dynamic typology - cysouw.decysouw.de/home/presentations_files/cysouwdynamictypology.pdf ·...
TRANSCRIPT
Dynamic TypologyMichael Cysouw
LMU Munich
Reasons for Typological Similarity
• Non-historical reasons‣ Universals‣ Chance
• Historical reasons‣ Common descent‣ Borrowing, substrate, etc.
The Importance of History
• The actual languages as attested in this world are not necessarily reflecting the possible human languages
• History (very probably) is still a significant factor for the explanation of current typological distributions
Order of Object and Verb(Mathew Dryer 2005 from WALS)
LETTERdoi:10.1038/nature09923
Evolved structure of language shows lineage-specifictrends in word-order universalsMichael Dunn1,2, Simon J. Greenhill3,4, Stephen C. Levinson1,2 & Russell D. Gray3
Languages vary widely but not without limit. The central goal oflinguistics is to describe the diversity of human languages andexplain the constraints on that diversity. Generative linguists fol-lowing Chomsky have claimed that linguistic diversity must beconstrained by innate parameters that are set as a child learns alanguage1,2. In contrast, other linguists following Greenberg haveclaimed that there are statistical tendencies for co-occurrence oftraits reflecting universal systems biases3–5, rather than absoluteconstraints or parametric variation. Here we use computationalphylogenetic methods to address the nature of constraints onlinguistic diversity in an evolutionary framework6. First, contraryto the generative account of parameter setting, we show that theevolution of only a few word-order features of languages arestrongly correlated. Second, contrary to the Greenbergian general-izations, we show that most observed functional dependenciesbetween traits are lineage-specific rather than universal tendencies.These findings support the view that—at least with respect to wordorder—cultural evolution is the primary factor that determineslinguistic structure, with the current state of a linguistic systemshaping and constraining future states.
Human language is unique amongst animal communication sys-tems not only for its structural complexity but also for its diversity atevery level of structure and meaning. There are about 7,000 extantlanguages, some with just a dozen contrastive sounds, others with morethan 100, some with complex patterns of word formation, others withsimple words only, some with the verb at the beginning of the sentence,some in the middle, and some at the end. Understanding this diversityand the systematic constraints on it is the central goal of linguistics. Thegenerative approach to linguistic variation has held that linguisticdiversity can be explained by changes in parameter settings. Each ofthese parameters controls a number of specific linguistic traits. Forexample, the setting ‘heads first’ will cause a language both to placeverbs before objects (‘kick the ball’), and prepositions before nouns(‘into the goal’)1,7. According to this account, language change occurswhen child learners simplify or regularize by choosing parameter set-tings other than those of the parental generation. Across a few genera-tions such changes might work through a population, effectinglanguage change across all the associated traits. Language changeshould therefore be relatively fast, and the traits set by one parametermust co-vary8.
In contrast, the statistical approach adopted by Greenbergian linguistssamples languages to find empirically co-occurring traits. These co-occurring traits are expected to be statistical tendencies attributable touniversal cognitive or systems biases. Among the most robust of thesetendencies are the so-called ‘‘word-order universals’’3 linking the orderof elements in a clause. Dryer has tested these generalizations on aworldwide sample of 625 languages and finds evidence for some of theseexpected linkages between word orders9. According to Dryer’s reformu-lation of the word-order universals, dominant verb–object orderingcorrelates with prepositions, as well as relative clauses and genitives
after the noun, whereas dominant object–verb ordering predicts post-positions, relative clauses and genitives before the noun4. One generalexplanation for these observations is that languages tend to be consist-ent (‘harmonic’) in their order of the most important element or ‘head’of a phrase relative to its ‘complement’ or ‘modifier’3, and so if the verbis first before its object, the adposition (here preposition) precedes thenoun, while if the verb is last after its object, the adposition follows thenoun (a ‘postposition’). Other functionally motivated explanationsemphasize consistent direction of branching within the syntactic struc-ture of a sentence9 or information structure and processing efficiency5.
To demonstrate that these correlations reflect underlying cognitiveor systems biases, the languages must be sampled in a way that controlsfor features linked only by direct inheritance from a commonancestor10. However, efforts to obtain a statistically independent sampleof languages confront several practical problems. First, our knowledgeof language relationships is incomplete: specialists disagree about high-level groupings of languages and many languages are only tentativelyassigned to language families. Second, a few large language familiescontain the bulk of global linguistic variation, making sampling purelyfrom unrelated languages impractical. Some balance of related, unre-lated and areally distributed languages has usually been aimed for inpractice11,12.
The approach we adopt here controls for shared inheritance byexamining correlation in the evolution of traits within well-establishedfamily trees13. Drawing on the powerful methods developed in evolu-tionary biology, we can then track correlated changes during the his-torical processes of language evolution as languages split and diversify.Large language families, a problem for the sampling method describedabove, now become an essential resource, because they permit theidentification of coupling between character state changes over long timeperiods. We selected four large language families for which quantitativephylogenies are available: Austronesian (with about 1,268 languages14
and a time depth of about 5,200 years15), Indo-European (about 449languages14, time depth of about 8,700 years16), Bantu (about 668 or522 for Narrow Bantu17, time depth about 4,000 years18) and Uto-Aztecan (about 61 languages19, time-depth about 5,000 years20).Between them these language families encompass well over a third ofthe world’s approximately 7,000 languages. We focused our analyses onthe ‘word-order universals’ because these are the most frequently citedexemplary candidates for strongly correlated linguistic features, withplausible motivations for interdependencies rooted in prominent formaland functional theories of grammar.
To test the extent of functional dependencies between word-ordervariables, we used a Bayesian phylogenetic method implemented in thesoftware BayesTraits21. For eight word-order features we comparedcorrelated and uncorrelated evolutionary models. Thus, for each pairof features, we calculated the likelihood that the observed states of thecharacters were the result of the two features evolving independently,and compared this to the likelihood that the observed states were theresult of coupled evolutionary change. This likelihood calculation was
1Max Planck Institute for Psycholinguistics, Post Office Box 310, 6500 AH Nijmegen, The Netherlands. 2Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Kapittelweg 29,6525 EN Nijmegen, The Netherlands. 3Department of Psychology, University of Auckland, Auckland 1142, New Zealand. 4Computational Evolution Group, University of Auckland, Auckland 1142, NewZealand.
5 M A Y 2 0 1 1 | V O L 4 7 3 | N A T U R E | 7 9
Macmillan Publishers Limited. All rights reserved©2011
Dynamic Typology
• It is not the actual frequencies that matter
• It is the stable distribution that matters
• A stable distribution is a situation in which just as many languages change from A to B as change from B to A.
• The extent to which the actual is different from the stable situation signals an effect of history
Type A Type B
50 200
probability of change: 20%
probability of change: 5%
1010
Stable distribution
Type A Type B
50 200
probability of change: 20%
probability of change: 5%
1010
Instable distribution
probability of change: 10%5
Type A Type B
83 167
probability of change: 10%
probability of change: 5%
8.38.3
Expected stable distribution
5/10+5×250= =10/10+5×250
Estimating Transition Probabilities
• Are transitions probabilities measurable ?‣ there is no assumption here why a change happens‣ might be internal or external (borrowing, substrate)‣ the question is whether specific features of language
are more amenable to change than others
• If yes: ‣ use group internal variation to estimate
transition probabilities
The world’s genera
A “sample” of the world’s languages
40% blue60% red
A real sample stratified for genealogical
grouping
How to get probabilities of change ...
Change from blue to red:
p = 2/12 = 17%
Change from red to blue:
p = 2/18 = 11%
blue: 17/11+17 = 61% (actual 40 %)
red: 11/11+17 = 39 % (actual 60 %)
Expected Stable Distribution:
Problem:
Problem:
‣ Originally red with one change to blue ?
‣ Originally blue with two changes to red ?
probability of any change happening
= α· frequency (blue) + β
For groups of three languages:
α = 3· (pblue→red − pred→blue)
β = 3· pred→blue · (1 − pblue→red)
Elena Maslova’s proposal
probability of contact triples = α· frequency (blue) + β
For groups of three languages A, B, C:– A+B related– B+C geographically close, not related– A ‘contact triple’ is a set in which A≠B=C
α = (pblue→red − pred→blue)
β = pred→blue · (1 − pred→blue)
Elena Maslova’s proposal for contact
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
WALS frequencies
Stab
le e
stim
ates
LETTERdoi:10.1038/nature09923
Evolved structure of language shows lineage-specifictrends in word-order universalsMichael Dunn1,2, Simon J. Greenhill3,4, Stephen C. Levinson1,2 & Russell D. Gray3
Languages vary widely but not without limit. The central goal oflinguistics is to describe the diversity of human languages andexplain the constraints on that diversity. Generative linguists fol-lowing Chomsky have claimed that linguistic diversity must beconstrained by innate parameters that are set as a child learns alanguage1,2. In contrast, other linguists following Greenberg haveclaimed that there are statistical tendencies for co-occurrence oftraits reflecting universal systems biases3–5, rather than absoluteconstraints or parametric variation. Here we use computationalphylogenetic methods to address the nature of constraints onlinguistic diversity in an evolutionary framework6. First, contraryto the generative account of parameter setting, we show that theevolution of only a few word-order features of languages arestrongly correlated. Second, contrary to the Greenbergian general-izations, we show that most observed functional dependenciesbetween traits are lineage-specific rather than universal tendencies.These findings support the view that—at least with respect to wordorder—cultural evolution is the primary factor that determineslinguistic structure, with the current state of a linguistic systemshaping and constraining future states.
Human language is unique amongst animal communication sys-tems not only for its structural complexity but also for its diversity atevery level of structure and meaning. There are about 7,000 extantlanguages, some with just a dozen contrastive sounds, others with morethan 100, some with complex patterns of word formation, others withsimple words only, some with the verb at the beginning of the sentence,some in the middle, and some at the end. Understanding this diversityand the systematic constraints on it is the central goal of linguistics. Thegenerative approach to linguistic variation has held that linguisticdiversity can be explained by changes in parameter settings. Each ofthese parameters controls a number of specific linguistic traits. Forexample, the setting ‘heads first’ will cause a language both to placeverbs before objects (‘kick the ball’), and prepositions before nouns(‘into the goal’)1,7. According to this account, language change occurswhen child learners simplify or regularize by choosing parameter set-tings other than those of the parental generation. Across a few genera-tions such changes might work through a population, effectinglanguage change across all the associated traits. Language changeshould therefore be relatively fast, and the traits set by one parametermust co-vary8.
In contrast, the statistical approach adopted by Greenbergian linguistssamples languages to find empirically co-occurring traits. These co-occurring traits are expected to be statistical tendencies attributable touniversal cognitive or systems biases. Among the most robust of thesetendencies are the so-called ‘‘word-order universals’’3 linking the orderof elements in a clause. Dryer has tested these generalizations on aworldwide sample of 625 languages and finds evidence for some of theseexpected linkages between word orders9. According to Dryer’s reformu-lation of the word-order universals, dominant verb–object orderingcorrelates with prepositions, as well as relative clauses and genitives
after the noun, whereas dominant object–verb ordering predicts post-positions, relative clauses and genitives before the noun4. One generalexplanation for these observations is that languages tend to be consist-ent (‘harmonic’) in their order of the most important element or ‘head’of a phrase relative to its ‘complement’ or ‘modifier’3, and so if the verbis first before its object, the adposition (here preposition) precedes thenoun, while if the verb is last after its object, the adposition follows thenoun (a ‘postposition’). Other functionally motivated explanationsemphasize consistent direction of branching within the syntactic struc-ture of a sentence9 or information structure and processing efficiency5.
To demonstrate that these correlations reflect underlying cognitiveor systems biases, the languages must be sampled in a way that controlsfor features linked only by direct inheritance from a commonancestor10. However, efforts to obtain a statistically independent sampleof languages confront several practical problems. First, our knowledgeof language relationships is incomplete: specialists disagree about high-level groupings of languages and many languages are only tentativelyassigned to language families. Second, a few large language familiescontain the bulk of global linguistic variation, making sampling purelyfrom unrelated languages impractical. Some balance of related, unre-lated and areally distributed languages has usually been aimed for inpractice11,12.
The approach we adopt here controls for shared inheritance byexamining correlation in the evolution of traits within well-establishedfamily trees13. Drawing on the powerful methods developed in evolu-tionary biology, we can then track correlated changes during the his-torical processes of language evolution as languages split and diversify.Large language families, a problem for the sampling method describedabove, now become an essential resource, because they permit theidentification of coupling between character state changes over long timeperiods. We selected four large language families for which quantitativephylogenies are available: Austronesian (with about 1,268 languages14
and a time depth of about 5,200 years15), Indo-European (about 449languages14, time depth of about 8,700 years16), Bantu (about 668 or522 for Narrow Bantu17, time depth about 4,000 years18) and Uto-Aztecan (about 61 languages19, time-depth about 5,000 years20).Between them these language families encompass well over a third ofthe world’s approximately 7,000 languages. We focused our analyses onthe ‘word-order universals’ because these are the most frequently citedexemplary candidates for strongly correlated linguistic features, withplausible motivations for interdependencies rooted in prominent formaland functional theories of grammar.
To test the extent of functional dependencies between word-ordervariables, we used a Bayesian phylogenetic method implemented in thesoftware BayesTraits21. For eight word-order features we comparedcorrelated and uncorrelated evolutionary models. Thus, for each pairof features, we calculated the likelihood that the observed states of thecharacters were the result of the two features evolving independently,and compared this to the likelihood that the observed states were theresult of coupled evolutionary change. This likelihood calculation was
1Max Planck Institute for Psycholinguistics, Post Office Box 310, 6500 AH Nijmegen, The Netherlands. 2Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Kapittelweg 29,6525 EN Nijmegen, The Netherlands. 3Department of Psychology, University of Auckland, Auckland 1142, New Zealand. 4Computational Evolution Group, University of Auckland, Auckland 1142, NewZealand.
5 M A Y 2 0 1 1 | V O L 4 7 3 | N A T U R E | 7 9
Macmillan Publishers Limited. All rights reserved©2011
TsouRukai
Siraya
Sediq
CiuliAtayalSquliqAtayal
Sunda
Minangkabau
IndonesianMalayBahasaIndonesia
Iban
TobaBatakLampung
TimugonMurutMerinaMalagasy
Palauan
MunaKatobuTongkuno DWunaPopalia
Wolio
Paulohi
MurnatenAluneAlune
KasiraIrahutuLetineseBuruNamroleBay
LoniuLou
ManamKairiru
DobuanBwaidogaSaliba
KilivilaGapapaiwa
MekeoMotu
AriaMoukSengseng
KoveMaleu
Lakalai
NakanaiBileki DMaututu
NggelaKwaio
Vitu
TigakTungagTungakLavongaiNalik
KuanuaSiar
Kokota
BabatanaSisingga
BanoniTeop
NehanHapeNissan
KiribatiKusaie
Woleai
WoleaianPuluwatesePuloAnnan
PonapeanMokilese
Rotuman
TokelauRapanuiEasterIsland
TahitianModern
HawaiianMaori
FutunaFutunaWest
FutunaEastNiueSamoanTongan
FijianBau
DehuIaai
SyeErromangan
LenakelAnejomAneityum
NgunaPaameseSouth
SakaoPortOlry
BumaTeanu
Giman
AmbaiYapenMor
NgadhaManggarai
Lamaholot
ESLewa D
ESKamberaSouthern DKambera
ESUmbuRatuNggai D
NiasChamorro
Bajo
BikolNagaCity
CebuanoMamanwa
TboliTagabiliWBukidnonManobo
AgtaIlokano
IfugaoBatadBontokGuinaang
Pangasinan
Kapampangan
PaiwanO
OO
O
OO
O
O
OO
O
OO
OO
O
OOO
O
O
OO
OOO
OO
OO
OOO
OO
OO
OOO
OO
O
OO
OO
O
OOO
OO
O
OO
OO
OO
OO
O
OOO
OO
O
OO
O
OO
OO
OOOO
O
OO
O
OO
OO
O
OO
O
OO
OO
O
O
OO
O
OO
O
O
OO
OO
OO
OO
O
O
O
Austronesian
Tocharian BTocharian A
Albanian G
Greek ModAncientGreek
Armenian Mod
FrenchProvencalWalloonLadin
Italian
Catalan
SpanishPortuguese ST
Sardinian CRumanian List
Latin
Old Norse
DanishSwedish List
Riksmal
FaroeseIcelandic ST
German STDutch List
FrisianFlemishAfrikaans
Pennsylvania DutchLuxembourgish
English STOld English
Gothic
Old Church Slavonic
SlovenianSerbocroatian
MacedonianBulgarian
Lusatian ULusatian LSlovak
CzechUkrainianByelorussian
RussianPolish
LatvianLithuanian ST
Nepali ListBihariBengali
AssameseOriya
GujaratiMarathi
MarwariSindhi
LahdnaPanjabi ST
UrduHindi
SinghaleseKashmiri
Kurdish
WaziriAfghan
TadzikPersian List
WakhiOssetic
Baluchi
Scots GaelicIrish B
Welsh N
Breton STCornish
Hittite
OO
O
OO
O
OOOO
O
O
OO
OO
O
O
OO
O
OO
OO
OOO
OOO
O
O
O
OO
OO
OOOO
OO
OO
OO
OOO
OO
OO
OO
OO
OO
OO
O
OO
OO
OO
O
OO
O
OO
O
Indo-European
PipilAztecTetelcingo
Cora
OpataEudeve
GuarijioTarahumara
Yaqui
NorthernTepehuan
PapagoPimaSoutheasternTepehuan
CahuillaLuiseno
MonoNorthernPaiute
Comanche
SouthernPaiuteSouthernUte
Chemehuevi
Hopi
OO
O
OOO
OO
O
OO
OO
OO
O
OO
O
OUto-Aztecan
Makwa P311
Cewa N31bNyanja N31a
Shona S10Ngoni S45
Xhosa S41Zulu S42
Hadimu G43cKaguru G12
Gikuyu E51Haya J22H
Hima J13Rwanda J611
Luba L33Ndembu K221
Lwena K141Ndonga R22
Bira D322Doko C40D
Mongo C613MongoNkundo C611
Tiv 802
O
OO
OO
OO
OO
OOO
OO
OO
OO
OOO
OBantu
Figure 1 | Two word-order features plotted onto maximum clade credibilitytrees of the four language families. Squares represent order of adposition andnoun; circles represent order of verb and object. The tree sample underlyingthis tree is generated from lexical data16,22. Blue-blue indicates postposition,
object–verb. Red-red indicates preposition, verb–object. Red-blue indicatespreposition, object–verb. Blue-red indicates postposition, verb–object. Blackindicates polymorphic states.
RESEARCH LETTER
8 0 | N A T U R E | V O L 4 7 3 | 5 M A Y 2 0 1 1
Macmillan Publishers Limited. All rights reserved©2011
LETTERdoi:10.1038/nature09923
Evolved structure of language shows lineage-specifictrends in word-order universalsMichael Dunn1,2, Simon J. Greenhill3,4, Stephen C. Levinson1,2 & Russell D. Gray3
Languages vary widely but not without limit. The central goal oflinguistics is to describe the diversity of human languages andexplain the constraints on that diversity. Generative linguists fol-lowing Chomsky have claimed that linguistic diversity must beconstrained by innate parameters that are set as a child learns alanguage1,2. In contrast, other linguists following Greenberg haveclaimed that there are statistical tendencies for co-occurrence oftraits reflecting universal systems biases3–5, rather than absoluteconstraints or parametric variation. Here we use computationalphylogenetic methods to address the nature of constraints onlinguistic diversity in an evolutionary framework6. First, contraryto the generative account of parameter setting, we show that theevolution of only a few word-order features of languages arestrongly correlated. Second, contrary to the Greenbergian general-izations, we show that most observed functional dependenciesbetween traits are lineage-specific rather than universal tendencies.These findings support the view that—at least with respect to wordorder—cultural evolution is the primary factor that determineslinguistic structure, with the current state of a linguistic systemshaping and constraining future states.
Human language is unique amongst animal communication sys-tems not only for its structural complexity but also for its diversity atevery level of structure and meaning. There are about 7,000 extantlanguages, some with just a dozen contrastive sounds, others with morethan 100, some with complex patterns of word formation, others withsimple words only, some with the verb at the beginning of the sentence,some in the middle, and some at the end. Understanding this diversityand the systematic constraints on it is the central goal of linguistics. Thegenerative approach to linguistic variation has held that linguisticdiversity can be explained by changes in parameter settings. Each ofthese parameters controls a number of specific linguistic traits. Forexample, the setting ‘heads first’ will cause a language both to placeverbs before objects (‘kick the ball’), and prepositions before nouns(‘into the goal’)1,7. According to this account, language change occurswhen child learners simplify or regularize by choosing parameter set-tings other than those of the parental generation. Across a few genera-tions such changes might work through a population, effectinglanguage change across all the associated traits. Language changeshould therefore be relatively fast, and the traits set by one parametermust co-vary8.
In contrast, the statistical approach adopted by Greenbergian linguistssamples languages to find empirically co-occurring traits. These co-occurring traits are expected to be statistical tendencies attributable touniversal cognitive or systems biases. Among the most robust of thesetendencies are the so-called ‘‘word-order universals’’3 linking the orderof elements in a clause. Dryer has tested these generalizations on aworldwide sample of 625 languages and finds evidence for some of theseexpected linkages between word orders9. According to Dryer’s reformu-lation of the word-order universals, dominant verb–object orderingcorrelates with prepositions, as well as relative clauses and genitives
after the noun, whereas dominant object–verb ordering predicts post-positions, relative clauses and genitives before the noun4. One generalexplanation for these observations is that languages tend to be consist-ent (‘harmonic’) in their order of the most important element or ‘head’of a phrase relative to its ‘complement’ or ‘modifier’3, and so if the verbis first before its object, the adposition (here preposition) precedes thenoun, while if the verb is last after its object, the adposition follows thenoun (a ‘postposition’). Other functionally motivated explanationsemphasize consistent direction of branching within the syntactic struc-ture of a sentence9 or information structure and processing efficiency5.
To demonstrate that these correlations reflect underlying cognitiveor systems biases, the languages must be sampled in a way that controlsfor features linked only by direct inheritance from a commonancestor10. However, efforts to obtain a statistically independent sampleof languages confront several practical problems. First, our knowledgeof language relationships is incomplete: specialists disagree about high-level groupings of languages and many languages are only tentativelyassigned to language families. Second, a few large language familiescontain the bulk of global linguistic variation, making sampling purelyfrom unrelated languages impractical. Some balance of related, unre-lated and areally distributed languages has usually been aimed for inpractice11,12.
The approach we adopt here controls for shared inheritance byexamining correlation in the evolution of traits within well-establishedfamily trees13. Drawing on the powerful methods developed in evolu-tionary biology, we can then track correlated changes during the his-torical processes of language evolution as languages split and diversify.Large language families, a problem for the sampling method describedabove, now become an essential resource, because they permit theidentification of coupling between character state changes over long timeperiods. We selected four large language families for which quantitativephylogenies are available: Austronesian (with about 1,268 languages14
and a time depth of about 5,200 years15), Indo-European (about 449languages14, time depth of about 8,700 years16), Bantu (about 668 or522 for Narrow Bantu17, time depth about 4,000 years18) and Uto-Aztecan (about 61 languages19, time-depth about 5,000 years20).Between them these language families encompass well over a third ofthe world’s approximately 7,000 languages. We focused our analyses onthe ‘word-order universals’ because these are the most frequently citedexemplary candidates for strongly correlated linguistic features, withplausible motivations for interdependencies rooted in prominent formaland functional theories of grammar.
To test the extent of functional dependencies between word-ordervariables, we used a Bayesian phylogenetic method implemented in thesoftware BayesTraits21. For eight word-order features we comparedcorrelated and uncorrelated evolutionary models. Thus, for each pairof features, we calculated the likelihood that the observed states of thecharacters were the result of the two features evolving independently,and compared this to the likelihood that the observed states were theresult of coupled evolutionary change. This likelihood calculation was
1Max Planck Institute for Psycholinguistics, Post Office Box 310, 6500 AH Nijmegen, The Netherlands. 2Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Kapittelweg 29,6525 EN Nijmegen, The Netherlands. 3Department of Psychology, University of Auckland, Auckland 1142, New Zealand. 4Computational Evolution Group, University of Auckland, Auckland 1142, NewZealand.
5 M A Y 2 0 1 1 | V O L 4 7 3 | N A T U R E | 7 9
Macmillan Publishers Limited. All rights reserved©2011
S1.4. Correlation Analysis To identify the relationships between the word order characters we mapped the WALS typological data onto these phylogenies using a continuous-time Markov model of trait evolution under a dependent and independent model of trait evolution. The linguistic characters were adapted from Dryerʼs word order features, published in the WALS database (Dryer 2005a-h; Section S2). The phylogenetic correlation analysis was carried out using DISCRETE, a Bayesian method of model comparison for discrete parameters implemented in BayesTraits (Pagel and Meade 2006). DISCRETE/BayesTraits estimates the posterior probability distribution of transition rates between the character states across the set of phylogenetic trees (Section S1.1) under a continuous-time Markov model of trait evolution (Pagel 1994, Pagel and Meade 2006). We calculated the fit of an independent and dependent model of trait evolution on each pair of characters. In the independent model, the linguistic features evolve separately (Figure S1). For example, a transition from having postpositions to having prepositions is not contingent on the background state of the second character (word order) and vice versa. Under this model there are four rates of change between states: Postpositions ⇒ Prepositions, Prepositions ⇒ Postpositions, Object-Verb ⇒ Verb-Object and Verb-Object ⇒ Object-Verb.
Figure S1: An independent model of linguistic trait evolution. Changes in character A (Adposition type, top) are not linked to changes in character B (word order, bottom). The magnitude of the estimated transition parameters is indicated here by arrow weight.
In contrast, the dependent model treats each pair of linguistic features as a single, four-state character, which can be represented by model with eight rates of change (Figure S2).
Figure S2: A dependent model of linguistic trait evolution. Changes in character A (Adposition type) are linked to changes in character B (word order). The magnitude of the estimated transition parameters is indicated here by arrow weight.
WWW.NATURE.COM/NATURE | 5
S1.4. Correlation Analysis To identify the relationships between the word order characters we mapped the WALS typological data onto these phylogenies using a continuous-time Markov model of trait evolution under a dependent and independent model of trait evolution. The linguistic characters were adapted from Dryerʼs word order features, published in the WALS database (Dryer 2005a-h; Section S2). The phylogenetic correlation analysis was carried out using DISCRETE, a Bayesian method of model comparison for discrete parameters implemented in BayesTraits (Pagel and Meade 2006). DISCRETE/BayesTraits estimates the posterior probability distribution of transition rates between the character states across the set of phylogenetic trees (Section S1.1) under a continuous-time Markov model of trait evolution (Pagel 1994, Pagel and Meade 2006). We calculated the fit of an independent and dependent model of trait evolution on each pair of characters. In the independent model, the linguistic features evolve separately (Figure S1). For example, a transition from having postpositions to having prepositions is not contingent on the background state of the second character (word order) and vice versa. Under this model there are four rates of change between states: Postpositions ⇒ Prepositions, Prepositions ⇒ Postpositions, Object-Verb ⇒ Verb-Object and Verb-Object ⇒ Object-Verb.
Figure S1: An independent model of linguistic trait evolution. Changes in character A (Adposition type, top) are not linked to changes in character B (word order, bottom). The magnitude of the estimated transition parameters is indicated here by arrow weight.
In contrast, the dependent model treats each pair of linguistic features as a single, four-state character, which can be represented by model with eight rates of change (Figure S2).
Figure S2: A dependent model of linguistic trait evolution. Changes in character A (Adposition type) are linked to changes in character B (word order). The magnitude of the estimated transition parameters is indicated here by arrow weight.
WWW.NATURE.COM/NATURE | 5
Independent Model
Dependent Model
-50
510
1520
Bay
es F
acto
r
NUM-SBV
NUM-REL
OBV-REL
NUM-OBV
ADP-REL
REL-SBV
DEM-SBV
DEM-NUM
GEN-REL
ADJ-OBV
DEM-REL
ADP-DEM
ADJ-SBV
DEM-GEN
ADP-NUM
DEM-OBV
ADJ-ADP
ADJ-REL
ADP-SBV
GEN-SBV
GEN-OBV
ADJ-DEM
GEN-NUM
ADJ-NUM
ADP-GEN
OBV-SBV
ADJ-GEN
ADP-OBV
Postpostion,object-verb
Postpostion,verb-object
Preposition,object-verb
Preposition,verb-object
Postpostion,object-verb
Postpostion,verb-object
Preposition,object-verb
Preposition,verb-object
Austronesian Indo-European
developed here could be used to explore the dependency relationshipsbetween a wide range of linguistic features. Nearly all branches oflinguistic theory have predicted such dependencies. Here we haveexamined the paradigm example (word-order universals) of theGreenbergian approach, taken also by the Chomskyan approach as‘‘descriptive generalizations that should be derived from principles ofUG [Universal Grammar]’’.1,27 What the current analyses unexpectedlyreveal is that systematic linkages of traits are likely to be the rare excep-tion rather than the rule. Linguistic diversity does not seem to be tightlyconstrained by universal cognitive factors specialized for language29.Instead, it is the product of cultural evolution, canalized by the systemsthat have evolved during diversification, so that future states lie in anevolutionary landscape with channels and basins of attraction that arespecific to linguistic lineages.
Received 10 June 2009; accepted 8 February 2011.Published online 13 April 2011.
1. Baker, M. The Atoms of Language (Basic Books, 2001).2. Chomsky, N. Lectures on Government and Binding (Foris, 1981).3. Greenberg, J. H. in Universals of Grammar (ed. Joseph H. Greenberg) 73–113 (MIT
Press, 1963).4. Dryer, M. in Language Typology and Syntactic Description Vol. I Clause Structure 2nd
edn (ed. Shopen, T.) 61–131 (Cambridge University Press, 2007).5. Hawkins, J. A. inLanguageUniversals (edsChristiansen,M.H.,Collins, C.&Edelman,
S.) 54–78 (Oxford University Press, 2009).6. Croft, W. Evolutionary linguistics. Annu. Rev. Anthropol. 37, 219–234 (2008).7. Chomsky, N. Knowledge of Language: its Nature, Origin and Use (Praeger, 1986).8. Lightfoot, D. W. The child’s trigger experience—degree-0 learnability. Behav. Brain
Sci. 12, 321–334 (1989).9. Dryer, M. S. The Greenbergian word order correlations. Language 68, 81–138
(1992).10. Mace, R. & Pagel, M. The comparative method in anthropology. Curr. Anthropol. 35,
549–564 (1994).11. Bakker, P. in Oxford Handbook of Linguistic Typology (ed. Song, J. J.) (Oxford
University Press, 2010).12. Cysouw, M. in Quantative Linguistics: An International Handbook (eds Altmann, G.,
Kohler, R. & Piotrowski, R.) 554–578 (Mouton de Gruyter, 2005).13. Pagel, M., Meade, A. & Barker, D. Bayesian estimation of ancestral character states
on phylogenies. Syst. Biol. 53, 673–684 (2004).14. Gordon, R. G. J. Ethnologue: Languages of the World 15th edn (SIL International,
2005).15. Gray, R. D., Drummond, A. J. & Greenhill, S. J. Language phylogenies reveal
expansion pulses andpauses inPacific settlement. Science 323, 479–483 (2009).
16. Gray,R. D. & Atkinson, Q. D. Language-tree divergence times support the Anatoliantheory of Indo-European origin. Nature 426, 435–439 (2003).
17. Guthrie, M. Comparative Bantu Vol. 2 (Gregg International, 1971).18. Diamond, J. & Bellwood, P. Farmers and their languages: the first expansions.
Science 300, 597–603 (2003).19. Campbell, L.American IndianLanguages:TheHistorical LinguisticsofNativeAmerica
133–138 (Oxford University Press, 1997).20. Kemp, B. M. et al. Evaluating the farming/language dispersal hypothesis with
genetic variation exhibited by populations in the Southwest and Mesoamerica.Proc. Natl Acad. Sci. USA 107, 6759–6764 (2010).
21. Pagel, M. & Meade, A. Bayesian analysis of correlated evolution of discretecharacters by reversible-jump Markov chain Monte Carlo. Am. Nat. 167, 808–825(2006).
22. Dyen, I., Kruskal, J. B. & Black, P. An Indo-European classification, a lexicostatisticalexperiment. Trans. Am. Phil. Soc. 82, 1–132 (1992).
23. Greenhill, S. J., Blust, R. & Gray, R. D. The Austronesian basic vocabulary database:from bioinformatics to lexomics. Evol. Bioinform. 4, 271–283 (2008).
24. Holden, C. J. Bantu language trees reflect the spread of farming across sub-Saharan Africa: a maximum-parsimony analysis. Proc. R. Soc. Lond. B 269,793–799 (2002).
25. Haspelmath, M., Dryer, M. S., Gil, D. & Comrie, B. The World Atlas of LanguageStructures Online (Max Planck Digital Library, 2008).
26. Raftery, A. Approximate Bayes factors and accounting for model uncertainty ingeneralised linear models. Biometrika 83, 251–266 (1996).
27. Cela-Conde, C. & Marty, G. Noam Chomsky’s minimalist program and thephilosophy of mind. An interview. Syntax 1, 19–36 (1998).
28. Reesink, G., Singer, R. & Dunn, M. Explaining the linguistic diversity of Sahul usingpopulation models. PLoS Biol. 7, e1000241 (2009).
29. Evans,N.&Levinson,S.C. Themythof language universals: languagediversity andits importance for cognitive science. Behav. Brain Sci. 32, 429–492 (2009).
Supplementary Information is linked to the online version of the paper atwww.nature.com/nature.
Acknowledgements We thank M. Liberman for comments on our initial results andF. Jordan and G. Reesink for comments on drafts of this paper. L. Campbell, J. Hill,W. Miller and R. Ross provided and coded the Uto-Aztecan lexical data.
Author Contributions R.D.G. and M.D. conceived and designed the study. S.J.G., R.D.G.and M.D.provided lexicaldataandphylogenetic trees.M.D. codedword-order data, andconducted the phylogenetic comparative analyses with S.J.G. All authors were involvedin discussion and interpretation of the results. All authors contributed to the writingwith S.C.L. and M.D. having leading roles; M.D., R.D.G. and S.J.G. produced theSupplementary Information.
Author Information Reprints and permissions information is available atwww.nature.com/reprints. The authors declare no competing financial interests.Readers are welcome to comment on the online version of this article atwww.nature.com/nature. Correspondence and requests for materials should beaddressed to M.D. ([email protected]).
Postposition,verb–object
Preposition,object–verb
Preposition,verb–object
Postposition,object–verb
Postposition,verb–object
Preposition,object–verb
Preposition,verb–object
Postposition,object–verb
Austronesian Indo-European
Figure 3 | The transition probabilities between states leading to object–verband adposition–noun alignments in Austronesian and Indo-European.Data were taken from the model most frequently selected in the analyses;probability is indicated by line weight. The state pairs across the midline of each
figure (postposition, object–verb; and preposition, verb–object) areGreenberg’s ‘harmonic’ or stable word orders. Nevertheless, each languagefamily shows tendencies for specific directions and probabilities of statetransitions.
RESEARCH LETTER
8 2 | N A T U R E | V O L 4 7 3 | 5 M A Y 2 0 1 1
Macmillan Publishers Limited. All rights reserved©2011
Austronesian Indo-European
0.0
0.2
0.4
0.6
0.8
1.0
Postpositions, Object-Verb
Pro
porti
on in
sta
ble
stat
e
Austronesian Indo-European
0.0
0.2
0.4
0.6
0.8
1.0
Prepositions, Object-Verb
Pro
porti
on in
sta
ble
stat
e
Austronesian Indo-European
0.0
0.2
0.4
0.6
0.8
1.0
Prepostions, Verb-Object
Pro
porti
on in
sta
ble
stat
e
Austronesian Indo-European
0.0
0.2
0.4
0.6
0.8
1.0
Postpositions, Verb-Object
Pro
porti
on in
sta
ble
stat
e
Looking forward
• Much more testing necessary‣ do different approaches converge on the same
estimates for transition probabilities ?
• Only very few types allowed‣ computations quickly get too large for the
amount of available data on the world’s languages