The Origins of Language Jordan Zlatev

1

The Origins of LanguageJordan Zlatev

Lecture 11Stages in the origin of language

2

The major puzzle in explaining the origins of language Evolution is based on natural (and sexual)

selection giving rise to adapations: (basically) continuous...

However, there is such a wide gap between animal (vocal) signaling and human language, that the evolution of language appears discontinuous!

=> The “continuity paradox” (Bickerton 1990)

Animal signals vs. language (after Zlatev et al. 2005)Feature Animal signals Language

(1) Degree of learning Mostly genetically determined

Mostly learned from experience

(2) Conscious control None, or highly limited High

(3) Contextuality Tied to a particular context (stimulus setting)

Flexible, relatively independent from specific context

(4) Interpretation (Relatively) fixed response Flexible, “negotiable”

(5) Communicative relations

Mostly dyadic:Subject-Recepient

Mostly triadic:Speaker-Addressee-Referent

(6) Systematicity None, or very limited High

?3

The need for “protolanguage”

If there indeed exists a more primitive variety of language alongside fully developed human language, then the task of accounting for the origins of language is made much easier. No longer do we have to hypothesize some gargantuan leap from speechlessness to full language, a leap so vast and abrupt that evolutionary theory would be hard to put to account for it.(Bickerton 1990: 128)

4

Animal signals vs. proto-languageFeature Animal communication Protolanguage

(1) Degree of learning Mostly genetically determined

Mostly learned from experience

(2) Conscious control None, or highly limited High

(3) Contextuality Tied to a particular context (stimulus setting)

Flexible, relatively independent from specific context

(4) Interpretation (Relatively) fixed response Flexible, “negotiable”

(5) Communicative relations

Mostly dyadic:Subject-Recepient

Mostly triadic:Speaker-Addressee-Referent

(6) Systematicity None, or very limited High

5

The need for “prerequisites”

There is no plausible evolutionary scenario according to which language could “evolve” from animal signals.

Even present-daytheorists of innate Universal Grammar require some “broad” pre-adaptations/exaptations or prerequisites.

Hauser, Chomsky & Fitch (2002)

6

7

A magical ”macro-mutation”

Animal signalsLanguage

?

8

The only (reasonable) alternative

Stages in semiotic evolution

Animal signalsLanguage

9

Stages in phylogeny and ontogeny When I speak of a “stage” in phylogeny, I do not have in

mind a all-or-none switch in the genotype that yields a discontinuous change in the phenotype, but rather the coalescence of a variety of changes that can be characterized as forming a global pattern that may emerge over the course of tens or even hundreds of millennia.

(Arbib 2005: 123)

By “semiotic stage” we here mean a (not necessarily stable) period in development characterized by the clear establishment of a novel semiotic capacity, which may “dominate” the communication of the child at this stage, but does not replace capacities from previous stages.

(Zlatev and Andrén 2009: 381)

10

(Semiotic) stage in evolution Each step in this sequence

corresponds to a functional communication system, if not as elaborate and rich as modern human language. And none of the gaps that need to be bridged when going from one step to the next looks anything like the huge chasm commonly pictured between syntax and non-syntax.

(Johansson 2005: 236)

11

Donald (1991, 2001)

Animal signals Language(and episodic cognition) (and symbolic

cognition)

(Bodily) mimesis

12

Mimesis ( Memetics)Mimetic skills or mimesis rests on the ability to produce conscious, self-initiated, representational acts that are intentional but not linguistic.

(Donald 1991: 168)

The important properties of individual mimetic acts include intentionality, generativity, communicativity, reference, autocuing, and the ability to model an unlimited number of objects. (Donald 1991 : 171)

An stage-based evolutionary theory

Donald’s (1991) model of evolution related to the notion of sign function (based on Sonesson 2007).

Problems in Donald’s theory

1.What is the precursor to mimesis?2. How did the ”sign function” (i.e.

”symbols”) emerge?3. How did grammar evolve?

?

15

How can we know?

Compare the closest we have to the “last common ancestor” – present day apes – and human beings (especially children) with respect to:1. Meaning (“symbols” and “pragmatics”)2. Expression

VocalizationGesture

3. Grammar

16

(1) “Symbols” (Heine and Kuteva 2007: 132)

The term symbol is used in a wide range of works … but it has been defined in a number of different ways:

(a) Signs that, unlike indexical and iconic signs, exhibit an arbitrary relation between a meaning and a form used for its expression;

(b) Objects whose reference is context-independent, including objects displaced in time and/or space (i.e. outside the here-and-now)

(c) A convention, or shared cultural understanding, whereby different symbol users interpret the symbol in the same way;

(d) Signs that are intentional, or at least “functionally referential”

(e) Signs that are connected to other signs of the same kind in a network of internal relations…”

Sign (use) or signification(rather than “symbol”, “displacement”, “construal” etc)

X stands for Y for subject(s) S, in a way that:

(a) The relation is asymmetrical (X Y, not X Y)

(b) X and Y are differentiated: X is qualitatively different from Y for S

(c) X and Y are connected: in perceiving or enacting X, S conceives of Y

18

So – can apes use signs? In the wild: Not that we know of.

Clearest example: bonobo branches as “path markers”

In captivity (spontaneously): Perhaps…Clearest examples: “imperative pointing” and Sherman and Austin’s “inventions”

“Enculturated” (with language): YesKanzi, Panbanisha, Koko, Chantek – but why so limited (functionally), and with little transmission?

19

So – what is “missing”?

A “language acquisition device”, UG?Hardly: the problem is not with “grammar” per se – but with spontaneous sign use

Pragmatics: “An infrastructure of shared

intentionality” (Tomasello 2008) “A human-specific urge to learn, or if

necessary, create … semiotic systems” (Axel Svahn)

20

The SEDSU project (Zlatev et al. 2006)

“Pre-sign” differences between human beings and great apes found above all in:

Imitative capacity: difficulty with do-as-I-do imitation

Intersubjectivity: lower rates of mutual gaze (Zlatev et al. forthcoming), contagion, food sharing (Call et al. forthcoming)

21

Children, 2 years

Chimps(captive)

Bonobos(captive)

Gorillas(captive)

Orangs(captive)

Picture comprehension

+ (+) ? - ?

Imitating actions

+ (+) ? ? ?

Food sharing + + + - -Yawning contagion

(+) + ? - -

Mutual gaze + (+) (+) - ?Recognizing being imitated

+ + + + +

Cross-species comparisons

22

Children, 2 years

Chimps(captive)

Bonobos(captive)

Gorillas(captive)

Orangs(captive)

Picture comprehension

+ (+) ? - ?

Imitating actions

+ (+) ? ? ?

Food sharing + + + - -Yawning contagion

(+) + ? - -

Mutual gaze + (+) (+) - ?Recognizing being imitated

+ + + + +

Cross-species comparisons

Tendency

23

(1) Meaning: conclusions

For spontaneous sign use to evolve, there were adaptations in Imitation learning Co-operative communication Intersubjectivity (beyond “cognitive

empathy”)ÞMotivational and cognitive adaptations

Selective pressures: tool manufacture, pedagogy, displaced communication (all matters of collaborative sociality)

24

(2) Gesture and vocalization Advantages of gestural protolanguage

More volitional than vocalization in apes Manual sequences – further selected by tools Greater transparency (iconicity) Proto-predications: point + mime Clearer brain homologues for modern “language

areas” Advantages of vocal protolanguage

“Simpler” route from ape vocalizations (?) The symbolic potential of ape vocalizations is

underexplored Modern languages are (in normal conditions) spoken

25

(2) Gesture and vocalization

Factors for “transition to speech”: Multimodal social interaction (hands: “showing”) More economical More “arbitrary” – more reliance on conventionality and normativity

For the adaptations for speech (control of the tongue, jaws and breathing, to be adaptive “there and then” (evolution does not plan ahead!) there needs to have been a communication system in place.

Was proto-language at first gestural, or gestural-vocal?

Proposal: Bodily mimesis (characterizing Homo erectus culture) was mostly gestural – and only after ca. 2 mil years of evolution were the adapations for “high speed phonology” complete: with Homo sapiens (200,000 YA)

26

Greenfield and Savage-Rumbough (1990)CHASE (lexigram) YOU (point at person)

(3) Grammar

Despite optimistic interpretations, there is no evidence for any grammatical structure in Kanzi’s productive sequences.

27

Whence grammar?

Exaptations (pre-adaptations): “where” (x) and “what” P pathways in the

visual system, giving rise to P(x): “proto-predication” (Hurford 2003)?

Social scripts: “structured generalized patterns of social behaviour” (Johansson 2005: 231), Aiello (1998)?

Tool-making scripts (Johansson 2005: 231)? Hierarchical structure in action sequences

(Greenfield 1991)?

28

Whence grammar? Adaptations?

A large protolanguage lexicon, places a burden on memory, and is less efficient for communication▪ “analytic strategy”: break down “holistic” utterances

into components (Wray 2001, Arbib 2003)?▪ “synthetic strategy”: combine signs for e.g.

actions/locations/properties and objects: predication Note: not even triadic mimesis is totally

“holistic”, but allows iconic gestures for actions and pointing gestures to be combined into “proto-predication”

29

Stages in the evolution of grammar

1. Utterance structure (“two word stage”)

2. Hierarchical structure, but no recursivity (e.g. no subordination)

3. Flexible structure: “different ways to express the ‘same’ meaning” (: 234)

4. Recursive structure: “flows naturally from the ability to handle nested predications” (: 237): “Are you sure that Mary thinks that Joe is impossible” (Harder 2004)

30

Stages in the evolution of grammar It is far from clear that there are biological

adaptations for stages 1-4 (“innate vs. learned”, lecture 9)

Hence, the evolution of grammar, at least from the “two word stage” can have proceeded through processes of historical, rather then biological evolution, during the last 90,000 years since the spread of our common ancestors out of (and throughout), Africa.

Grammaticalization theory (Heine & Kuteva 2002, 2007) can help explain, and chart, this process of cultural evolution.

31

(Cultural) evolution of grammar Languages spoken 10,000 years ago,

typologically not much different from present languages

Push back in historical time, using generalizations concerning processes of grammaticalization (e.g. ‘want’ > FUT) From lexical to grammatical, and “even more

grammatical” Basically unidirectional Processes of grammaticalization have been similar in

the past: process, not structure, “uniformitarianism”

32

Grammaticalization ‘body’ (NOUN) > PRON > PASS ‘back’ (NOUN) > ADV > ADP ‘here’ (AVD) > DEM > REL ‘say’ (VERB) > COMP > SUBORD

Typical functions/meanings emerging Question Location specification (“relative”, “absolute”) Possession Personal pronouns

33

“The Earliest Conceivable Language Structure” Thing-words + Process-words No morphology or “grams” Only word order for syntax Perhaps concepts for location,

possession – but not grammatically expressed

No personal pronouns, but names, roles

A very “minimal syntax” indeed!

34

Stages: from LCA to present LCA: vocal signals and “flexible”, but not human-like

gestures “Lucy”: Larger groups: vocal grooming, mutual gaze

(extended mother-child communication)? Homo erectus: pantomime and pointing,

“vocomimesis”? Homo hielderbegensis/neanderthalensis:

gestural-vocal proto-language Homo sapiens: vocal-gestural protolanguage (“Stage

X”) Modern human beings: complex, grammatical, and

multimodal language (speech, gesture) Technological human beings: writing, internet etc.

35

Next stage?