Language, Literacy, and Sociocultural Studies
Deacon Chapter Summaries
"The Size of Intelligence"
A Gross Misunderstanding
The aim of this chapter is to determine what happened to human brains to make it possible for our ancestors to break the language learning barrier. According to popular books and magazines, it is a well-established fact that the human brain is a better, more powerful computer because it is proportionately bigger and can process and retain more information. In other words, "Bigger is smarter". However, researchers do not agree on how best to measure the effective size increase. But there is broad agreement that in some way an increase in brain size does correspond with our increased mental abilities. Rather than a single trait (bigger = smarter), brain size may reflect a delicate balance of the various brain functions.
Human brains are unusually large, in both absolute and relative terms. Larger brains may mean more storage capacity or discriminatory ability, but they may also mean changes in rate of processing, changes in relative excitatory or inhibitory influence over other linked systems, or a difference in intrinsic signal production tendencies. So it appears that differences in brain size can influence its functions.
However brain functions are not uniformly distributed throughout the brain. "We can't afford to assume that the human brain is just bigger, and put all our efforts toward discovering what selected its size" (p147).
We rank people's mental abilities for various purposes, in work, school and casual conversations. We assume that intelligence has a fixed value determined very early in life. The comparatively large size of human brains seem to provide a satisfying confirmation that the amount of mental ability one has is related to the amount of tissue one devotes to producing it.
Our brains are not the biggest, nor do they have the most neurons or connections. We do have large brains in relation to our body size but in simple ratio terms mice are brainier. How to assess brain size is not obvious. Perhaps we can gain a clearer idea of the problem by considering some less complex organ system: the musculature.
Larger muscles are capable of generating more force. By analogy a larger brain should be capable of greater computing power, of processing more information per second, and of producing more complex mental representations and communications than smaller brains. However, even though a large animal with greater total muscle mass can exert more force with its limbs than a smaller one, it may not be able to jump proportionately higher as a result. Larger bodies need more muscle mass to move them with the same facility as smaller ones. This is essentially the difference between gross strength and net strength. Gross strength probably correlates well with total muscle mass, but net strength depends on many other factors. There is only one measure of gross strength, but there are innumerable different measures of net strength, even for a single individual. This is because net strength is a comparison of part to whole. Whether gross or net strength is important depends on the context. Often both are important in different aspects of an activity and both may interact to determine performance in complicated ways. It appears to be the same in measuring gross and net brain function.
The presumed correlation of use with the evolutionary enlargement or reduction of organs is seldom questioned. Large size means more used; small size means less used. In the case of muscles, bones, and glands, we have a correlation between functional demand and size as a result of physiological changes during a lifetime. Physical exercise can cause muscle mass to increase and inactivity can cause muscles to atrophy. The analogy between changes in a lifetime and changes in evolution appears to work in this case.
Unlike muscle mass, brain size does not increase or decrease with use during the lifetime. There is no way to appeal to a physiological adaptation where habitual brain use is associated with enlargement or reduction in brain size.
Brains in Bodies
Some fraction of brain function must always be devoted to handling the information-processing demands and is therefore unavailable for other cognitive uses. The gross function could be divided into visceral (feelings) and cognitive fractions. Therefore if larger brains have to serve information-processing demands of larger bodies, they will not necessarily offer any net increase in cognitive power. Most very small mammals have comparable brain/body ratios to humans. Ratio steadily decreases with increasing body size. This does not suggest that mental abilities decline with increasing size.
One major cost that will scale with respect to the size of the brain is its metabolic demand. The brain is the most metabolically expensive organ at rest. Because such noncognitive costs must be considered, we should not expect that the average scaling of brain size to body size reflects anything like a line of balanced intellectual capacity. In other words, brains may have to devote bigger portions of their information-processing functions to management-like functions because a larger body will demand more managerial functions just to break even.
An assessment of net brain function is further complicated by its nonlinearity.
Like an expanding business, brains may have to devote an increasingly larger
proportion of the information-processing capacity to managementlike functions,
just to maintain equivalent levels of integration and control in the face
of increased size and complexity. Larger brains tend to gain a greater
gross computational capacity, but with decreasing net computational efficiency.
With larger size there will be more computational overhead that must be
subtracted with respect to input-output capacity. There is reason
to suspect that larger brains will need to be differently organized from
smaller brains with different proper masses required for different functions,
further complicating comparisons across ranges of sizes.
Very large differences in absolute brain size seem to have a correlation with some aspects of mental ability that we recognize as intelligence. One facet of the brain size/intelligence problem is the fact that different-sized animals live in very different worlds. Natural forms are subjected to quite different types of forces and physical constraints if they differ significantly in size. Fleas must be designed very differently from the dogs they live on. Within mammals we find a range of sizes nearly as extreme as that between the fleas and dogs. But the body plans of all mammals, including their brains, are surprisingly similar. However information-processing makes very different demands on large and small species brains.
Small animals' reflexes must be quicker to control smaller limbs and respond to rapid locomotor feedback. Their higher metabolic rates, and low energy reserves offer little leeway in foraging, defense, and mating activities. A short life span allows little time for learning from experience. In contrast, larger animals can get by with slower reflexes, can vary mating and foraging activities and thus optimize their behaviors. Being longer-lived puts a greater emphasis on learning and memory and less on automatic programmed behavior. It also allows for opportunities to travel long distances. This will expose an animal to significant changes in environment. Therefore, they must be able to assess the effects and adapt their responses accordingly.
Scaling brain functions up or down with size and life span is not simply a matter of more or less computing power. In brains that differ by millions or billions of neurons, maintaining a comparable degree of functional connections among them would require astronomical increases in connections, well beyond any hope of being housed in one body. It is impossible to meet the scaling-up demand in any real brain.
In real brains, connections per neuron increase slightly with size. But the proportion of connections to and from any one area is reduced. Increasing size means an increasing fragmentation of function. It also means loss of speed because of the increased distances and increased number of nodes that must be traveled, by a signal in order to reach removed sites in the network. Thus, even if size confers great information-carrying capacity, these gains may be balanced by significant costs in other areas of function.
Where size should matter most is in differences in reliance on alternative learning strategies and the extent and organization of mnemonic storage. This capacity which enables animals to produce novel responses by reusing information in new context, has been shown to correlate with brain size but not encephalization.
This pattern of size-correlated differences in the importance of different learning strategies is of particular relevance to special learning problems posed by language. Flexible and indirect learning strategies can only be of use if there is sufficient time to use them. Since language learning is an extreme example of a highly distributed learning problem, smaller-brained, short-lived species would likely be far more biased against the appropriate learning strategies than larger-brained, longer-lived species. Thus absolute brain size might have played an important limiting role in language evolution irrespective of any increase in computing power.
The associative relationships that support symbolic reference are unlearnable by brute force. This explains why even simple languages are nearly impossible for nonhuman brains to learn. Other species have intrinsic learning biases that undermine the process before it can even get started. Evolving a more powerful learning device is not the solution to the language-learning problem. Evolving a larger, more powerful learning device is not the answer to the language-learning problem. Perhaps we cannot assess the evolution of human intelligence in global terms.
There is no escaping the fact, however, that human brains are unusually
large in absolute and relative terms. It appears that something having
to do with the sizes of brain structures is central to the origins of the
human mind. We need to find the changes in brain organization that
correlate with this global change in brain size and the functional consequences
of such changes.
Glossary of Terms
axon -- long threadlike part of a nerve cell; conducts impulses from the cell
cognitive -- pertaining to knowledge; awareness
Chapter Six: "Growing Apart"
". . . nothing is great and little otherwise by comparison."
The Chihuahua Fallacy
As we learned in Chapter Five, encephalization is the degree to which an animal's brain size is larger than would be predicted for an animal of its size. Encephalization is, therefore, the relationship between the brain and the body, and can be affected by changes in either one.
This is most easily demonstrated in domestic breeds of dogs. Small dogs have a large ratio between brain and body size, and large dogs have a small ratio between brain and body size. Does this fact, then, mean that small dogs with a larger ratio are intelligent and large dogs with a smaller ratio are relatively stupid? There is, of course, no pattern like this for dog intelligence.
In terms of dog breeding, body size is more variable than brain size within the same species. Smaller dogs within a particular species have only a little less brain size than the larger dogs in that species. Thus, as Terrence Deacon says, "small dogs appear hyper-encephalized and large dogs appear hypo-encephalized compared to typical mammals."
Although there are differences in the smartness and dispositions of various breeds (some of which are developed intentionally by breeders), no one thinks that breeding for a genius dog would lead to small dogs or that breeding for stupidity would lead to large dogs.
Differences in dog encephalization, however, are the result of breeding for certain body size effects or body proportions (leg length, head shape, etc.) and not the result of breeding for intelligence. In fact, because of breeding, dogs and other domesticated animals have, on the average, smaller brains than undomesticated animals.
Deacon says that the problem with talking about brain/body relationships is that we often think the relevant part is the brain size when instead it's really the body size that makes the difference. Chihuahuas and human dwarves are encephalized because their bodies are small, not because their brains are large.
When we see a dachshund, too, we see right away that its legs are shorter than other dogs' legs because we have a frame of reference to compare it to, i.e., an average dog. We can then distinguish the figure from the ground. By looking at a single body part, like the brain, to compare to the size of the rest of the body is biased because it looks at this one body part out of context.
Fortunately, there is another way to talk about brain size and its evolution. We know that primates on the whole have developed larger brains for their size than most other mammals. Most researchers have thought this meant that primates are smarter than other mammals and that humans are the smartest of all primates. But, as Deacon asks, does larger mean smarter? And, more importantly, is the larger brain size a result of brain growth or the result instead of reduced body size? In other words, which is the figure and which is the ground?
The truth is that primates don't have brains that develop faster, they have instead bodies that grow slower. As Deacon says, "If primates were selected for increased intelligence, why should this produce a change in body growth but not brain growth? . . . Are we justified in inferring anything about brain evolution from a reduction in body growth with respect to brain growth?"
Deacon then looks at how and why primate bodies have been reduced. He asks, "Does primate brain and body growth differ from other mammal patterns the way Chihuahua brain and body growth differs from the growth patterns of other dogs?" (pgs. 171-2)
In fact, the path to reduced body size without a corresponding reduced brain size is different in humans and Chihuahuas. The bodies of dwarfed animals, like Chihuahuas, show slow growth near the end of development, but human fetus brains and bodies grow at nearly identical rates, but their whole body is smaller at every stage of gestation.
Although in the past it was thought that primates developed bigger brains than other mammals because of a more "cognitively demanding niche," now researchers see that humans begin by growing according to the normal primate plan, but our brains continue to grow for longer than expected.
As Deacon says, "Our brains grow as though in a primate with an adult body size that well exceeds 1,000 pounds, and yet our body growth pattern is quite similar to that of a chimpanzee. Both the human brain and the rest of the human body grow according to expected trends for their target adult sizes. These trends just don't normally belong together in the same individual. It's as though the 'designer' got the parts mixed up."
In other words, it's as if the human brain doesn't know it's developing in such a small body.
Deacon says that it's our preoccupation with encephalization and brain size that has led us to not distinguish between three different processes going on: "dwarfism in small mammal breeds, embryonic reduction of body growth but not brain growth in primates, and brain growth prolongation in humans without an extension of body growth."
Now, we need to look at how such differences have come about and researchers are just beginning to understand this.
Using Fly Genes to Make Human Brains
What has helped us to understand these differences are the small fruit fly and something called homeotic genes.
Homeotic genes are a group of genes in the embryo that control the development of the brain, heart, stomach, limbs, and other organs which are repeated along a line in the body. These genes, named because "their expression patterns correlate with the repetition of similarly organized segments along the body axis (homeo meaning 'similar;' sharing a common underlying plan), are important because of their evolutionary conservatism."
In recent research, a homeotic gene that is important for the development of the human head and brain has been transplanted to mutant fly embryos whose growth has been stunted and this human gene has partially helped to make the fly grow normally, showing in fact that the homeotic gene is doing what it's supposed to do.
Homeotic genes have one major similarity: "They each encode a DNA-binding region that enables the protein molecule they produce to bind to other sites on a chromosome and so regulate the expression of other genes." In fact, one of the most common versions of these DNA-binding regions is the homeobox which has "the ability to control whole suites of other genes." In fact, there is a sort of "hierarchic feedback" which allows these genes to cause similar genetic events at different parts in embryo development and at different stages, too.
Homeotic genes set off pattern formations in the embryo because different genes become active in separated groups of cells. Certain cells, in fact, inherit "developmental fates" from other cells and, as a result, the bands in human embryos look like the segments of a worm or caterpillar. What happens then in human embryos, however, is that the segments become blurred because each segment becomes different and the segments don't all line up after development like they would in a worm.
This is especially true in the head, but it's taken scientists a long time to prove that the development of vertebrate body segments continues similarly in the human head. This is where the fruit fly has come in handy. Scientists could easily see in the fruit fly that the structures of the head were similar to those in the rest of the body. In the fruit fly, "the segmented structure of the antennae resembles the segmented structure of the limbs. . . This suggests that the same basic program of genetic expression is slightly modified by homeotic controller genes to make both limbs and antennae." Scientist now see that a similar thing occurs in humans.
Homeotic genes are important, too, in the development of the human head and brain. "The early nervous system begins as a long tube," says Deacon, "running down the back from the head to the tail end of a wormlike, undifferentiated body. Hox genes, a type of homeotic genes, produce a segmented pattern in humans that is almost as regular as the ones in the fruit flies and in roughly the same order as in humans."
The discovery of homeotic genes has caused a revolution in the study of brain development. Scientists have discovered that even small adjustments in these genes in mice, frog, and fly embryos have caused changes in the structure of their bodies, including adding or deleting parts of the brain. Scientists think, therefore, that the changes in primate and human bodies may well be traced to homeotic genes.
Scientists have also discovered that individual genes in humans can affect the body and not the head, and vice versa. This may help to explain why human brains continue to grow after body growth has basically stopped. It may also explain why in selective breeding of dogs, the animals' bodies are affected but their heads and brains are basically not.
Scientists have also been looking for the basic divisions in the brain, trying to find out where and why segmental shifts have occurred. They still haven't found out where the different growth fields in the brain are and haven't determined "the nature of the underlying segmental shifts in proportions." As Deacons says, "To measure things appropriately, it helps to know the locations of the natural developmental divisions." (p. 182)
The other problem with brain size is that certain parts of the brain make up a larger portion of the brain than others and then proportional comparisons can be incorrect because each part is not treated separately. An additional problem is not recognizing that certain growth in the brain is within a growth area and others are between growth fields.
Another thing to look at is how human brains are different from other primates. First, the cerebral cortex is twice as large as predicted for other forebrain structures. Also in fact, the cerebellum and cerebral cortex, two of the most deviant brain structures, originate in humans from the back side of the developing neural tube. This variation from primate patterns suggests that the growth relationship between major dorsal and ventral segments of the developing forebrain has been somehow altered, whereas the many relationships within each of these broad divisions have remained relatively constant.
Perhaps the main discovery here for the brain relates to those homeotic
genes again because when we look at areas in the brain that are significantly
enlarged and those that are only slightly enlarged, these areas correspond
to the existence or lack of existence of these important genes.
The Developmental Clock
Cells that make up different sized animals don't differ very much in size; what makes certain animals and certain organs in those animals larger is the larger number of cells produced.
The decision about how long cell division will continue in a specific species looks as if it is made relatively early in the development of any animal. It has also been shown that individual organs have a "developmental clock" which makes them grow to the appropriate size even if transplanted into another animal. Thus, a pig brain is always a pig brain, i.e., it always develops to the size of a pig brain no matter what animal it's been placed in.
As we've seen before, the growth of the human brain is prolonged;
in other words, its developmental clock has been extended even though the
rest of the cells in the body have stopped proliferating.
Return to Top
Chapter 7 -- A Darwinian Electrician
by Jane Glickman-Bond
Musical Chairs: The Evolution of the Brain
Darwinian explanations require that incremental stages in adaptation must be ends in and of themselves. If brains evolved piecemeal, then how could new structures be added, given that it must be linked in a highly systematic fashion with millions of neurons in dozens of brain regions. For changes, even on the periphery to be effective, correlated change must occur in the way the brain handles input.
Morphogenesis is the result of local cell-cell interactions in which signaling neurons from one cell affect neighboring cells. The developmental assignment of neural functions to different regions of the brain is systematically determined (the brain as a whole participates in the designing of it's parts).
Axons establish functional connections (synapses) with target neurons.
Axonal extension allows distant cell populations to directly interact and
communicate. This superimposes non-local developmental logic on top
of local regional differentiation.
Brain development is complicated and counter-intuitive. The nervous system as a whole participates actively in it's own construction. Classical neuro-anatomical theories assumed that differentiation of each brain structure was an independent trait, i.e. different parts of the brain could be subject to independent evolutionary influences.
Neuro-biology assumes that non-growth processes play a major role in determining size, organization, and function of brain regions. Many of these functions are actively self-destructive. Cell death, (spontaneous or driven by competition for resources) is important for developmental sculpting of the neuron system, serving to match the proportions of one to another.
The Darwinian aspect is that an overproduction of random variants would be followed by the selective support of some and the elimination of most. While this appears wasteful, this system is the most efficient in its use of information. This circumvents the need to plan ahead and it allows development with a minimum of design or regulation.
Programmed neuronal death: Matches different but inter-connected cell populations to one another. In studies, the amputation of a limb or a muscle to an embryo led to a loss of larger and proportional motor neurons. A graft to an embryo led to fewer neurons lost. Nature prefers to overproduce and then to match rather than to monitor and coordinate.
Guidance mechanisms for fetal axons: Guide filaments, regional differences in cell-surface adhesions, spatial patterns of attraction and repulsion molecules. Mechanical properties of tissues, specific growth factors released by cells in target regions that help support axons upon arrival.
An outside source of brain wiring led to constancy in genome size.
Larger brains are less able to rely on genetic mechanisms to determine structural
differences. They are more prone to proportional perturbations due
to extrapolated growth processes.
Interspecies neural transplantation
It had been assumed that transplanting neural tissue between species (in this study pigs and rats) would lead to misrouted connections due to different guidance signals. The study actually found that donor neurons made appropriate and functional connections.
Neurons also use selective discrimination. Changes in bias affect
initial growth and variety or biases the selective process that culls connections.
Biases that influence axonal selection:
Cerebral cortex-cell structure, functions of specificity, connections to other brain structures. The brain tissue from the rat that was transplanted from one area in the brain to another was uninfluenced by the site of origin. Later in the rat's development however, strict regional differences in connectivity remain.
Complementary bias- comes from input projections in the thalamus.
The Cerebral cortex and the thalamus are equally likely to grow linking connections. Growth appears more affected by physical proximity and the physical constraints of timing of axonal growth, molecular gradients and proximity all offer subtle growth bias.
Phylogenic differences in size and functions of particular cortical or nuclear regions can not be attributed to the addition of cells to that area or to changes in gene expression in that area. Systemic change affects a number of brain regions when connections happen to converge in it.
Construction of Neural circuitry biases developmental timing, neuronal numbers, correlated activity patterns. If determination of which connections persist and which are retracted depends on correlated activity of other axons to the target region, then the relative quantities of projections that arrive in any target region are an important selective bias affecting which connections will be eliminated or not.
Among competing structures, the structure that sends the greatest number of axons to a particular target will tend to drive the activity patterns in that target more effectively. This means that modifications of the relative proportions of peripheral and central nervous system structures can significantly alter connective patterns.
Genetic biasing at the level of whole cell populations can result in reliable shifts in connection patterns. Displacement means that regional brain and peripheral nervous system size effects play a major role in mammalian brain evolution.
Relative enlargement of one target will attract connections away from the smaller. The competitive structures are fiercer with in smaller structures than with in the larger.
In newborn ferrets, the tactile center was cut from the thalamus and the thalamic target for visual stimulation was damaged. The undamaged visual projection was re-routed to undamaged tactile receptors. The tactile area of the cortex became responsive to visual stimulation. This has been repeated with rats, ferrets and hamsters.
The input information is essential for sculpting output connections. Neither input nor outputs are predetermined intrinsically. Transient connections that are eliminated during development of ancestral species might not be eliminated under differently based conditions that could arise in later lineage.
Rat fibers (fornix) also project beyond normal targets to reach central mid brain regions during transient period in early development. These are culled as the rat matures. Some retain it into adulthood. Elephants retain these fibers, as do some human brains. Size difference in these species may contribute to different competitive milieau that favors retention of this connection and biases against it in smaller brain species.
Sensori-motor specialization achieved in different species by the enlargement or reduction of peripheral structures. Cave dwellers and subterranean species have a reduction in eyes. In the blind mole, the visual area of the brain has been taken over by somatic, motor and auditory regions. Axonal displacement cascades throughout the developing brain spontaneously matching the central organization to the peripheral specialization.
An Alien Brain Transplant
The ability to speak and the ability to learn symbolic associations and
specifically the ability to produce skilled vocalizations can be traced
to changes in motor projections to the midbrain and brain stem. The ability
to overcome the symbol-learning problem can be traced to the expansion of
the prefrontal cortical region and the preeminence of its projections in
competition for synapses throughout the brain.
The concept of "proper mass" is invoked along with the assumption that larger brain structures are more powerful analyzers or more spacious storage devices than smaller ones. However, shifts in relative proportions will be translated into functional trade-offs. The class of neural computations supported by an enlarged region will tend to have more influence over final global outputs than that supported by a reduced region. From an adaptation perspective, the source of the competitive bias is irrelevant. The final connectional proportions are what will determine function.
Neural functions will be modified in response to proportional changes by a kind of functional displacement of some computational tendencies by others not just increase or decrease of localized functional capacities.
The displacement theory offers a predictive tool for interpreting the significance of quantitative differences in brain structure. The shifted pattern of regional parcellation in human brains can be interpreted as though enlarged systems were deluged with massive new set of peripheral influences. This input was supplied not from the periphery but internally as a result of shifts in the early production of neurons. It is an evolutionary response to virtual input with increased processing demands.
Morphogenesis: <cell biology> The process of shape formation: the processes that are responsible for producing the complex shapes of adults from the simple ball of cells that derives from division of the fertilized egg.
Axons: <cell biology> Long process, usually single, of a neuron, that carries efferent (outgoing) action potentials from the cell body towards target cells. Each nerve cell has one axon, which can be over a foot long. A nerve cell communicates with another nerve cell by transmitting signals from the branches at the end of its axon.
Dendrites: A long, branching outgrowth or extension from a neuron, that carries electrical signals from synapses to the cell body, unlike an axon that carries electrical signals away from the cell body. Each nerve cell usually has many dendrites. This classical definition, however, lost some weight with the discovery of axo-axonal and dendro-dendritic synapses.
Synapses: <physiology> A connection between excitable cells, by which an excitation is conveyed from one to the other. Chemical synapse: one in which an action potential causes the exocytosis of neurotransmitter from the presynaptic cell, which diffuses across the synaptic cleft and binds to ligand gated ion channels on the post synaptic cell. These ion channels then affect the resting potential of the post synaptic cell.
Neurons: An excitable cell specialised for the transmission of electrical signals over long distances. Neurons receive input from sensory cells or other neurons and send output to muscles or other neurons. Neurons with sensory input are called sensory neurons, neurons with muscle outputs are called motoneurons, neurons that connect only with other neurons are called interneurons. Neurons connect with each other via synapses. Neurons can be the longest cells known, a single axon can be several metres in length. Although signals are usually sent via action potentials, some neurons are nonspiking.
Functional differentiation: 1. Pertaining to, or connected with, a function or duty; official. 2. <physiology> Pertaining to the function of an organ or part, or to the functions in general. The distinguishing of one thing or disease from another.
Genome: <genetics, molecular biology> The total set of genes carried by an individual or cell.
Somatic: 1. pertaining to or characteristic of the soma or body. 2. pertaining to the body wall in contrast to the viscera.
Phrenology: 1.The science of the special functions of the several parts of the brain, or of the supposed connection between the various faculties of the mind and particular organs in the brain. 2. In popular usage, the physiological hypothesis of Gall, that the mental faculties, and traits of character, are shown on the surface of the head or skull; craniology. Gall marked out on his model of the head the places of twenty-six organs, as round inclosures with vacant interspaces. Spurzheim and Combe divided the whole scalp into oblong and conterminous patches.
Theory of localized brain functions proposed at the turn of the 19th
century. Reasoned that the shape of brains and the effects of
localized brain damage suggested that brain functions were located in local
centers. The individual differences corresponded to differences in talents
and traits. Therefore, head shape reflected underlying enlargements
or reductions of various centers.
Deacon, stumbles upon another language phenomenon study opportunity, while walking near the aquarium with his wife. This is where Deacon heard Hoover, a talking seal. In this chapter, Deacon investigates how Hoover makes such a human sound. He learns that a fisherman found Hoover as an orphaned, sickly young pup. The fisherman nursed Hoover back to health. Hoover was then adopted by the Boston Aquarium. He was sick several times, including one probable case of encephalitis. When Hoover approached puberty he began to talk. He vocalized words and phrases probably learned from the fisherman. He did not use increased speech at times such as mating season, but used more seal vocalizations then. His speech was different from seal vocalizations, not a modification. Deacon raises the question of how this came about. Did Hoover learn to speak? He didn't produce speech in the same manner as humans. Hoover reportedly died of natural causes and the question of his speech remains unanswered.
Deacon sees the Hoover case as a potential clue to Homo sapien speech. He raises the question of whether there is a common factor of brain organizations in species which can articulate speech. This is integrated into this chapter contrasting the brain functions and language usage development between mammals, birds, cetaceans, and humans.
Deacon begins to differentiate between the visceral or automatic processes, and the voluntary processes common to skeletal movements. In analyzing various systems in the human processes there are both visceral and voluntary processes involved in aspects of sound formation. These contrast to mammal processes as more visceral. Association of vocalization to the midbrain systems involves interaction of oral and respiratory tracts-visceral (automatic) processes. The larynx is in a pivotal role as the gateway to respiratory and oral tracts, which include a functional conflict of ingestion and respiration. It doesnít work well to swallow and breath at the same time, right?
Organization of the nuclei side by side in the brain stems is a key to
breathing, chewing, and swallowing processes. Groups of neurons in
these brain stems control muscles for these functions. These include visceral
motor columns (automatic) and skeletal motor columns (voluntary) on either
end with branchio motor functions columns in between these. The branchio
motor functions involve muscular functions and both voluntary and visceral
motor functions. Visceral motor systems were probably ëcooptedí
for communication due to their indication of arousal. Hissing sounds
could have been one of earliest forms of these arousal sounds.
Researchers used electrical stimulation of the central gray area in cat, squirrel monkey, and macaque monkey brains to determine their range of vocal calls. This worries me. These researchers determined that vocalizations are only one component of this process. Hypothalamic and limbic systems also can arouse. Vocalization is one outward manifestation of integrated emotional and behavioral arousal. Other outputs may be posture, autonomic, and hormonal changes. Researchers also used electrical stimulation experiments on monkeys to decide that when a lesion is farther from the midbrain there is less impact on vocalization than if closer.
(Note: I am concerned about what may or may not have occurred in all these experiments. Did these monkeys volunteer for this; was there pain involved; and who determines whether it was painful (the monkeys or the scientists)? I felt a need to pose these questions, which you may delete or ignore as you wish.)
In most mammalian vocalization is highly automatic and invariant. Deacon infers that learning plays little or no role in these motor programs. Different limbic circuits have different pathways and an activity signature or code which signals which vocal program to run depending upon which arousal or emotional state. Infectious laughter is an example of these reflex links.
Why can't mammals sing like birds? I will try to summarize how Deacon addresses this question. Human vocalization depends upon the cerebral cortex and rapid, skilled movements of oral and vocal muscles. This is different than cetaceans and birds but their are some parallels in skilled vocalizations. A key factor may be that humans, cetaceans, and birds avoid leaving vocalization to the control of visceral motor systems of the brain.
Many systems of the brain are involved in skilled vocalization. Visceral muscle systems most directly involved in mammals are not suited to this because they operate as programs of autonomous operations not affected by other systems, but running a stereotypic course. These must operate flexibly and involve skeletal muscle systems which may be developed and modified. Skilled behaviors can be as rapid and autonomous as innate behaviors, once it is programmed.
The cerebral cortex and cerebellum of mammals are some of the most critical
structures for conscious movement and for the development and modification
of skilled behaviors. For skilled vocalization the cerebellum
and basal ganglia play complementary role to the cortex and seem to be for
fine tuning and automating skilled movement patterns. Damage to primary
motor cortex of the mammal brain can cause paralysis, and affect the use
of mouth and tongue.
The key difference in birds and mammals is related to the tongue and laryngeal muscle control in mammals. The larynx is viscerally controlled, but the tongue is controlled by visceral and skeletal motor systems. The tongue can move both stereotypically and deliberately. The hypoglossal nucleus controls the tongue muscles. Deacon, experimenting on monkeys, found that each class of intrinsic and extrinsic tongue muscle had itís own hypoglossal column which makes speech and eating possible in humans and monkeys.
One key tongue muscle which is significant in comparing mammals to birds is the geniohyoid muscle. This muscle isnít attached to the tongue and is controlled by a separate hypoglossal column. What is key is that the subnucleus controlling this muscle is in the same relative. I think what Deacon is saying is that birds recruited their respiratory muscles and nerves to control the syrinx instead of the tongue in order to save space for a more compact body conducive to flight. The tongue controlled nucleus in birds, not a visceral system, controls the syrinx, thus controlling sound and breath. This leads to more deliberate and monitored air control in birds.
Cetaceans also exhibit vocal flexibility and learning ability. They demonstrate the rule of visceral versus skeletal motor control. Many of their sounds are probably not produced by the larynx, but by an elaborate system of sinuses in the front of the skull. Deacon says the blowhole muscles, like facial muscles are probably controlled by the skeletal motor nuclei of the brain stem. Deliberate control of air flow to blowholes is important to these aquatic mammals, possibly related to an adaptation that evolved from their need for air and water control.
Human phonation is not entirely skeletal. Most, like other mammals are by laryngeal muscle constriction. Most sounds involve both muscular systems, including skeletal control of jaw, lips, and tongue. Some are entirely controlled by oral muscles. Deacon says that the parallel in humans,cetaceans, and birds is that the skeletal muscle control must be involved for flexibility, learning, and intentional control of sound production.
Human vocalization is a dual process involving the orchestration of oral and laryngeal sound production. Singing exemplifies application of directive laryngeal control, while intonation may be visceral controlled. We must learn choreographed combinations of tongue, lip, and laryngeal movements in order to apply them in a vast array of contexts. Therefore, human speech ability involves more than the shift to oral muscles. It involves our capacity to have voluntary motor control of laryngeal movements, and higher brain neural control of the human larynx.
Deacon presents evidence in novel vocalizations dissociated from particular affective states. He says that human visceral-emotional and skeletal-muscle systems compete or complement each other in vocalization. Human vocalization involves both. As an example, he compares Jane Goodall's chimp observation, in which the chimp couldn't suppress his visceral food call to the human difficulty with controlling inopportune laughter.
Deacon states that the production of human speech involves intentional cortical motor control over autonomous subcortical vocal behaviors in an externalized model of an internalized neural relationship. He calls this the bridge from the primate communication to human speech and key to deconstructing language into evolutionary antecedents.
To explain this further, he presents that although mammal calls are not learned, the cortical regions appear to be critical for learning circumstances for using or inhibiting calls. Cortical motor damage doesnít stop the production of calls in monkeys, but affects them indirectly. One reason that mammal calls are largely unaffected by cortical motor damage is that calls do not generally involve complex movements of mouth and tongue. Air flow and frequency shifts involved in most innate animal and human calls (ie:laughter and sobbing). Only in humans does damage to or stimulation of the cortical auditory and motor cortex produce a disturbance of the structure of vocalizations. Speech is continuous, with breaks for breath or thought and sentence pauses, but unnatural breaks may be related to aphasia. Speech involves coordinated rapid tongue, lip, palate, jaw changes.
Calls have foreground visceral motor programs with background stable oral facial postures. Speech inverts this to skeletal motor systems with oral behaviors as foreground and respiratory systems as background. He puts humans somewhere between birds and chimps as to visceral versus skeletal motor control. Humans use oral muscles to control visceral calls, whereas the chimps used hand muscles to cover it up in the Goodall example. This represents a significant differentiation of human vocalization from mammal calls.
A Leveraged Takeover
All mammals have neurons that send output axons to deeper brain structures. In the motor cortical region they send output to the brain stem and spinal cord. In monkeys the projections are more extensive than other mammals. Monkey projections likely contact motor neurons directly, which probably relates to their increased voluntary control over hand and finger movements, food preparation, and facial communicative gestures. In nonprimates few cortical projections contact the output motor neurons directly.
Deacon relates the displacement process in the example of changes in the brain of Spalax, the blind mole rat with a big head. In primates there were many descending axons needing space, so more numerous cortical axons were displaced with less local connections. I think he's saying it was a body space problem with not enough room for all those axons and control centers. As human cortical axons expanded, where could they go? He says they went to the face and tongue muscle nuclei and to the nuclei-contolling visceral muscle systems. Thus we have human voluntary control over visceral systems. Isn't that an oxymoron?
Evidence of this human cortical control over the larynx is in muteness occurring from cortical damage. Neurologists say that cortical damage does not eliminate monkey vocalization. Besides control it takes coordination with breathing, and with muscle movements of the tongue, lips, and jaw. Probably they're all under common cortical control.
Laughter studies showed that humans usually only laugh or vocalize when exhaling, while primates laugh while inhaling and exhaling. Speech involves top down control, not bottom up control (visceral). This is evidenced by coordination of rapid inhalation to conduct coordinated complex lengthy speech patterns. Predominance of cortical projections to visceral motor systems also explains infant babbling of diverse phonemes. Other infant species donít produce near as many different phonemes. Human infants donít have to get excited or upset to babble. When upset more innate stereotypic crying interferes with babbling. I think heís showing how babbling is more voluntary or intentional than mammal sounds. This babbling doesnít occur until a few months old when cells begin to encase cortical motor tracts. This implies cortical control over babbling.
This brain to body proportion relationship may give insight into fossil ancestor speech sounds. When the brain size increased in Homo habilus, his cortical control over vocals also probably increased. Also the spinal cord in our thoracic cavity increased to enhance breath control. The question may be which came first, speech as a result of modifications, or modifications as a result of communicative need? These questions should be addressed in chapter 11. Deacon concludes that it was a slow evolutionary process, but that skeletal motor control of upper vocal tract articulation came before more visceral laryngeal control. Therefore early speech would involved more oral sounds such as clicks, instead of tonal variations or vowels.
His conclusion in the Hoover case? Hoover's probable brain
damage combined with need for breath control as an aquatic mammal, could
have led to cortical control over tongue and larynx. Yet, Hoover's
not here to tell us, so this part of our story remains a mystery.
To paraphrase Deacon; A being, human or non-human, may be perfectly capable to form sounds into words, however, this is NOT enough to sufficiently create and maintain a language. Language is dependent on learning and memory abilities and NOT on motor skills. The human brain has evolved into an organized mechanism that allows for SYMBOLIC ACQUISTION. All other useful adaptations in language development and articulation are a result of this ability for symbolic acquisition.
Our human brains are different from the brains of other animals because they (our brains) have ADAPTED to the highly complex demands imposed by our SYMBOLIC LEARNING. These adaptations have allowed our brains to develop differently from the brains of other primates in extreme ways. A closer look at these differences offers us clues to how our own human brains compute.
Symbol and The Prefrontal Cortex
Humans have large PREFRONTAL CORTEXES. What does the prefrontal cortex do? This is not a simple question to answer, because doctors and researchers have a hard time tying any specific area of the prefrontal cortex to any one specific sensory or motor function. This region of the brain represents a more general, and less specific, area than other parts of the brain "map". But we do know, and research has shown, that this prefrontal region is the part of the brain that works to PLAN complex behaviors. Although, to complicate matters, many people who have had injuries to this region of the brain often show no signs of impairment. If this part of our brain is damaged, little of our basic language abilities; producing speech, comprehending speech, or analyzing grammar, are ever disrupted, even though much of our information processing relies on the guidance of prefrontal directions. This reliance (dependence) on prefrontal organization of perceptions, actions, and learning is what sets us apart. We are "front-heavy" in our thinking.
So, if the prefrontal cortex does not influence the ability to
speak, or understand speech, or analyze grammar, how is it important in
understanding language and symbolic meaning? Tests conducted on humans
with prefrontal damage and monkeys with prefrontal damage have indicated
that damage to our really big prefrontal cortex does hamper our ability
to analyze higher-order relationships. AND it is the ability to realize
and understand complex higher-order relationships that affects our ability
to acquire the ability to LINK associations. When we link associations
with experiences and possibilities in an abstract manner we construct SYMBOL.
Remember, we begin creating symbol with the iconic relationship, proceed
to the idexical, and finally arrive at the symbolic. Humans connect
"reference" to language in ways unlike other animals. In
tests with primates we can encourage a kind of language system to develop
by supplying the symbolic reference----externally----but it is our unique
human qualities that allows us to internalize this stimulation.
The Case of the Williams Syndrome Children
A study done on children born with Williams syndrome seems to point
to the interesting issue that language learning can be divided into two
distinct areas, (1) grammar and syntactic ability, and (2) cognitive problem
solving abilities. Williams syndrome children can often be very adept
at storytelling and recitation of verbal information, while at the same
time exhibit great problems with the understanding of contextual language
knowledge. Physiologically speaking, most of their brain is underdeveloped,
but their frontal cortex and cerebellum are quite normal. Their prefrontal
region of the brain steps in and controls many processes that the other,
underdeveloped areas usually control. They can learn symbolic
associations even though their ability to learn nonsymbolic associations
is greatly limited. They can even demonstrate a great ability to use
vocabulary while at the same time have difficulty with problem solving issues.
Their problem lies in their inability to make indexical relationships, thus
they cannot make the higher-order connections required for successful cognitive
The cerebellum seems to play a role in the regulation of automatic
movements as well as the articulation of speech. The cerebellum is
extremely activated during difficult word association tasks, and in humans,
more than any other species, it is involved with sound analysis. It
is this process of sound analysis that allows for cognitive speech processes.
Deacon mentions this to illustrate his point that the critical aspects of
modern language functions---enabling symbolic construction, vocalization,
and hearing and speech analysis, depend on a special array of links between
the prefrontal areas of the brain.
In conclusion, Deacon sums his point up nicely by saying that it
is special to our human nature that we have enlarged prefrontal cortex regions
of the brain. This enlargement was part of a co-evolutionary process
between the brain and language. As the prefrontal cortex evolved and
enlarged, the ability to create symbolic relationships increased, as these
abilities to create symbolic relationships increased so did the means for
speaking, hearing, and the ability to analyze speech and sounds. As
these abilities developed, and became successful, the brain areas responsible
for their functions responded by evolving into elaborate and highly efficient
mechanisms. The process and the mechanism evolved interdependently,
each providing the essential 'ingredients' for the other.
In a nutshell, this chapter describes the anatomy and function
of specific areas of the brain which are used to produce and process language.
Deacon begins his discussion of brain anatomy by discussing the two fundamental areas of the brain utilized for speech and language. The first is Wernicke's area, located along the superior and middle temporal gyrus within the temporal lobe. (Just above your left ear, and a little further towards the back or your head.) It forms the back of the region of the brain known as the auditory complex. Any guesses as to what types of language processes occur there? Wernicke's area is responsible for the receptive skills involved in language, such as the comprehension of words and the selection of words when an individual is producing sentences (Jannedy et al, p. 253). Kind of makes sense that it's located in the auditory complex, if you're trying to listen and comprehend speech.
Aphasia is a term used to describe the effects of damage to a particular area of the brain. People who have had an aneurysm in the brain, (possibly due to stroke or traumatic brain injury) tend to exhibit disturbances in their language abilities. Damage to this area of the brain results in Wernicke's aphasia (how original!). This type of aphasic has "lost hold of the cognitive tools to map the sounds of words to what they mean" Deacon, p. 280. There is "a profound disturbance of the ability to understand speech, along with a tendency to speak fluently but with anomalous words and words combinations" Deacon, p. 281. (Because I wasn't sure of the exact meaning: anomalous- deviating from the common order; irregular - thank you Webster.) Deacon's language can be a challenge, so bear with me.
I've had the opportunity to listen to Wernicke's aphasics, through numerous videos in an audiology class. It is sad, but funny to hear them talk. This type of aphasia is probably better than most (if that's really possible) as the individual is extremely willing to talk and can seem "chatty". They do not understand most questions and they speak in long rambling streams, which make no sense to the person listening. What's important is that the Wernicke's aphasia has correct use of syntax and other grammatical features of language, though they are producing these expression in a way that makes no logical sense to the listener. Also they do not understand that they aren't making sense. This has provided support for the phonological loop theory of language. The gist of which here would be that the person is unable to monitor their own speech because once the language leaves the mouth and the sound of it enters the ear, the cycle begins over again, and they are unaware that there is anything wrong with what they are trying to tell you.
The second language area discussed by Deacon is Broca's area. It is located in the lower part of the frontal lobe (cortex) of the brain. (Again start at your left ear and go up about an inch or so.) It borders the frontal and motor cortices. Again, anyone care to guess what functions are carried out in this area? Broca's area is primarily responsible for "organizing the articulatory patterns of language and directing the motor cortex. ." for speech (Jannedy et al, p. 253).
Where Wernicke's aphasia is primarily related to errors in comprehension, Broca's aphasia results in errors of expression, or an "inability to plan the motor sequences used in speech or sign" (Jannedy et al, p. 256). Broca's aphasia is much more frustrating for the individual who is afflicted. Their speech is typically marked with a "telegraphic" flow and little to no affect or inflection. They become very angry because they understand the questions asked of them, but upon answering they are unable to produce the output which the brain has formulated. Unlike the Wernicke aphasic, they are acutely aware of their own errors in speech, and typically become so frustrated that they give up on their attempt.
Deacon notes that language disorders caused by aphasia do not result in an individual's return to child-like speech. Rather, "language tends to break down along distinct componential lines, in which the functional losses reflect specific difficulties, and not some generic diminution of language ability or complexity" (Deacon, p. 281). OK, so here he's basically saying that when language disorders occur due to aphasia, the individual doesn't talk in simple sentences and two-word utterances, rather specific parts or functions of language are targeted by the aphasia. Therefore "deterioration of a function is not just the reverse of development, . . . and a loss of a brain region does not result in a loss of function" (Deacon, p. 283). Deacon expresses the importance of Wernicke and Broca's research by stating that they were the first to analyze the brain in terms of a circuitry metaphor, rather than a region by region metaphor.
Deacon does present some other perspectives about aphasia which challenge Broca's conclusions, but Deacon dismisses most of these by saying that these theories are not as reliable when there is "extensive or bilateral damage" (Deacon, p. 285). Some other challenges which are discussed relate to the differences of particular patients who have Broca's aphasia but have very different symptoms. These patients are native speakers of highly-inflected languages, which are very different than English. These individuals exhibit a lesser rate of "agrammaticality" than individuals whoa re native English speakers. Basically what Deacon decides is, "we need to stop conceiving of localization of language functions and instead try to understand how language functions map onto brain functions" (Deacon, p. 285). This makes a lot of sense to the functionalist in me. I think that the way the brain functions cannot be explained in any simple terms. I think there is so much about cortical communication within the brain that we still don't understand. But it may be possible to establish some norms for language and brain function, and their areas of compatibility. Deacon suggests to think of these processes as "composite behavioral products or logically defined outcomes, as opposed to neural operations" (Deacon, p. 286). In fact he also suggests that it is pointless to have a one-to-one mapping for brain region and linguistic operations. He does acknowledge the fact that most of the current theories and research are based on damaged areas of the brain which may not reflect the day-to-day "normal" processing patterns of the brain.
The one point Deacon makes which really hits home to me, is the idea that when researchers make distinctions in cognitive function, they are analyzing a product which was produced by the whole brain. By studying only a discrete area they are not taking into consideration the effects of compounding processes occurring in other areas of the brain which get "added" to the discrete area and function being assessed. He summarizes by saying that, "language functions are dependent on the interactions among a number of separated regions within the brain" (Deacon, p. 288). Deacon then begins to discuss the results of brain-language data, rather than brain damage theories. This research shows a pattern of brain activity which seems more fluid and less discrete than the previous Broca and Wernicke localized hypotheses would suggest. As Deacon states, "what these stimulation studies demonstrate is that the regions where stimulation (of the neurons) disrupts language function fan out from the frontal mouth area into the prefrontal lobes, and from around the auditory area back into the temporal and parietal areas" (Deacon, p. 289).
He also concludes that there appear to be "tiers radiating outward from these two foci" (Deacon, p. 289). This basically means that there are larger portions of the brain involved than previously thought, and that rather than compartmentalized, they seem to function in radial/regional patterns. These multiple regions are different for "auto-pilot" or repetitive speech, simple word perception, and generating word lists. Deacon's greatest emphasis in this section is the notion of brain function in terms of parallel processing rather than sequential processing.
So, if you are producing speech, (talking) not only will your brain be highly active in the speech production and planning areas, but also in the motor cortex areas. Seems simple enough right? But, this allows us to see that the brain is processing all of those intricate functions at the same time, in parallel. So here is Deacon's rundown of which areas are responsible for unique processing, take notes!
"First, though different modalities of stimuli can be used, the differential activity patterns produced by the more complicated tasks using these inputs are very similar. Second, the ventral prefrontal area again seems activated when word analysis is required. Third, areas on the midline are active. Repeating words seems to involve the anterior cingulate cortex. The cingulate cortex seems essential for most tasks that require intense attention, and so may not be uniquely part of linguistic processing. Fourth, both motor activity and word analysis independently produce intense activation of the cerebellum." (Deacon, p. 295).
Wow, quite a mouthful! Deacon suggests that analysis of the working brain provides us a unique perspective into the hierarchical organization of brain processing structures, and that "the classic language areas are not unitary modules, but rather complicated clusters of areas, each with different component functions" (Deacon, p. 297). In other words, our brain attacks the process of coding and decoding with a "divide and conquer" philosophy.
If you are at all interested in the results of the brain-language research Deacon has some very interesting sketches of where certain features and elements of language are processed. Check out page 296, 304, 305, & 308. These are all quite self-explanatory and his captions are good summaries of the chapter content.
The next section focuses on lateralization and cross-hemispheric functioning. One of Deacon's more interesting points is that of lateralization as a result (instead of cause) of brain-language co-evolution. Lateralization has been a big issue for brain researchers, as traditionally language areas of the brain were located in a specific and discrete location, the left hemisphere. What we've discovered recently shows that language processing is not universally located in the left-hemisphere. There has been some research with children who must have large portions of their left-hemisphere removed due to tumors, or traumatic brain injury, and they are still able to acquire language. Deacon also states that "10% of people are not left-lateralized, and others are ambiguously lateralized" (Deacon, p. 310). One thing from Deacon's research review that is for sure, if you're going to suffer a traumatic head injury, try and do it at a very young age. The brain demonstrates high a level of plasticity, the ability to relocate certain functions to other processing areas. I suppose it's too late for most of us, so wear your bicycle helmets!
Deacon also shows how though for most people the left hemisphere is dominant for language functions, the right hemisphere is simultaneously utilized for such language functions as organization and logic tracking. Basically, if you have right hemisphere damage your speech will be OK, but you'll have no sense of humor, because you won't be able to follow a joke, and understand that the punchline is indeed a punchline. The right hemisphere also seems to carry out the processing of the prosodic features of language. Prosodic features are things like rhythm, pitch, stress, and pauses. The "little" things hidden in our speech to help guide and focus listeners.
You can see that there are indeed very important functions of language processed by the right hemisphere, and therefore we can no longer say that "speech" or "language" is located in the left hemisphere. What we can say, is that certain features of language are processed in regions located in the left or right hemisphere. We must now approach brain function as a system of parallel operations occurring across both hemispheres of the brain.
So that was a lot of information to process, but I knew you could all get through it. If anyone has any specific questions, they can email me and the group, and we'll get a little dialogue going (using both hemispheres).