This excerpt from
I of the Vortex.
Rodolfo R. Llinás.
© 2001 The MIT Press.
is provided in screen-viewable form for personal use only by members
of MIT CogNet.
Unauthorized use or dissemination of this information is expressly
forbidden.
If you have any questions about this material, please contact
cognetadmin@cognet.mit.edu.
Kiki Smith, Tongue and Hand, 1985. Painted plaster, 5-1/2 3-1/2 3 (14
8.9 7.6 cm) tongue unit; 5-1/2 3 3-1/2 (14 7.7 8.9 cm) hand unit.
Photograph by Ellen Page Wilson. Courtesy of PaceWildenstein.
11 Language as the Child of
Abstract Thought
The Beginnings of Abstraction
Let us begin by agreeing on a consensus deªnition of abstraction and/or
abstract thinking. An abstraction generally refers to something that only
exists in the mind: an idea, a conception, a mental representation of
something that may (or may not) exist in the outside world. Abstraction,
or the collection of neural processes that generate abstraction, is a funda-
mental principle of nervous system function. The nature of these pro-
cesses owes its origin to the phylogenetic wiring patterns acquired by the
nervous system on its evolutionary course. As such, it is more than likely
that abstract thinking probably began long ago in very primitive nervous
systems. This view emerges from considering the nervous system as being
geared toward predictive movement. In order to place movement within
the context of the whole animal, the animal must ªrst be capable of gen-
erating some type of internal image or description of itself as a whole,
and this image must support the strategy around which to organize the
tactics of what the animal will do.
At ªrst glance, the generation of a voluntary internal sensory-motor
transformation (see chapter 7) does not correspond directly to the more
obvious neuronal connectivity required to transform, for example, a
226 Chapter 11
stubbing of the toe into a leg ºexion. The newer type of wiring represents
more than the segmental reºex. It is metasegmental, representing a global
function such as the coordinated walking of an elongated multisegmental
animal rather than just the stepping supported by a given local segment.
By elongated animal, I refer to any encephalized creature with a head
end and a tail or foot end with a column or chain of nervous tissue tra-
versing (and subserving) the length of the body. This description encom-
passes a broad range from the more lowly creatures with very primitive
notochords to those with quite sophisticated spinal cords. That the nerv-
ous system was selected in evolution to be organized in a segmented fash-
ion was probably driven by the neurobiological practicality of optimizing
body surface area to volume, in order to minimize the distance a nerve
signal must travel either to or from the external world. Elongated ani-
mals are basically made of a horizontal stack of coins, where the neu-
ral wherewithal subserving each coin is organized to know about its
respective segment and relatively little else. In order to make a complete
working animal out of these segments, there must be a portion of the
nervous system that is not exclusively segmental in its organization. This
portion of the nervous system can put the many segments together into
something that beforehand did not exist: a uniªed whole. As stated
above, we may consider this the beginning of abstract function as this
portion of the nervous system does not relate directly to the connectivity
of the nervous system at any particular, segmental level. The central nerv-
ous system abstracts the fact that the animal is composed of a series of
unit segments; ipso facto, the process of intersegmental integration is an
abstraction, and represents the beginning of abstraction as a naturally se-
lected biological process. That this is the evolutionary direction is sup-
ported by the observation that the central nervous system mushrooms
out in front of the spinal cord, polarizing encephalization. We see some-
thing important happening: from the animal s very neurological becom-
ing is the fact that the animal can have an internal representation of itself
not only as a set of parts but as a whole entity. It is here, from this germi-
nal metaevent, that abstraction begins and the self emerges.
How does this relate to prediction? Well, beyond the description of the
animal itself and the description of the input that comes to this animal,
the intrinsic circuits of the nervous system are capable of generating a
Language as the Child of Abstract Thought 227
premotor representation of what is going on outside. From this, self-
referentially in the presence of that motor image, the animal is capable of
deciding what to do. The animal is capable of prediction. Run, ªght, ªnd
food or whatever, functionally the animal is the circuit that represents its
sensory motor attributes, and this central event is an abstract entity.
Now of great importance is the event that happens between the stimu-
lus that evokes movement and the liberated motor FAP. The prompting
stimulus may be of external origin (a ferret is climbing up my pants!) or
internal (I left the stove on at home!). Either way, these stimuli, if granted
appropriate internal signiªcance (by and within the momentary status or
context of the cycling thalamocortical system), are ampliªed into an
emotional state. We have already seen that the nervous system is wired in
such a way that under normal conditions FAPs are only liberated into ac-
tion by the generated emotional states that precede them. Such internal
events, emotions, are then by deªnition premotor states.
We can take this further. Emotions or emotional states are events that
do not exist in the outside world; they are purely internal events and
would remain completely hidden to us (as observers of others) were it not
for motricity. Precisely which emotion may be occurring is inferred only
through the expression of the FAP that is liberated by that given emotion.
The dog is snarling and baring its teeth at me; what is going on inside of
this dog is most likely not that he is happy to see me. Just how it is that I
came to know or infer this is another important issue that we will discuss
shortly. The point here is that emotions, being themselves purely internal
events, are simply invented states on the part of the central nervous sys-
tem and as such are clearly abstractions. It is fair to say that just as emo-
tions are intrinsic products of central nervous system function, so are
abstractions.
Intentionality
Returning to prediction, it is clear that prediction must have a goal, oth-
erwise it is not referentially based on anything; purposeless movement is
not only wasteful, but can be quite dangerous as well. The goal or object
of movement must be well deªned, and we may deªne it here as that
which one intends to do in relation to that object or goal. Also an
228 Chapter 11
abstraction, intentionality is the premotor detail of the desired result of
movement through which a particular emotional state is expressed: the
choice of what to do before the doing of it.
Consider the following: if our brains are capable of planning move-
ment strategies that are implementable if desired, then they should also
be able to outwardly express intentionality as a motor representation of
what is happening inside our heads. I am suggesting that the outward ex-
pression of premotor activity precedes and predicts the activation of spe-
ciªc motor patterns. An example of such a process is shouting run!
when in danger before we actually begin to run. This raises a crucially
important point about language itself. I suggest that our ability to vocal-
ize the different aspects of intentionality developed ªrst as the ability to
separate the properties of things from the things themselves. This process
of abstraction would over time engender what we may consider to be a
mental catalog, much like an alphabet, that would allow us to generate
inside our heads events that would be reentered admixtures of the pri-
mary events that went into the generation of language to begin with. We
have arrived at our ªrst corollary: even before language was sufªciently
well structured to be communicable, its genesis must have had as a pre-
requisite foundation the nervous system s capacity to generate the
premotor imagery required to abstract the properties of things from the
things themselves. That is, it required the premotor imagery to make ab-
stractions of universals.
And so we see that there are two very important issues we must bear in
mind when considering the evolution of language. One, that abstract
thinking must have preceded language during evolution, and two, that
the premotor events leading to expression of language are in every way
the same as those premotor events that precede any movement that is ex-
ecuted for a purpose. From these two points, which are so similar as to be
almost the same, we may gather that language is simply an element
within a much larger, more general category of function.
Prosody: The First Cracklings of Language
Let us now dig into the plausible origins of language and how this indis-
pensable tool must have evolved. Just as in the case of the rather mean-
Language as the Child of Abstract Thought 229
dering evolution of the eye, language may be difªcult to trace cleanly
backwards through evolutionary time (see, for discussions and concepts,
MacNeilage 1994, 1998; Verhaegen 1995; Gordon 1996; Ujhelyi 1996;
Aboitiz and Garcia 1997 a, b; Honda and Kusakawa 1997; Ganger and
Stromswold 1998; Gannon et al. 1998; Kay et al. 1998; Doupe and Kuhl
1999; Nowak and Krakauer 1999). As in previous examples, evolution-
ary steps prior to the emergence of a specialized organ may not necessar-
ily function or appear in any fashion similar to what we see in the present
day organ. As we learned from the lifetime of lifetimes that brought us
the eye, we may perhaps expect an unexpected genealogical path in the
ancestry of language.
Now before proceeding further, we must gather a few clarifying
deªnitions. Just what do we mean by language ? The ªrst thing that
typically comes to mind is human language, its wide variety of types, that
it is often written as well as spoken, and that languages other than our
own are at once fascinating and opaque. If you feel that language is ex-
clusively a human capability or that we humans invented language, then I
must tell you that I wholeheartedly disagree, although you may not be
alone in this thinking. The reasons for which I disagree are straightfor-
ward: language clearly exists in many species far, far older in the evolu-
tionary sense than we Homo sapiens; furthermore it is too general a trait
throughout the animal kingdom ever to be seriously considered as exclu-
sive to humans. Although it is most likely true that we exhibit the richest
and most complex of languages, we are neither the origin nor sole pos-
sessor of language.
Let us deªne language as the given methodology by which one animal
may communicate with another. In this regard, language is a rather large,
generic category, for this deªnition says nothing about the intent to com-
municate, only that some level of communication is ultimately achieved.
We have so far been saying that language is a logical product of the in-
trinsic abstracting properties of the central nervous system, or simply of
abstract thought. But this I would say is a subcategory within language
that I would call biological prosody. Prosody is a more generalized
form of motor behavior, an outward gesturing of an internal state, an
outward expression of a centrally generated abstraction that means
something to another animal. For us, smiling, laughter, frowning, the
230 Chapter 11
lifting of one s eyebrows are all forms of prosody, for they convey one s
internal, momentary state in a way that is recognizable and understand-
able to someone else. Prosody is language, but it is not spoken language.
Nonetheless, it is purposeful communication. Prosody is by no means
conªned to humans; it is widespread throughout the animal kingdom,
and in the evolutionary sense is very old. Darwin, in his brilliant text on
facial expression, studied prosody in animals in terms of moods and faces
and how faces and postures represent the momentary internal states of an
animal. These are representations of the internal abstractions such as
emotions and intentions. So a prosodic event is an abstraction coupled
with a motor expression that conveys to another animal what its internal
state is like at that moment.
If prosody represents a subcategory within language, what would be
an example of language without prosody? Well, there are types of lan-
guage that are extraordinarily speciªc and although they carry very sim-
ple messages, they are nevertheless essential for the survival of the
species. The pheromone delivery and reception systems of a moth are
known to be effective for distances of several miles, and so they clearly
represent communication at a distance. The pheromone released by the
female is recognized speciªcally and exclusively by the male of the same
species and is communicatively effective enough for the couple to ªnd
each other in an otherwise crowded niche (Willis and Arbas 1991;
Hildebrand 1995; Roelofs 1995; Baker et al. 1998). But although this
language is critical for the species survival, it has nothing to do with the
outward expression of an internally generated abstraction. This is there-
fore not prosody, but rather simply a behavior-modifying event through
the release and reception of a speciªc molecule.
In most cases, however, language circumscribes a prosodic event. We
ªnd that language as such may be observed at many levels in evolution,
where it serves very different functions. One of the ªrst nonhuman lan-
guages to be understood sufªciently by people, one that communicates
simple orders, is the language of bees. This language is basically a dance,
a rhythm and orientation performed in space. These dances, each one
speciªc to bees of a particular species, give information about the quan-
tity and location of food with respect to the beehive. In this way all bees
of the colony can know about and help procure the food (von Frisch
1994; Gould 1976, 1990; Hammer and Menzel 1995; Menzel and Mul-
Language as the Child of Abstract Thought 231
ler 1996; Waddington et al. 1998). Languages such as these have also
been studied in vertebrates and other invertebrates. By deªnition and in
all cases, these forms of communication require social order so that the
information conveyed may be used to some purpose by the receiving
organism.
There are variations of language that convey information quite differ-
ent from just the issue of food for one s family. For instance, when most
animals are attacked, they generally posture in a manner that is clearly
recognizable as a defensive stance by the attacker or as a counter-punch
ready to occur. Such posturing may be as simple as a puffer ªsh increas-
ing its size to appear more formidable, or the very common showing of
teeth and growling we see in most vertebrates. Animals with horns, such
as the rhinoceros or buffalo, take a stance where the horns are directed at
the threatening or attacking animal. These are all languages, granted
with very limited repertoire, but this form of prosody must be at the very
foundation of all types of purposeful communication between and within
species.
Moving higher up the evolutionary scale, we can look at languages
that convey a higher level of organization. The language of wolves is an
excellent example where by means of prosody, wolf packs express rela-
tively complex, socially structured attack and defense behaviors. In this
case, the relations between the differing animals are not only of simple
prosody but also represent a (social) context within which this prosodic
language is exercised.
This form of prosody in wolves is quite sophisticated, utilizing a num-
ber of different motor avenues for overall expression, including vocaliza-
tion, eye contact, head gesturing, and whole body communication. For
instance, the establishment of dominance, or which wolf is to be the al-
pha male, is by means of communication not just of one s physical might,
but by those males subordinate to the alpha male showing their social po-
sition by expressing submission. They will roll over onto their backs and
offer the alpha male their neck. This form of language leads to an estab-
lishment of social hierarchy that is central to the strategies of the pack as
a whole. However, during hunting other factors arise, and so:
True leadership is not strongly apparent in a wolf pack, as any animal may initi-
ate its movement. However, the pack is highly cooperative in hunting and in the
care of the young. Dominance organization is not strongly apparent in wild wolf
232 Chapter 11
packs, but in the captive packs in zoos, where the principle occupation of hunting
is made unnecessary by artiªcial feeding, wolves spend a large amount of their
time threatening one another and enforcing dominance. (Dewsbury and
Rethlingshafer, 1973)
Notwithstanding this caution about the risks of using behavior ob-
served in captivity to make sweeping generalizations about behavior in
the wild, the social organization of wolves does have a hierarchy, so the
language or the prosodic events would be understood differently at its
different levels. The prosody that cues one adult male to know that it is
his turn within an attack sequence may to a young pup be simply the
learning of the global differences between attacking and defending. The
main point here is that the social hierarchy itself could not exist of there
were no elements of commonality existing at all levels. In this case we see
that language is developed in the context of a particular social order as a
means of binding animals into a single working entity for the beneªt of
all.
Similarly, there are very interesting hunting behaviors that one sees in
wild dogs of Africa. They often attack where there is tall grass, for that is
usually where the smaller animals of prey are found. These dogs have
white-tipped tails that they keep straight up in the air. They move their
tails back and forth in particular patterns that stimulate peripheral vi-
sion, so each dog, without having to look around that much, can see
where the other tails are, and thus receive a constantly updated status re-
port on the structure of the pack as it ºushes out and/or corners its prey.
It is fascinating to ponder here the reasonable conclusion that anatomy
has evolved to be related to the language of strategy.
These prosodic, ensemble properties of the canine species have made
possible the very particular relationship that dogs have with man. This
relation is especially clear in situations where dog and man form a work-
ing team, as in shepherding or hunting. Here the dog is simply expressing
properties (abstraction and prosody) that were genetically determined
and is imprinting the relations of the hierarchy it already knows save
for the fact that here, the hierarchy is with an animal other than another
dog.
Language as the Child of Abstract Thought 233
Mimicry: The Origin of Meaning Between Organisms
What seems to be emerging is that language must have evolved from a
prelinguistic type of attribute, mostly concerning prosody, punctuated by
either particular sounds or particular gestures. But we must not overlook
a very crucial element to language. If prosody is the outward expression
of a momentary internal state, what is its purpose unless it is understood
by another animal? Communication without consensus meaning simply
is not communication. So what we are really asking is: How did the
meaning part come to communication?
Here is something I would have to refer to as the infectious nature of
brain activity. Laughter is the perfect example it is infectious between
people. Someone begins laughing, you hear (and/or see) this, and soon
you cannot help but laugh. Put another way, laughter is generated and
when you receive it, you create the similar state in your head. It is as if the
abstraction itself is infectious an intrinsic property of neuronal circuits
that seems to get outside of itself. We may think that if laughter in infec-
tious, and so is yawning, then maybe showing your teeth and growling is
infectious as well. Let s examine this point further.
Consider the kookaburra, a bird of Australia. These birds hang out to-
gether, usually spotting the branches of two or three neighboring trees.
From the silence, one of the birds starts making the characteristic kooka-
burra sound, which happens to sound, to us, exactly like distorted hu-
man laughter. Another kookaburra joins in, mimicking the same sound
as the ªrst, and within seconds, the whole ºock is laughing. We see a
similar event with ªreºies: one male lights up, the others entrain. And so
the female, far off in the distance, gets a real ºash!
What does this mimicry tell us about language? One can imagine that
if the nervous system were to acquire by accident the ability to recognize,
through sensory input, the FAPs displayed by others, this would be a very
useful property if the animals are going to live in groups. In fact, this
property makes the group de facto birds of a feather! And so animals
with the ability to mimic each other would immediately tend to form a
family, because what has been engendered is indeed a sense of familiarity:
Hey! You re one of us!
234 Chapter 11
Figure 11.1
Ant ªshing learned by mimicry. (Photo by T. Nishida, from Whiten et al. 1999.)
Recognition of one s kin is as old as time. Fair enough, but this recog-
nition is fostered through mimicry, through a creature s repeating the
FAPs of others that it gleans through its senses (ªgure 11.1). But what can
one say of communicative meaning across species? I know that the dog
still snarling at me from a few pages back means trouble for me. It means
trouble to another dog as well. How does it know? How do I know?
Rather, how did we come to know? Let s wind evolution backward and
think it through.
An animal bares its teeth at another animal; the other animal does not
recognize this FAP. Nonetheless, this FAP evokes a show teeth back or
a run FAP. Why? Because all of those who didn t at once recognize the
baring of teeth FAP died! They became lunch, and thus over time they
were naturally selected out of the gene pool! But those that remained, in
the natural selection sense, evolved to have such FAP recognition become
an intrinsic a priori. Thus, out of phylogeny comes the ability to recog-
nize certain things as dangerous without necessarily memorizing them in
real time, in ontogenetic time. For instance, ªsh know from birth that
Language as the Child of Abstract Thought 235
very brightly colored (other) ªsh are dangerous. At birth, they just know
it. Those brightly colored ªsh over there are poisonous; they advertise it!
And this is as wired as is the fact that ªsh have tails!
The above is the evolutionary downstream endpoint, if you will, but
mimicry is at its origin. Of course we agree that language would make no
sense if the receiver did not understand what language means. How does
understanding occur? The easiest way for the receiver to understand
what language means is for the receiving animal to somehow associate its
own production of the motor event with the sensory reception of the mo-
tor event. That is, an outward motor expression of the internal sensor-
imotor image is matched with that which the receiving animal is actually,
sensorially, experiencing: monkey see, monkey do. It can be done
through learning, but it has to be more powerful than that. It would have
to be done by understanding in a sensory way the consequences of this
mimicked motor behavior. This is a bit complicated; there is a hurdle to
get over.
Let s say I am an alligator now and I bare my teeth. Simple enough, but
here s the problem: I most likely will never see myself baring my teeth.
However, I might hear my own growling and so when a growling occurs I
may recognize it as what I do when I am in a certain mood. If I see that
growling is often associated with teeth baring I will soon associate the
two. This is why when another alligator bares its teeth at me, I don t just
ignore it, or simply look the other way as if nothing happened. We are at
an evolutionary crux. Coming to know universals must have come from
not knowing universals. This can happen in only one of two ways. Either
the nervous system absolutely knows in advance what it is doing or it
doesn t, in which case it is natural selection that has determined my re-
sponse: when another alligator bares its teeth at me, I bare mine right
back. The answer must be, of course, the latter, natural selection. How
do I know? Think of how people act when they see themselves on TV for
the ªrst time, or in a ªlm. They are always amazed at how silly they ap-
pear to themselves. Something internal does not match with the external
presentation of the same event, so just the (sensorimotor image of the)
doing of it is not enough for a complete understanding. There must be
an understanding, in a sensory way, of the consequences of the motor
behavior, and the only way is through mimicry: I do not see myself baring
236 Chapter 11
my teeth but I know that I do that when I am angry and/or ready to at-
tack. That is my emotional state and that is the FAP it liberates. Now I
see this other animal baring its teeth at me: it could be another alligator,
or it could be any animal that bares its teeth. Do I put it together that that
FAP is the outward expression of the same internal state of anger that I
know, or do I not put it together? Hopefully I do put it together my life
depends on it but the point here is that this is the only way in which an
animal can come to recognize and know what the momentary internal
state of another animal may be. Furthermore, understanding must be ac-
quired by trial and error and it must be acquired through mimicry, which
means attempting to replicate the motor behavior one notices being ex-
pressed by another. Such understandings are generalized by trial and er-
ror. It may take me forever to understand that a particular gesture by
another animal means danger, because I may not know that. Instead I
may simply have lived around those alligators that do know, long
enough, and may simply have mimicked the behaviors displayed by them
after something bared their teeth and snarled at them. They walked away
and so I walk away with them with not so much as an inkling of an idea
as to what the precipitating event might have been! But my chance of sur-
vival is increased because I receive the beneªts of the knowledge of my
fellow alligators. I am naturally selected because I behave as if I know
that if something bares its teeth at me it means danger. This is what I
mean by the generalization of these understandings and that they too are
at the behest of trial and error. We may look at these as providing each
animal s speciªc informational niche, what those around me do and what
these things come to mean to me.
It is pattern recognition, of course, but this pattern recognition is to-
tally context-dependent. This is why you can t take an alligator from the
Nile and put it in the Amazon and expect it to survive. There is no longer
any familiarity of patterns, any matching of the internal with the per-
ceived external. The alligator won t recognize anything in its environ-
ment, and its system expects a set of external features that are
unfortunately completely different from what surrounds it! In other
words, abstraction has been bamboozled!
A brief footnote. It appears as if abstraction is seeking a match up of
the internal with the external. The system accepts as recognizable pat-
Language as the Child of Abstract Thought 237
terns that are only close at best, if that. Recall our discussion about im-
printing. Clearly this phenomenon employs pattern recognition and thus
abstraction to engender and associate understanding or meaning. An-
other example: recently there has been the creation of an artiªcial bee
from a microchip. It does the dance, communicates with the other, real
bees, and off they all race together looking for the food (Montague et al.
1995). And so it doesn t really have to be a bee: just a reasonably close,
dynamic, four-dimensional geometric pattern!
Role of Mimicry in Language Development
Let us return to mimicry and its role in the development of language.
What routes might natural selection have taken to allow for mimicry to
be an intrinsic property? The path of least resistance what else? We are
moving from the ability to mimic to the desire to mimic, a crucial next
step. The instinct to mimic is ampliªed by the fact that for most animals,
mimicry is very easily accomplished through the auditory system. Why?
Because if you hear a sound you can produce your own sound over and
over until it matches. The visual system is more difªcult, as I pointed out
before, because one usually does not see oneself doing something. So you
get mimicry evolving fastest and running rampant in systems where it is
very easy for it to ºourish.
Mimicry of Sounds
Clearly animals can make sounds. The nice thing about making sounds is
that the animal can hear the sounds it makes. The FAP of vocalization
makes a huge difference, because the animal can now match from a mo-
tor point of view whatever sensory tape it may have, that is, whatever
sounds it may be attempting to mimic (remember the bird song issues).
Vocalization is quite a fascinating subject. We tend, for some reason, to
lump vocalization and language together, typically in the sense of human
language, but vocalization is much older. The true richness of vocaliza-
tion as we know it evolved once intention or prosody was coupled with
that effector. But that effector, vocalization, came ªrst, probably as some
motor accident that caused something beneªcial and was thus naturally
selected for.
238 Chapter 11
Vocalization was itself part of a rather sophisticated FAP to begin with.
Which one, you ask? When you hurt, you cry out. And you cry out be-
cause it hurts, but this vocalization is also a form of defense. When you
cry out you tend to startle for an instant the attacking animal, whether
human or other, and this distracts it momentarily from its offensive. If
you scream, the attacker may stop and possibly even go away. You have
increased your chances of survival. Now think in the evolutionary sense.
You scream when you are bitten, when you are attacked. From this it
evolves to be that you also may scream when you think you are being bit-
ten and/or attacked. Next you simply scream whenever you feel pain, and
now it has nothing to do with being attacked. A sharp pang in your stom-
ach, you get a cramp in your leg, you stub your toe on a rock. It general-
izes from that.
Vocalizing moves from just representing the responses to external ac-
tivity to that of internal activity. Vocalization is a motor reºection of
arousal. Now one should not think that arousal is only a response to
something from the outside world. To be sure, your alarm clock is a very
arousing object. But arousal is an internal state and is generated from
within as much as from without. Some hours later you realize that you
locked your keys inside your car by mistake: Oh no! This is arousal, ab-
solutely, and this state is generated purely from an internal stimulus, that
the concept of locking the keys in the car is a disturbing one. The concept
is the stimulus. Or a dream may be so intense that it actually rouses you
from sleep. This is again arousal purely from the inside. Stimuli of the ex-
ternal world are beating on the nervous system and stimuli from the in-
ternal world are beating on the system; they are in this regard quite
equivalent. Thus it is that an animal cries out both because it has been
bitten and because it has been bitten from the inside.
And so an animal hears noises, makes noises, hears its own noises be-
ing made, and comes to learn what the noises mean when it makes the
noise and when it hears another making the (same) noise. This form of
mimicry is probably the best way to associate things, as we have said, be-
cause when you are angry and you make an angry noise, you will come to
recognize as angry these noises in other animals. You come to recog-
nize that they are having the same sort of internal experience that you
had when you made that particular noise. It is the same as the laughter
Language as the Child of Abstract Thought 239
we spoke of earlier. You come to hear and recognize laughter because you
associate it with your own.
Mimicry ºourished quite well through these systems because we can
hear the sounds we make and make the sounds we hear. If you also pos-
sess an apparatus capable of generating sufªciently complex sound and
sound patterns, as we have with our sophisticated laryngeal mechanism
(Hirose and Gay 1972; Passingham 1981; Doupe 1993; Zhang et al.
1994; Davis et al. 1996; Jurgens and Zwirner 1996; Jurgens 1998;
Doupe and Kuhl 1999) and as birds have with their broad-ranged sound
shaping syrinx (Goller and Suthers 1996a,b; Goller and Larsen 1997a,b;
Wild 1997a,b; Suthers 1997; Suthers et al. 1999), then the complexities
of internal meaning and their outward expression will ºourish and grow
as well. A cow, on the other hand, can moo, and that s about it; it
doesn t have many other voices, so to speak. In humans and many birds,
the range of the phonating apparatus is enormous and so naturally be-
comes selected as a very good medium for communication.
Visual Mimicry
Also quite common in the animal kingdom but not so easy to bring about
is mimicry through the visual system. Consider the ºounder; although
perhaps not the most elegant and beautiful of creatures, the ºounder is
nonetheless fascinating in its intent. Think of a ºounder: it has its two
eyes on one side of the body, the dark side. The other side, the underbelly,
is light colored. A ºounder on the bottom of the ocean settles in and dis-
appears into an image of the surrounding external world. What is so in-
triguing here is that in completing an image of its universe, this particular
animal is creating with its body a pattern over a patch of sea bottom that,
by deªnition, it will never see. Its eyes, facing upward, are not positioned
to see the terrain the body is covering. The image, made for others to see,
is an attempt to mimic the visual context of whatever else is surrounding
the ªsh where it has come to rest. The ºounder is creating a pattern in its
skin of what the sea ºoor underneath it should look like to others given
the surroundings. This is an odd animal. Obviously, if one covers its eyes,
it cannot make the (given) pattern. If a ºounder is placed on top of a
chessboard, it will attempt to copy the pattern to ªt in visually and con-
tinue the pattern. The important thing is that this ªsh has to be able to
240 Chapter 11
Figure 11.2
Photographs of ºounders, illustrating their capacity for camouºaging themselves
by changing their appearance to mimic the color and texture of the background.
The adaptive changes take place within 2 8 seconds. Left and middle images are
of the same ªsh. (Left from website www.richmond.edu/ ed344/webunits/verte-
brates/camouºage.html; middle and right from Ramachandran et al., 1996,
ªgure 1, p. 816.)
generalize, to abstract. It won t manage a great likeness to the chessboard
pattern one can see the clear problems with the optical system of this
ªsh but it certainly gives it a sporting try. This is a very beautiful case of
mimicry from a visual system point of view. The animal has the ability to
create a bit of reality that does not exist but that is nonetheless close
enough to the terrain for another animal to not be able to discern a dis-
continuity in what it is seeing (ªgure 11.2). This camouºaging behavior
on the part of the ºounder can only be explained by the fact that abstrac-
tion must be involved in this type of pattern generation.
Another animal in which we see relatively more evolved mimicry
through the visual system is the cuttleªsh Sepia. These cephalopods
have specialized cells in the skin called chromophores that can expand or
contract by nervous activation and thus appear to change from white
to black (Ferguson et al. 1994; Loi et al. 1996; Shasher et al. 1996).
Through the use of the chromophores, the cuttleªsh will make all sorts of
neonlike patterns, lines and geometries of lines that go this way and that
across their bodies. A quite striking effect, this is another example of a
probably very complex prosodic language, in this case a purely visual
one, a semaphoric type of language, if you will. This is also a very rich
language, because the cuttleªsh can make these dissipating patterns very
rapidly and with a great range of complexity. That there is meaning con-
Language as the Child of Abstract Thought 241
veyed by this language between the two animals is clear; from the exact-
ness and speed with which they mimic each other s patterns and the
patterns of patterns, it is obvious more is going on than idle reºex. Some-
thing must drive the necessary sophistication and control to be able to
signal with such an incredible amount of complexity, and that something
has intention to do written all over it.
Let us summarize brieºy. Language must clearly be born of abstraction
or abstract thinking, or, to put it another way, the processes generating
the abstracting properties of the nervous system must have preceded
what we call language, most particularly prosodic language. We have
deªned prosody as the outward expression of a momentary internal state
that by way of this outward expression means something to another crea-
ture. Such internal states, emotions and intentions alike, do not exist in
the outside world and so by deªnition must be abstractions. How such
internal states have come to mean the same thing or close enough to the
same thing to be useful between animals must have evolved through the
conduit of mimicry. Mimicry provides the commonality of same behavior
so that the associations between internal states and perceived behavior in
others may occur. I do this when I feel this way; I now see you doing this
and so maybe you feel the same way when you are doing it. And so, over
eons of trial and error, meaning between organisms evolved.
Mimicry has occurred in two primary ways, both of which are
reºections of internal abstraction, though of differing types. There is
mimicry that is copying, such as the I hear this sound, I make this sound
until it matches method, and there is extrapolation, as the ºounder does
by abstracting itself into a visual pattern that others see. Globally speak-
ing, for meaningful communication, copying sets the stage of commonal-
ity by which the nuances of extrapolation may be discerned between
creatures.
Of Human Language
From all of the above, it should be clear that language of any kind could
not have materialized suddenly out of nowhere, a lightning bolt of ad-
vancement in biological evolution. This will not sit well with those who
242 Chapter 11
feel that there is no form of true language except for that found in hu-
mans, but it is reasonable to consider that abstraction and prosody,
whose intra- and interspecies meaning developed slowly out of mimicry,
must be the prerequisite elements, the evolutionary preamble, for that
which we know as human language.
Language, and in particular human language, arose as an extension of
premotor conditions, namely those of the increasing complexities of
intentionality as abstract thinking grew richer. It simply became neces-
sary for us humans to do more with what we had, with our motor capa-
bilities, and by increasing the sophistication of purposeful movement
through modifying and overriding the existing FAPs. The expansion into
a much wider range of motor expression probably occurred at the same
time as the ability to override FAPs.
The reader will recall from chapter 7 that the ability to override a liber-
ated FAP is an ability born out of an increasingly elaborate thalamo-
cortical system. It is also most clearly supported by the development of
corticospinal connectivity (the so-called pyramidal tract). In a word, we
see the addition of increasingly sophisticated intentionality behind pur-
poseful movement. The evolution of such specialized systems as the pyra-
midal tract, which is related to toe and ªnger movements, as well as to
the activation of the cranial nerves activating lips, tongue, and pharynx-
larynx, supported the ability to overide particular FAPs, removing inher-
ent constraints. It allowed for the incredible dexterity we see in higher
mammals, particularly in simians and humans. It is fair to say that the
evolution and enrichment of the cortical mantle, as occurred in the motor
system, is the most important overall message to be gathered from the
evolution of the cortex. It has been the ability of the nervous system to in-
crease the number of possible functional states without violating the
FAPs upon which voluntary movements must ultimately ride. We cannot
override FAPs to the point where they simply do not exist or are com-
pletely dormant. We have already discussed in depth the critically impor-
tant need they ªll in reducing the nervous system s computational
overhead. Rather, it is the ªnely achieved evolutionary balance between
automatic, computational efªciency and the ability to generate necessary
nuance to our movements that characterizes our brain as the most exqui-
sitely capable of all brains. These issues are so standard, so taken for
Language as the Child of Abstract Thought 243
granted that we hardly ever realize the incredible coordination of FAPs
that is necessary for something so commonplace as giving a public
speech. One must be able to maintain an erect, vertical stance, or, per-
haps, depending on circumstances, a constant pacing movement, while
seamlessly executing the very complicated synergies that allow for respi-
ratory, laryngeal, and orofacial mechanisms to act synchronously to pro-
duce even one recognizable word! Add to this prosodic movements such
as gesturing with the arms and facial expressions, and we are now de-
scribing a rather complex motor event. Yet, for most, it is considered as
just walking and talking at the same time.
Let us take the example of the evolution of increased eye/hand coordi-
nation, where the new sophistication of cortical connectivity produces a
wider variety of possible movements, and yet does so without violating
the FAPs that underlie all voluntary movements. We may say here that
this new system interlaces with the present FAPs and uses them as col-
umns or buttresses on which to support the very new. Metaphorically
speaking, the richer intentionality demands a stage upon which to enact
particular plays that have never been played before. This should sound
familiar with respect to the processes underlying abstraction, for they are
one and the same. In fact, this is precisely where abstract thinking gets
particularly puriªed, as discussed in chapter 6. But, take note, vocaliza-
tion became with usage a FAP! The new system, bringing increased
intentionality and prosodic capabilities, used and expanded the FAP of
vocalization, taking advantage of its wide range of (sound) pattern gener-
ation. And so, just as the increased complexity of intentionality de-
manded an increase in the richness of eye/hand coordination, so too did
the increased need to express internal abstractions take advantage of and
increase the dexterity of the vocal mechanism of the animal. It did so not
only with the laryngeal and orofacial structures that shape the different
sounds, but also with the entire respiratory apparatus, so necessary in
generating the patterned airºow required to make these very speciªc
sounds (Wild 1994; Davis et al. 1996). These sounds, known to us all
through the processes of mimicry and repetition, are the phonemes that
comprise our language. They represent the very granularity of our lan-
guage and they are the basis for human language independent of the par-
ticular language used. In complicated eye/hand coordination in which the
244 Chapter 11
numbers of motor patterns activated are very large (Jeannerod 1986;
Miall 1998), the possible patterns are nevertheless ªnite in number. It is
the same with phonemes or the letters of an alphabet: much can be done
as these ªnite grains of motricity become mixed into incredible mosaics
of expression. When we add our already rich prosodic capabilities to this
language, it is difªcult to argue that we, as a species, have evolved to ex-
press our internal states more thoroughly than any other species.
With respect to linguistic theory, one major theoretical proposal has
been a source of much controversy in this century. This is the concept of
modularity as it relates to brain function. As originally stated by
Chomsky in his book on the neurological basis of language (Chomsky
1972), it was his view that the unique ability of the human nervous sys-
tem to generate complex language was produced by a very special func-
tion in the brain, probably subserved by a very specialized region. This is
not necessarily so. Certainly the presence of Wernicke s area (language
comprehension, or auditory association area) and Broca s speech area,
and the problems caused by lesions to these areas (alexia, inability to
read, anomia, word-ªnding difªculties, aphasias, speech disorders) pro-
vided a strong impetus and justiªcation for this particular view. And yet
this answer is unsatisfying, because of the nervous system s limited ability
to reorganize itself beyond a certain point, the imprecise localization of
these areas, and the possibility of the migration of such functions from
one part of the brain to another (for instance as in epilepsy). These
ªndings question the very simplistic view of a phrenological type of mod-
ular organization that permeated neurology for many years and that
again seemed to be supported by the misuse of noninvasive imaging tech-
niques in what may be called neophrenology. But the fact that it may be
difªcult to pinpoint a brain event to less than a few cubic centimeters of
tissue is not sufªcient evidence to discard modularity entirely, especially if
such modularity is considered a functional structure, even if a dissipative
one.
My own particular reason for accepting this view has to do with pa-
tients such as those described in chapter 7 and the issue of FAPs. Not only
phonemes but also particular words can be generated by an individual
whose brain has been damaged to such an extent that only the most mini-
mal neurological correlate to the module remains (ªgure 11.3). This indi-
Language as the Child of Abstract Thought 245
Figure 11.3
Positron emission tomography (PET) scan of a woman who has been in a coma
for more than 20 years in a permanent vegetative state. She spontaneously but in-
frequently produces isolated words unrelated to any external stimulation. Her
condition developed as a result of a succession of three massive strokes that col-
lectively destroyed all but the basal ganglia, some parts of the thalamus, and a re-
gion of the cortex known as Broca s area that controls the motor generation of
speech. The highlighted areas indicate those few brain regions of signiªcant meta-
bolic activity. (See chapter 7, p. 152.) This case history indicates that selected ce-
rebral circuits can support modular motor expression. In the case of spoken
words, this requires proper phonological articulation, and the properly timed ac-
tivation of various muscles (including the diaphragm) in addition to the vocal
cords. (Adapted from Schiff et al. 1999, ªgure 4).
cates that the motor FAP is necessary and sufªcient for the behavior of
that particular module but not sufªcient for other aspects of speech, such
as being able to produce the thought required behind the words or even
the appropriate context within which those words should be used.
This excerpt from
I of the Vortex.
Rodolfo R. Llinás.
© 2001 The MIT Press.
is provided in screen-viewable form for personal use only by members
of MIT CogNet.
Unauthorized use or dissemination of this information is expressly
forbidden.
If you have any questions about this material, please contact
cognetadmin@cognet.mit.edu.
Wyszukiwarka
Podobne podstrony:
CHAP11chap11worksheets chap11CHAP11CHAP11mcga shs?pt guide chap11chap11chap11chap11 (2)Chap11 (2)więcej podobnych podstron