Transcript Language

Lecture 5
Body and brain for language
1





Spell out your main question(s). Make sure
that it is relevant to language origins!
Define your main terms: “language”,
“gesture”, “cognition”, “adaptation”,
“culture” – at least provisionally
Describe previous answers – theories: briefly.
Evaluate against evidence.
Conclusions – and if necessary, revise
(discuss) definitions again!
2

“The uniform language capacities of all
human populations today prove that all
adaptations for language… must have been
in place” not LATER than 100,000 years
ago…”

But: when, how and why did they appear?
3
4

In bodily anatomy
In brain structure

Main method

 look for differences between “us and apes” that
seem to be adaptations for language - and other
specifically human forms of communication
 Try to trace their origins on the basis of paleoanthropological evidence (“stones and bones”)
5
1.
2.
3.
Speech production
Speech perception
Brain size and organization
1. Lateralization
2. “Modularity”
3. Extended mirror neuron system
(Arbib 2005; Zlatev 2008)
6

“The larynx works like a valve, opening and closing to
let air pass. When it is shut, food can pass into the
esophagus at no risk to the lungs. The best place for
such a seal is right at the top of the trachea so that no
food or drink accidentally goes even a little ways
down it, but humans have a second use for the valve.
We work it like a musical instrument shaping the
sounds made by passing air as we speak. The musical
valve works best if we pull it a bit down into the
trachea so that the air wave shaped by the larynx can
resonate before leaving the mouth.”
http://ebbolles.typepad.com/babels_dawn/2006/10/the_human_laryn.html
7
“At birth the human larynx is in the normal, animal
location, enabling babies to nurse without risk of
choking.
 The larynx typically begins to move lower at about
three months of age and reaches its final position by
age four. People familiar with children’s speech will
notice that the start of the relocation is also when
infants start to coo. The end is about the time the
children finally become clearly intelligible to wellmeaning strangers.
 The lowered larynx lets humans produce a much
wider variety of sounds, particularly vowel sounds,
than apes can generate.”

http://ebbolles.typepad.com/babels_dawn/2006/10/the_human_laryn.html
8

“a two step process” (:78)
 The laryngal skeleton relative to the hyoid bone: also
in chimpanzees
 The decent of the hyoid bone (possibly unique to H.
sapiens – but only among the primates! (Fitch 2002)

Actually: a three step process!
 Sexual selection, in part – a second descent in males
in puberty (as in some deer)
 Either exaptation or adaptation for speech (to
outweigh the risk of choking)
9

When?
 Neanderthal hyoid bone similar to sapiens =>
back to the common ancestor: H. heidelbergensis
or H. Ergaster? 0,8 or 1,7 MYA?
 If at least in part an adaptation for speech (i.e. not
only an exaptation):
“our ancestors had some form of spoken language
before they had a human vocal tract” (: 80)
(evolution does not plan ahead)
10

Pathways from neocortex to vocal cords in H. sapiens,
but not in other apes: “removal of selection for [only]
innate automatic vocalizations, leaving room for vocal
learning to develop” (:81, cf. Deacon on devolution)

Enlarged hypoglossal canal: present at least in “later
Homo”, but very possibly earlier - unclear conclusions

Motor neurons down the spine to the thorax (region
from the neck to the diaphragm), but absent in H.
ergaster (MacLarnon & Hewlitt 1999)
– most important concrete evidence
11

“Specific language impairment” (SLI), now being
often called “Familial language impairment” (FLI)
Defective gene FOXP2: “the language gene”
More recently: “severe impairment in the selection
and sequencing of fine orofacial movements which
are necessary for articulation” (:83)
 Two functional mutations in FOXP2 in the human line
(different from the chimpanzees) – but when?
 Broca’s area, basal ganglia and cerebellum are
involved in non-speech sequencing (action, gesture,
song…)

12
Descended hyoid bone (second step in the lowering of
the larynx)
 Enlarged hypoglossal canal (better control of the
tongue)
 FOXP2 gene identical to H. sapiens (articulation and
other sequences), (Krause et al. 2007)
… are all present in H. Neanderthalensis
=> “some type of speech must have been present in our
last common ancestor with the Neanderthals, 500,000
years ago or so, though fully human speech with all our
articulatory capacity need not be much older than
100,000 years” (:85)

13



Categorical perception for calls, and even
human phonemes is observed in apes and
monkeys: similar to children – not uniquely
human.
Speech perception involves neural patterns of
activation different from other sounds
(though not a “module”).
Some “fine tuning” for improved perception
of 2-4 kHz: typical for speech. Present in H.
heidelbergensis (0,4 MYA) => Neanderthals
14



Absolute and relative (brain/body ratio)
increase of brain size: first in H. ergaster, then
in H. neanderthalensis and H. sapiens)
Disproportionate increase in some areas:
Prefrontal lobes, auditory-parietal areas
(“Perisylvian cortex”), cerebellum (see p.92)
Neotony: rapid-growth rate is prolonged to
childhood (“premature born apes”), favouring
neocortex and (social) learning
15
The brain: high metabolic cost and dangerous child
birth – what is the evolutionary benefit?
By-product of growth and “weaker jaw muscles”??
“Cooling device”??
Surplus energy and smaller gastrointestinal tract?
“Environment driven” (fruit, navigation, tools)
“Socially driven” (“politics and coalition-building
are important for a primate’s success”, :97)
 Language/communication-driven: language-brain
co-evolution over 4 MY (Deacon 1997)





16





Is language “located” in the left hemisphere?
RHS is involved in prosody, word learning,
discourse processing, gesture, even grammar
+ plasticity (epigenetic development)
Still: evidence for anatomical (large
“pyramidal neurons”) and behavioral
(handedness, pointing, signing) asymmetries
But: such asymmetries are to some extent
also in found in chimpanzees
Endocast of H. ergaster: enlarged Broca’s area
17
No evidence whatsoever of “encapsulated” modules
(Fodor 1983), No “Big Modules”
 Low-level sensory processing in specific areas, but:
interconnected, interactive not genetically
determined!
 The classical “language area”, Broca’s area (Brodman
areas 44, 45) in the left premotor cortex: involved in
control of action, imitation, gesture… “convergence
of audio-visual-motor processing streams” (: 110):
multimodal coordination?
 The “language” or even “grammar” gene FOXP2… see
video

18

Observe the 6 minute excerpt from “What
Makes Us Human? Part 2” BBC

Do the difficulties of the members of the Kfamily (and the boy who helped find FOXP2)
seem to do specifically with “syntax”, or even a
module within syntax?

From the little said about FOXP2 – does it seem
a likely candidate for a “language gene”?
19

“The discovery of mirror neurons in the frontal lobes of
monkeys, and their potential relevance to human brain
evolution … is the single most important "unreported" (or at
least, unpublicised) story of the decade [i.e. the 1990s].

I predict that mirror neurons will do for psychology what
DNA did for biology: they will provide a unifying framework
and help explain a host of mental abilities that have hitherto
remained mysterious and inaccessible to experiments.”
(V. Ramachandran
HTTP://WWW.EDGE.ORG/3RD_CULTURE/RAMA/RAMA_P1.HTML
20

Canonical neurons:
active only during the
monkey’s own
movements/actions

Mirror neurons:
active both during
execution and
observation of similar
movements/actions
(Rizzolatti et al. 1996)
21
F5: area in premotor
cortex
AIP: anterior
intraparietal area
22

It is amazing how these cells have been proposed as
a solution to just about every mystery in the human
mind: from empathy to imitation, mind-reading,
language (evolution), autism, and even sexual
preferences!

It is not surprising that their role has been regarded
as much over-rated by some researchers in the field
(Preson & de Waal 2002; Donald 2005; Csirba 2007).
23
 Directly recorded only in monkey brains, with
only indirect evidence for human brains.
 Not sufficient for either ”simulation” (on any
level) nor representation (or signification):
X stands for Y for subject S
 Present in macaques, while monkeys can
neither imitate, gesture, nor use language…
24

From ”mirror neurons” to neural circuits,
involving multimodal perception-action
cycles...

Basic idea: gradual evolution of the monkey
mirror system for manual actions (similar to
Arbib 2005, 2008) – but with a few important
differences (more on this when discussing
”Stages”, Lecture 11)
25

“Mimetic skills or mimesis rests
on the ability to produce
conscious, self-initiated,
representational acts that are
intentional but not linguistic.”
(Donald 1991: 168)

Mime, gesture, imitation, skill,
mimetic imagination

A domain-general adaptation (in
Homo ergaster), possibly initially
for tool use, and then extended to
communication
26
Level
Acts and states
Language
… dividing (semi)compositionally into meaningful
sub-acts that systematically relate to other similar
acts (as in grammar)
Proto-language
… that are conventional-normative
Triadic mimesis
(Intentionally
communicative)
… intended to stand for some action, object or event
for an addressee (and for the addressee to recognize
this intention)
Dyadic mimesis
(Not intentionally
communicative)
… under conscious control and corresponding to –
either iconically or indexically – some action, object
or event, and at the same time differentiated from it
Proto-mimesis
… involving cross-modal mapping between
exteroception (normally dominated by vision) and
proprioception (normally dominated by kinesthetics)
27
 Mirror neurons are part of a frontal-parietal-
temporal system in the monkey brain
 Responding to an open-ended set of manual
actions
 Basis for emulation: anticipating the results
of others’ actions (without ”theory of mind”)
 Possibly: basis for (simple) empathy, and
contagion (mimicry), see video
28
The perisylvian cortex (and the prefrontal cortex)
have expanded most in the human brain compared
to apes (see Deacon 1997)
 Moneky F5 -> Human BA 44 (pars opercularis of the
inferior frontal gyrus)
 Moneky PF -> Human inferior pariatal lobule
(involved in tasks of imaginatory reenactment)
 Moneky ST -> Human Superior Temporal Sulcus
(involved in the analysis of biological motion)
-> And on the basis of behavioral and anatomical data:
in apes too!

29
IPL
Broca
STS
30


Segregation of BA 44 into a dorsal part (for action) and a
ventral part, source of an ”efferent copy” active only during
imitation (Iacoboni 2005)
Lateralization
 right IPL active when imagining the motion of others, left IPL when
imagining self-motion
 Superior temporal sulcus (STS): analysis of other’s motion (left HS)
and ”in relation to the self” (right HS)

Quite possibly: prior to the evolution of language: some
degree of handedness in chimps, and evidence for
lateralization in Homo ergaster
31
Moneky F5 mirror neruons do not repsond to
”intransitive” (non-object related) actions
 By supporting imitation, the extended ape
(common ancestor) MN system would support this,
at least to some extent (apes better than monekys
in imitation, less so than us)
 Further extending, and differentiating between
expression and content along with lateralization:
iconic gestures (pantomime): the first true signs?

32
McNeill (2005): the content (imagery) of gestures is
based on RH-activity, their ”orchestration” mainly
on LH (”Broca’s area”)
 Gestures for non-present actions and objects
would presuppose extended lateralization (at least
by Homo heidelbergensis)
 Combining iconic gesture and pointing:
”proto-predication”: (Point-X, Iconic-gesture-X)
(but not ”gestural language”)

33




Segregation of BA 45 (heteromodal) and BA 44
(primarily for speech)
An important function of vocalization: to
”disambiguate” and conventionalize iconic gestures
BA 4a and BA 6 in ventral pre-motor cortex (PMv) –
active during both production and perception of
meaningless syllables (Wilson et al 2004), support for
the ”motor theory of speech perception” (Leiberman
et al 1967)
Wernicke’s area (= superior and middle temporal):
homologue of monkey PF (Arbib 2005)? Not really...
34
”…Wernicke’s area as
combining
capabilities for
recognizing
protosign and
protospeech to
support a languageready brain that is
capable of learning
signed languages as
readily as spoken
languages” (Arbib
2005)
IPL
Broca
Wernicke
STS
35
An extension of control for bodily
mimesis to “vocomimesis” and
eventually phonology (Zlatev
2008b)
36
BA 44, 45 = “Broca”
BA 22, 39, 40 = “Wernicke”
Overlap extensively with
the “human mirror neuron
system” (Arbib 2005;
Iacoboni 2005; Decety &
Chaminande 2005): in tasks
of perception-action
matching, imitation,
imagination, pantomime…
BA 4, 6 = perceptionproduction of
“meaningless syllables”
(Wilson et al. 2004)



Exaptations for hierarchical structure for action and
imitation (involving not only the extended MNS, but
basal ganglia, pre-SMA and cerebellum)
Grammar proper: from protolanguage (over the last
100 000 years) on the basis cultural and linguistic, ”postbiological” evolution (Arbib 2005)
All form-classes (e.g. adjectives, prepositions, affixes)
can be potentially traced back to either object-words
(nouns) and actions-words (verbs) (Heine and Kuteva
2002, 2007)
37

A product of a longer period of brain-culture co-evolution
(6 000 000 – 200 000 YA)
+ a shorter period of cultural-linguistic evolution
(200 000 – present): are all languages equally complex?

”…no powerful syntactic mechanisms need have been encoded
in the brain of the first Homo sapiens. Rather it was the
extension of the imitation-enriched mirror system to support
intended communication that enabled human societies, across
many millennia of invention and cultural evolution, to
achieve human languages in the modern sense” (Arbib 2005:
123)
38

There may be other adaptations, less directly
related to language, but preparing the (long)
road towards it…
39