Stromswold, Genetics & Evolution of Language Montreal, 2010 Genetics & the Origins of Language Karin Stromswold (karin@ruccs.rutgers.edu) Psychology Dept. & Center for Cognitive Science Rutgers University - New Brunswick . Portions of this work were supported by grants from NSF (BCS-9875168, BCS-0002010, BCS-0042561, BCS- 0124095, BCS-0446838), NIH (HD37818), the Charles & Johanna Busch Biomedical Research Fund and the Bamford-Lahey Children's Foundation. K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 With many thanks to … • Ellyn Sheffield (former postdoc, now at Towson U.) • Lab Manager: Diane Molnar • A small army of research assistants and undergraduate students • 1400+ parents & children who participated in the PEGI (Prenatal Environment & Genetic Interaction) project • NSF (BCS-0446838; BCS-9875168, BCS-0002010, BCS-0042561, BCS-0124095)), Charles & Johanna Busch Biomedical Research Fund, Bamford-Lahey Children's Foundation K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 1
Stromswold, Genetics & Evolution of Language Montreal, 2010 Outline • How genetic findings can inform evolutionary theories of language • Genetic factors & language • What we know – [Molecular genetics - Enard - 6/24] – Family aggregation studies – Adoption studies – Twin studies • [What we don’t know] • Evolutionary implications of genetic findings K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 Genetics vs. Heritability of Language • Genetic of language : If innate predispositions or neural structures enable us to use/acquire language, they must be encoded in our DNA • Heritability ( h 2 ) of language: estimate of the extent to which genetic factors account for observed variance in a trait • Why we might fail to find evidence for significant h 2 ’s – Genetic factors not important for language – Environmental factors swamp genetic factors – People are linguistically identical: if there is no variation, h 2 = 0, even if genetic factors determine a trait Chomsky (1980): Language is like number of fingers Lieberman (1984): Language is like height K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 2
Stromswold, Genetics & Evolution of Language Montreal, 2010 Variability in Language Acquisition • Specific Language Impairment (SLI): – 5-10% of children have unexplained spoken language impairments • Variability in rate of acquisition in normal children – Phonology (e.g., Sander, 1972, Vihman, 1990) – Vocabulary (e.g., Goldfield & Reznick, 1990, Fenson et al., 1994), – Morphology (e.g., Brown 1973, deVilliers & deVilliers, 1973), – Syntax (e.g., Stromswold 1990, 1995, Snyder & Stromswold 1997) K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 Variability in Adult Language • SLI in adults: – 50% of SLI children grow up to be language impaired • Variability in normal adults’ linguistic proficiency : – Verbal fluency (e.g., Day, 1979) – Interpretation of compound nouns (e.g., Gleitman & Gleitman, 1970) – Sentence processing (e.g., Bever et al., 1989, King & Just, 1991; Corley & Corley, 1995) – Grammaticality judgments (e.g., Ross, 1979; Nagatu, 1992; Cowart, 1994). – Second language acquisition (e.g., Fillmore, 1979; deKeyser, 2000, Dornjei, 2005) K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 3
Stromswold, Genetics & Evolution of Language Montreal, 2010 Genetics & the Evolution of Language • If language h 2 ’s are significant, then genetic factors play some role in language acquisition or linguistic proficiency obtained – It does not mean that genetic factors do play a role for “normal” people – It does not mean that the genetic factors are specific to language or subcomponents of language Significant language h 2 ’s mean that genetic variance exists now • • Genetic variance for language means something that contributes to linguistic ability could have been selected for. – It does not mean that it was selected for – Even if it was selected for, it does not say what was selected for K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 Pathological & Normal Variance: Fingers Number of fingers: significant h 2 for pathological variance only Polydactyly Oligodactyly Length of fingers: significant h 2 for pathological & normal variance Arachnodactyly Brachydactyly Long fingers Stubby fingers K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 4
Stromswold, Genetics & Evolution of Language Montreal, 2010 Normal & Impaired Language h 2 ’s • Possibility 1: SLI h 2 = normal language h 2 • Possible interpretation: Same genetic factors for SLI & normal language • Possibility 2: SLI h 2 > normal language h 2 • Possible interpretation: Genetic factors affect both SLI and normal language, but some that only affect SLI language (“length of fingers”) • Possibility 3: only SLI h 2 is significant • Possible interpretation: Genetic factors only account for SLI variance (“number of fingers”) • Evolutionary implications of Possibilities 2 & 3 If SLI h 2 ≠ normal language h 2 , this is consistent with unique neural underpinnings and/or evolutionary history for language in two groups K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 Selectivity of Language h 2 ’s Do genetic factors selectively affect some aspect(s) of language? – Spoken vs. written language – Expressive vs. receptive language – Subcomponents of language (phonology, lexicon, morphology, syntax, discourse etc.) – Competence vs. performance: (e.g., the rat the cat chased fled) Evolutionary implications – If different h 2 ’s for different aspects of language, different genetic factors may be involved (see “Specificity”) This is consistent with different aspects of language having unique neural underpinnings and/or evolutionary history K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 5
Stromswold, Genetics & Evolution of Language Montreal, 2010 Selection & Competence/Performance K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 Specificity of “Language” h 2 ’s Do the same genetic factors affect linguistic & nonlinguistic abilities? Genetic overlap for linguistic & nonlinguistic abilities is consistent with them having shared neural underpinnings and/or evolutionary history Do the same genetic factors affect different aspects of language? Genetic overlap for subcomponents of language is consistent with them having shared neural underpinnings and/or evolutionary history Are there genetic factors that are specific to specific aspects of language? Lack of genetic overlap for subcomponents of language is consistent with them having unique neural underpinnings and/or evolutionary history K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 6
Stromswold, Genetics & Evolution of Language Montreal, 2010 [Molecular Genetics of SLI] • FOXP2-CNTNAP2 gene (see Enard’s talk on Thursday) – Autosomal dominant disorder with core symptom being speech dyspraxia – All and only affected KE family members have aberrant form of FOXP2 – Not a common cause of spoken language impairment: 10+ studies haven’t found it – [In Vernes et al., core endophenotype for CNTNAP2 was nonword repetition] • Other loci linked to spoken language impairments (13q21: 16q24; 19q13: 1p36, 2p15, 6p21, 15q21, 2p22; 1p22; 2q31 ; 15q13] Loci are not specific – Same loci linked to different “phenotypes” of SLI – Most loci are also linked to written language disorders – Most loci are also linked to other neuropsychological disorders (autism, Tourette’s, ADHD, schizophrenia, bipolar) Are these bi-linkages merely coincidental (the loci encompass 1000s of genes)? Or do they indicate that the same genotype may be expressed in different ways? What are the evolutionary implications of these bilinkages? K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 KE family: AD Oral-Motor Disorder The KE Family: Clearly AD, but disorder not specific to language • Core deficit: Difficulty with articution of speech •? Grammatical deficits • Low nonverbal IQ • Nonverbal learning disorders K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 7
Stromswold, Genetics & Evolution of Language Montreal, 2010 AS: Another Case of AD Speech Dyspraxia? Stuttered as a child Key: Circle = female, Square = male, Diamond = not known Black = impaired, White = normal, Hatched = learning disability • AS's paternal grandfather has a severe speech/language dysfluency, presumably from birth. Short sentences, frequent • AS’s paternal grandfather has 2 sisters with possible mild dysfluency and 1 brother with intact speech/language. • AS's father is an only child and has no history of speech/language or learning disorder. (very mild stutter?) • AS's mother has 1 brother and 1 half-brother. No history of language/speech or learning disorder on maternal side • AS has an older sister who has normal speech/language development, but may have a mild learning disability. • AS has 4 older half-siblings (same mother), all with no history of speech/language or learning disability. From Stromswold (2009) K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 Other Families with High Rates of Language Impairment (from Stromswold, 1998, Human Biology ) K. Stromswold, Summer Institute in Cognitive Science, Montreal 2010 8
Recommend
More recommend
