Animal Communication Animal Communication – – Focus on Vocal Learning Focus on Vocal Learning Focus on Vocal Learning Focus on Vocal Learning Stephanie A White, PhD Dept of Integrative Biology and Physiology UCLA UCLA June 2010 June 2010 UQAM Summer Institute on the Origins of Language
Language Language • A major component of what makes A major component of what makes humans unique • Sensory motor conceptual intentional • Sensory-motor, conceptual-intentional, recursion (Hauser, Chomsky & Fitch; Science, 2002) • Preadaptions for subcomponents must P d ti f b t t exist in other species.
The vocal learning sub-component • The ability to modify motor programs to make new vocalizations • Used in mimicry, but can be creative • Accomplished via hearing others and auditory feedback of self • Not to be confused as association of innate sounds with objects in the environment • Not ‘uniquely human’ yet, critical t to a sense of well-being f ll b i
The KE Family Speech Deficit The KE Family Speech Deficit • videos 'Friends' Timmy w/speech therapist Timmy w/speech therapist You Tube clip Need for non-human animal models
How can we prove that a given species is capable of vocal species is capable of vocal learning?
Brief Communication Brief Communication Nature (2005) A i Animal behaviour: Elephants are capable l b h i El h t bl of vocal learning Joyce H. Poole, Peter L. Tyack, Angela S. Stoeger-Horwath and Stephanie Watwood “T “Two animals coin unexpected sounds as i l i t d d a surprising form of social communication ” communication.
Sonograms of vocalizations Mlaika Calimero z ncy, Hz equen Fre Time, sec Sonograms are visual depictions of sound Sonograms are visual depictions of sound X axis = frequency (pitch), y = time, darkness = amplitude
Evidence for elephant vocal learning Scatterplot of frequency vs. duration of calls. Ml ik ' Mlaika's imitations (green triangles) resemble truck sounds (light blue i it ti ( t i l ) bl t k d (li ht bl triangles) and differ from her normal calls (yellow (yellow triangles) triangles), which African elephants not only sound different resemble calls of other African elephants (dark blue: stars, adult female; diamonds adult male; squares female calf; hexagons male female; diamonds, adult male; squares, female calf; hexagons, male f from Indian elephants… I di l h t calf). Calimero makes chirp-like sounds (pink circles) similar to the chirps of the Asian elephants (red circles) who lived with him.
…they look different too! Indian elephant African elephants
Songbirds are ideal for the study of vocal learning
O N S D A D O N S
Developmental timeline/Critical periods/Auditory dependence
Relative timing of human vs. zebra finch vocal learning vocal learning Babbling/ language-specific speech perception p p p First words! uman Universal speech perception 0 3 6 9 12 months nch H 20 60 100 days Fin Sensory acquisition Sensorimotor learning Crystallization Hatchling 13days 17days Fledgling Adult
Adult singing: practice vs Adult singing: practice vs. performance Undirected Directed adapted from Jarvis et al (1999)
Normal song development Normal song development Crystallization Sensorimotor Learning Sensory Acquisition Sings A A B A A B B Sings B>A Sings B>A 0 50 100 Post-hatch day
Deafening Crystallization Sensorimotor Learning Sensory Acquisition Deafening 65 15 90 35 Post-hatch day
Vocal learning depends on hearing Eastern phoebe Zebra finch intact deaf The Eastern phoebe in contrast to The Eastern phoebe, in contrast to the zebra finch, is not a songbird and doesn’t need to learn its song Wilbrecht & Nottebohm, 2003 Kroodsma & Konishi, 1991
Isolation from tutor Isolation from tutor Crystallization Sensorimotor Learning Sensory Acquisition Isolation 0 100 100 50 50 Post-hatch day
I solates learn new notes when controls do not CONTROL I SOLATE Pre- tutor 65d TUTOR TUTOR Post- tutor 86d 86d
Two similar birds, but w m , u Memorizes Memorizes Does not memorize Does not memorize
Summary of auditory experience Summary of auditory experience • Females prefer songs of normal > isolate > deaf at 35 > deaf at 20 days • Isolation can create two birds of the same chronological age, but of different learning states • Enables identification of molecules unique to the memorizing state • A similar postponement of sensorimotor learning can be done (Funabiki & Konishi, 2003)
Neural substrate Neural substrate
Zebra finch brains are sexually dimorphic HVc D RA T h m e s t y s g n s o e u s e a l y d i m o r p h i c x i s H V c e l a e m F R R A M a l e D D R Female Female Male Male (courtesy of Eugene Akutagawa)
The avian song circuit Cortex HVC HVC HVC HVC HVC HVC “sing s g this!” Cortex RA RA RA RA RA RA X X Motor to motor neurons to to neurons neurons syrinx syrinx syrinx syrinx vocal motor pathway vocal motor pathway controlling the syrinx
The avian song circuit Cortex HVC HVC HVC HVC HVC HVC “modify!” S triatum LMAN LMAN LMAN C H V C H V H V C L MA N L MA L MA N N s y R A R A x n r r s y i M R A i D L x n D L M M D L X X X RA RA RA RA RA RA X X T halamus DLM DLM DLM anterior forebrain pathway
Comparison of avian and human circuitry HUMAN SONGBIRD Cortex Cortex S triatum S triatum Area X T halamus T halamus
Songbird model system Songbird model system • Vocal learner that captures some of the Vocal learner that captures some of the complexity of human speech • Socially-learned and influenced • Parallel phases and neuroanatomy p y • Advantage: Identified circuit • Advantage: Identified circuit
The KE family Deficits in spoken and written language • Production and comprehension Production and comprehension • • Normal non-verbal IQ but low verbal IQ • Normal peripheral control of orofacial • musculature musculature orofacial dyspraxia Vargha-Khadem et al (1995) PNAS
The basal ganglia is a major site of abnormality abnormality Frontal view Transverse view striatum striatum Watkins et al (2002) Belton et al (2003)
FoxP2 - a molecular entry point to language dimerization DNA binding poly-Q 1 1 152 152 209 209 504 504 587 587 715 715 R553H • FOX genes are transcription factors, so act by regulating other molecules molecules • FOX family members act during embryogenesis • FOXP2 mutations cause orofacial dyspraxia Lai, Fisher et al. (2001) Nature
FoxP2 DNA Binding Domain: R553
Working model for FoxP2 in vocal learning Formation vs function Practice vs performance 0 30 60 90 120 Sensory acquisition Sensorimotor learning Maintenance ? ? xP2 Fox
Sagittal vs. frontal orientation HVC LMAN RA X X X cortex DLM D D striatum R L thalamus
Detection of ‘gene of interest’ bird 4 slide A slide A GENE GENE mRNA: a a g c c t probe: u u c g g a
Parallel expression of FOXP2 in songbirds and humans 1d zebra finch 22wk human embryo sense sense antisense antisense antisense antisense sense sense Teramitsu, Kudo, Geschwind, White (2004)
Working model for FoxP2 in vocal learning Formation vs function Practice vs performance 0 30 60 90 120 Sensory acquisition 2 FoxP2 Sensorimotor learning Maintenance ? ? rgets Tar
Adult singing: two behaviors in one Adult singing: two behaviors in one Directed Undirected adapted from Hessler & Doupe (1999)
FoxP2 expression in Area X is regulated by the social context of song by the social context of song Non-singing Undirected Directed mid mid 3’ 3’ mid mid 3’ 3’ mid mid 3’ 3’ nal xP2 sign 1.2 alized Fox 1.0 Norma 0.8 * * Kruskal-Wallis, p < 0. 0005 Teramitsu & White (2006)
Working model for FoxP2 in vocal learning Formation vs function Practice vs performance 0 30 60 90 120 Sensory acquisition Sensorimotor learning Maintenance ? ? xP2 ets Fox Targe
Deafening prior to sensorimotor learning leads to abnormal songs Tutor Sham-control son (60d) Deafened son (60d)
Motor-driven down-regulation of FoxP2 Non-singing Singing Control Control Deaf Deaf ratio 1.2 g striatum m density r Non-singing vs. 1.1 singing: p < 0.02, ge pixel d X/outlying both conditions both conditions 1.0 Averag Area X Teramitsu, Poopatanapong, Torrisi & White (2010)
Hearing links song to FoxP2 levels Hearing Deaf R 2 0 7 R 2 0 08 R 2 =0.7, p<0.025 R 2 =0.08, p=0.43, ns 0 025 0 43 1.1 tio 1.0 1 0 xP2 rat 1.0 0 9 0.9 Fox 0.9 0 8 0.8 200 600 1000 0 1000 2000 Amount of singing seconds Amount of singing, seconds Teramitsu, Poopatanapong, Torrisi & White (2010)
Our working model for FoxP2 in vocal learning Formation vs function Practice vs performance 0 30 60 90 120 Sensory acquisition Sensorimotor learning Maintenance xP2 ets Fox Targe
Recommend
More recommend
