What is Life?
Cybernetic Evolution.
By Mike Gene (
Biologists find themselves in the uncomfortable position of studying something that is difficult to define. Traditionally, they define life by listing several features characteristic to life, including metabolism, growth and development, responsiveness, and reproduction. Because of its central importance in evolution, many emphasize reproduction. Yet Daniel Koshland tells this story:
What is the definition of life? I remember a conference of the scientific elite that sought to answer that question. Is an enzyme alive? Is a virus alive? Is a cell alive? After many hours of launching promising balloons that defined life in a sentence, followed by equally conclusive punctures of these balloons, a solution seemed at hand: "The ability to reproduce--that is the essential characteristic of life," said one statesman of science. Everyone nodded in agreement that the essential of a life was the ability to reproduce, until one small voice was heard. "Then one rabbit is dead. Two rabbits--a male and female--are alive but either one alone is dead." At that point, we all became convinced that although everyone knows what life is there is no simple definition of life. [1]
Koshland himself defines life by citing seven features that are both universal and essential to life: “P(rogram), I(mprovisation), C(ompartmentalization), E(nergy), R(egeneration), A(daptability), S(eclusion), PICERAS, for short--are the fundamental principles on which a living system is based.” [1] That life resists attempts to define it in an efficient and simplistic manner may speak to manner in which reductionism fails to account for it.
In 2001, Bernard Korzeniewski used a cybernetic approach to come up with a definition of life that turns out to be quite helpful when thinking of evolution as a process that was designed.
Korzeniewski sets out his definition:
Life (a
living individual) is defined as a network of inferior negative feedbacks (regulatory
mechanisms) subordinated to (being at service of ) a superior positive feedback
(potential of expansion). [2]
First, let’s define negative and positive feedback. Negative feedback occurs when a change is detected and a system responds by reversing the change. Think of the manner in which your thermostat controls the temperature in your room during a cold winter day. When the temperature drops (the change), the thermostat senses the change and triggers the furnace, which blows hot air into the room to reverse the change. Positive feedback can be envisioned by a crowd of people in a theater. One person smells smoke and shouts “fire!” Others nearby then notice and likewise shout “fire.” The shouting (and panic) then spread quickly throughout the theater. Positive feedback thus works when a reaction serves to enhance the change. With that said, let’s now turn to Korzeniewski’s argument.
Korzeniewski notes that in order to formulate a definition
of life, we need to “find properties that can be
attributed exclusively to living individuals, which will allow life to be
clearly separated from inanimate matter.” He pays tribute to the attempts to define
life in terms of its properties and explains the relative success/appeal of
this approach as follows:
There are
many properties characterizing the presently known living forms. The set of
features usually mentioned in such contexts generally allows one to separate
fairly well a phenomenon of life from inanimate phenomena, i.e. distinguish the
former in a rather univocal way. There are two main reasons for this fact.
Firstly, in spite of the enormous diversity of life forms, the most fundamental
principles of biochemical construction and function are astonishingly similar
(in fact: identical) for all organisms existing presently on the Earth.
Secondly, even the simplest live entities, to which some bacteria and
archeabacteria belong, still exhibit a very high degree of complexity (some
much simpler forms, viroids and viruses, being parasites, are not able to live
independently). Therefore, the complexity itself already seems to be a good
determinant of life.
He then (unconsciously) pays tribute to the importance of teleology:
One of the
very typical properties of living organisms consists in the “purposefulness''
of their function that manifests itself in the form of various regulatory
mechanisms (mostly negative feedbacks). Nevertheless, negative feedbacks are
also built into many devices constructed by people, thermostats or robots, for
instance.
Purposefulness is not accepted as the criterion since it would also ensnare manmade machines and devices as “life.” Yet from a teleological perspective, this might very well suffice as a definition of life: Life is a non-manmade phenomena that nevertheless demonstrates the purposefulness as seen in manmade phenomena. Yet this is a tangent, so let’s get back to Korzeniewski.
He then settles for three universal biotic properties:
Life on
the Earth (and, in the author's opinion, life in general) seems to possess
three properties (strongly related to each other and in fact being different
aspects of the same thing) which are absent in inanimate systems. Namely, life
is (1) composed of particular individuals, that (2) reproduce (which involves
transferring their identity to progeny) and (3) evolve (their identity can
change from generation to generation).
After exploring these, he offers the following observation:
The third
aspect of life - impossible to be considered separately from the other two
aspects – is the fact that life is composed of particular individuals, and does
not constitute a certain continuous mass. It is an individual that is the
“carrier'' of identity, that reproduces and evolves (an individual is
understood here to be the whole life cycle of a given organism, with all its
stages from birth to death). An individual, reproduction and evolution are
three strictly connected faces of life. Therefore, in order to define life, one
has also to define a living individual. The classical paradigm considers an
individual to be a structurally and functionally integrated entity.
To explain the “individual,” Korzeniewski turns to cybernetics:
In the
present article, attempting to formulate possibly a minimal definition of life,
a living individual is defined within the cybernetic paradigm, as a system of
inferior negative feedbacks subordinated to (being at service of ) a superior
positive feedback…..The full set of negative feedbacks (regulatory mechanisms),
working on different hierarchical levels and representing the cybernetic aspect
of the function of a living individual, has the “purpose'' of sustaining the
identity of the individual. In turn, the only “purpose'' of this identity is to
reproduce itself in as many copies as possible….. Cybernetics (Wiener, 1948)
defines the negative feedback as the main regulatory mechanism, in which any
deviation of the value of some parameter from the assigned value exerts an
effect which counteracts this deviation and leads eventually to maintaining the
parameter value on a more or less constant level. The thermostat in a
refrigerator is a typical example of the negative feedback in the inanimate
world. From the point of view of purposeful action (involving purposeful
regulatory mechanisms), living organisms are in fact “built of '' a huge amount
of hierarchically organized negative feedbacks.
He provides many examples of this at work on a molecular, cellular, and organismal level and notes:
All these negative
feedbacks are organized hierarchically and mutually interconnected (directly or
indirectly). For example, insulin and glucagon are proteins. Therefore, their
synthesis involves (among others) the above-mentioned negative feedbacks
responsible for the regulation of amino acid synthesis, ATP production and
protein synthesis. A similar connection exists, for instance, between the
muscle contraction system and the energy (ATP) production system. Plenty of
analogous examples can be easily found. In general, all the negative feedbacks
within an individual form a hierarchically organized network, where all
negative feedbacks are mutually dependent (directly or indirectly).
He then explains how this hierarchy of negative feedback systems serve a higher positive feedback system (with added emphases below). At this point, a new perspective on evolution can emerge.
Therefore, the “purpose'', in the functional sense, of each
particular negative feedback is determined by the context of the whole complex
of negative feedbacks, constituting the identity of a particular living
individual. In the cybernetic sense, it is
exclusively by an appropriate network of negative feedbacks (its functional
identity) that the function of an organism can be described. The only
“purpose'' (in the biological sense) of this identity is to preserve its own
existence in time, that is to survive in current, specific environmental
conditions, as well as to produce as many copies of itself as possible. The entire network of negative feedback
mechanisms is ultimately directed at the latter task. Within the cybernetic paradigm, however, reproduction is nothing but a
positive feedback…..In the positive feedback, faster the rate of increase
of a given parameter value is, the greater this value already is. For example,
in the case of a nuclear chain reaction (or a stony avalanche), the increase in
the number of free neutrons (rolling stones) is proportional to the current
number of free neutrons (rolling stones). The same property characterizes the
reproduction of living organisms, e.g. bacteria in a medium or freely
reproducing rabbits: their number increases exponentially in time (the increase
in the numerical force of rabbits in time is described by the famous Fibbonacci
sequence). Of course, the increase in the number of living individuals is
normally limited by the capacity of the environment. In stable ecosystems, their number is
approximately constant in time. Therefore, in this case, there is no expansion
of living individuals (increase in their numerical force), so characteristic
for the positive feedback. Nevertheless,
even here the positive feedback (resulting from the network of negative
feedbacks underlying it) sustains a potential of expansion (imposed by the
pressure of reproduction), which can express itself as a competition of a given
individual with individuals carrying different identities (of either the same
or different species), or as the process of populating new areas (or habitats).
Therefore, even in populations remaining in stasis, the realization of the
superior positive feedback remains the main “purpose'' of the regulatory
mechanisms that underlie the function of living organisms.
The (identity of the) cybernetic
individual-defined above as a system of negative feedbacks at the service of a
superior positive feedback (potential of expansion)-automatically becomes the
subject of evolution.
The key to Korzeniewski’s cybernetic definition is in coupling the negative feedback systems to positive feedback (with emphases added):
Many examples of positive
feedbacks and negative feedbacks can be found among inanimate phenomena. On the other hand, they never occur there
in the above-quoted combination. The positive feedback is present in both
natural phenomena (stony avalanche, nuclear chain reaction) and artifacts built
by people (amplifiers). Let us consider the stony avalanche. Here, the
parameter the value of which increases exponentially in time (at least at the
beginning) is the number of rolling stones. However, rolling stones do not have
any purposeful regulatory mechanisms (negative feedbacks) to sustain their
movement and eventually the whole avalanche decays at the bottom of a slope.
Moreover, the “identity'' of rolling stones adopts only a rudimentary form: it
is just their movement. Therefore, of course, a stony avalanche is by no means
alive in terms of the cybernetic definition quoted above….Negative feedbacks do not seem to occur in the realm of the inanimate.
However, they are quite frequent in
artificial devices built by people, such as robots or a mere thermostat in a
refrigerator. Robots can even possess a hierarchically organized network of negative feedbacks. All
such devices perform a purposeful
function directed at some task: maintaining constantly low temperature,
manipulating objects and so on. However, they lack the superior purpose-the
positive feedback. Instead, they serve aims imposed on them by people, and not
their own interests. Although they have some identity (at least robots), they
cannot propagate this identity by themselves. As such, they are not autonomous
in their purposeful action-they are a part of some bigger system (human
technical civilization).
The description above amounts to a subtle, yet distinct, paradigm shift. It doesn’t contradict a non-teleological view of life and evolution; it simply renders them myopic. If life is a designed cybernetic system, front-loaded to carry out future objectives, evolution (even Darwinian evolution) becomes a function of the way life was designed. That is, Darwinian evolution becomes a part of a intended positive feedback loop, unleashing life’s inherent potential to modulate its systems of negative feedback systems to adapt to any changes in an immensely complex and noisy environment. This process of modulation may then have been rigged as part of front-loading, such that a “direction” to evolution is possible. It would be interesting to consider the possible machinery (itself under feedback control) that works to tap into the positive feedback of expansion. In other words, the cell is not a passive player simply reacting to random mutations. It is an active player in its own evolution. More on this later…..
In the meantime, there is something in Korzeniewski’s article that may be of interest to fans of William Dembski’s approach to design. Let’s set the stage.
Life and Possible Routes to Determining Specifications
One of the complaints against Dembski's design inference is that we can't apply the concept of specification to biology. The problem, it seems, is that common examples of specification point to such things as words and games (cards, dice) where we know the specifications before hand as a function of human experience with such things as language and games. But when it comes to life, we don't know such things. For example, if letters were arranged by an ETI, chances are that we could not determine if the arrangement was due to chance or ID (since we don't know ETI language).
However, I think this problem fades somewhat when we turn from the static arrangement of letters to the dynamic reality of machine-like things. In the case of letters, the significance of arrangement depends on our knowledge of language. But when it comes to machines, the significance of the arrangement does not have this dependence, but instead shows itself in what the machine does. Because the machine will do something regardless of what we already believe or think. And we can figure out what it does and why it does what it does.
Say we land on another planet and find something that looks like a machine. If the criticism against Dembski applies, we can never ever hope to understand this thing as a machine because we have no experience designing non-human artifacts. So is humanity thus forever blinded from finding evidence of ETI? Might not we instead study the thing and eventually stumble upon a way to activate it, such that various parts begin to move in a coordinated fashion so that something happens (that didn't happen when the machine was still)?
From here, we might speculate what the machine does as its function. And if this is the case, what the machine does depends on the number of parts, their shapes, and their arrangement. What's more, it is the function which explains the existence of the parts, their shape, and their arrangement. The function thus becomes that which serves as a basis for specification.
Okay, we turn to life. But by life, I do not mean some abstract vital force. I mean the concrete representation of life, things we know and observe as cells. Cells can be viewed as roughly analogous to finding an alien machine on another planet. Why? Just as the machine is "out of place" in the context of the planet's natural surroundings, so too are cells "out of place" in the context of the non-living world. Just as the machine does something as a consequence of moving parts, so does the cell. Just as the existence of the machine's parts depends on what it does, so too do the parts of the cell.
With this in mind, let's begin to outline possible specifications involved in making cells (so that cells do what cells do). I'll approach this topic with increasing resolution.
Low Resolution Specifications
As I see it, the specifications of life have to do with channeling at least three flows: the flow of information, the flow of material, and the flow of energy (this insight is acquired from the knowledge of how engineers go about designing their products). To get life, I would propose that these three lines of flow must communicate in a coherent and specific fashion. And flow and communication presume compartmentalization. Thus, understanding the nature of how things are compartmentalized and how they communicate through these three lines of flow gets us not only an understanding of life, but also the specifications employed by life. We are beginning to understand parts of this communication and flow thanks to the increased sophistication of our own designs (which we can use as crude models to understand life processes). And as with engineering, the three flows often blur into each other, but can be specified such that one flow is primary.
So let me spell this out further. As noted above, life is about channeling flow between compartments where such channeling is best interpreted as communication. The theme of compartmentalization is actualized in what we call a cell (all life is built upon cells; there are no exceptions). The chemical infrastructure needed by cells demands a core level of highly constrained ensembles of molecules, which more often than not, need to coexist at the right time, place, proportions, and positions. This channeling both sustains and employs this infrastructure such that a population of cells is quite different from a biochemical extract formed from the same population. Such a state is far from an equilibrium state and is maintained through the channeling of the above three flows.
Now, we turn to the wisdom of Richard Dawkins, who observes:
"however many ways there may be of being alive, it is certain that
there are vastly more ways of being dead…the islands of suvivability, however
large and however numerous they may be, are miniscule in size and infinitesimal
in number when compared with the ocean of dead unworkability"
If we consider all the parts and processes that make up the non-living world, I think Dawkins' logic holds here: however many ways there are of having these parts and processes give rise to the Three Flows, there are vastly more of ways they can exist such that the Three Flows do not exist.
Medium Resolution Specifications
To mediate and integrate each of the Three Flows, multiple components are needed such that they interact in ways to maintain the flows. Information flow requires a sender-encoder, which emits a coded message, a transmission channel, and a receiver-decoder-amplifier-responder. Energy flow requires components that extract, convert, store, and channel energy. Material flow requires components that synthesize material, transport cargo and construct architectures. All flows don't exist independently, but instead communicate with each other. Finding the minimum number of components sufficient to carry out and integrate these flows would entail further specifications for cells/life. And again, Dawkins' logic would hold: no matter how many ways you can hook up things to mediate these flows, there are a vastly higher number of ways to hook up things such that these flows do not exist.
High Resolution Specifications
At this level, the components are considered to determine how they carry out their role in each respective flow. How does one store the energy of an ion gradient into a chemical form (ATP)? How does one move one component from one part of the cell to another part of the cell? How does one decode the information that is encoded? At this level of resolution, we confront such concepts as molecular machines, proof-reading, and quality control. And what it interesting is that you don't find much variability at this level: whether we are talking about cell division, translation, or ATP synthesis in bacteria or humans, the machines, proof-reading, and quality control mechanisms used to maintain these cellular flows are basically the same. In this case, we can say that however many ways we can hook up macromolecules to work as machines, proof-reading devices, and quality control managers, such numbers are miniscule in size and infinitesimal in number when compared with the ocean of dead unworkability.
Return to Korzeniewski
A common theme among the three levels of resolution is communication. Thus cybernetics, the science of control and information, is a natural place to look for more insight. And the cybernetic perspective offered Korzeniewski is quite helpful in these efforts. He explains:
Being dissipative structures
(Prigogine, 1980; Prigogine & Stengers, 1984), living organisms are systems
essentially displaced from thermodynamic equilibrium. They are characterized by
low entropy, and therefore by a large amount of information (
The purposefulness of living
individuals is strictly correlated with the network of regulatory mechanisms
(negative feedbacks) serving their main purpose: survival and reproduction, as
it was discussed above. Therefore, the author would like to propose the
following expression for the amount of the biologically relevant (purposeful)
information:
IB= log2(n - 2)
where IB denotes the
amount of biologically relevant (purposeful) information (in bits) and n stands
for the number of the negative feedbacks present in a given biological
individual. The number 2 in the brackets corresponds to the minimal number of
states a negative feedback loop can distinguish (and adequately respond to):
the value of the regulated parameter differs (state 1) or not (state 2) from
the assigned value. It is possible to extend this definition and to include a
greater number of different values of the relevant parameters recognized by
particular negative feedbacks.
Fans of Dembski may appreciate this attempt to quantify biologically relevant information because of the criteria of detachability. Paul Agutter et al. comment:
To maintain a constant internal environment requires control mechanisms: sensors, effectors, information processing and feedback systems. These terms were imported into the language of twentieth century physiology from control engineering in the 1940s – but they do not belong to the language of physics (or, indeed, of nineteenth century physiology). [3]
Cites: