. Bio emology. Cu en Biology 26 (5), R187–R191
(2016). DOIS: /10.1016/j.cub.2016.01.054 .
4. Na ins, P. M. in Vib a ion Communica ion in Ve eb a es (eds Ba h, F. G. &
Schmid, A.) Ecology o Sensing 127–148 (Sp inge Be lin Heidelbe g, Be lin,
Heidelbe g, 2001). DOIS: 10.1007/978-3-662-22644-5 7.
5. Mason M.J., and Wenge L.M.D. (2019) Mechanisms o ib a ion de ec ion in
mammals in Bio emologyS: S udying Vib a ional Beha io , Animal Signals and
Communica ion 6, PSM Hill e al (eds) Sp inge Na u e Swi ze land AG 2019, DOIS:
10.1007/987-3-030-22293-2_10
6. Fo a, N.G., G i in, M. J., and Mo ioka, M. Di e ence h esholds o ib a ion o
he oo : Dependence on equency and magni ude o ib a ion. Jou nal o Sound
and Vib a ion 330 (4), 805–815 (2011). DOIS: 10.1016/j.js .2010.08.039
7. Sieden op , C., Heubach, K., Ischebeck, A., Gallasch, E., Fend, M., Mo aghy, F.,
Koppels ae e , F., Haala, I..A., K ause, B., Felbe , SS.
e al
(2008). Va iabili y o
BOLD esponse e oked by oo ib o ac ile s imula ion: In luence o ib a ion
ampli ude and s imulus wa e o m. Neu oImage. 41. 504-10.
10.1016/j.neu oimage.2008.02.049.
8. S zalkowski, N.D.J., Ali, R.A., and Ben , L.R. The i ing cha ac e is ics o oo
sole cu aneous mechano ecep o a e en s in esponse o ib a ion s imuli.
Jou nal o Neu ophysiology 118 (4), 1931–1942 (2017). DOI:
10.1152/jn.00647.2016
9. Kekoni, J., Hmlinen, H., Rau io, J. & Tuke a, T. Mechanical sensibili y o he sole
o he oo de e mined wi h ib a o y s imuli o a ying equency. Expe imen al
B ain Resea ch 78 (2) (1989). DOIS: 10.1007/BF00228915
Decla a ion o In e es s
The au ho s decla e no compe ing in e es s.
6
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
Figu e 1. Humans can pinpoin he di ec ion o a g ound ib a ion
sou ce.
(A) Sensibili y o subs a e ib a ions is known in a ious animals (blue
and ed), bu he abili y o localize ib a ion sou ce has been
expe imen ally con i med in only a ew ( ed). Human localiza ion abili y
was p e iously unexplo ed. (B) Expe imen al se -up. Blind olded
pa icipan s, wea ing noise-cancelling headphones, s ood ba e oo (one o
wo ee ) on a pla e and epo ed he pe cei ed di ec ion (F on , 0°; Back,
180°; Le , -90°; Righ , +90°) o a g ound ib a ion sou ce (1 m away; 30
Hz o 125 Hz sine wa es; ~10 dB abo e de ec ion h eshold). (C)
Pe cen age o co ec localiza ion esponses ac oss condi ions.
Pe o mance signi ican ly su passed chance (25%, dashed ed line), wi h
highes accu acy o wo ee and 125 Hz s imulus. (D) Dis ibu ion o
epo ed di ec ions unde op imal condi ions ( wo ee , 125 Hz) o each
sou ce loca ion (pu ple a ows). Do s indica e he mean numbe o ials
epo ed pe di ec ion (max 5); g ey a eas ep esen SD ac oss
pa icipan s. Responses gene ally aligned wi h he ue sou ce di ec ion,
excep o back condi ion.
7
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
SUPPLEMENTAL INFORMATION
Supplemen al Table S1. E ec o expe imen al condi ions on
localiza ion accu acy. Resul s o Bayesian one sample - es assessing
whe he accu acy in each condi ion exceeded he 25% chance le el (H1).
Bayesian ac o s and associa ed e o s a e shown. BFs > 10 = s ong
suppo o H1S; > 3 = mode a e suppo o H1S; < 0.3 = mode a e
suppo o H0S; < 0.1 = s ong suppo o H0.
8
203
204
205
206
207
208
209
210
211
212
Supplemen al Table S2. De ec ion h esholds and in ensi y le els
used o he expe imen s.
9
213
214
215
216
217
218
