Fake Banner
    What Is Life?
    By Steve Davis | January 11th 2009 03:29 PM | 29 comments | Print | E-mail | Track Comments
    About Steve

    gadfly: noun (1) fly that stings horses and cattle. (2) (derog) annoying person, esp one that provokes others into action by criticism, etc....

    View Steve's Profile
    In 1943 the eminent physicist Erwin Schrodinger gave a series of lectures in Dublin that were later published in book form under the title What Is Life? Its success was considerable as it kick-started the new field of molecular biology, but Schrodinger deliberately avoided an investigation into a definition of life, believing that the time was not ripe.

    In more recent times, Fred Adams, Professor of Physics at Michigan University, in The Origins of Existence – How Life Emerged in the Universe, wrestled manfully with this question, but he eventually concluded that “Achieving a universal definition of life is unquestionably of fundamental importance, but no such definition has yet been forthcoming.”

    There is a noticeable reluctance among scientists to grapple with this question of life. All are happy to speculate about the conditions necessary for life to originate, but none seem inclined to actually define life itself. In The Selfish Gene for example, Richard Dawkins devoted a page or so to explaining the conditions necessary for its origin, then stated that “At some point a particularly remarkable molecule was formed by accident.” He then went on to speculate about the further development of this molecule he calls a replicator, but failed to explain to his readers what life actually is. A strident critic of Dawkins, Professor Gabriel Dover, in his wonderfully quirky and scientifically illuminating Dear Mr Darwin, described the conditions necessary for life from a galactic viewpoint, but like Dawkins he omitted a definition. Professor Freeman Dyson, another critic of selfish gene theory, in his excellent Origins of Life, did go so far as to provide the characteristics of life, as did Fred Adams, but these provide a description, not a definition. These approaches seem to typify the attitude of the scientific community to what appears to be perceived as a difficult subject, but as we press on I hope to show that perception to be misplaced.


    Adams specified reproduction and metabolism as the characteristics of life, while Dyson took a similar position but quite pointedly differentiated himself from the selfish gene view of life, first stating that “the essence of life from the beginning was homeostasis based on a complicated web of molecular structures…The tyranny of the replicators was always mitigated by the more ancient cooperative structure of homeostasis that was present in every organism,” and with an admirable touch disclosed a personal aspect to the question with the following: “I have been trying to imagine a framework for the origin of life, guided by a personal philosophy that considers the primal characteristics of life to be homeostasis rather than replication, diversity rather than uniformity, the flexibility of the genome rather than the tyranny of the gene, the error tolerance of the whole rather than the precision of the parts.”


    If we take the three specifics highlighted by Adams and Dyson, homeostasis reproduction and metabolism, (homeostasis is the ability to maintain a constant chemical balance in a changing environment, metabolism is the chemical processes that occur in cells in particular the consumption of energy, reproduction can be as simple as cell division but can involve a cell dividing in accordance with a code of instructions) we see that all of the characteristics of life are founded on cooperation, either within a cell between its parts, or between combinations of cells. If we are to examine this from a philosophical viewpoint as Dyson suggests and as I believe we should, then instead of just using physiological processes to define life we must go one step further and look in turn to their essence. But first it would be useful to examine Richard Dawkins’ description of life after origin.


    After postulating his remarkable replicating molecule, or gene, in the opening chapter of The Selfish Gene, Dawkins returned to its origins in the final chapter to examine its progress thereafter, and asked the questions “Why did genes come together…Why do they cooperate?” He then spent a full page in not answering the questions, instead explaining how they cooperate. He continued “Nowadays much of this cooperation goes on within cells. It must have started as rudimentary cooperation between self-replicating molecules in the primeval soup.” It’s undeniable, from Dawkins’ own pen as he continued, that:



    ·     Those replicators that cooperated had greater biological fitness: “it (a gene) flourishes only in the presence of the right set of other genes.”



    ·     Only those that acted collectively underwent further development: “But cooperation between genes did not stay limited to cellular biochemistry. Cells came together to form many-celled bodies.”



    ·     Those replicators that created walls to protect the society of genes had greater biological fitness: “Cell walls perhaps arose as a device to keep useful chemicals together…”



    ·     Those that acted as though the society was greater than the individual, and could undertake specialised functions that contributed to the social good had greater biological fitness: “The cells in the club can specialise, each therefore becoming more efficient at its particular task. Specialist cells can serve other cells in the club and they also benefit from the efficiency of other specialists.”


     

    It’s clear from the work of Dawkins and all other biologists’ descriptions of the endless and complex cooperative arrangements that exist in nature, from the level of molecules and compounds through to organisms and societies, that natural selection has ensured that cooperation is the principal contributor to biological fitness. It’s also clear that the cooperation engaged in by those first replicating molecules is evidence that group selection is not just a theoretical possibility that lacks evidence from the natural world as Dawkins maintains; rather it is the very mechanism that facilitated the initial evolution of life forms. Group selection would appear to be the dominant evolutionary principle.


    When we think of the truly amazing cooperative processes that make up the immune systems of modern vertebrates for example, it’s tempting to think that cooperation is more important in complex life forms than in their original ancestral molecules, but those first co-operators must surely remain at the pinnacle of biological significance.


    Can we extract from this a useful definition of life? Can we go deeper than a description of the physiological processes outlined by Adams and Dyson? It seems inescapable that life at the molecular level is actually a remarkably simple and basic concept – life is cooperation.


    The key here is to return to the very beginning and consider the process that saw lifeless molecules assume life. At what point we must ask, did they assume life? Unquestionably, when the first molecules began coalescing, then began cooperating, they began living, for it’s at that point that they began performing those functions that we generally consider to be the characteristics of life.


    There seems to be no good reason why a definition of life at the molecular level should not hold true as life forms slowly increased in complexity. Indeed, there is nothing about organisms, nothing about the myriad social arrangements existing in the natural world that undermines this definition. All are wholly dependent on cooperation for survival, and all demonstrate homeostasis, metabolism and reproduction in one form or another. (The forms may differ, but all have mechanisms for maintaining stability, for intake and consumption of energy, and for reproduction.)


    If the definition of biological life is cooperation among molecules, then possibly a useful general definition is that life is just simply cooperation. We can test this by asking the reverse question. If life is cooperation, is cooperation therefore life? It strikes me that such a case could be mounted as long as conditions are attached. For instance, an engine demonstrates cooperation between its various parts to achieve a particular end, but we would only refer to it as being alive in a metaphorical sense, as an engine needs external inputs for initiation and for sourcing of energy. The limitations of the engine example however, lead us in the right direction for a general definition. We can conclude as a general rule that independent spontaneous cooperation is life.


    There’s one important fact that stands out like a beacon from this discussion. Cooperation preceded evolution. All else is commentary. (Apologies to Wilson and Wilson!)


    Keep in mind that after the first molecules assumed life, consciousness did not come into the picture for perhaps billions of years, despite our natural inclination as conscious beings to confuse the two and equate consciousness with life. Life is far more basic and easier to explain than consciousness, and the very simplicity of the definition given here has consequences that might have contributed to the reluctance of some to delve too deeply into its meaning. But that is another story.

    Comments

    Gerhard Adam

    Steve, as you well know, I agree with you regarding cooperation as a driving force in biology, but I am also inclined to suggest that there is a third description which plays a role; neutrality.

    Many complex phenomenon occur because there is neither cooperation, nor competition, but rather a simple rule (almost algorithmic) with respect to the action that occurs.  By analogy, we could never predict the existence of an operating system, simply by examining a computer's instruction set, nor could we predict the internet by examining device protocols.  Yet, in all these cases we obtain emergent properties from relatively simple rules that define the interactions.

    While one could argue that a router in a network is "cooperating", it would be more apt to suggest that it is neutral in that it doesn't promote, nor does it demote any particular data packet, but rather it simply follows the rules it knows and as a consequence the data eventually gets to its intended destination.

    Similarly, with chemistry, we have a set of fundamental rules that govern how atoms interact to form molecules and by following those rules we obtain a wide range of chemical compounds and macromolecules that are capable of phenomenal actions.  As a result, might it not be simply that life is an emergent property of complex chemistry?

    Often this has been compared to the idea of a 747 spontaneously assembling itself, but this is incorrect since the 747 is a purely mechanistic process while chemistry is dynamic and self-activating.  DNA can be induced to replicate itself as we see with a PCR test.  A 747 could never be induced to produce another 747 because that is not intrinsic in its properties. 

    As has been discussed before, where competition occurs primarily within a group, I would suggest that cooperation is also largely a group dynamic, where the prevalent form of interaction is neutrality since that is the least costly in terms of energy and risk. 

    Mundus vult decipi
    Hatice Cullingford

    Very timely, thanks for writing this, Steve. Did you see this



    Joyce says that only when a system is developed in the lab that has the capability of evolving novel functions on its own can it be properly called life.
    where Gerald Joyce is a chemist at the Scripps Research Institute? It is about the RNA world hypothesis: In Vitro Evolution - RNA That Replicates Itself Indefinitely Created, Say Researchers
    Steve Davis
    might it not be simply that life is an emergent property of complex chemistry?
    I think that idea fits well with the view I presented Gerhard, it's a useful extra facet. Or to put it another way, same concept different angle. My dictionary definition begins; "ability to function and grow" which implies some type of activity as its essence. That activity would be cooperation for the reasons outlined.
    Thanks for your interest Hatice. I think that the standard for evaluating life given by Joyce is just a slightly more complex one than I have given, I think the two are quite compatible. I'm not sure what is meant by "novel functions." I would be inclined to require versions of homeostasis metabolism and reproduction to be base functions, but that's just a personal view. It will be interesting to see if they come up with something substantial.
    To the image of the 'cooperative molecules', the same arguments as those against the 'selfish gene' metaphor can be opposed.

    Steve Davis
    To the image of the 'cooperative molecules', the same arguments as those against the 'selfish gene' metaphor can be opposed.
    Not really. You need to state your objections. If, as is generally understood, cooperation is working together, then cooperation among molecules is a fact of life. It's not clear from your comment whether you support selfish gene theory or oppose it. 
    ricochet17
    For those who have read Shrodinger's book, check out Ed Regis's book entitled What Is Life?: Investigating the Nature of Life in the Age of Synthetic Biology.A short book, and well worth the read.
    The problem I have with these definitions of life is that they are too narrow. What about an aritifically created "life" in the form of an artificial intelligence? It is not clear how the biological features of these definitions apply. Yet a "thinking" entity that passes the Turing test is certainly likely within our some of our lifetimes. Does such an entity deserve to be considered as alive, and accorded some measure of "rights"? Josh Storrs Hall covers many of the ethical issues in his paper "Ethics for Machines". But I think that a more meaningful definition of life awaits a broader perspective informed by experience with dramatically different kinds of life forms. Maybe some of the physicists were properly reluctant to limit their view of life to just biology as we know it now?

    Gerhard Adam
    "What about an aritifically created "life" in the form of an artificial intelligence?  Yet a "thinking" entity that passes the Turing test is certainly likely within our some of our lifetimes."

    I don't believe that's particularly relevant.  This presupposes that a criteria for life is some vague definition of "intelligence".  Regardless of how sophisticated a machine may be, it isn't the "intelligence" that matters in measuring artificial life, but rather whether it possesses self-motivation.

    There are machines that can beat me at chess or a variety of tasks, but there aren't any that care whether they do or not.  Without that aspect of it, there's no point is worrying about artificial life.
    Mundus vult decipi
    Gerhard Adam
    Does such an entity deserve to be considered as alive, and accorded some measure of "rights"?

    I'm also not clear why such a question is even considered.  This illustrates one of the most common biases in these discussions, in presupposing that being "alive" is somehow synonymous with being human.

    We don't worry about rights for viruses, or insects, or the millions of organisms that populate this planet.  Why should we engage in a discussion to put machines on a parity with humans?
    Mundus vult decipi
    I found Gerhard Adam's responses to be fairly short-sighted. His first response creates a straw man that is deliberately limited in capability, and then says it is not alive. Why not consider the possibility of a less limited artificial intelligence? Capabilities are expanding rapidly. His second response is the amoral one. We have no need to consider the rights of other living beings. He should read the article I referenced (easily found on the web). Then he might understand why an artificial intelligence in the not-too-distant future might have as much respect for human capabilities as we do for many other animals. If there is "life" elsewhere in the Universe, it is likely to broaden our definition of what life really is, not narrow it. Why start with a definition that is likely to be out-moded in less than 50 years?

    Gerhard Adam

    I'm sorry, but I did read the paper you referred to and it's simply science fiction and fantasy.  The author proposes the following fantastic conclusion:

    "These are crucial questions for us, for not too long after there are computers as intelligent as we are, there will be ones that are much more so. *We* will all too soon be the lower-order creatures. It will behoove us to have taught them well their responsibilities toward us."

    This is such complete utter nonsense, that it isn't even logically consistent.  The view suggests that (1) we know how to "teach" such values (which we have never succeeded in doing even amongst other humans), (2) that we are intelligent enough to convey such a concept to something MORE intelligent (despite the fact that we still don't know what that is), and (3) that somehow such a higher intelligence would view us as the benevolent creator.

    This is complete rubbish.  By definition, if such a machine would be alive, it certainly wouldn't look to a "lower organism" for guidance any more than you ask your dog for advice.

    Furthermore the paper proposes actions for concepts that aren't well defined nor understood (intelligence, conscience, etc.) and then makes the fantastic recommendation that these should be incorporated into computers (lest they develop and evolve without such guidance).

    "His second response is the amoral one. We have no need to consider the rights of other living beings."

    I'm sorry, but I am not arrogant enough to suggest that it is up to me to bestow "rights" on other living organisms.  A "right" is a political definition and has no bearing in dealing with other species that are (1) incapable of participating of their own volition and (2) dependent on our largesse for recognition of this quality.

    Whatever attitudes exist towards animals, your use of the word "amoral" already restricts it to a human-only experience.  There can be no discussion of morality between members of different species that are incapable of formulating what that is.

    The most obvious problem with the whole scenario (which the AI folks never seem to get) is that unless they can reduce intelligence to an algorithm they are hoping to cobble together enough elements to result in an undetermined emergent property.  If it is determinable then it is algorithmic, if it's indeterminable then how can we assume to know how it would manifest (and it would be insane to try - sort of like randomly combining genes in viruses just to see if something cool happens).

    Mundus vult decipi
    Gerhard Adam
    In thinking about the issue of AI some more, I can't help but marvel at the idea that a supposedly intelligent species would contemplate and even look forward to the idea of building a superior competitive species.

    Words fail me....
    Mundus vult decipi
    Steve Davis

    Does such an entity deserve to be considered as alive?
    I don't think so, as it needs external inputs for initiation and for sourcing of energy.
    But I think that a more meaningful definition of life awaits a broader perspective informed by experience with dramatically different kinds of life forms...If there is "life" elsewhere in the Universe, it is likely to broaden our definition of what life really is, not narrow it.
    But how will we know if dramatically different life forms are actually alive without a general, workable, acceptable definition or theory of life?

    Steve Davis says:
    Does such an entity deserve to be considered as alive?
    I don't think so, as it needs external inputs for initiation and for sourcing of energy.

    All the entities you include in your definition as being alive seem to also require some external sourcing of energy. People don't live long without food, and plants require light for energy...

    And the "initiation" is an interesting metaphor. Isn't DNA an "initiation" or a program. Is it so different than software?

    Writing this off as fantasy or science fiction is once again short sighted. It is like someone calling the moon landing a fantasy after Sputnik, and after JFK announced the plan to accomplish it. The means and the will have already been demonstrated. Significant steps have already been achieved. It is more a question of when than if.

    I agree with the concept that consciousness is not required for life. But can an entity that is conscious not be considered alive?

    Gerhard Adam
    "All the entities you include in your definition as being alive seem to also require some external sourcing of energy"

    Without speaking for Steve, I will surmise that his point was that they cannot obtain their energy means by their own volition (or motivation).

    "Isn't DNA an "initiation" or a program. Is it so different than software"

    While it may be viewed in a analogous manner, it is a vastly different process. 

    "It is like someone calling the moon landing a fantasy after Sputnik, and after JFK announced the plan to accomplish it. The means and the will have already been demonstrated. Significant steps have already been achieved. It is more a question of when than if."

    Sorry, but this is where you lose me.  The moon landing is hardly on par with creating an independent, conscious, intelligent life-form.  In the first place, the physics and processes involved in the moon landing were well understood (even if they had to be developed), while even the concept of "intelligence" is barely coherently defined.  While there is arguably "the will", the means have by no means been demonstrated by any definition.  The steps that have been achieved are trivial (compared to even the most limited forms of intelligence) and involve little more than brute force usage of algorithmic processes.  The problem of creating real artificial intelligence (versus just a fast encyclopedia) is potentially intractable.

    AI has always been hopelessly optimistic about solving a problem that they can't even define.
    Mundus vult decipi
    "I think, therefore I am" - René Descartes - 1637.

    Steve Davis

    All the entities you include in your definition as being alive seem to also require some external sourcing of energy. That's correct, but they collect the energy themselves, as in a plant collecting energy from the sun, or a human or other animal society organising the external collection and internal distribution of energy.
    And the "initiation" is an interesting metaphor. You're right, this is probably the most awkward part of the definition, but keep in mind the process I followed to arrive at a definition. The first complex molecules that came together and began the cooperative processes we now consider to be the characteristics of life, initiated that process independently of other entities. That cannot be the case for "artificial life forms". Now it could be argued that this excludes modern organisms from the definition as we/they were initiated by parents, but as we/they represent an uninterrupted flow of life that came from those first cooperators,  I think the definition holds up OK. 

    All the entities you include in your definition as being alive seem to also require some external sourcing of energy. That's correct, but they collect the energy themselves, as in a plant collecting energy from the sun, or a human or other animal society organising the external collection and internal distribution of energy.

    One of the common themes in artificial intelligence, and the related area of (programmable) nanotechnology is the issue of replication and construction. So if the artificially intelligent (life) form gets its energy from solar panels it makes itself does that bring it in line with this part of your definition?

    I also think that you could extend your concept of an uninterrupted chain of development to artificial intelligences in a way similar to your inclusion of "modern organisms". They are the creation of people. Just as modern artificial life forms like human-animal chimeras are the creation of people, not a "natural" product of evolution, and are clearly alive.

    Considering a thinking and conscious artificial intelligence as being "alive" seems so obvious to me that any definition of "life" that excludes it seems deficient. I am convinced that current concepts of molecular biology are insufficient to circumscribe all possible life forms, and that a definition that limits life to that set is way too narrow.

    Gerhard Adam
    "One of the common themes in artificial intelligence, and the related area of (programmable) nanotechnology is the issue of replication and construction. So if the artificially intelligent (life) form gets its energy from solar panels it makes itself does that bring it in line with this part of your definition?"

    Actually it doesn't.  You are equating life with something akin to an emergent property that "appears" to be alive.  When a living thing reproduces, every aspect of it is alive from the single cells engaged in fertilization to the developed offspring.

    A machine building another machine isn't alive at any stage.  The solar panels aren't alive, and neither are any of the mechanical components.  In short, you cannot claim life if none of the components are themselves alive.  Therefore the only claim to life that can be made is that somehow the appearance of consciousness or intelligence would bridge that gap. 

    Which brings us back to the original problem since you would have to have some means of assessing intelligence or consciousness as an independent property.

    Another point to consider is that if intelligence or consciousness exists, then an extension of that is "free will".  For such a thing to exist in an intelligent being, then that means the being must be capable of being uncooperative or deceptive.  Should such a thing be possible, then it would be insane for us to build it (if it were possible) and it would render it useless to humans in any other capacity.
    Mundus vult decipi
    Steve Davis
    Anonymous, your questions are good ones, and I think Gerhard's responses are nicely adequate. But I cannot see the urgency in the thrust of your argument. Surely the entities you have in mind can be referred to as artificial life forms, no matter what level of complexity they can be taken to. Why do they need to be granted the status of biological life forms? And maybe there are good practical reasons for not allowing a blurring of the differences.
    In general I'm sympathetic to the points made here. My only concern is that we're reducing one definitional problem to another one. What does it mean for an entity to be cooperative, or for multiple entities to cooperate? Wiktionary gives (inter alia) "To work or act together, especially for a common purpose or benefit" There's a teleological aspect to cooperation that I'm not quite comfortable with.

    Gerhard Adam
    Consider that there doesn't really need to be any particular intent behind cooperation initially.  Aggression is risky and dangerous, so the primary mode of interaction is generally going to be indifference in the absence of any particular competitive element.

    Simple "cooperative" or social organizations may involve nothing more than animals congregating together and deriving advantage by their larger numbers which reduces the probability of individual predation (like herd animals). 

    Even among predators it isn't difficult to see how multiple predators could participate in pursuing such a herd.  At some point, such predators would have to develop some sense of awareness regarding their mutual interest and it's even possible that the initial point of sharing or cooperating could be nothing more than being too tired to fight or challenge each other, especially if resources are plentiful.

    From these simple beginnings, it would simply be a matter of those animals that developed a more cooperative spirit may have been more successful that those that didn't.  Especially in cases, where such cooperation may have facilitated obtaining larger game, etc.

    In any case, it doesn't require any more formal purpose than such small developments.

    Mundus vult decipi
    @Gerhard -- I agree with everything you're saying, BUT...SD somewhere in the middle of the post/article says
    It seems inescapable that life at the molecular level is actually a remarkably simple and basic concept – life is cooperation.

    And really it's the notion of cooperation at this cellular/molecular level which I think is wrong-headed, or at least in need ot clarification.

    Gerhard Adam
    While the terminology may seem awkward, we have to consider that at some point, single-celled organisms have the ability to share genetic information and ultimately formed colonies (corals, jellyfish, etc.).  From this, cells eventually formed multicellular organisms which are definitely the product of cooperation.  It could certainly be argued what form of cooperation this was, or whether it was "coerced" behavior, but in the end, a multicellular organism is only viable because the cells are cooperating to fulfill the obligations of the full animal.

    Cells that do not cooperate are detrimental and usually characterized as cancerous.

    The difficulty is that we tend to think of cooperation as something that requires a conscious decision, but perhaps the difficulty really lies in "us" thinking that we have more control over this than we actually do.  Just as we can acknowledge cooperative behavior in higher animals, it makes sense that single-celled organisms that became capable of dividing the work and maximizing the benefit per energy expended through cooperation would tend to be favored in niches that might otherwise be too difficult to exploit.

    It also doesn't require any special agency, any more than considerations like electrons tending to move to their lowest stable energy state.  In other words, it simply becomes the "path of least resistance" using the standard laws of physics and chemistry.

    Another thought that struck me was envisioning a group of people putting sandbags along a riverbank to avoid flooding.  We could certainly take the long-view and say they were cooperating in this effort, but we could also argue that it would be impossible to distinguish cooperation versus self-interested behavior, simply because they were working towards the same objective.  In other words, in this example, is it really necessary for people to be cooperating versus each simply operating in their own "world" to place sandbags?

    I don't know if that sounds like gibberish since it's pretty late here, but it's a thought.
    Mundus vult decipi
    Steve Davis

    That's a perfectly reasonable query Fred. I think the answer is in the definition you gave, which is very similar to the one I made use of in writing the article. While it would be a stretch to talk of complex molecules working together for a common purpose, it's reasonable to suggest a common benefit, and that would satisfy your definition of cooperation.

    The benefit in this case is survival. Those first molecules that came together to begin the functions we now consider to be characteristics of life were the molecules that survived.

    Steve
    Thanks for all the thinking you've done to reveal the common sense of life. My question is this, would you agree with the statement that the biologic purpose of all sentient life forms is to survive, thrive, and evolve, the method we are genetically predisposed to use to accomplish our purpose is Cooperation?

    Steve Davis
    Sorry for the delay in responding; I've been away.
    In general terms "yes" to your question, except that I think the purpose of life is a little simpler than your summary, I think the purpose of life is the creation and/or nurturing of life.
    And I see cooperation as preceding evolution, therefore preceding genetics, but I agree that cooperation is the means we use to accomplish that end.
    Steve
    thanks for the reply, I have gone back and read most of your articles--really great insights and contributions to our collective foundation of logic on the patterns of human evolution. Your insights compliment many of the lessons and principles I am sharing in my new book (my 2d) and I would be honored to talk over the phone or ask you a few questions via email to elaborate on a few of these points--if you agree, I would of course quote you in the final manuscript
    Thanks again
    Pete

    Steve Davis
    Sure, there's an email link at the top of my profile page.