11
8/05/15 1 Think Big! The bright future of linguistics Red-crested Cardinal Paroaria coronata Biology today Tree of Life Encyclopedia of life Tree of Life

Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

  • Upload
    others

  • View
    6

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

1  

Think Big! The bright future of linguistics

Red-crested Cardinal Paroaria coronata

Biology today

Tree of Life

Encyclopedia of life

Tree of Life

Page 2: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

2  

Genbank taxonomy browser Genbank - cytochrome b sequence

Linguistics today Think big – scaling up…

1. Big data 2. Big methods 3. Big questions 4. Big teams

Max Planck Institute for the Science of Human History

GlottoBank

Page 3: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

3  

GlottoBank

GramBank LexiBank PhonoBank ParaBank

26

NEW DATABASES: QUALITATIVE AND QUANTITATIVE

GLOTTOBANK: world-scale databases, specifically for quantitative applications…

• GRAMBANK Harald Hammarström, Hedvig Skirgård

• LEXIBANK Simon Greenhill

• PHONOBANK Mattis List

• IELEX and URALEX Michael Dunn

• Syncretism in paradigms Nick Evans

D-PLACE a global database of cultural variation linked to language trees and ecological

data

Following the Comrie model

The data deluge requires computational tools Think big – scaling up…

1. Data 2. Methods

Page 4: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

4  

What value can computational methods add?

•  Dating language divergences

•  Phylogeography

•  Functional dependencies

•  Networks

“linguists don’t do dates” April & Robert McMahon (2006)

Austronesian Basic Vocabulary Database http:/language.psy.auckland.ac.nz

1201 Languages, 210 Words, 237,921 entries

Cognacy

John Lynch Laurent SagartBob Blust Jeff Marck Malcolm Ross

Cognate coding

Language “father” cognacy binary Paiwan tjama 1 1 0 Itbayaten qamaq 1 1 0 Mangarrai ema 1 1 0 Motu tama-na 1 1 0 Fijian (Bau) tama-na 1 1 0 Tongan tama i 1 1 0 Rarotongan metua 2 0 1 Maori matua 2 0 1

Page 5: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

5  

Data

•  400 well-attested languages –  No creoles, obvious borrowing removed

•  Outgroup –  Old Chinese (controversial) –  Buyang (less controversial)

•  Binary Coding –  presence/absence of cognates –  34,440 cognate sets –  Covarion model

Bayesian Phylogenetic Inference

1.  Data 2.  Model (and priors) 3.  Tree search 4.  Dating (without a

strict clock)

Uncertainty in tree estimation

Gray et al (2009) Science

PHIL 1.0 CPA 1.0

PMP 1.0

1.0

POC 1.0

0.8 POL 1.0

CEMP 0.8 EMP 0.58

0.99

Western Malayo-Polynesian

Central Malayo-Polynesian

Western Malayo-Polynesian

Micronesian

Admiralties

North & Central Vanuatu

South Vanuatu

South East Solomonic

Temotu

Papuan Tip

SHWNG

Papuan Tip

North New Guinea

Central Malayo-Polynesian

Philippines

Lexicostatistical

Eastern PolynesianElliceanFutunic

Micronesian

North & Central VanuatuSouth Vanuatu

S.E. SolomonicSouth VanuatuTemotuAdmiralties

Meso-Melanesian

North New GuineaPapuan Tip

SHWNG

Central Malayo-Polynesian

Western Malayo-Polynesian

Philippines

Formosan

PCP

OcWOc

MP

CEMP

EMP

PCP

OcEMP

CEMP

MP

Meso-Melanesian

AdmiraltiesTemotu

S.E. Solomonic

Ethnologue Phylogenetic

Austronesian phylogram

Page 6: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

6  

Constraint Min Max

East Polynesian 1.15 1.8

Tuvalu/Tokelau 1.0 2.0

Micronesian 1.9 2.2

Proto-Reefs Santa Cruz 3.0 3.15

Proto-Oceanic 3.2 3.6

Proto-Javanese 1.1 1.3

Proto-Chamic 1.8 2.5

Malayic-Chamic 2.0 3.0

Old Javanese 0.7 1.2

Proto-Malagasy 1.1 1.3

Proto-Malayo-Polynesian 3.6 4.5

Favorlong 0.346 0.384

Siraya 0.346 0.384

Old Chinese 2.3 2.9

Calibrations Prediction 2. Age of Proto-Austronesian

Gray et al (2009) Science

3 predictions: Sequence, timing, pulses and pauses

Gray, Drummond & Greenhill. 2009. Science, 323, 479-483.

Bouckaert et al (2012) Science.

X X

X

Page 7: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

7  

AfghanWaziri

Persian_ListTadzik

BaluchiKurdishOld_PersianAvestan

Digor_OsseticIron_OsseticWakhi

AssameseOriyaBengali

Bihari

HindiLahndaUrdu

MarwariSindhi

GujaratiMarathi

Nepali_List

KashmiriRomaniSinghalese

Vedic_Sanskrit

Albanian_CAlbanian_K

Albanian_GAlbanian_Top

Ancient_Greek Greek_MLGreek_ModArmenian_ListArmenian_Mod

Classical_Armenian

Breton_ListBreton_SEBreton_STCornishWelsh_CWelsh_N

Irish_AScots_GaelicOld_Irish

CatalanPortuguese_STSpanishFrenchWalloonProvencal

FriulianRomanshLadin

Italian

Sardinian_CSardinian_NSardinian_L

Romanian_ListVlach

LatinOscanUmbrian

DanishSwedish_ListSwedish_UpSwedish_VL

Riksmal

FaroeseIcelandic_STOld_Norse

Dutch_ListFlemishFrisian

German_STLuxembourgishOld_High_German

English_STOld_English

Gothic

BulgarianMacedonianSerbocroatianOld_Church_SlavonicSlovenianByelorussianUkrainianPolishRussian

CzechCzech_ESlovakLusatian_LLusatian_U

LatvianLithuanian_STOld_Prussian

Tocharian_ATocharian_B

Hitt i teLuvian Lycian

Think big – scaling up…

1.  Data 2.  Methods 3.  Questions

What are the Hilbert problems in linguistics?

David Hilbert Martin Hilpert https://www.youtube.com/watch?v=X4OaN39sNAI&feature=youtu.be

Explain this!

http://www.worldmapper.org/display_languages.php?selected=583

Some suggestions…

1.  Why are there approximately 7000 languages? 2.  Why is language diversity distributed so

patchily? 3.  What drives the evolution of linguistic disparity? 4.  When did spoken language evolve? 5.  How far back can we push the time barrier for

detecting language relationships?

Lineage through time plots

David N. Reznick & Robert E. Ricklefs, Nature 457, 837-842(12 February 2009)

Page 8: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

8  

−6 −5 −4 −3 −2 −1 0

12

510

2050

100

200

500

time

lineages

Austronesian

−8 −6 −4 −2 0

12

510

2050

time

lineages

Indo-European

Austroasiatic Languages Through Time

12

510

2050

N

8000 6000 4000 2000 0

Paul Sidwell, ANU Simon Greenhill, ANU

Do rainfall and group size drive the diversity and distribution of Australian languages?

Think big – scaling up…

1.  Data 2.  Methods 3.  Questions 4.  Teams

Page 9: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

9  

Big interdisciplinary teams

Think Big! The bright future of linguistics

Vanuatu – the Galapagos of language evolution Why do “Remote Melanesians” not look like Polynesians?

Page 10: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

10  

Blust (2005, 2008)

1.  Phenotypic differences in “Remote Melanesia”

2.  Cultural similarities 3.  Language typology - serial verb

constructions 4.  Loss of decimal system - switch to

various quinary systems 5.  Large amount of sound change

Distribution of numeral systems

decimal

“quinary” other

Circumstantial evidence for extended contact Distribution of retention rates Variation in rates of sound change

Number Paiwan Cebuano Maori Nengone

1 ita usa tahi sa

2 dusa duhá rua rewè

3 tjelu tulo toru tini

4 sepatj upát whaa ece

5 rimáʔ lima rima sɛduŋ

Vanuatu PNG highlands Chimbu Valley, New Guinea Mek warrior, Irian Jaya Pentecost Island, Vanuatu Tanna, Vanuatu

Page 11: Think Big!€¦ · PHIL 1.0 CPA 1.0 PMP 1.0 1.0 POC 1.0 0.8 POL 1.0 CEMP 0.8 EMP 0.58 0.99 Western Malayo-Polynesian Central Malayo-Polynesian Western Malayo-Polynesian Micronesian

8/05/15  

11  

Vanuatu – the Galapagos of language evolution

Max Planck Institute for the Science of Human History

Leipzig linguistics library Jena prize for historical linguistics