Hoffmeyer - 2014 - The Semiome From Genetic To Semiotic Scaffolding

DOI 10.
1515/sem-2013-0099 Semiotica 2014; 198: 11 – 31
Jesper Hoffmeyer
The semiome: From genetic to semiotic
scaffolding
Abstract: The fact that agency is an essential aspect of life introduces new ex
planatory avenues into the map of evolutionary thought. There is hardly any pro
cess in animate nature that is not, in one way or another, regulated communi
catively, i.e., through the ability of living systems to read and interpret relevant
signs in their environment. Semiotics – the science of signs – therefore ought to
become a key tool for the “life sciences” in general and biology in particular. The
paper analyzes the ways semiotic interactions in nature have been developed
to scaffold the web of physiological, developmental, and ecological pathways.
Semiotic scaffolding is only very indirectly based on genetic scaffolding. The gene
products, the proteins, are not just molecules, but are always also semiotic tools,
and what the genes really do is to specify the efficiency of semiotic modulators. In
addition to the concept of the genome we need in biology a concept of the semi
ome: the entirety of an organism’s semiotic tool set: i.e., the means by which the
organisms of this species may extract significantly meaningful content from their
surroundings and engage in intra- or interspecific communicative behavior. The
semiome thus defines the scope of the organism’s cognitive and communicative
activity. The theoretical question raised in this paper is the question of the inter
connectedness between genomic and semiomic changes.
Keywords: agency; semiome; genome; semiotic scaffolding; evolution;

biosemiotics
Jesper Hoffmeyer: University of Copenhagen. E-mail: [email protected]
1 Agency
In chapter 3 of The Origin of Species, Darwin discusses the Malthusian aspect of
natural selection, and sums up his analysis thus:
In looking at Nature, it is most necessary to keep the foregoing considerations always in

mind – never to forget that every single organic being may be said to be striving to the utmost
to increase in numbers; that each lives by a struggle at some period of its life; that heavy
Brought to you by | CRAI Universitat de Barcelona

Authenticated | 161.116.100.92
Download Date | 3/11/14 4:04 AM
12 Jesper Hoffmeyer
destruction inevitably falls either on the young or old, during each generation or at recurrent
intervals. Lighten any check, mitigate the destruction ever so little, and the number of the
species will almost instantaneously increase to any amount. (Darwin 1972 [1859]: 71, italics
added)
Darwin, of course, did not take this term “strive” to be meant in a psychological
sense – and yet, it can hardly be denied that his words imply an ascription of
agency of some sort to every single organism. Agency may be defined as the ca
pacity of an agent to act in the world, and it is obvious from reading Darwin that
he simply took for granted the agency of living beings. In the final paragraphs of
his book, he famously declares that “There is grandeur in this view of life, with its
several powers, having been originally breathed by the Creator into a few forms or
into one” (Darwin 1972 [1859]: 463). The reference to the Creator (in whom Darwin
hardly believed in any longer at this late time in his life, the paragraph being
added in the sixth edition of 1872) allows Darwin to escape the hard question of
how life could possibly possess these “several powers” – and among them,
“agency.” That organisms were indeed “striving” was an obvious fact to Darwin
(as it is to me), so he did not see any need to discuss it further.
In modern geno-centric versions of Darwinism, however, such talk of “striv
ing” has, of course, long been absent. Genes are fragments of DNA-molecules and
nobody would claim that molecules are capable of “striving.” And if organisms
are seen as machines programmed by the genes to serve the purpose of the genes’
own survival, then organisms cannot “strive” either – for although these machines
are teleological devices serving a definite purpose, their teleology is not inherent
to them, but is built into them by their constructor. Analogously: a printer serves
the purpose of printing and a lawnmower serves to cut the grass; in both cases,
the functionality is established via the construction of the machine. But in life we
find an inconsistency: If living machines (organisms) are constructed by genes,
then these genes must obviously exhibit “agency.” So may molecules “strive”?
The response to this question, I suppose, will be that natural selection “explains
away” the apparent teleology, replacing it with good old mechanistic explanations.
This explanation will not do, however, as I shall discuss (see Short 2002, also,
for a detailed philosophical analysis of this point). Natural selection presupposes
competition among units (whether genes, organisms, species, etc.), and “compe
tition” presupposes agency or striving: if such “units” did not attempt to escape
dangers and to provide for their own necessary resources, then no “competition”
would ensue, and life itself would, in all probability, go extinct as soon as its pre-
given source of resources were depleted. In other words, without agency, “natural
selection” could not work . . . and for this reason, “natural selection” therefore
cannot be used to explain “agency.”

13
The semiome: From genetic to semiotic scaffolding
This now raises the question that Darwin did not answer (or which he dele
gated to the Creator): Where does “agency” come from? And how has life evolved
to possess such a strange power that no non-living entities possess? A power that
is most intimately associated with what life really is, as becomes evident the
moment a person dies. What is lost here is exactly the property we are talking
about: agency. The most conspicuous trait of a dead organism is that it exhibits
no agency.
Important work is presently being done to confront this challenge (Kauffman
2000; Kauffman and Clayton 2006; Hoffmeyer 2010; Deacon 2011). Here I shall not
pursue the question any further apart from observing that if agency is not a prod
uct of natural selection, but a necessary condition for this mechanism to work at
all, then the question of agency remains a central unsolved problem in biology.
Considering how intimately connected agency is to “being alive,” it appears that
agency must be part of the complex puzzle of the origin of life – and that, by im
plication, the origin of life question cannot be treated in purely chemical terms.
What should be explained is not only how macromolecules were first formed,
in other words, but also how and why living systems would come to possess this
teleodynamics, as Terrence Deacon has called it: this strange “interest” in things
around them, i.e., the environment. How could inanimate nature create such a
property? Deacon does propose a preliminary model to begin thinking about this
problem (Deacon 2011), but until further clarification is obtained, I shall content
myself to observe that there must be a threshold zone around the formation of the
first genuinely living systems on Earth such that below this threshold zone, func-
tionality, agency, and semiosis (sign processes) does not occur, while above this
threshold zone all three of these interconnected aspects of being alive are now
present (Kull et al. 2009).
Considering the inflamed climate of recent public debates on Darwinism, it
might be wise to underline that the conclusion to be drawn from this discussion
is not that natural selection does not work. Natural selection remains a central
factor in our understanding of the evolutionary process. But the acceptance of
agency as an essential aspect of life does, of course, introduce additional explan
atory avenues into the map of evolutionary thought. First and foremost, organis
mic agency implies that the activity of organisms does interfere in non-predictive
ways in the outcome of evolutionary events. And since organismic activity is
largely controlled and regulated by semiosis, i.e., by sign processes, then
phytosemiotics, fungal semiotics, cytosemiotics, and zoosemiotics – in short,
biosemiotics – should be taken as major explanatory tools in modern evolution
ary theory. Indeed it may often be the case that natural selection acts by stabilizing
and fine-tuning scenarios that were already pre-established through the formation
of eco-semiotic interaction structures.

14 Jesper Hoffmeyer
2 Semiosis
Sign processes are neither forces nor things. They are process-relations that orga
nize activities. The causality of signs thus differs from the causality of forces.
While signs are frequently misunderstood or ignored, forces always exert their
power with merciless efficiency. Thus the concept of semiosis does bring a novel
element to the scientific tool set for, by definition, a sign-process requires an
interpretative agency.
The meaning of the term semiosis unfortunately is not well known. Most
people, even in academia, tend to burden the concept with mentalist connota
tions that it does not, in fact, imply. In semiotics proper, the term semiosis simply
means “sign action,” i.e., a process whereby a sign induces a receptive system to
make an interpretation. Thus a rat, for instance, quickly learns to lift its left hind
leg when approaching a distinct spot in the cage where it has repeatedly received
an electric shock through the paw of that leg. The rat’s avoidance reaction dis
closes that touching this spot now signifies pain.
Three processes are involved here: 1) the rat senses a distinct spot (the sign
vehicle); 2) the rat senses an electric shock in its hind leg (the object); 3) the rat
avoids putting its hind leg on this spot (the interpretant). The capacity of the rat to
integrate these three processes so that (1) provokes the release of (3) referring to
the (now non-actualized) shock (2) constitutes the essential core of what an inter
pretative act is all about. Since even de-cerebrated rats have been shown to per
form this interpretative act, no mental processes need be implied for semiosis to
occur.
Human linguistic utterances undoubtedly reflect mental processes, and one
reason why the term semiosis is so often taken to carry mentalist associations
may be the conception of “semiotics” as a branch of linguistics. Yet unnoticed by
most people, apparently, semiotics as a science has taken a long step ways away
from the tradition of Swiss linguist Ferdinand de Saussure (1857–1913) that came
to predominate the field in the 1970s. Instead, contemporary semiotics considers
human language as only one – admittedly, very peculiar – sign system among the
millions of sign systems that have evolved in natural systems. Songs of birds,
croaking of frogs, or spawning behavior of fish are other well known cases of sign
systems, but millions of less conspicuous semiotic interactions are also active as
regulators of organismic activity. Moreover, semiosis is an essential regulator not
only of intraspecific behavior, but also of interspecific behavior, as is dramati
cally illustrated in Nature’s many symbiotic interactions, which are all dependent
on the operation of multiple semiotically controlled interactions.

The semiome: From genetic to semiotic scaffolding 15
3 Scaffolding
The term scaffolding usually refers to a temporary structure used to support
people and material in the construction or repair of buildings and other large
structures. Since children, just like houses, are in a certain sense “under con
struction,” the concept of scaffolding may be used as a metaphor for the tools that
support children’s development. Thus, in the 1950s, the cognitive psychologist
Jerome S. Bruner began to use the term scaffolding to refer to any temporary
framework that is put up for support and access to meaning as needed, and then
taken away when the child secures control of success with a task. Predating
Bruner, the Russian psychologist Lev Vygotsky dealt intensively with the ways
that children internalize the challenges of the cultural milieu they inhabit. In
1930, Vygotsky famously gave the example of a young child playing with a hobby
horse as a means for the child to master the relationship between words and real
ity (Vygotsky 1978). Here, when the counterfeit, that is the object, replaces the
real horse, the object takes on the same psychological place as the word (horse)
does, and through the play the child thereby intuitively appropriates a mastery of
the (pestering) relation between words and reality, i.e., “horse” as a word and
“horse” as an animal. The learning process of children – and of adults – is of
course always a cognitive process, but it is eased a lot by being supported through
such bodily activity.
Andy Clark, for instance, gives the example of an equation such as
7,222 × 9,422, which may be solved in the heads by some autistic individuals, but
which most of us must resort to pencil and paper to manage (Clark 1997: 61). The
trick is to piece the task into smaller chunks that are assembled into the finished
result. Using an external medium, the paper, we can archive the results of short-
term operations, which can then be “added” after the proper procedure. The
paper-and-pencil procedure thus acts as a scaffold for the thought process.
Such scaffolds are semiotic tools. The stick is a sign that means horse for the
toddler and the signs on the paper refer to numerical information. Stick and char
acters on paper now appear in lieu of something else in reality (the horse and the
figures, respectively) – and by this “translation” to another medium, we are more
easily able to manipulate them. Through the appropriate motoric expressions,
the stick may enter into the child’s inner story of riding, fencing, fight, etc.; while
paper signs placed in a certain manner by vertical addition lead the person doing
the calculation to the correct answer. The significance of things or events in this
way becomes palpable – and thus “grasped,” as it is quite naturally expressed
when a person “catches” the meaning through “motoric activity.”
From my own school years, I remember how our German language teacher
told us strange or even stupid stories designed to help us remember some less

16 Jesper Hoffmeyer
obvious features of the German language. For example he told us the following
story so as to memorize which nouns would get umlaut in plural:
In der tiefen, tiefen Nacht (die Nacht, die Nächte),

im schwartzen Wald (der Wald, die Wälder),
gab es ein kleines Dorf (das Dorf, die Dörfer).
Denn plötzlich kam eine Wurst (die Wurst, die Würste)1
The fact that I still today – fifty-six years later – remember at least these first four
lines of the story, testifies to the efficiency of such instructional scaffolding.
4 Genetic scaffolding
In addition to the operational genes that most people equate with the very con
cept genes, the genome contains huge stores of hidden, never sought-after “menu
items” – what scientists call pseudogenes – that potentially may be brought into
circulation, and thus may come to play a role in evolutionary change. Indeed,
there exist quite a number of mechanisms whereby sequences on the DNA-
molecule may be duplicated and even change positions on the chromosome.
Such mechanisms have led to the formation of gene families – i.e., families of nar
rowly related functional genes, as well as pseudogenes. Pseudogenes are areas of
the genome that are nonfunctional in the sense that they are not available to the
transcription process – but that nevertheless exhibit remarkable similarity to
known functional genes in their base sequences. Unlike functional genes, pseudo
genes are not expressed and, therefore, are not subject to natural selection. Since
they do not in any way contribute to the success or failure of their carrier organ
isms, they are free to undergo mutation without incurring the selectional con
sequences that normal genes do. Pseudogenes therefore represent a tacit resource
base of latent proteins with unexplored properties – from which new functional
genes may eventually be picked up through the operation of natural selection.
Pseudogenes are thought to be the most important tools for the evolutionary
build up and maintenance of functional properties. Functional properties of
organisms, such as efficient oxygen supply to tissues and organs, depend on the
presence of specific proteins at specific times – and this, in turn, depends on
a finely controlled and coordinated processing of available gene resources in
1 In English: “In the deep, deep night, in the black forest, there was a little village. Then sud
denly came a sausage . . .”

17
a huge number of single cells. Stem cells in the bone marrow of mammalian
species, for instance, persistently divide and differentiate into the many differ
ent kinds of blood cells, and – in the human body – an estimated 2.4 million
hemoglobin-rich red blood cells are produced per second. A series of coordi
nated physiological and cytological events are involved in the fine-tuning of this
system, but at the most basic level, the decisive factor is the availability of the
necessary gene resources appearing in an appropriate regulatory unit on the
genome.
The hemoglobin gene family nicely illustrates this dynamic. Since oxygen dif
fuses through tissue slowly, animals beyond a certain size must find a way to
solve the challenge of getting enough oxygen into their tissues. Insects have
solved the problem by developing small air channels (trachea) – but this solution
only works because insects’ body mass is so small (conversely, this may be the
reason why insects never become big); in vertebrates, however, the solution to
this same problem is hemoglobin – a highly specialized protein molecule with
a capacity to carry forward a small iron-containing molecule (a porphyrin) that
will bind oxygen reversibly – meaning that it will take up oxygen wherever there
is plenty of it (for instance, in the lungs) and give it off where it is lacking (for
instance, in tissues). The delicate biochemical task that the hemoglobin mole
cule performs is to keep the highly reactive oxygen molecule safe inside the 3-D
protein structure while it is transported around the body. In molecular models
of hemoglobin, the iron porphyrin (called heme) can be seen to consist in a
nearly flat structure enclosed within a hydrophobic depression of the protein
molecule.
Human hemoglobin is in reality a complex (a tetramer) of four subunits, each
of which is a protein chain of just under 150 amino acids. All four subunits are
derived from the same ancestral protein chain, but pairwise they are now slightly
different, and are called alpha and beta chains, respectively. Significantly, the
early embryo doesn’t form any of these two subunits, but instead forms ksi and
epsilon chains that deviate somewhat from the alpha and beta chains of the
adult hemoglobin. Towards the end of embryonic development, a fifth chain
is made called the gamma chain – this is because the oxygen supply for the
embryo comes through the placenta, and it therefore needs a special hemoglobin
tuned to this temporary – but critical – condition. And for a short period around
birth, the newborn baby even produces a sixth chain of proteins called the delta
chain.
Such variety in the kinds of hemoglobin chain formation found through the
embryogenesis of one individual is a relatively recent phenomenon, in terms of
evolution – as is disclosed by comparing the hemoglobin chains from a number
of different species, as in shown in Figure 1. This figure shows that the original

18 Jesper Hoffmeyer
oxygen-carrying protein was a monomer not very different from the protein,
myoglobin – a protein that nowadays is used for storing oxygen in muscle cells.
Approximately five hundred million years ago, the gene involved in the protein
synthesis of myoglobin underwent a duplication, where one of the gene copies
became the ancestor of the myoglobin genes found in all higher organisms today,
while the other gene copy, in time, developed (probably as a pseudogene) to be
come a proper oxygen transporter – and this gene thus became the progenitor of
all the world’s hemoglobin genes today. Thus, all later species in this lineage have
separate genes for myoglobin and hemoglobin.
In retrospect, it is easy to see this evolutionary development as a reflection of
a new “size strategy” in evolution. In the small animals of early evolutionary time
(e.g., the protozoans and flatworms), there was not much need for an oxygen
transporter, since the distance to the surface of the animal would always be insig
nificant. But as animals increased in size, and began feeding on smaller animals
for their nourishment, oxygen transportation became a major challenge. Four
hundred million years ago, a second gene duplication occurred, through which
Fig. 1: An evolutionary tree for human hemoglobin. Shown above is the arrangement of human
hemoglobin genes over three different chromosomes. Sequences that are still transcribed are
shown in dark gray. The very light areas are no longer transcribed in humans, but are now
pseudogenes. Shown below is a time axis (in millions of years) indicating approximately when
in evolution different hemoglobin genes appeared (modified from Mathews et al. 1999: 234)

19
the ancestral forms of alpha and beta chains were established. Since then, all
organisms in the lineage have both alpha and beta chains in their hemoglobin. In
natural history, this corresponds with the divergence into sharks and bony fishes,
and it is the evolutionary line of the latter that led to the reptiles, and eventually,
to the mammals.
Similarly, the gene duplications behind the divergence of the specialized
fetal hemoglobin chains occurred some two hundred million years ago, which
roughly corresponds with the appearance of placental mammals – since the par
ticular oxygen-binding properties of these chains would be needed for the special
tasks of ensuring an adequate supply of oxygen to the embryo. Like the musical
motifs of a classical symphony that are interwoven in a variety of ways to form the
experience of a unified whole, so we see how evolution little by little has managed
to weave the hemoglobin themes into one another to form a functionally optimal
unity.
Evolution is a tinkering process as Francois Jacob has famously put it, and at
the most basic level, it consists in the building up of more and more intricate
genetic scaffolding devices in an emergent kind of process that allows evolution
access to the “upper floors” – and thus to “invent” new higher-level patterns that
subsume aggregates of lower level patterns under their umbrellas. Placental
mammals, for instance, would not have been an evolutionary option, had it not
been for the appropriation by the genetic system of new types of fetal hemoglobin
genes. The birth of a new organismic function thus takes the lucky conjunction of
two events: 1) an already existing nonfunctional gene (a pseudogene) acquires a
new meaning (through integration into a functional, i.e., transcribed, part of the
genome), 2) the resulting gene product (i.e., the changed protein component) fills
an unfilled gap in the semiotic needs of the cell or the embryo. In this way, new
genes become scaffolding devices supporting new kinds of interaction that imbue
some kind of semiotic advantage upon their bearer.
5 The genome
The genome is normally defined as the entirety of an organism’s genetic informa

tion, but it is rarely specified what exactly should be meant by the term “informa
tion” in this particular context. Judging from the way the term genetic information
is used in modern biology, however, it is rather obvious that genetic informa
tion is normally understood simply as a set of “instructions.” Now, since “instruc
tion” is a meaningless term if not addressed to “somebody” that might eventually
effectuate this instruction, “information” remains a highly questionable term

20 Jesper Hoffmeyer
in the context of genes; who exactly is it that should understand and effectuate
the alleged “instruction”?2 An instruction must be interpreted by some agent –
otherwise, it will not have any effect. And yet, the problem of interpretation never
enters into the discourse of modern biology. As a result, it is hard to see any dif
ference between what is meant by “information” in this sense, other than what is
meant by the simple word “cause.” This raises the question of why the term “in
formation” is so broadly used in biology. What is obtained by this strange change
of terminology from causal connections to informational relations? Might it be
that the term information – perhaps at some unconscious level – appeals to the
teleological intuition about life that scientists share with other people? Or, in
other words, that the seemingly objective term “information” – falsely, as we have
seen – lends credibility to the belief that life can be exhaustively explained by
mechanistic biology?
From a semiotic point of view, there is nothing mysterious about the term
genetic information, since in this view, “information” is simply a set of signs to be
interpreted by the cell (or the body) in which the gene is located. The genome may
then be defined as the entirety of genetically (i.e., digitally) coded signs that the
fertilized egg has inherited from its parent organisms (or has otherwise acquired).
Historically, however, the gene concept was formed under the influence of the
early twentieth century experiments with Drosophila genetics, where deviations
in the normal development of flies (mutations) were mapped to distinct loci on
the Drosophila chromosomes. Consequently, “the gene” came to be seen as a
material entity – and as the ruler of inherited traits. The fact that a snail genome
will lead to the appearance of a snail and a wolf genome will lead to the appear
ance of a wolf, gradually implanted the conception in the minds of biologists that
the genome controls the ontogenetic process.
Genes, however, do no such thing. Molecules, even DNA molecules, are for
good reasons deprived of cognitive skills, and thus they cannot “control” things
or events. A much better way to understand what genes really do is to see them as
semiotic modulators. For illustration, let us consider the growth of an embryonic
tissue layer in the virtual organism Scitoi mesoib (figure 2).
We shall specifically observe the dynamic growth of a neural tissue layer
called sirap that normally is regulated by the concentration of a signal molecule,
2 Originally the concept of information was introduced into biology through inspiration from
information theory, i.e., in the sense of “Shannon information.” But this purely probabilistic
concept of information has not proved of much help in biological theory, and gradually the con
ception of information slid away towards the semantic concept of information now tacitly taken
for granted by nearly all biologists.

Fig. 2: Fetal development of sirap tissue in the virtual organism Scitoi mesoib (see text). From
Hoffmeyer (2008: 132)
s irapin, which gradually builds up due to its excretion from an underlying embry
onic cell layer (Figure 2, top and middle). Normally, when sirapin reaches a cer
tain concentration inside the cells of the sirap layer, further cell divisions – and
thus further growth – will come to a stop.
Now, suppose that a mutation hits the gene sir which specifies the enzyme
(sirapin synthetase) that catalyzes the formation of sirapin in the underlying
tissue layer. Such a mutation might eventually cause a slight decrease in the
catalytic activity of the enzyme, and the mutated embryo would therefore pro
duce sirapin at a slower rate. Consequently, the sirapin concentration in the

22 Jesper Hoffmeyer
s irap layer will stay below the threshold value for an extended period and, ac
cordingly, the relative size of the sirap neural tissue will be increased (Figure 2,
bottom).
This simple yet realistic model easily explains how small mutational modifi
cations might provoke severe changes in morphology or instinctual behavior in
an individual or a species. Large-scale morphological change may therefore take
place without any new genes being involved at all – and in cases where such
changes affect the eventual formation of brain architecture, it follows that new
behavioral patterns may appear simply through the modification of already exist-
ing genetic material.
The key here is the idea that the ontogeny of morphological and behavioral
traits is secured through a sophisticated mechanism of semiotic scaffolding – and
this is only in a very indirect and mediated way, based essentially on genetic scaf
folding. For to modify genes is to change the way that different processes are
“tuned to one another” – and the product that results from such a process bears
no simple relation to the mutation itself. From a semiotic point of view, what
happens in the sir-mutant is a weakening of the signal involved in fitting the
growth of one cell layer to another. Thus, in the case of the mutant Scitoi mesoib,
the cells in the sirap tissue layer do not recognize the signal in time, and this de
lay in the interpretive response has the effect of allowing the growth of this layer
to continue for an extended period of time.
Genes, then, do not specify traits in the adult organism, and in a way they do
not even specify proteins. They do, of course, mechanistically determine the ami
no acid backbones of given proteins – although it should be remembered that a
plurality of processes, such as RNA editing, protein folding, and protein target
ing, all contextually interfere in the simple causality of this otherwise seemingly
deterministic process. However, since proteins are not just molecules, but are
always also semiotic tools, what the genes really do specify is the efficiency of semi-
otic modulators. They serve to adjust the screws, as it were, in the biosemiotic
machinery of the organism. Therefore, in order to adjust the predominant meta
phors by which we think about the relations between genetic regulation and the
semiotic reality of life, let me suggest a model of the genetic system as an
inventory-control system – i.e., a system for ascertaining that the appropriate
stock of molecular resources are always present when the cell or organism needs
them for its never-ending semiotic interactions. In this inventory-control system,
individual genes appear like the ever-present (though largely unseen) application
options that only become available for use through the activation of different
menu bars in our computer software applications. In these “menued” gene en
sembles, there are application options for chromatin structure, application op
tions for the proteins involved in the cell-division apparatus, application options

23
for catabolic and anabolic enzymes, application options for membrane related
proteins and other structural components, etc.
Decisions on which application menus should be activated, as well as which
options made available through the activation of these “menued” gene ensem
bles should actually be realized at any given time, will depend on the intricate
semiotic interplay of the total cellular system as it may be biased in any given
moment by the relative concentrations of important modulatory proteins. This
context-dependent event is most properly referred to as a decision in this sense,
for that term highlights the fact that there is no one element deterministically
causing one thing to happen over another, but that the selection is a function of
the system state and its needs at any given moment.
In this heuristic, every activated application option functions as a release
mechanism whereby the production of a new resource is initiated and stored in
the proper department of the cell. Mechanistically, this initiation corresponds to
the RNA polymerase reaction whereby a particular gene is transcribed into RNA.
As long as the given menu option is activated, the corresponding mRNA produc
tion will continue and protein synthesis will go on. But the moment it is inter
rupted, mRNA degradation will naturally ensue and this will rapidly bring any
further synthesis of that particular protein to an end.
We can now see how the view of the genome as controlling a life-long ontoge
netic process tends to blind us to the semiotic competences of the cellular and
organismic system. Genes, like human words, do not directly influence the world
around them: We do not believe in the power of spells; the rock will not fall from
the cave just because Ali Baba says “open sesame.” Likewise we should not be
lieve in the power of naked genetic information. As Harvard geneticist Richard C.
Lewontin has famously expressed it: “First, DNA is not self-reproducing, second,
it makes nothing, and third, organisms are not determined by it” (Lewontin 1992:
33). The causal efficacy of genes presupposes that some living system interprets
them; and just as inert, intrinsically meaningless words serve to support human
activity and communication, so do inert, intrinsically meaningless genes support
cellular activity and communication.
The genome is not controlling ontogeny, it scaffolds it (just as books do not
determine cultural patterns, but they certainly scaffold them).
6 Semiotic scaffolding
Life depends on the fine-tuned co-ordination of an astronomical number of bio
chemical reactions taking place inside or across different kinds of membranous
structures (Hoffmeyer 1999). The total area occupied by cell membranes in a

24 Jesper Hoffmeyer
uman body for instance has been calculated as one third of a km2 and the area
h
of membranes filling up the internal space of cells, i.e., the membranes around
mitochondria, endosplasmatic reticulum, Golgi apparatus, and the many smaller
cell organelles, are probably several orders of magnitude larger than the area of
the cell membrane itself, giving us a total area of membranes in the human body
of perhaps 30 km2. These membranes generally are fluid structures that are ac
tively upheld under the expense of metabolic energy. Nothing in this complex
system has any inherent stability under the conditions prevailing in living sys
tems, and the maintenance of a living system thus requires a very intricate system
of dynamic interactions. While this system is of course driven by metabolic ener
gy it must be controlled by semiotic means, i.e., local processes must obey the
need for global function, and this result can only be obtained through communi
cative activity connecting distant parts and different functional domains of the
body to each other.
The semiotic coordination of the processes described here makes up the
branch of biosemiotics called endosemiotics, i.e., the semiotics of processes tak
ing place inside the organism. Exosemiotics, on the other hand, is the term used
for biosemiotic processes going on between organisms (whether or not these
belong to the same species) as well as for semiotic processes connected to the
interpretation of abiotic markers, such as migratory birds making use of stellar
configurations to trace their way. That the endo- and exo- prefixes have thus come
to refer to the organismic level, i.e., the borderline around bodies, is strictly a
matter of technical terminology and should not be taken to signify any privileged
role in biosemiotics of this particular boundary. In fact, semiotics is in principle
always connected to some kind of inside-outside interaction.
Through the totality of life processes in the world, a semiosphere is created,
surrounding the earth in much the same way the atmosphere, hydrosphere or
biosphere envelops the planet (Hoffmeyer 1996). The concept of the semiosphere
adds a semiotic dimension to the more well-known concept of the biosphere, em
phasizing the need to see life as belonging to a shared semiotic universe through
which cells, organisms and species all over the planet interact in still hardly
understood ways. Every single species (humans included) has only a limited ac
cess to the semiosphere, for each one’s capacity for sensing and interpreting cues
in its surroundings has evolved to fit a particular ecological niche – in the termi
nology of Jakob von Uexküll, each species is confined to its own limited Umwelt,
understood as the internal model that the individuals of a given species constructs
of its surroundings. The semiosphere poses constraints or boundary conditions to
the Umwelts of populations since these are forced to occupy specific semiotic
niches – i.e., they will have to master specific sets of signs of visual, acoustic, ol
factory, tactile, and chemical origin in order to survive in that semiosphere.

25
The network of semiotic interactions by which individual cells, organisms,

populations, or ecological units are controlling their activities can thus be seen as
scaffolding devices, assuring that activities become tuned to needs, i.e., tuned to
the task of coping with challenges confronting such agents. Just as the scaffold
raised to erect a building will largely delimit what kind of building is raised, thus
the semiotic controls on biological activities delimit when and how such activi
ties should take place. To analyze and conceptualize the myriad of semiotic scaf
folding mechanisms operative at different levels in natural systems is the core
subject matter of biosemiotics. The semiotic demands to populations will always
be a decisive challenge to their success, and ecosystem dynamics, therefore, shall
have to include a proper understanding of the semiotic networks operative in
ecosystems.
The emergence of new scaffolding devices act like stepping stones in a river
to lead the evolutionary process forward one step at a time – and, for the most
part, farther away from the brink for each step. In themselves, such semiotic scaf
folding patterns may take many forms, and rely on many different principles, but
the core property of a scaffold remains that of focusing the energy flow (behavior)
of the concerned system or subsystem upon a constrained repertoire of possibili
ties, or in guiding the systems behavior to obey a definite sequence of events. A
receptor molecule at the surface of a cell may be tuned to open a neighboring
channel if, and only if, it is hit by one out of a small number of protein molecules
carrying specific domains at their surfaces, just as the offspring of a bird may be
tuned to learn only one or a very narrow band of sequences of sounds. When the
proper cue arrives, the receptor opens the channel or the bird young learns the
species’ song. The receptor may be fooled, however – as when a HIV virus iconi
cally mimics the surface domains of one of the organism’s own proteins – and the
bird may be fooled if it’s been deposited in the nest of a foster bird, without fur
ther contact with adult birds of its own species. Semiotic scaffolding mechanisms
thus depend on interpretation, and interpretation always runs the risk of being
“wrong” (relative to the scaffolded function).
For example, let us consider the case of infertility in the so-called eyeless
mutant of the axolotl, as discussed by Leo W. Buss (1987). Under normal condi
tions, the amphibian eye will be produced by the chemical interaction between
the newly formed optic vesicle and the embryonic ectoderm layer. An inducer
produced by the optic vesicle is used for scaffolding this interaction. What hap
pens in the eyeless mutant is that this step (the step from “a” to “b” in Figure 3) is
disrupted, because the ectoderm of the mutant does not respond properly to the
inducer, meaning that no eye will be formed and the mutant becomes blind. In
addition to being blind, however, this poor creature also lacks the capacity to
leave offspring, because the region of the brain called hypothalamus will only

26 Jesper Hoffmeyer
Fig. 3: The normal sequence of steps in the development of the amphibian eye (modified from
Coulombre 1965)
develop properly in the presence of induction from tissues of the nascent eye. In
the eyeless individual, no eye is formed to direct the development of the hypo
thalamus, the hypothalamus therefore cannot produce gonadotropin hormones
– and in the absence of these hormones, the individual becomes sterile.
The deficiencies of the eyeless mutant clearly illustrates the tinkering ways in
which ontogeny has become scaffolded by evolution. There is presumably no
other reason why the development of hypothalamus should depend on the pres
ence of a functional eye – other than the fact that the eventual formation of the
eye is in a location that happens to be anatomically close to that region of the
brain where hypothalamus is normally developed in this lineage. The eye just
happened to be in the neighborhood of the nascent hypothalamus-region in nor
mal individuals, and probably for no other reason than this, evolution managed
to exapt the eye for the new role as an ontogenetic switch for the initiation of

27
proper development of a hypothalamus. As Buss says: “Ontogeny must reenact

the interactions which gave rise to it” (Buss 1987: 97). In the terminology of this
paper, ontogeny is safeguarded by myriad of semiotic scaffolds, each depending
on another, in long chains of successive steps.
While the evolutionary emergence of subtle scaffolding devices may be seen
as a result of natural selection, it should be noticed that a fundamental aspect of
the process is the capacity – or talent one might say – of individual organisms,
cells, and cell assemblies to change their internal settings in integrated waves
under the influence of new and/or external molecular cues. The semiotic logic of
localized dynamic biochemistry in a given embryonic tissue thus will tend to tell
us as much or more about the actual causality behind semiotic emergence as will
explanations in terms of “natural selection” that cannot well monitor the incipi
ent differences between individual organisms, and that only become relevant at
later stages of the emergent process.
Shifting the focus from the endosemiotic domain to the exosemiotic domain,
the same basic dynamics are at work. Organisms are, as we say, very semiogenic:
they have a surprising capacity to take advantage of any regularities they might
come upon in their environment as signifying vehicles, or signposts. And animate
nature exhibits an infinite number of subtle interaction patterns between organ
isms belonging to species that may even be as remotely related so as to belong to
different kingdoms. A well-known example is the mutualistic symbiosis between
the clown anemone fish (Amphiprion percula) and the multicolored sea anemone
with which it shares its life. The skin of the clown fish is covered by a layer of mucus
that makes it immune to the anemone’s lethal touch, but before taking residence
between the arms of the anemone the fish nevertheless perform an elaborate
dance with the sea anemone in which they gently touch the tentacles with differ
ent parts of their bodies until they are acclimated to their new host. The cohabita
tion between two species is, in principle, extremely dangerous for both of them,
and a successful result obviously requires that both parts confirm their identity to
one another through elaborate biochemical and/or behavioral schemes. The in
teraction between anemone and fish seems to be mutualistic: the fish gets protec
tion from the anemone and access to leftover food scraps from its meals, while it
reciprocates by removing parasites and perhaps also by bullying off intruders.
Summing up this analysis we can delineate the mechanism behind such
semiotic emergence in the following three steps:
1. Communicative patterns arising by chance in groups of animals or plants will
tend to stabilize themselves whenever they work.
2. New communicative patterns will tend to appear on the top of already estab
lished patterns, thereby strengthening the scaffolding of the initial interac
tive pattern.

28 Jesper Hoffmeyer
3. If interactive patterns endure for many generations, mutational events may

stepwise take over the task of scaffolding. This makes the scaffolding more
reliable, but less flexible.
7 The semiome
We can define a semiome in analogy with the genome as the entirety of an organ-
ism’s semiotic tool set: i.e., the means by which the organisms of this species may
extract significantly meaningful content from their surroundings and engage in
intra- or interspecific communicative behavior. The semiome thus defines the
scope of the organism’s cognitive and communicative activity. When a new organ
ism is emanating from a fertilized egg, the very first event is a transformation of
the highly differentiated gamete nuclei (from sperm and egg) to one totipotent
nucleus able to give rise to a whole organism. In mammals, the newly formed
embryonic nucleus is transcriptionally silent through the first few cell divisions,
and the initial developmental events are instead triggered by a maternally en
coded gene expression program inherent to the cellular machinery of the egg. The
expression of the genome is thus from the very beginning controlled by epigene
tic factors and this epigenetic control function remains implicit – although rarely
considered – throughout all genomic activity.
Likewise, the semiome of an organism will only gradually develop to unfold
to its full potential – and this process again is implicitly dependent on the inter
action of the developing organism with its biotic and abiotic environment. That
environmental factors are implied very early in this process was shown by Gilbert
Gottlieb, who discovered that ducklings will not develop a preference for the
mother’s species-specific call-signal if they are prevented from themselves vocal
izing while still in the egg. Apparently, duckling embryos have to be exposed to
their own prenatal call sounds before they can develop a preference for the quite
different sounds of their mother’s call (Gottlieb 1981). It is important to underline
here, as Gottlieb did, that “these sorts of facts do not show that all traits are
‘learned,’ as opposed to innate. They show that reliable developmental outcomes
occur because of reliable interactions between the developing organism and its
environment” (Griffiths and Gray 1994: 280). The semiome thus develops as both
instrument for, and product of, the interactive dynamics between genome and
environment: both resources and challenges present in the environment feed
back into genomic processes that reciprocally assure the provision of the protein
resources needed for launching the proper semiotic repertoire of the organism,
i.e., the semiome.

29
The concept of the semiome is closely related to Jakob von Uexküll’s concept
of Umwelt, but introducing the term semiome is meant to stress the analogy to the
genome and thus also the necessity of separating out of two autonomous –
although narrowly intertwined – dynamic systems inherent to organismic devel
opment and evolution: one for scaffolding the “body machinery” and one for
scaffolding “cognition and communication.” Doing so, we may be better posi
tioned for studying the internal interactive dynamics connecting genome and
semiome without subsuming one under the other.
Biological evolution has often enough been juxtaposed with change in the
gene pool of populations or with changes over time in one or more inherited
traits. Evolution, in other words, is thought to be primarily about genomes. And
yet, semiomes have changed over time as well: a moth semiome, for instance, is
very different from the semiome of a parrot or a whale. The moth lives in a silent
world only interrupted by the bat’s fateful frequencies of approximately 20,000 Hz,
whereas parrots and whales are engaged in lively – although very differently
structured – oral “conversations.”
The theoretical question raised here is the question of the interconnected
ness between genomic and semiomic changes. When the hypothalamus of a sala
mander, as we saw, can only be developed through inductive processes emanat
ing in the nascent eye, the implication is that communicative processes in the
neural tissue of the growing vertebrate embryo are indeed scaffolding “genomic”
action. Recent studies of the cascade of events taking place during vertebrate eye
development, however, makes clear that not only are many different genes in
volved in controlling the process but these genes are interacting with each other
in multiple direct and indirect ways (Bayley et al. 2004), implying that the genetic
scaffolding is strongly dependent on semiotic interactions. Again: the genome
does not control development – it only scaffolds it, and it does so in a narrow in
terplay with a diversity of semiotic scaffolding loops. The question of whether
new mutations preceded the establishment of new semiotic control loops or vice
versa seems futile, considering how deeply integrated the two scaffolding devices
are.
8 Conclusions
We have seen that changes in the semiome of a population or lineage is as much
involved in the establishment of evolutionary change, as is change in the genome.
The belief that hereditary mechanisms assure stability down through generations
such that phenotypic mechanisms could never take the lead role, is true only
so far. For we have seen that genes are not really independent units at the

30 Jesper Hoffmeyer
evelopmental level, but are only exerting their effects through complicated se
d
miotic control loops. A range of environmental factors such as temperature, diet,
physical stress, presence of predators, etc., may trigger changes in these loops in
the absence altogether of any genetic change – and such changes may have quite
dramatic effects on the phenotypic outcome. To focus only on the genome and
disregard the equally causal and consequential loops of the semiome is to dis
regard the very facts of biological evolution as such, i.e., to confuse life with the
tools that life depends on.
Acknowledgments: Some minor parts of this paper are based on modified ex
tracts from Hoffmeyer 2008.
References
Bayley, Travis J., Heithem El-Hodiri, Li Zhang, Rina Shah, Peter H. Mathers & Milan Jamrich.
2004. Regulation of vertebrate eye development by Rx genes. International Journal of
Developmental Biology 48. 761–770.
Buss, Leo W. 1987. The evolution of individuality. Princeton, NJ: Princeton University Press.
Clark, Andy. 1997. Being there: Putting brain, body, and world together again. Cambridge, MA:
MIT Press.
Coulombre, Alfred J. 1965. The eye. In R. L. DeHaan & H. Ursprung (eds.), Organogenesis,
219–251. New York: Holt, Rinehart and Winston.
Darwin, Charles. 1972 [1859]. On the origin of species by means of natural selection or the
preservation of favored races in the struggle for life, 6th edn. London: J. M. Dent.
Deacon, Terrence. 2011. Incomplete nature: How mind emerged from matter. New York &
London: W. W. Norton.
Gottlieb, G. 1981. Roles of early experience in species-specific perceptual development. In R. N.
Aslin, J. R. Alberts & M. P. Petersen (eds.), Audition, somatic perception, and the chemical
senses (Development of perception: Psychobiological perspectives 1), 5–44. New York,
Academic Press.
Griffiths, P. E. & R. D. Gray. 1994. Developmental systems and evolutionary explanations.
Journal of Philosophy 91(6). 277–304.
Hoffmeyer, Jesper. 1996. Signs of meaning in the universe. Bloomington, IN: Indiana University
Press.
Hoffmeyer, Jesper. 1999. Order out of indeterminacy. Special issue, Semiotica 127(1–4).
321–343.
Hoffmeyer, Jesper. 2008. Biosemiotics: An examination into the signs of life and the life of
signs. Scranton & London: University of Scranton Press.
Hoffmeyer, Jesper. 2010. Semiotic freedom: An emerging force. In Paul Davis & Niels Henrik
Gregersen (eds.), Information and the nature of reality: From physics to metaphysics,
185–204. Cambridge: Cambridge University Press.
Kauffman, Stuart A. 2000. Investigations. Oxford: Oxford University Press.

Kauffman, Stuart A. & Philip Clayton. 2006. On emergence, agency, and organization. Biology
and Philosophy 21. 501–521.
Kull, Kalevi, Terrence Deacon, Claus Emmeche, Jesper Hoffmeyer & Frederik Stjernfelt. 2009.
Theses on biosemiotics: Prolegomena to a theoretical biology. Biological Theory 4(2).
167–173.
Lewontin, Richard C. 1992. The dream of the human genome. New York Review, May 28. 31–40.
Mathews, C. K., K. E. van Holde & K. G. Ahern. 1999. Biochemistry, 3rd edn. San Francisco:
Addison-Wesley.
Short, Tom L. 2002. Darwin’s concept of final cause: Neither new nor trivial. Biology and
Philosophy 17. 323–340.
Vygotsky, Lev. 1978. Mind in society: The development of higher psychological processes.
Harvard, MA: Harvard University Press.
Bionote
Jesper Hoffmeyer (b. 1942) is Professor Emeritus at the University of Copenhagen
〈[email protected]〉. His research interests include biosemiotics and philoso
phy of nature. His publications include A legacy of living systems: Gregory Bateson
as precursor to biosemiotics (2008); and “A Natural History of Intentionality”
(2012).


Hoffmeyer - 2014 - The Semiome From Genetic To Semiotic Scaffolding

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hoffmeyer - 2014 - The Semiome From Genetic To Semiotic Scaffolding

Uploaded by

Copyright:

Available Formats

DOI 10.

1515/sem-2013-0099 Semiotica 2014; 198: 11 – 31

Keywords: agency; semiome; genome; semiotic scaffolding; evolution;

Jesper Hoffmeyer: University of Copenhagen. E-mail: [email protected]

In looking at Nature, it is most necessary to keep the foregoing considerations always in

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

In der tiefen, tiefen Nacht (die Nacht, die Nächte),

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

The genome is normally defined as the entirety of an organism’s genetic informa­

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

The network of semiotic interactions by which individual cells, organisms,

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

proper development of a hypothalamus. As Buss says: “Ontogeny must reenact

Brought to you by | CRAI Universitat de Barcelona

3. If interactive patterns endure for many generations, mutational events may

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

Brought to you by | CRAI Universitat de Barcelona

You might also like

The genome is normally defined as the entirety of an organism’s genetic informa