Content uploaded by Till Nikolaus von Heiseler
Author content
All content in this area was uploaded by Till Nikolaus von Heiseler on Dec 17, 2016
Content may be subject to copyright.
Till Nikolaus von Heiseler
Independent Researcher
formatlabor.net@gmail.com
How language evolved as a backchannel between
two feedback loops
Abstract. Language is what makes us human. It is the basis of human knowledge,
culture, and society. Despite its importance, how language evolved is still a mystery.
Various recent studies suggest that humans developed through a “super-fast”
evolutionary process found nowhere else within the animal kingdom. This suggests
a discontinuity in the evolutionary process itself. We propose the following model:
Humans evolved in a unique evolutionary system consisting of two feedback
loops, there being a backchannel between them; the lower loop producing
the variations needed for selection in the upper loop to take place. What is meant
by the “backchannel” here is a structure enabling the selection of the lower loop
to “anticipate” the selection of the upper one. The content of this backchannel
is displaced action encoded in narration. We show that not only the human brain
and language but also most of the unique human faculties (including theory of mind,
episodic memory and the unique human altruism) are adapted almost exclusively to
developing the functioning of the backchannel (narration) at a super-fast evolutionary
pace.
Keywords: biolinguistics; evolutionary tipping point; extended founder effect;
human altruism; human evolution; language evolution; Pullo-Vorenus-Hypothesis.
1. Introduction
Language is what makes us human forming the basis of human knowledge,
culture, and society. Despite its importance, however, the evolution
of language remains largely a mystery, and has been described as being “the
THEORIA ET HISTORIA SCIENTIARUM, VOL. XI
Ed. Nicolaus Copernicus University 2014
http://dx.doi.org/10.12775/ths-2014-008First published online 31.01.2015
130 Till Nikolaus von Heiseler
hardest problem in science today” (Christiansen and Kirby 2003). Ground-
breaking studies in neuroscience, primatology, anthropology, and cognitive
science – based to a large part on new technology (including electronic data
processing, DNA - analysis, neuroimaging technology) – have revealed data
that were never accessible before. These have yet, however, to be brought
together and synthesized into a larger picture. Perhaps the time is ripe now to
develop from existing evidence an entirely new hypothesis, regarding how
language evolved.
The starting point for our hypothesis outlined in this paper is the
idea that systems construct their own elements. We will claim that there
is something special about human evolution (cf. Lahn et al. 2004) and
that a fundamental system change occurred before humans evolved. What
is proposed is a discontinuity at the system level that came before the origin
of our species and we think of humans and language being not the cause but
rather the result of it. Our model thus fuses the question of the evolution
of language with that of the origin of our species. In terms of the hypothesis
presented here – both of these are the effects of a system change – one that
can be referred to as a major transition in evolution (Maynard Smith and
Szathmáry 1995).
The system that evolved basically consists of two feedback loops and
a backchannel between them. The lower circle produces the variations
needed for selection in the upper loop to take place. By “backchannel” we
mean a structure whereby the selection of the lower circle “anticipates”
the selection of the upper circle by implementing a medium that converts
the differences of the upper selection into differences in the selection of the
lower circle. The backchannel develops into an efcient medium in which
language plays an essential role. We assume that most unique human faculties
evolved for – and in favour of – this backchannel and were selected through
competition for effective transmission. This needs to be claried.
2. Background
2.1. Super-Fast Evolution: A sign that there is something special
about human evolution
The development of the walking foot, the changes in the anatomy of the
hands, the S-curved spine, the development of the human brain, the speech
apparatus and cognitive abilities, and the like require so many fundamental
changes that within the current established framework it seems close to
131
How language evolved as a backchannel between two feedback loops
miraculous that all these developments have occurred in such a short
evolutionary period of time. Also, there is genetic evidence revealing
gaps in our current understanding: our ancestors in fact having undergone
fundamental genetic changes within the last 6 million years (Britten 2010;
Hughes et al. 2010; Lahn et al. 2004). The changes found are very specic,
indicating there to have been a strong and sustained selective pressure and
the development of fundamentally new traits. Genomic regions have also
been identied that are conserved in vertebrates in general but in the human
lineage have accumulated substitutions at a markedly accelerated rate (Bird
et al. 2007). The increased rate of substitutions found for the human lineage
suggests that their function may have changed entirely (Burbano et al. 2012;
Bush and Lahn 2008; Pollard et al. 2006; Prabhakar et al. 2006). It thus
becomes clear that the notion that there is only little genetic difference
between humans and chimpanzees (Diamond 1991) is nothing more than
a popular myth. The large phenotypic divergence between humans and
chimpanzees has been driven mainly by changes in gene regulation rather
than by altered protein-coding gene sequences (single base substitution),
by duplication (gene amplication), deletion, exchange between intron
and exon, changes concerning the “frames” (Sibley and Ahlquist 1987).
Genetic changes of all of these types are exponentially more powerful than
single-base mutations are. Such fundamental genetic changes can only
be positively selected if there are possibilities for evolutionary change
suggesting the development of fundamental new traits. As Bruce Lahn has
declared: “To accomplish so much in so little evolutionary time – a few
millions of years – requires a selective process that is perhaps categorically
different from the typical processes of acquiring new biological traits.” Since
Lahn found the pace of evolution here to have been about 16 times as fast as
the development found in New World monkeys, he concluded, “Our study
offers the rst genetic evidence that humans occupy a unique position in the
tree of life” (Lahn et al. 2004). This could be an indication for a discontinuity
in the evolutionary process itself. To understand how discontinuities occur
we introduce two concepts: the concept of major transitions in evolution
(Maynard Smith and Szathmáry 1995) and the concept of the evolutionary
tipping point.
2.2 Major transitions in evolution
A discontinuity in the evolutionary process is required for the shift from
one working evolutionary system to another. These major transitions
in evolution (Maynard Smith and Szathmáry 1995) are highly improbable
132 Till Nikolaus von Heiseler
events of strong impact that change how the evolutionary process works;
their often involving such matters as the unit of selection, how information
is transmitted, how variations emerge (heritability), and the frame
of selection (the selection circumstances and the selector). Examples of such
transformations are the emergence of the cell nucleus and the development
of sexual reproduction. The large-scale acceleration of an evolutionary
process is often an indication of a discontinuity of this sort. In taking account
of this, one can hypothesize that the super-fast evolution of hominini –
starting with the complete redesign of the locomotion system – could be an
effect of a previously overlooked major transition in evolution.
2.3 The Evolutionary Tipping Point
The evolutionary tipping point (ETP) is the point at which a function becomes
positively selected for the rst time. Once this point has been reached, it may
be relatively easy to explain how a certain trait developed further. Since
the ETP is the point at which a function is positively selected for the rst
time, it is evident that a trait which was selected at the tipping point was
not shaped by the selection for that particular function before. The ETP has
in general the following prerequisites: a) the existing organism (constructed
by evolution); b) the environment in question, and c) often though not
always, a certain temporary context. At the tipping point a completely
new function appears, one that generates reproductive advantages for
the organism. The rst positive selection of a new function that appears
at the ETP is not necessarily based on a mutation alone, but can be based
on the whole context in which the new function is positively selected. An
ETP is triggered, therefore, by a fortuitous mixture of a genetic foundation,
of behaviour, of circumstances, and of chance. This is especially crucial for
the problem of the evolution of language, because language use is always
situated in a specic social context.
3. Language – its structure and its unique function
3.1. Linguistic genius and footprints of evolution
It is widely agreed that language is unique to humans (making “human
language” a pleonasm) and that it has a genetic component. To illustrate this
point, Chomsky often gives an example of his granddaughter and her pet.
Although both grew up in an English-speaking environment, only one of them
learned to speak English. This point is also echoed by one of Chomsky’s most
133
How language evolved as a backchannel between two feedback loops
ardent opponents, Tomasello (2003): “Everyone agrees that human beings
can acquire a natural language only because they are biologically prepared
to do so.” The process of language acquisition (at the level of grammar)
is quite distinct from learning by imitation, its instead being “rather like
theory construction” (Chomsky 2010): the infant identies certain sounds as
language and then uses these data as evidence for a syntactic structure that
allows it to generate an innite variety of expressions. As a result, humans
(including young children) can utter sentences they never heard before.
If a trait is complex and has specic features related to some particular
function connected to its complex design, it is likely that the trait was
the target of a selective process (Pinker 2010). Many features of language
ability are so specialized that it seems highly implausible that they evolved
for reasons other than for furthering linguistic abilities. This is especially
true for powerful innate language-learning mechanisms, such as the ability
of neonates to identify patterns of tokens (nding word boundaries) on
the basis of statistical probability of the phonetic structure in uent ongoing
speech (Pelucchi et al. 2009). Neonates prefer language to all other acoustic
stimuli (Shultz and Vouloumanos 2010) and are fascinated by language
more than by anything else (Mehler et al. 2006); even foetuses recognize
speech with sufcient clarity for this to inuence the melody of their rst
cry (Mampe et al. 2009). Children are born with the presumption that
language is structured in words that relate to each other to build propositions
(sentences).
Most of these abilities seem to be so closely related to language that
it would be difcult to imagine that they could have developed for non-
linguistic reasons. In addition, it has been possible with the help of certain
innovative techniques to detect those “footprints of selection” in the human
genome that appear to most likely be connected with language (Przeworski
et al. 2000; Bustamante 2010; Enard et al. 2002). In contrast to apes, which –
if they are laboriously trained – use language for instrumental purposes (such
as fullling their needs) humans “just love to talk” (Corballis 2011: 163).
This is especially true of the innate language faculty suggesting a strong and
lasting selective pressure on linguistic abilities.
3.2. Language did not evolve from animal communication
The common preconception that animal communication is the precursor
of language is one of the major obstacles to understanding the evolution
of language. Language and animal communication are separate and distinct
phenomena that evolved separately and for different evolutionary functions
134 Till Nikolaus von Heiseler
(von Heiseler 2014). The parts of the brain in which the two are coded
also differ. Instead, there is evidence that the mirror neuron system found
in primates is the precursor of certain parts of the language faculty (Arbib
2005). This makes it likely that action reasoning is an important prerequisite
to language. But what is the new exclusive function language evolved for,
a function that animal communication systems do not full, and why was
it put under such strong evolutionary pressure?
3.3. The unique structure of language and its exclusive function
A trait always only evolves so far that it can full its most challenging
function, and in turn to construct a convincing evolutionary story of a trait
we need to ask why and how it’s most complex feature evolved. The most
complex feature of language is syntax. In its simplest form a verb produces
– depending on its valency – slots for different thematic roles (agent,
patient, instrument etc.). The verb displays the relationship between
the diverse thematic roles (marked by inection, by word order, by a pre- or
a postposition; in sign language: by role-taking, by line of sight, direction
of signing the verb, by a generalized preposition etc.). Without the verb there
would be no syntax. Yet what does the verb refer to? It generally represents
an action. Syntax, and consequently language, is thus adapted to describe
action. In turn there is no way to describe a displaced action or an absent
event without use of syntax: a single utterance will always be interpreted as
information about the present: if someone cries out “Fire!” she does not mean
that the Bibliotheca of Alexandria burned down more than two thousand
years ago, but that there is a re here and now. Thus, the displacement of an
action is possible only if an utterance is given a context through other words
within a sentence.
From a cognitive point of view, propositions seem to be more
fundamental than words. We do not simply put words together to build
a sentence, but rst we think of an event that we express in words. Words
can thus be seen as ssion products of propositions. The sentence “An ape
grips the grape”, for example, symbolizes a holistic sensorial experience,
the perception of an action. The distinctions used are not “in the world” but
are the result of categorization that implies a distinction between apes and
non-apes, gripping and actions of all other possible sorts and between grapes
and all other types of objects. Thus we could say: the perception of an action
is decomposed by means of syntactic structure. Yet why should an individual
decompose reality? We can think of at least two reasons: (a) for the purpose
of reasoning (to understand “what is going on”), and/or (b) to tell someone
135
How language evolved as a backchannel between two feedback loops
about something that happened earlier and out of sight – to narrate displaced
actions or events. The marvel of language is that it can describe an innite
number of events through use of a limited number of discrete elements
(lexemes).
Goodall argues for the most important unique function of language being
that of the displacement of action, that is, the communication of events that
are not present, pointing out at the same time that “Chimps […] are unable to
communicate about things that aren’t present” (2010). Corballis (2011: 113–
114), in turn, writes that “grammatical language evolved primarily to enable
us to share episodes. […] Language is exquisitely designed to communicate
who did what to whom, where, when and why.” This suggests that language
is adapted to describe action. We can further infer that the exclusive function
is the communication of an absent action, because the existence of a present
action could be communicated simply by pointing. “Absent” would
mean something that happened in the past and out of sight: storytelling.
Consequently there are good reasons to believe that language evolved for
narration (for more evidence also see: von Heiseler 2014).
There is evidence that primates use a social reasoning system to
understand the actions of others (Rizzolatti and Craighero 2004) and that
they use this knowledge for social strategies, to build alliances, to manipulate
conspecics, and so on (Humphrey 1976; de Waal 1982; Byrne and Whiten
1988; Dunbar 2003). Yet why should an individual tell anyone about
anything? Understanding why it could be a benet to refer to something
that happened earlier and out of sight could be the key to understanding
the origin of language. Why did storytelling play such a prominent role
in our evolution? All that would be needed is that a narrative sentence
is uttered and would give the sender a reproductive advantage. The key
to language evolution is therefore to nd a scenario in which a narrative
sentence would be strongly positively selected. This would mean that the rst
utterance (no matter how primitive) that referred to a past action conferring
an evolutionary advantage on the sender would probably start an escalating
evolutionary process, as a result of which the narrative ability would evolve
even further. Since the understanding of action implies imitation and
internalization (suppression of the physical imitation reex) (Rizzolatti et al.
1996), the main challenge of signing a verb is therefore not the signing itself,
but remembering a past action at the particular moment of language use.
The challenge for the receiver would be to understand a simulation within
the framework of a narration (understanding that signs refer to past actions).
At the same time, the rst narrative utterance that occurs should be a tipping
point in evolution and would need to be connected somehow to a major
136 Till Nikolaus von Heiseler
transition in evolution, a point at which the agent (selector) or the object
(the unit selected), and/or the transmission of information, changes. Yet why
was such an immense selective pressure placed on the narrative abilities?
What could be a context in which symbolizing an absent action would confer
a reproductive advantage?
4. The model
4.1. Agents and objects of selection
We can classify different forms of evolutionary process according to
the agent of selection (selector): natural selection (here dened as a selection
by the environment!) and selection by conspecics. Furthermore, the object
of selection (Mayr 1997) can be either the individual (or, from a more
abstract perspective, the allele of a gene pool), or – and let’s propose (as
a working hypothesis) if the gene pool of each group is closed and group
hostility exists – groups. We can now cross cut these four categories with
the three basic functions: nutrition, defence, and reproduction, and therefore
construct twelve different categories of selection and thus concepts of the
evolutionary processes (Table 1).
Without our consideration of a positive feedback loop and from an
intuitive, evolutionarily naïve perspective, some aspects described in the
second column (groups selected by the environment) – such as cultural skills
to utilize the environment: weapons for hunting – appear promising, because
the success of humans is closely related to cultural transmission and social
achievements including cooperation within groups. This is why it is not
surprising that many researchers use unwittingly concepts of this category,
while on the other hand they ofcially deny – for good reason – naïve group
selection. This does not suggest that what is denoted by the table’s second
column does not exist (that the cultural development of hominini groups
and the collaboration of individuals do provide an advantage to the group),
but that it can only explain why humans maintain their predominance over
other vertebrates (showed e.g. by the fact that humans build and visit zoos,
while other animals are kept in them), but not how their cognitive abilities
and their unique altruism evolved in the rst place. To suggest anything as
the cause of an evolutionary process without explaining how the gene pool
of a population changes its allele frequency is a teleological illusion, that is,
it confuses cause and effect.
137
How language evolved as a backchannel between two feedback loops
Table 1. The 12 selection scenarios: this table categorizes the concepts of selection
that could be used to explain the human evolution: the rst two columns
refer to natural selection (in our denition: the selector is nature or
the environment) and the last two columns recursive selection (selection
of hominini by hominini). In the rst and the third columns the individual
(or its genes) is the object of selection, while in the second and fourth
columns the group is the object of selection. All existing theories of language
evolution can be categorized using this table.
Table 1. The 12 selection scenarios: This table categorizes the concepts of selection that
could be used to explain the human evolution: the first two columns refer to natural
selection (in our definition: the selector is nature or the environment) and the last
two columns recursive selection (selection of hominini by hominini). In the first and
the third columns the individual (or its genes) is the object of selection, while in the
second and fourth columns the group is the object of selection. All existing theories
of language evolution can be categorized using this table.
Without our consideration of a positive feedback loop and from an intuitive,
evolutionarily naïve perspective, some aspects described in the second column (groups
selected by the environment) – such as cultural skills to utilize the environment: weapons
for hunting
–
appears promising, because the success of humans is closely related to
cultural transmission and social achievements including cooperation within groups. This is
why it is not surprising that many researchers use unwittingly concepts of this category,
while on the other hand they official deny – for good reason – naïve group selection. All
this does not suggest that what is denoted by the table’s second column does not exist (that
the cultural development of hominini groups and the collaboration of individuals do
provide an advantage to the group), but that it can only explain why humans maintain their
predominance over other vertebrates (showed e.g. by the fact that humans build and visit
zoos, while other animals are kept in it), but not how their cognitive abilities and their
unique altruism evolved in the first place. To suggest anything as the cause of an
Most monkeys and apes live in groups. They developed most of their
cognitive abilities for social reasons to compete within groups (originally
introduced by Humphrey (1976)), thereby triggering a cognitive arms race
(de Waal 1982). An action reasoning system and a prototype of episodic
memory – both necessary for the capacity for narration – are good
candidates for being just the types of abilities that could have developed
within the framework of this so-called Machiavellian Intelligence (Byrne
and Whiten 1988). As we can see in our table: this already is a recursive
adaptation process, because the selector is not the natural environment, but
the other individuals of the group. However, the difference in the speed of this
development between apes and hominini suggests that the competition for
social intelligence cannot be the exclusive cause of the development.
Faced with the fact of super-fast evolution and the development
of “capabilities that have no parallels in the animal kingdom” (Saxe
138 Till Nikolaus von Heiseler
2013), it seems plausible to assume that an entirely new function emerged.
The pace of the evolutionary development especially suggests a more
powerful positive feedback loop.1 Our table shows: there are more recursive
adaptations possible beside the rank competitions within groups. The two
most promising recursive selective scenarios seem to be: group conicts
(groups selected by groups) and sexual selection (an individual selected
by another individual). Are they detectable in human evolution?
4.2. Recursive Adaptations
(1) War before Civilisation. There is evidence for war before civilization: most
tribe societies engage in warfare, also our closest relatives, the chimpanzees,
wage wars (Goodall 1986). The assumption that our ancestors engaged
in conicts is also well supported by a broad literature (among others: Darwin
1871; Keith 1948; Dart 1953; Bigelow 1969; Wilson 1975; Hamilton 1975;
Van der Dennen 1995; Keeley 1996 and Pinker 2011). The most convincing
evidence, however, shows that the recombination of male and female lines
(the haplogroups of the Y-chromosome and the haplogroups of mtDNA) can
be explained only by bellicose interactions and by the integration of females
of the inferior group into the superior group (see 4.4).
(2) Sexual selection. The basis of (inter)sexual selection2 is that males
and females tend to differ in their level of parental investment. The sex
with higher parental investment will be choosier, while the sex with lower
investment is courting and advertising itself (Trivers 1972). Since usually
the parental investment of the female is higher – beginning with the larger
gamete and the gestation of the offspring – females often choose their mating
partners more carefully than males do (Bateman 1948). Fisher (1930) showed
that sexual selection can start by chance and then escalate.3 In this context
1
A positive feedback loop could be a recursive adaptation process. This would occur if
the agent and the object of selection – the selector and the selected – were the same: if the se-
lective environment of a species were themselves. Under these circumstances, super-fast evo-
lution seems possible. Because they produce their own selective pressure, with every deve-
lopment whatsoever the challenges grow, and the selective pressure tends to stay on the same
level – depending on the asymmetry of reproduction of the individual within a group.
2 There are two kinds of sexual selection: intrasexual and intersexual. In intrasexual
selection individuals of the same sex compete conictual for access to members of the op-
posite sex. In intersexual selection, one individual is chosen by a sexual partner. We will use
the term “sexual selection” only to refer to intersexual selection.
3 If in a given population a female preference for a certain quality of a male emerges
through mutation, then the male with this attribute will gain a reproductive advantage with
that female, while not reducing his attractiveness to other females. This male’s average mat-
139
How language evolved as a backchannel between two feedback loops
traits can evolve that could not be selected “by nature” (by the natural
environment). A strong dimorphism (differences between the male and
female phenotype) is not a general sign of female choice – as is sometimes
misunderstood,4 but only of a particular type of sexual selection sometimes
referred to as handicap principle (Zahavi 1975). The handicap works as
a costly and therefore true sign of tness: the male shows with a costly signal
(e.g. with a big train that male blue peafowl develop), that he is t enough to
survive despite the handicap. The key to the success of this mating system
is that males will inherit the handicap along with the tness to survive with it,
thereby ensuring them good mating chances, while the females will generally
not inherit the handicap but will receive only the genes proven to be t (by
the handicap) from their (handicapped) fathers. Since the handicap will grow
only to the point with the optimal cost-benet-ratio it will nd its equilibrium
depending on the costs of the handicap, the nature of the female preference
and the asymmetry of the male reproductive success. The costs are paid to
the environment and the benets depend on the female choice. The selective
pressure through female choice is much more effective and will lead to
a faster evolutionary process than any selection by the natural environment
and even for social intelligence. The crucial point concerning human
evolution is that theoretically anything can be the target of sexual selection.
The only prerequisite is the perceptibility of the trait for the choosing female.
Neither group conicts nor sexual selection alone can explain human
evolution. Sexual selection is indeed a very powerful mechanism, but as
Fisher (1930) explained: anything could happen, and its outcome is entirely
unpredictable (Miller 2000). In contrast the selection between groups could
guide the development in a certain direction (towards bellicose competence),
ings with all females plus his exclusive matings with the females who found his appearance
desirable will sometimes lead to the favoured quality spreading through the population, be-
cause the offspring of the male with the favoured quality will not only inherit the trait from its
father but also the female preference for it from its mother. Thus the genetic basis for the fe-
male preference is spread piggyback with the desired quality in the evolutionary process.
4 For example: Deacon (2010) writes: “sexual selection inevitably produces comple-
mentary divergence of male and female traits” and takes this as a reason to doubt the relevance
of female choice for the development of the most distinctive of human traits as language and
other cognitive abilities: “Therefore, accounting for the extravagant complexity of language
in terms of sexual selection requires explaining why it lacks this otherwise-ubiquitous mark
of extreme sexual dimorphism.” Deacon confuses sexual selection in general with a special
type of sexual selection, namely the handicap principle. Only the latter will produce a strong
dimorphism. This dimorphism develops in two steps: rst the trait (handicap) evolved due
to the sexual selection and then this trait gets suppressed in females by natural selection on
the level of gene regulation.
140 Till Nikolaus von Heiseler
but would not be very powerful, because here an evolutionary mechanism
is missing on the gene level. Furthermore there would be no motive to engage
in the bellicose interactions if they were risky. Likewise the unique cognitive
ability, especially language, could not be selected by group conicts. But
what would happen if both mechanisms would intertwine?
4.3. The ideal(ized) scenario
If the two recursive processes – intersexual selection and the selection
between groups – interlock, then a dynamic system emerges in which
sexual selection creates random variations on a group level in a runaway
process, and these variations then get selected through bellicose interaction.
Both sexual selection and bellicose interactions can select traits that
the environment cannot. To kick off, this doubly recursive process requires
the closing of the gene pool (no casual intermingling between groups).
The transformation of the gene pool proceeds now at the speed of sexual
selection, a speed more similar to that of articial selection than of natural
selection. But since the groups (only if their gene pools are closed) as a whole
are objects of selection, cultural elements and group structures can also be
selected by and for bellicose interaction. Both the speed of the development
and the kind of traits that can be developed change fundamentally. All this
allows us to categorize the emergence of this unique evolutionary scenario
as a major transition in evolution (Maynard Smith and Szathmáry 1995),
alongside, for example, the emergence of sexual reproduction or even
the genesis of eukaryotes.
4.3.1. The Extended Founder Effect
The evolutionary process depends on the production of variations between
the selected entities. If groups are the target of selection, then the speed
of the process depends on the ability of the system to produce differences on
the group level. The speed of development depends on two factors: the sexual
selection as a runaway process and the separation of new founder groups
from a source group. The latter is based on what we call the extended founder
effect (introduced in this paper). The extended founder effect is the hypothesis
that in the hominini line the splitting-off of new founder groups from a source
group includes – beside what is known as the founder effect5 – also non-
5 The founder effect occurs when a small number of migrants split off from a larger
population. The new founder group always has less variation than the source group – because
141
How language evolved as a backchannel between two feedback loops
random elements: related individuals of the same sex could join the same
group, based on social bonding or kinship in the source group, assortative
mating – the choice of mating partners with similar phenotypes (MacDougall
and Montgomerie 2003) – could also play a crucial role. The strongest impact,
however, would occur when the formation of new founder groups of hominini
is inuenced by sexual attraction. In other words, the female preference and
male traits must t. If a subgroup of males with strong social bonds were to
separate from a bigger group, then the females who were attracted to them
would follow them – or vice versa. And since the reproductive success of any
trait depends on the female choice for that trait, the most important factor
of group formation is the nature of female preference. In other words: the
extended founder effect will isolate different female preferences with the
effect that every isolated group is dominated by a slightly different female
choice. Most of the features facilitating the formation of non-random founder
groups will increase in line with growing cognitive abilities. This is to say:
the expanded founder effect increases during the evolutionary process. Up to
now we observed our scenario on the system level the groups being the
variations and the object of selection. But how could such a scenario be
possible on the level of the individual and evolutionary stable?
4.3.2. Bravery as a Handicap
If different female preferences prevail in different groups, the object
of selection in intergroup conicts is – indirectly – the predominant female
choice. This is a two-step-process: rst a preference develops by chance
changing the allele frequency of the gene pool of a group and then the groups
compete for territories. The selection on the group level is a selection between
blind and contingent variants. In this process those female choices will survive
that will produce groups that replace others. For this system (variations
produced by sexual selection and the extended founder effect that get selected
by group conicts) to function effectively, the groups must regularly engage
in group conicts. Since participation in such conicts requires the risk of self-
sacrice for the good of the group, this leads us to the problem of altruistic
behaviour. If there were an imaginary group of the brave, the individuals
of the group would not use an evolutionarily stable strategy (ESS), because
it carries only a cut out (a random section) of the original gene pool of the source group (James
1970) – and can, by chance, be distinctively different from the parent population from which
it derived. In smaller groups, genetic drift can also play a signicant role (Mayr 1942).
142 Till Nikolaus von Heiseler
an individual member of it who was a coward would manage to benet from
the replacement of other groups without risking his own life.
The super-fast evolutionary process of hominini legitimates
the construction of an ideal scenario that would make this extraordinary
development possible. In a second step we will ask how this idealized
scenario is possible. The ideal scenario would be that the differences
in males’ victory related behaviour in the bellicose interaction are translated
into differences in reproduction. Here we face a classic mapping problem:
we have differences on one side (difference in the war-like behaviour) and
differences on the other (difference in reproduction inuenced by female
choice), and we need a medium for the transmission of these differences
over time and space. This is to say: behavior in war-like interactions needs
to be translated into mating frequency: the females would need to choose
their mating partner according to their perception of absent actions. If this
system works, bravery would be a “handicap” (an expensive and therefore
true sign of tness (Zahavi 1975)), and would play the similar role as
the train of a peacock. As mentioned earlier, one requirement of a handicap
is its perceptibility. In this case the “perception” of the male behavior
in the war-like interactions is not observed directly by females but only
through a channel that transmits differences of the male behaviour in the
bellicose interactions to the female consciousness inuencing their choice.
Also the channel needs to overcome space and time. Space: the bellicose
interactions will be in most cases not in a visible distance to the breeding
area. Time: the bellicose interaction will not be followed by mating instantly.
A good medium for the transmission of male behaviour in bellicose
interaction would be narration. In other words, females would need to love
war heroes as they appear in narrations.
On the system level this would equate with a backchannel. Before the back
channel is implemented the sexual selection and the extended founder effect
would produce variants on the group level that are blind for the selection
in bellicose interactions. Now by the emergence of the back channel
the selection of the lower level (female choice) can somehow orientate itself
towards the selection of the higher level (the bellicose interactions). This
makes an altruistic behavior in form of bravery evolutionary stable. In other
words: a behavior that appears to be altruistic becomes the best strategy to
spread the genes of the individual. The size of the handicap (the optimal
degree of braveness) would depend on the variation in male reproductive
success and the value placed on the handicap of bravery in the female choice.
The groups therefore compete in the variation of male reproductive success
and in the “clearness” of the backchannel.
143
How language evolved as a backchannel between two feedback loops
Since one’s reputation is about one’s exploits of war, the social structure
is indirectly (through the production of bravery as a handicap as transmitted
by the medium of narration) oriented towards the outer challenges of the
group. This is to say: because the female choice is the most important factor
of the development of a group the changing of the gene frequency can
orientate indirectly on the outer challenge.
In the ideal situation, the individual providing the greatest benets for
the superiority of the group would reproduce most often. This ideal scenario
obviously never can be achieved because the contribution to a victory of an
individual will not always be perceptible and the medium of narration always
includes interference (noise). The perfect channel, however, would seamlessly
connect the behaviour relevant to ensuring victory and reproductive success.
In reality even the orientation to heroic deeds could be quite “noisy”
(corrupt), because the victory-relevant behaviour is not always visible for
anyone – and even if it is, certain crucial qualities can hardly be successfully
transmitted by narration. However, the major noise source – decoupling
successful war behaviour and reproduction – would lie in the narration itself:
if males could make females believe that they are braver than they actually
are, it would give them a considerable advantage, because they would not
need to put themselves as much in danger in order to reproduce equally.
Every male will try to nd a way to “cheat” and thereby corrupt the system:
not in a conscious way, but through the evolutionary logic of variation and
selection on the individual level. If a male could propagate his genes without
constantly risking his life, this strategy would spread through the population;
therefore, self-propaganda would be one of the biggest interference sources
in the backchannel (relating bravery in war and reproduction), as it could
minimize the costs of bellicose bravery and increase the benets from female
choice, thereby decoupling reproduction from victory-critical behaviour.
This form of corruption could escalate because females that would choose
linguistic genius would probably give birth to great talkers, which will be
disproportionately successful.
All this leads to a dialectical development: the language ability would be
on one hand a show-off behaviour that would disconnect reproduction from
bravery – and therefore be the main interference in the backchannel – while
on the other hand, in the long run, the selection of the great talker would
improve the resolution capacity of the channel (how precise a narration
about the bellicose interactions can give a picture of what really happened).
Paradoxically, the channel in this autopoietic system improves itself by its
own noise.
144 Till Nikolaus von Heiseler
The aggressiveness of a group does not reect the variation in male
reproductive success in general, but only the part of the female choice (which
produces the variation of the male reproductive success) that is guided
by the attraction level to bravery. Because the optimal quantity of bravery
– as a real risk taken in the bellicose interactions – becomes minimized
by the female choice for the great braggadocio, the level of aggressiveness
of groups in which this female preference rules will decline, and the groups
in which the females care more about the great talkers will avoid other more
aggressive groups and may settle in less attractive habitats or even immigrate
to uninhabited regions.
Figure 1. The feedback loop structure within a single group. There are two
feedback loops both depending on the nature of aspects of the female
choice: A = positive selection of the bellicose performance (outer loop);
B = positively selection of narrative performance (inner loop). (A)
The outer loop includes two selective mechanisms. (A1) The female
choice selects the hero how he appears in the medium narration and
thereby selects the bellicose talent and the bravery positively. (A2) In the
bellicose interactions the talent is selected positively while the bravery
145
How language evolved as a backchannel between two feedback loops
is selected negatively. This complex outer loop results in: (1) a regulation
of the degree of bravery and (2) a positive selection of the bellicose skills.
(B) In the inner loop the male narrative performance (based on the narrative
competence) is selected by the female choice. This is a positive loop
limited by the positive variations on the gene level. The outer loop (A) and
inner loop interact in two ways: on one hand both loops compete, because
the female choice for heroic behavior (as transmitted by narrations)
and narrative performance compete; on the other hand the result of the
escalation of the inner loop (narrative competence) is the basis of the media
(narration) of the outer loop. Every improvement of narrative competence
improves the media of narration and thereby the transformation of victory-
relevant behavior in the bellicose interaction into reproduction. Because
the inner loop selecting the narrative competence is positive, while
the outer loop (positive selection of bellicose skills, regulation of bravery)
includes a negative selection (bravery in the conicts between groups),
the system itself shifts towards the inner circle. As a result the females
will be less and less impressed by bellicose skills and more and more
aroused by narrative performance. However, in bellicose interactions
the groups compete in the strength of the outer loop. For this four
aspects play a part: (1) the asymmetry of the male reproductive success;
(2) the quality of the backchannel primarily premised on the narrative
skills; (3) the extent to which the female choice is inuenced by heroic
deeds as they appear in narrations in contrast to other preferences
(including narrative performance) and (4) the danger of bellicose
interactions depending on the bellicose skills, the culture of a group and
relative strength of the competing groups. The strength of the outer loop
conforms to the averaging (and evolutionary stable) level of bravery and
therefore corresponds with the aggressiveness of the group. Time scale:
the outer feedback loop is a regulator that can adapt to the optimal level
of bravery and aggressiveness of the group in a few generations. The fastest
chance concerns the danger of the bellicose interactions depending on
the strength of other groups. An expanding superior group will therefore
turn more and more aggressive. The calibration of the bravery level can
also include personal experience and epigenetic effects. Another result
of this system is the escalation of bellicose culture, including weapons,
tactic of ambushes, war paint, techniques of signal transmission, methods
of synchronisation, practises of motivation (such as war dance), formation
of combat units etc. The bellicose skills adapt to the bellicose culture.
Furthermore the narrative competence escalates in this system and thereby
many heterogeneous sub-skills. This escalation is limited by the positive
mutations concerning the narrative competence.
146 Till Nikolaus von Heiseler
In isolation the system tends to the inner circle because a good narrator
does not need to put his live in jeopardy. Because the great narrators will
be more successful than the great heroes, the female choice for the great
narrators will be more successful as well. This will reduce the optimal level
of bravery. In contrast: in bellicose interactions the groups compete in the
strength of the outer circle. Groups governed by female choices that prefer
great talkers would be replaced by groups ruled by female choices that prefer
the great war hero. This makes it likely that victorious groups by chance
develop a culture that blocks a fast decrease of bravery through female
preference for good and well performed narration.
4.4. Problem of the reduction of variability
and the integration of females
If a dominant group were to displace all other groups and spread throughout
all habitats, genetic variability would be extremely reduced: the variability
of all populations of all habitats would be reduced to the original
variability within the dominant group. Since evolution always depends
on genetic variability, the victory of a small group over all others would
limit the possibility of further development. Moreover, it is likely that
in all groups some benecial mutations could emerge; all of which would
be destroyed by total replacement. In fact, there seems to be a mechanism
that not only has a distinct advantage for overall development (selected
in competition with developments in other regions) but which also pays
off at the level of the individual (or its genes). This is to integrate healthy,
young, attractive females from the losing group into the victorious group.
The advantage on the genetic level is on the side of the males of the
victorious group. For the overall process, however, it is important that all
mutations that can be the object of later sexual selection be retained, even if
they develop in a losing group. As the female preference may simultaneously
be for different characteristics in different places, it would be possible for
different qualities to develop in different groups, which then get recombined
by bellicose interactions and the appropriation of females from the losing
into the winning group. The only prerequisite for the survival of these new
qualities introduced by the females of the losing group is for these attributes to
be a target of positive selection by female choice (either this preference could
be already found in the predominant group – say for narrative abilities – or
it could be introduced by the integrated females). Because the narrative skills
do not produce as many costs as the brave bellicose performance, the genes
concerning the narrative abilities will spread through the dominant group.
147
How language evolved as a backchannel between two feedback loops
This recombination is particularly important in terms of the development
of complex traits. This is especially true for those concerning the formation
of narrative ability, since this includes various skills on many different levels
that can be selected by female choice (see 4.5). Every aspect that produces
perceptual differences can be chosen by sexual selection, including linguistic
complexity, and narrative clarity (intelligibility of the narration). In the next
generation, the backchannel therefore could be improved and differences
between groups concerning the clearness of the backchannel could get
selected on the group level.
abilities – or it could be introduced by the integrated females). Because the narrative skills
do not produce as many costs as the brave bellicose performance, the genes concerning the
narrative abilities will spread through the dominant group. This recombination is
particularly important in terms of the development of complex traits. This is especially true
for those concerning the formation of narrative ability, since this includes various skills on
many different levels that can be selected by female choice (see 4.5) Every aspect that
produces perceptual differences can be chosen by sexual selection, including linguistic
complexity, and narrative clarity (intelligibility of the narration). In the next generation, the
backchannel therefore could be improved and differences between groups concerning the
clearness of the backchannel could get selected on the group level.
There is evidence for the recombination of male and female lines that lie in our
genes. The different structure of the Haplogroups of the Y-Chromosome and the
Haplogroups of mtDNA can only be explained by bellicose interactions and by the
integration of females of the inferior group into the superior group.
4.5 The Speed of Evolution and the development of the unique human traits
The speed of evolution depends on the production of variation, on the probability of the
variation being an improvement, on the number of fields in which improvements are
possible, on evolutionary pressure, and on the number of individuals involved in the
evolutionary process. Bigger changes will be based on mutations of the gene expression
and include insertions and deletions (Britten 2010). But those bigger changes can be
positively selected only if there is something to improve. Most traits are already optimized
in a long evolutionary history. Furthermore, the problem with most traits is that any change
Figure 2. Simplified tree of the
female blood line. Black =
African Lines
Figure 3. Simplified tree of the male
blood line. Black = African
Lines
Figure 2. Simplied tree of the
female blood line. Black
= African Lines
Figure 3. Simplied tree of the male
blood line. Black = African
Lines
There is evidence for the recombination of male and female lines
that lie in our genes. The different structure of the Haplogroups of the
Y-Chromosome and the Haplogroups of mtDNA can only be explained
by bellicose interactions and by the integration of females of the inferior
group into the superior group.
4.5. The Speed of Evolution and the development
of the unique human traits
The speed of evolution depends on the production of variation, on
the probability of the variation being an improvement, on the number
of elds in which improvements are possible, on evolutionary pressure, and
on the number of individuals involved in the evolutionary process. Bigger
changes will be based on mutations of the gene expression and include
insertions and deletions (Britten 2010). But those bigger changes can be
positively selected only if there is something to improve. Most traits are
148 Till Nikolaus von Heiseler
already optimized in a long evolutionary history. Furthermore, the problem
with most traits is that any change can be destructive. Narrative abilities,
however, can improve on many levels simultaneously and if a certain
quality is improved in the next generation, the demand will also improve,
while the selective pressure will be constant, depending on the variation
of the male reproductive success. There are different elds that could be
positively selected if the narrative performance would confer a reproductive
advantage for such heterogenic qualities as: vocal characteristics, humor,
theory of mind, episodic memory and lexicon size could develop. In other
words: narrative skills are comprising many widely different sub skills
and all of them could be tested by narrative performances. It is likely that
the female choice will be oriented around the qualities that produce signicant
perceptible differences (based on the female’s own cognitive abilities). First
females will only be interested in males that full the minimum standard
in all perceptible elds (size, symmetry, signals of health, beauty of the gait,
status, smell, etc.), than they focus on their preferred eld, but would also
appreciate signicant improvements (differences to other males) in all other
elds. In the case of hominini, positive mutations can be combined through
the back ow through the female germline into the surviving group.
Alongside episodic memory, having a theory of mind is an important
prerequisite for being able to tell a complex story. First, a good storyteller
always needs to keep two things in mind: the whole story (his own knowledge
of the narration) and what he has related of it so far (the knowledge of the
receiver of the story). Second, to understand a story, both the narrator and
the receiver need to understand the motivations and beliefs of the hero
(because to understand a story means to understand the motives of the hero).
The receiver needs to assume that the hero acts according to his knowledge
and not according to objective facts. However, the narrator has to make sure
not only that he himself understands the motives of the hero, but also that
the receiver understands them at every given point in the story. Thus, to
understand whether the receiver grasps the story, the narrator needs a second-
order theory of mind (i.e., he needs to understand what the receiver believes
the hero believes). From this we can conclude that the theory of mind,
the ability to understand the beliefs and motivations of others, will be put
in the adaptation process to narration under a strong selective pressure.
149
How language evolved as a backchannel between two feedback loops
4.6. Evolution and cultural development
In our scenario, language is neither simply adapted to the brain, nor
the brain simply to language, instead, both language and brain adapt to
storytelling. Some of the abilities will be useful mostly for language (e.g.
specialized learning algorithms, phonology, syntax, lexicon) others can be
used for other social and cultural aspects (e.g. episodic memory, theory
of mind, prosody).
After the evolutionary tipping point, – the rst narrative utterance
that conferred a reproductive advantage – human brains and language
are both selected for the beauty of narrations (beauty is here dened as
the attractiveness of a narration: the qualities that make the narration
pleasant and the narrator attractive). Every adaptation of the brain and its
cognitive capacities changes the selective pressure on language (as a cultural
entity), and every development of language and its use slightly alters
the selective pressure on the cognitive capacities and the brain. The human
brain structure can be explained on one hand by older features, which
developed for other reasons (in the long evolutionary history of animals,
vertebrates and apes) and on the other hand by the adaptation to narration,
structures specically developed in our scenario. Likewise, the structure
of narration is based on cognitive capacities and in particular the language
abilities, and the cultural transmission including lexicon, syntax, narrative
conventions and so on. This is to say, the springboard of this recursive
evolution is the pre-existing structures of both brain and language (as part
of the culture in a certain group), but will be subdued under the selection
for the beauty of narration.
5. Discussion
5.1. Selection on two levels and the closing of the gene pool
Sometimes it is said, that if there are two levels of selection the lower level
(e.g. competition between individuals) would be stronger while the higher
level (e.g. competition between groups) would be much weaker. But this
is misleading because “strong” and “weak” would presuppose that both
would work on the same level. A complex trait can never be the result
of the group level and the selections are not comparable on a one-
dimensional scale. For example: it is likely that the Homo sapiens displaced
the Neanderthals in Europe about 35.000 years ago, because Homo sapiens
developed a mimetic culture. But it would be impossible that the Homo
150 Till Nikolaus von Heiseler
sapiens developed mimetic culture in order to displace the Neanderthals.
That we populated the planet and not Neanderthals could be therefore caused
by a selection between species – what makes it a big difference (at least for
us). Such an effect can hardly be called “weak”. But “weak” would be also
wrong in another sense: no complex trait can ever be formed by any selection
of the group level – no matter how slow.
These two levels of selection are generally found on the level of the
individuals (or its genes) and on the level of species that compete for the same
niche. However, in this competition not only the individual tness plays a role,
but also the social structure, the mating system, the level of collaboration etc.
The level of collaboration can never be a direct result of the selection between
species, but if the replacement process would happen often, it is likely that
the surviving species has a benecial social structure. In the regular evolution
the selection on the higher level (on the level of species) is very rare and
could only happen in some millions of years. If the individual of a species
– as suggested by our model – would not interbreed arbitrarily with other
individuals of the same species within a habitat, but would build groups that
would interbreed only exceptionally, the pace of the selection on the higher
level would increase enormously: every split off of a new founder group
would provide the selection on the higher level with a new variation.
This would implicate that the proof that hominini in a habitat are
not interbreeding arbitrarily (but are building groups to do so), would be
indispensable for our model to work. The evidence for this special social
structure of our ancestors comes from different elds: (1) the comparison
of the structure of the human male germline (Y-Chromosome line) and
the female germline (mtDNA-line) shows that groups are not intermixing
with other tribes arbitrarily and makes it seem likely that groups displace
other groups by killing the males and integrating some of the females (see
4.4). (2) The different tribes on the Andaman Islands do not interbreed for
thousands of years and keep their very diverse appearances. The same is true
for the tribes in New Guinea (Diamond 1991).
5.2. How could the transition in evolution emerge–leading
to the development of humans?
The starting point of our scenario would be the closing of the gene pool.
There is evidence that this change into a new mating system was triggered
by a climatic change. While in western and central Africa the tropical
rainforest remained, volcanic activities led to the development of a natural
barrier, as a result of which the region east of the Great Rift Valley dried
151
How language evolved as a backchannel between two feedback loops
out continuously (Coppens 2004). In an open and clear territory consisting
of isolated gallery forests surrounded by open grass elds – such as developed
increasingly due to climatic shifts – having intercourse with individuals from
another population without being discovered appears practically impossible,
especially because the territories in the savannah are much more expansive
than in the rainforest (e.g. while a group of chimpanzees in the forest has
a home range from 5 to 40 square kilometres, their territories in the savannahs
are 120 to 560 square kilometres). If now in one population the males guard
their territory, kill all males that enter it to mate with their females – and
at the same time mate with females of other groups – this behaviour will
spread through all groups through the male germline or by the replacement
of the groups with other strategies. The system that emerges here has two
positive feedback circles but no backchannel between them: the selection
on the lower level is blind to the selection on the higher level. The lower
level (sexual selection) can only indirectly – by trial and error – adapt to
the selection of the higher level (group conict). On this stage genes that
help males survive in bellicose interactions will spread through every single
surviving group. However, female preference cannot identify the good
warrior directly.
On a more abstract level, we could say that the target of selection is the
predominant female choice of a group (selected in bellicose interactions).
A good female choice would be enhanced bipedalism to free the hands for
weapons (e.g. rough stones to strike and to throw, clubs and spears from
perishable material) and to pick the good hunter. If the female choice
fortuitously favours the upright gait or position as part of the courtship
behaviour (e.g. the male showing its genital), the upright gait would evolve
with maximal speed. This process could be reinforced by the use of sticks
in status competitions within groups. One group making critical progress
would replace all others. With this replacement the female preference for
the upright gait would spread, and therefore improvements concerning
upright locomotion could happen in different groups simultaneously.
The best way of choosing the good hunter would probably be to nd males
attractive that present hunting trophies. If now the females would orientate
on the distinctive signals between males they would choose rare hunting
trophies. This would make the animals to hunt bigger, harder to hunt and more
dangerous. This is to say: because female choice is successful when it detects
distinguishing (distinction creating) qualities the hunted animal grows bigger
and more dangerous in the evolutionary process until the most dangerous
animal would be the hominini himself. This all would be possible without
any representation or any understanding that a war or hunting trophy signies
152 Till Nikolaus von Heiseler
an absent action and could be explained just by the logic of the system itself.
This already would be a back channel on the system level, even if it would
not include any consciousness of an absent deed. The consciousness of the
displaced action is therefore a result (and not the basis) of the back channel.
The advantage of understanding a trophy as a sign lies in the improving
of the detection of tness and is more exible than the instinctive preference
for a rare trophy. Now, what would be the most effective technique to present
a trophy? A good way would be, to present it in an engaging way, to make some
noise to attract attention and maybe repeat a movement showing the killing.
This “showing the killing” would be a mimetic gesture, a sign that would
refer to a displaced action. This would make it a narrative proposition. If
this narrative proposition were positively selected, this would be the tipping
point for the evolution of language and the starting point for a selection for
narration. The simplest possible narration (SPN) could be:
>I< >kill[ed]< >[this] enemy< >[with a] stone<
This miniature narration (SPN) is, we imagine, only one gesture
(likely to be repeated) added to the presentation of the war trophy, but
the crucial point is that this gesture would signify a past action. In this
rst signed proposition there is no chronology, because the signing only
contains one movement. The agent would be an implication (the narrator
himself); the verb would be a mimetic movement (signifying a past
deed); the instrument the real weapon and the patient itself: the dead
body (the signier) stands for the living enemy or animal (the signied).
If this simplest narration (SPN) is understood and creates a reproductive
advantage, the narrative abilities would get into a positive feedback loop
and escalate. Because the index, in the form of the trophy, needs to be
always present, there is no danger of lying.
Since the understanding of action implies imitation and internalization
– suppression of the physical imitation reex (Rizzolatti et al. 1996) –
the main challenge of signing a verb is not the signing itself, but remembering
a past action at the particular moment of narration. Thus the major cognitive
challenge would not be the imitation itself, but the reference to the past event.
To become narration the action must be stored as a memory, which must then
be accessed in a narrative situation. On this stage the remembering of the
past event is always triggered by the presenting of the trophy. However, an
even bigger cognitive challenge lies on the side of the receiver and requires
the understanding that the mimetic gesture is representing a past action. With
the rst narrative utterance that had a reproductive advantage to the sender
153
How language evolved as a backchannel between two feedback loops
the existing abilities are put under a selective pressure for narration. After
this ETP, the escalation of the system includes the development of all
cognitive features that are necessary for a good narration. Another critical
point in this development would be the point on which the narrations would
circulate within a population. This can occur only if someone else retells
someone’s narration. For this to happen there must be a motive to tell
a story about someone else. The motive is obvious in the case of rst person
narrations, since they valorise the narrator. But why should anyone tell
a story of anyone else? The only reason to tell a story of someone else would
be if this behaviour would generate a reproductive advantage. This could
happen when females would be fascinated by good narration and the narrative
skills itself. Therefore the question would be: Why is it a superior female
choice (and therefore an evolutionary stable strategy) to choose the male
with impressive narrative abilities? A narrative genius would gain a greater
reproductive success with the same bellicose performance. This advantage
would be massive because while bravery in the bellicose interactions could
be costly, the narration is not (but limited by the narrative competence).
When narrations are circulating, the burden of proving truth switches from
the indication of the trophy to the controlling of a narration through other
narrations. In other words, in the I-narration world, a narration can only be
told if the narrator can present a trophy. This makes bellicose interactions
a competition to acquire war trophies and also makes later conicts about
the possession of the war trophy probable. The change to a system with
circulating narrations, in which every narration is controlled by other
narrations, suggests a more cooperative strategy. When narrations are
the exclusive backchannel, the female choice is attracted to great narrators
and to the hero as he appears in the circulating narration (as females today
are turned on by heroic deeds and entertaining narrative performances).
Furthermore not only does the genetic basis of the language ability evolve,
but language itself adapts to its narrative function, which in turn gives rise
to new challenges for the narrator. The cultural aspects of language develop
due to their adaptation to the narrative conventions of a special culture and
thereby modify slightly the selective pressure (also depending on the female
preference concerning narrations). Thus the challenges could vary between
different groups and could therefore produce different adaptations that can
later be combined through the female germline. If the female choice would
be orientated on distinctive linguistic qualities, it could trigger an escalation
of complexity even beyond communicative efciency. From this we can
assume a super-fast evolutionary pace of both the narrative ability and
narrative conventions on different levels including syntax.
154 Till Nikolaus von Heiseler
5.3. Methodological reection
For heuristic reasons we rst developed an ideal scenario. That does not mean
we claim humans to have evolved in a perfect or God-given evolution, but
that our scenario is simplied and idealized. The method is rst to develop
a scenario that would solve the given problems (such as the development
of the unique human abilities in a super-fast evolution) and then to examine
how it could possibly be implemented – presupposing it would work much
messier in reality than in our concept. The background of this simplication
is abstract cybernetic modelling (see Wiener 1948). This is to say that
in the modelling of a self-regulating system (Foerster 1981) we would
rst search for possible feedbacks, analyse the logic of transmissions
of signals dictated by the nature of the media, etc. (see von Heiseler 2008).
The benet of this perspective is that it can identify differences that would
be unrecognizable in direct observation. Two examples: (1) at a certain state
in the development of our ancestors our scenario predicts a male acquiring
a reproductive success by presenting a trophy. This could be a gazelle,
a fang of a hippopotamus or a hand of hominini from another group. For
a direct observation there is no big difference between all three trophies.
For a cybernetic modelling there would be a vast difference between
a hunting trophy and a body part of a hominini, because the war trophies
would create an endless positive feedback (as an arms race) and would
implement a backchannel between female choice and intergroup conicts.
(2) If there would be a direct observer of our early ancestors she probably
would nd that a main inuence on the reproductive success would be
based on inner group competition, alliances, status etc. and that even
the female choice would mainly be affected by the status of an individual.
But once in a while the female choice is inuenced by the presenting of war
trophies and narrative abilities accompanying the presentation. Here again
we detect a positive feedback and realize that this kind of female choice
would be the foundation of an orientation of the social system of a group
as a whole on an outer challenge (the intergroup competition): The lower
feedback circle (the female choice) gets informed by the higher feedback
circle (group competition). As with all positive feedback there will be an
escalation. Thus the cybernetic modelling can predict the development
in this case much better than direct observation.
155
How language evolved as a backchannel between two feedback loops
5.4. The consequences
With our scenario we can explain the development of most unique
human abilities, such as theory of mind, episodic memory and language:
everything that could be selected by selecting a good narrative
performance. Furthermore the circulating narrations could be the basis
of the unique human altruism: if reputation is based on the circulating
narrations, behaviour that could be observed by any possible narrator will
be inuenced by the anticipation of the narration about the behaviour.
In other words: Acts will be shaped by possible narrations about them.
Individuals transform into society members. Every act is controlled by the
imagination of one’s reputation. The consequence: a reputation-economy
based on narration is the basis of the unique human social order.
For further research it could be productive to investigate the importance
of narrations for our lives and the cultural development of narration.
The essence of an object often relies on its origin (such as the difference
between an original piece of art and forgery). This means it is constructed
by a narration. Myths give people an identity, our self-concept is based on
our autobiographic narration, ctive and religious narration can give us role
models and our reputation is encoded in gossip and other forms of circulating
narrations in different media. Everything that is essentially meaningful to us
is connected to narrations.
References
Arbib, M. A. (2005). From money-like action recognition to human language:
An evolutionary framework for neurolinguistics. Behavioral and Brain
Sciences 28: 105–167.
Bateman, A. J. (1948). Intra-sexual selection in Drosophila. The Journal of Heredity
2: 349–368.
Bird, C. P. et al. (2007). Fast-evolving noncoding sequences in the human genome.
Genome Biology.
Britten, R. J. (2010). Transposable element insertions have strongly affected human
evolution. Biological Sciences - Evolution 107: 19945–19948.
Burbano, H. A. Green, R. E. and Pääbo, S. (2012). Analysis of Human Accelerated
DNA Regions Using Archaic Hominin Genomes. PLoS One.
Bush, E. C. and Lahn, B. (2008). A genome-wide screen for noncoding elements
important in primate evolution.” BMC Evol Biol. 8:17.
Bustamante, C. (2010). Genomic footprints of natural selection. Proc Natl Acad Sci
USA.
156 Till Nikolaus von Heiseler
Byrne, R. W. and Whiten, A. (1988). Machiavellian intelligence: Social expertise
and the evolution of intellect in monkeys, apes and humans. Cambridge:
Cambridge University Press.
Chomsky, N. (2010). Some simple evo devo theses. In The evolution of human
language, 45–62. Cambridge: Cambridge University Press.
Christiansen, M. H. and Kirby, S. (2003). Language Evolution: The Hardest Problem
in Science? In Language Evolution. s.l.: s.n.
Clark, A. (2006). Language, embodiment, and the cognitive niche. Trends
in Cognitive Sciences.
Coppens, I. (2004). Geotektonik, Klima und der Ursprung des Menschen. Spektrum
der Wissenschaft, Dossier: Evolution des Menschen.
Corballis, M. C. (2011). The Recursive Mind: The Origins of Human Language,
Thought, and Civilization. Princeton : Princeton University Press.
de Waal, F. (1982). Chimpanzee Politics. Baltimore: John Hopkins University Press.
Deacon, T. W. (1997). The Symbolic Species: The Co-Evolution of Language and
the Brain. New York London : W.W. Norton and Co.
Deacon, T. W. (2010). A role for relaxed selection in the evolution of the language
capacity. Proceedings of the National Academy of Sciences of the United States
of America 107: 9000–9006.
Diamond, J. (1991). The Rise and Fall of the Third Chimpanzee: How Our Animal
Heritage Affects the Way We Live. London: Vintage.
Dunbar, R. 2003. The Social Brain: Mind, Language, and Society in Evolutionary
Perspective. Annual Reviews Anthropology 32: 163–81.
Enard, W. et al. (2002). Molecular evolution of FOXP2, a gene involved in speech
and language. Nature 418: 869–872.
Fisher, R. A. (1930). The Genetical Theory of Natural Selection. Oxford: Clarendon.
Foerster, H. v. (1981). Observing Systemes. s.l.: Seaside Cal.
Gärdenfors, P. (2012). The Cognitive and Communicative Demands of Cooperation.
In J. v. e. a. Eijck (ed.), Games, Actions, and Social Software. Berlin/Heidelberg:
Springer.
Goodall, J. (1986). The chimpanzees of Gombe: Patterns of Behavior. Cambridge
Mass.: Harvard UP.
Goodall, J. (2010). Conversation with Freddy Gray. Spectator.
Hughes, J. F. et al. (2010). Chimpanzee and human Y chromosomes are remarkably
divergent in structure gene content. Nature 463.7280: 536–539.
Humphrey, N. (1976). The social function of intellect. In: P. Bateson
and R. Hinde (eds.), Growing Points in Ethology, 303–317. Cambridge:
Cambridge University Press.
James, J. W. (1970). The founder effect and response to articial selection. Genetical
research 16.3: 241–250.
Lahn, B. et al. (2004). Accelerated Evolution of Nervous System Genes in the Origin
of Homo sapiens. Cell 119: 1027–1040.
157
How language evolved as a backchannel between two feedback loops
MacDougall, A. and Montgomerie, R. (2003). Assortative mating by carotenoid-
based plumage colour: a quality indicator in American goldnches, Carduelis
tristis. Naturwissenschafte 90: 464.
Mampe, B., D., F. A., Christophe, A. and Wermk, K. (2009). Newborns’ Cry Melody
Is Shaped by Their Native Language. Current Biology 19.23.
Maynard Smith, J. and Szathmáry, E. (1995). The Major Transitions in Evolution.
Oxford / New York: Oxford University Press.
Mayr, E. (1942). Systematics and the Origin of Species from the Viewpoint
of a Zoologist. New York City: Columbia University Press.
Mayr, E. (1997). The objects of selection. Proceedings of the National Academy
of Sciences 94: 2091–2094.
Mehler, J. Nespor, M., Shukla, M. and Peña, M. (2006). Why is language unique to
humans?. Novartis Found Symp. 270: 251–280; discussion 280–292.
Miller, G. (2000). The mating mind: how sexual choice shaped the evolution of human
nature. London: Heineman.
Pelucchi, B. Hay, J. F. and Saffran, J. R. (2009). Statistical learning in a natural
language by 8-month old infants. Child Development 80: 674–685.
Piattelli-Palmarini, M. 2012. Three Models (and a Half) for the Description
of Language Evolution. Talk at Evolang IX Kyoto | The 9th Evolution
of Language Conference, s.n.
Pinker, S. (2007). The Language Instinct: How the Mind Creates Language. New
York: HaperCollins.
Pinker, S. (2010). The cognitive niche: Coevolution of intelligence, sociality, and
language. PNAS 107 suppl. 2: 8993–8999.
Pollard, K., Salama, S., N., L. and Lambo, t. M. C. S. (2006). et al. An RNA gene
expressed during cortical development evolved rapidly in humans. Nature 443:
167–172.
Prabhakar, S., Noonan, J., Pääbo, S. and Rubin, E. (2006). Accelerated evolution
of conserved noncoding sequences in humans. Science.
Przeworski, M., R. R, H. and A, D. R. (2000). Adjusting the focus on human
variation. Trends Genet 16: 296–302.
Rizzolatti, G. and Craighero, L. (2004). The Mirror neuron System. Annual Review
of Neuroscience 27: 169–192.
Rizzolatti, G., Fadiga, L., Gallese, V. and Fogassi, L. (1996). Premotor cortex and
the recognition of motor actions. Cognitive Brain Research 3: 131–141.
Saxe, R. (2013). The Story of a Study of the Mind [Interview] 2013.
Shultz, S. and Vouloumanos, A. (2010). Three-Month-Olds Prefer Speech to Other
Naturally Occurring Signals. Language Learning and Development 6: 241–
257.
Sibley, C. G. and Ahlquist, J. (1987). DNA Hybridization evidence of hominoid
phylogeny: result from an expanded data set. Journal of Molecular Evolution
99–121.
Tomasello, M. (1995). Language is not an instinct. Cognitive Development 131–156.
158 Till Nikolaus von Heiseler
Tomasello, M. (2003). Constructing a language. Boston, MA: Harvard University
Press.
Tomasello, M. (2009). Why we cooperate. Cambridge, MA: MIT Press.
Tomasello, M. et al. (2005). Understanding and sharing intentions: The origins
of cultural cognition. Behavioral and Brain Sciences 28: 675–691.
Tooby, J. and DeVore, I. (1987). The reconstruction of hominid evolution through
strategic modeling. In W. G. Kinzey (ed.), The Evolution of Human Behavior:
Primate Models. New York: Suny.
Trivers, R. (1972). Parental investment and sexual selection. In B. Campbell (ed.),
Sexual Selection and the Descent of Man, 136–179. Chicago: s.n.
von Heiseler, T. N. (2008). Medientheater. Berlin: Kadmos.
von Heiseler, T. N. (2014). Language evoloved for storytelling in a super-fast
evolution. In R. L. C. Cartmill (ed.), Evolution of Language, 114–121. London:
World Scientic.
Wiener, N. (1948). Cybernetics, or Control and Communication in the Animal and
the Machine. Cambridge, Mass.: MIT Press.
Zahavi, A. (1975). Mate selection – a selection for a handicap. Journal of Theoretical
Biology 53: 205–214.
