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. 
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.”  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). 
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.
- Koshland, DE. 2002. The Seven Pillars of Life. Science 295: 2215-2216
- Korzeniewski, B. 2001. Cybernetic Formulation of the Definition of Life. J. Theor. Biol. 209: 275-286