A Short Course on Biosemiotics: 1. Functional approach to the origin of life: emergence of signs in autocatalytic systems

Alexei A. Sharov
Laboratory of Genetics
National Institute on Aging (NIA/NIH)
Baltimore, USA
Presented in the Embryo Physics Course, March 14, 2012


Biosemiotics is a new discipline that integrates biology with semiotics, which is a theory on signs, meaning, and knowledge. The main idea of biosemiotics is that life has informational nature and it is coextensive with semiosis. Thus, biologists need to understand the nature of information. Biosemiotics rejects dualism and attempts to explain the origin of life and mind. Ideas and even logic are tools developed by evolving agents. It assumes that organisms develop their own models of the world (Umwelts) which may be different from human models. My approach to biosemiotics is based on the methodology of pragmatism (constructivism/instrumentalism) that assumes that meanings are useful conventions of evolving and communicating agents. Alternative approaches include objective idealism, hermeneutics, and materialism. In the first lecture I show how ideas of biosemiotics can be applied to the problem of the origin of life. The definition of life needs to be adjusted to include agents with their functional envelope (i.e., computers, ribosomes, etc.). Agents use functional information (signs) to preserve, control, and communicate their functions. They often create or recruit subagents or slave-agents to perform some of their living functions. The origin of life can be associated with the origin of first signs. This functional approach to the origin of life is an alternative strategy to the traditional structural approach, which assumes that a proto-organism gets assembled by chance from multiple components (membrane, several catalytic RNA species, coenzymes, etc.). I proposed that simple coenzyme-like molecules (CLMs) performed hereditary functions before the emergence of nucleic acids. Autocatalytic CLMs modified (encoded) surface properties of hydrocarbon microspheres, to which they were anchored, and these changes enhanced autocatalysis and propagation of CLMs. Heredity started from a single kind of self-reproducing CLM, and then evolved into more complex coenzyme autocatalytic networks containing multiple kinds of CLMs. Polymerization of CLMs on the surface of microspheres and development of template-based synthesis is a potential evolutionary path towards the emergence of nucleic acids.




Origin of life => Sharov AA. 2009. Coenzyme autocatalytic network on the surface of oil microspheres as a model for the origin of life. Int. J. Mol. Sci. 10(4): 1838-52. PMID: 19468342. http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=2680650&blobtype=pdf

Petri nets => Sharov, A. A. 1991. Self-reproducing systems: structure, niche relations and evolution. BioSystems, 25: 237-249.

Sharov AA. 2009. Role of utility and inference in the evolution of functional information. Biosemiotics 2: 101-115.
Sharov, A. A. 2001. Umwelt-theory and pragmatism. Semiotica 134: 211-228.
Sharov, A. A. 1999. The origin and evolution of signs. Semiotica 127: 521-535.
Sharov, A. A. 1998. From cybernetics to semiotics in biology. Semiotica 120: 403-419.

Papers on biosemiotics:


Alexei Sharov is a Staff Scientist at the Genetics Laboratory, National Institute on Aging (NIA/NIH) in Baltimore, MD. His reserach includes bioinformatic analysis of gene expression and epigenetic mechanisms that affect gene regulatory networks. He developed several online software tools for bioinformatics that include NIA Array Analysis (http://lgsun.grc.nia.nih.gov/ANOVA), CisFinder (http://lgsun.grc.nia.nih.gov/CisFinder), NIA Mouse Gene Index (http://lgsun.grc.nia.nih.gov/geneindex), and CisView (http://lgsun.grc.nia.nih.gov/cisview). He started his career as ecologist and entomologist and earned his degrees (B.S. in 1976; M.S./Candidate of science in 1980; and Ph.D./Doctor of sceince in 1988) at Moscow State University, Russia. After emigration to USA in 1990, he  worked on modeling insect population dynamics at the Department of Mathematics, West Virginia University (1991-1992), and on modeling of population spread in space at Virginia Tech (1992-2002). He also designed a novel strategy for slowing the spread of the gypsy moth with mating disruption via pheromones applied on the basis of large-scale monitoring with pheromone traps. This strategy was implemented by the Slow-the-Spread project (http://www.gmsts.org/). On-line lecture course on Population Ecology developed by Alexei (http://home.comcast.net/~sharov/PopEcol/popecol.html) is widely used by students from many countries. One of his projects was developing 3D visualization tools for teaching (http://home.comcast.net/~sharov/3d/3dinsect.html). Early in his career, Alexei Sharov became interested in theoretical biology and continued publishing theoretical papers, which were beyond his official duties. In 1988 he organized a weekly seminar on biosemiotics at Moscow State University, and two first conferences on biosemiotics.


11 responses to “A Short Course on Biosemiotics: 1. Functional approach to the origin of life: emergence of signs in autocatalytic systems”

Comments are now closed
  1. Alexei,

    Once more, thank you for your lecture.

    I have browsed your slides again. It would be interesting to compare the description of your hypothesis at the end of the talk made first in the normal language of chemistry and then biosemiotics. It would give a better opportunity to understand the relationship between two descriptions made in two different languages.

    In the middle of your talk, you have promised to make a defintion of information. Could you please repeat it as I am not sure if I understand what you mean by it.

    You have mentioned that some agents could have free will. It would be interesting if you could discuss it in your future lectures.

    Finally a question to your statement from slide 4: “Biosemiotics assumes creative capacity of life, which comes neither from external divine powers nor from internal supernatural forces. Instead, creativity is fully compatible with physics and chemistry (but not explained by them) and is a product of evolution.”

    Currently I do not quite understand the relationships between mathematics and physicalism. It seems to me that existing of mathematics contradict to the physicalism hypothesis. To this end, an experiment

    Two Mathematicians in a Bunker and Existence of Pi

    Can biosemiotics resolve it?

  2. Alexei Sharov says:

    Thank you for your comments and questions!
    (1) The possibility of double descriptions of simple agents using the language of physics/chemistry on one hand, versus semiotics, on the other hand, indeed raises an important question: “does the nature of a system depend on how we describe it?”. Is the system still semiotic if we apply the language of physics and chemistry for its description? I think that double descriptions are important because they can show the origin of a qualitatively new system.
    (2) Double descriptions are not fully equivalent. Physical description captures the structure and short-term dynamics, whereas semiotic description captures long-term dynamics, evolution, goal-directed functions. So, in a way, theu are complementary.
    (3) Functional information is a set of signs used by agent(s) to encode and control its functions.
    (4) The simplest version of a free will is “useful variation”. In general, variation is detrimental in agents as it can disrupt functions. However, at a certain level of complexity agents can compensate bad consequences of variations. Then variations become useful as they allow agents to discover new resources and functions. Higher levels of free will exist in agents with mind who can classify and model objects. These agents can model variations in their activity before they act.
    (5) On the relationships between mathematics and physicalism. I don’t believe in the Platinic existence of mathematical entities like numbers and Pi. I think it is better to consider them as TOOLS developed by intelligent agents. Mathematical tools are applied to ideal objects that exist in the minds of these agents. Mathematical tools can be communicated in the same way as we communicate other functional information. So is Pi in the mind of one person the same as Pi in the mind of another person? Pragmatism criterion is the following: the Pi is the same if it allows to solve the same set of problems with comparable efficiency and practical output.
    (6) Ideas of pragmatism are not easy to understand. Also religion pushes people towards objectve idealism which in not compatible with pragmatism. But traditional pragmatism (e.g., James, Dewey) needs an update because it underestimates the role of logic. In my next lecture I will talk about the balance between utility and logic.

  3. Thank you for the answers.

    I see the next problem with your position about mathematics. If mathematics is a tool developed by intelligent agents, why it is possible to employ it to describe the Universe when there were no intelligent agents yet? How would you answer this?

    Now your definition of free will. If we look at this from a mathematical viewpoint, then there are either deterministic or stochastic models. Free will then after all seems to be just an illusion in the head of an intelligent agent. His action is a just preprogrammed reaction to cope with variation. How it could make a free choice?

  4. Alexei Sharov says:

    (1) Why it is possible to employ mathematics to describe the Universe when there were no intelligent agents yet?
    Answer: because our models are designed for extrapolation in space and time. The World is real and have certain regularities, but agents may have different representations and models of the world. Then why our models work? Because agents keep and elaborate only those models that work.
    (2) Free will then after all seems to be just an illusion.
    Answer: Not all illusions are bad. Why not to use good illusions? Our difference is that you would like to know what the world really is (realism, objectivism), whereas I am interested in how to live in this world and pursue my goals (pragmatism, instrumentalism).

  5. I like your answer about free will. I have to read about pragmatism. It would be nice if you describe it a bit in your lectures.

    I would agree with you about modeling in general. This means however that we must not take physical laws literally, they are then just models.

    Note however, that the statement “The World is real and have certain regularities” already contains pretty strong assumptions.

    As for mathematics, I am not that sure though. I still see a contradiction between our physical models and existing of mathematics.

  6. Alexei Sharov says:

    (1) “we must not take physical laws literally, they are then just models”.
    Yes, fully agree!
    (2) the statement “The World is real and have certain regularities” already contains pretty strong assumptions
    Answer: This statement is supported by our ability to develop consistent models of the world that give us very reasonable predictions most of the time. Somebody may argue that the world is a virtual reality engineered by god or aliens. Although it is impossible to prove that reality is not virtual, our reality is highly consistent and generally does not show signs of personal conflict with potential “creators”. So we can take it as it is.
    (3) Mathematics makes abstractions from our activities. For example, counting objects is based on a physical operation that includes picking an object from one pile and putting it into another. Although it works with pencils, it does not work with atoms or with bacteria which replicate faster than we can count. The notion of “energy” is based on our ability to replace one resource by another (in some proportion) and still achieve the same result/action. Abstract notions are meaningful only as far as we know how to interpret them in our activities. It is reasonable to assume that intelligent aliens can count objects because counting is important for advanced functions. Thus, they may have a similar core arithmetics to ours. However, math foundations and interpretations may appear entirely different.

  7. Alexei,

    If I understood you correctly, your answer to my problem with two mathematicians is that Pi exist when mathematicians are talking about it and when they die, Pi is not there anymore. The predication of a mathematical model may be correct though. Well, one possibility.

    As for models developed by agents in general, I see a problem. If you see how science is working, a theory is not reduced to empirical data. Rather a theory organizes empirical data. Hence a question, where a theory come from.

  8. Alexei Sharov says:

    (1) I am not interested in “objective” ontology because there is no such ontology. Each ontology is a product and a tool of some agents. In your case, Pi existed within the ontology of those mathematicians but it was not transferred to other agents (outside). I fully agree that “a theory is not reduced to empirical data. Rather a theory organizes empirical data”.
    (2) Where a theory come from?
    Theories come from other theories in the same way as organisms come from other organisms. The question how first theories appear is similar to the question on the origin of life. But first we need to agree on the definition of “theory”. If theory=logic, or theory=ideal object then theories appeared long before humans. However, we may think of a theory as a model formulated and disseminated via language (i.e., tertiary modeling system), then we should look for the origin of theories within human species.

  9. Valery Anisimov says:

    Thank you for presentation. I think that your approach that all live systems may be described in some universal terms of the agents and information which they managed is quite promising. In fact per example some scholars now think that best model of the consciousness and mind is some net of the “agents” which cooperate and/or fighting with other “agents”. From the other hand each agent consists of the other (lower level) set of the agents and so on. We can extend this also on the upper levels (families, companies and other organizations, nations and so on).
    But concerning more concrete question of the physical (or chemical if you want) objects from which evolution was started few billions years ago on the moment I still see some gap between your “random coenzymes world” and RNA world. For evolution to be able to take off you need some way for making a copies of the coenzymes (or short chains of coenzymes). In fact we have nearly the same problem in the classical RNA world model because as far as I know for the moment all attempts create a RNA-s which will be able to make itself replication was failed.

  10. Alexei,

    Thank you for your answers. I am looking forward to your next lectures. Just a small last note now. It yould we good if you explain what happened during the development of the Universe when there was no life. Also at what point agents and theories have emerged and what was before that point in time.

  11. Alexei Sharov says:

    Here are my responses to Valery and Evgenii:
    (1) Making copies of coenzymes is indeed a serious problem. So far nobody suggested a reasonable pathway.
    (2) “What happened during the development of the Universe when there was no life?”
    Answer: All information about that time is a result of our extrapolation using our current physical models. We describe it “as if we were there”. But models change, and so our extrapolations. It may easily happen that the “theory” of Big Bang will be abandoned.
    I am not against building ontologies, but I don’t claim them universal. Ontologies are very useful for communication between humans and for efficient modeling.