Genitum, Non Factum
Thoughts on a Biological Turing Test
Page 3 of 3
Table of Contents
A Sketch of Life
A cartoon sketch of a single cell serves as a central reference point. In order to make some kind of progress in figuring out what, if anything, differentiates Mr. Sparkles from the tumblebot, we need a picture of what a living organism is. We can draw this as a minimal sketch. This basic picture recurs within the enactive literature and I will present just enough of it to help us along with our goal. Consider a cartoonishly impoverished single cell. It is alive, which means that it is made up of a set of interlinked processes that sustain its being. If the processes of metabolism end, the living cell ends. It is thus not co-extensive with any specific set of material elements, but with a suite of mutually sustaining processes. Those processes are critically dependent on the regulated exchange of material with its surround. Raw material for metabolism needs to be found and incorporated, and the useless end products need to be jettisoned. All life works like this, if we take a sufficiently abstract view.
A simple cell in a petri
dish, observed by a near-omniscient scientist.
Next we need to provide the cell with a means of getting
around. This will serve as a general and highly abstract way of
characterising the activity of the living, irrespective of the
specific form of sensorimotor embedding. The usual description
employed in the literature is based loosely on an E. Coli
bacterium. It makes use of small flagellae at the rear which can
operate in a coordinated fashion to move it in a straight line, or
in an uncoordinated fashion, in which case movement is random.
Suppose further that there is a glucose source somewhere a bit
away from the cell. Glucose, in this small account, will play the
role of everything that is necessary and desirable in life, or
somewhat more prosaically, it is a nutrient. The concentration of
glucose will increase as the cell gets nearer the source, and it
is thus in the cell's interest to approach it. Finally, let the
probability of switching from the random mode of movement to the
directed mode (or vice versa) depend on the rate of change of the
local concentration of glucose. If we have recently registered an
increase in glucose, then the chances of switching from directed
to random mode should be reduced, as things are going well.
Conversely, if we have registered a decrease, then the chances of
switching from directed to random mode should increase, as a new
trajectory might help.
At this point, our cell is nothing but a mechanism. What we have described is, in fact, a machine, and not a living being. It is a machine, because we designed its components and their interaction, and we put them together in just such a way that the cell would move towards the source. That is, we introduced an engineering goal that we conceived, and we made it manifest in our toy cell. But imagine, if you will, for the sake of argument, that this cell had somehow evolved, and it really was this simple. Look at the various parts: There is "perception" in that the cell senses or measures the ambient glucose concentration. There is action, in that the cell moves in one of two modes, and changes its behaviour every now and then. Perception and action are inextricably linked, by tying the probabilities of switching from one mode to the other to the ambient glucose concentration. And finally, this little sketch has been drawn in order to provide us with a picture of how a purely mechanical implementation can appear, to an outside observer, to act in its own interests. [Table of Contents]
Our cartoon cell encounters a cartoon world. In the spirit of Von Uexküll, let us consider the world this cell encounters. The minimal construction of the example allows us to state with precision what the world is that arises for the cell, and we can do so using either the vocabulary of perception, or by considering the propensity of the cell for action. Either way, a single direction ("Glucose, Ho!") is picked out of an otherwise undifferentiated totality. Movement is preferred, and hence more likely, in this direction rather than that. We made a machine that registers precisely one kind of distinction in the world, and that exhibits one kind of goal-directed behaviour.
The enactive framework allows us to appreciate the values of the cell, while recognizing them as the cell's own. But now we take this example and play with it a little. First off, we might contrast the scene that the observer meets (a cell in a petri dish on the lab bench in the lab on Wednesday afternoon) with the "experience" of the cell (no lab, no dish, no Wednesday). To the observer, glucose is one chemical among many, neither intrinsically good nor bad. To the cell, glucose is valenced, that is, it is of positive value to it, because of the role it plays in metabolism, which in turn is the sine qua non of the cell's existence. In this constrained and principled manner, we see the biological insistence that values are part of life, and that recognition of values is necessary if the living and their goings on are to be understood.
In matters to do with living beings, there is no view from nowhere we can avail of. Secondly, we need to notice that the dispassionate, remote view of the observer is, itself, something of a fiction, as the observer is likewise a living being for whom a world arises as a function of his or her constitution and history. Only in the narrative framework of our little sketch can we pretend that the observer is neutral, and has access to "reality" without qualification. In fact, no neutral position ("the view from nowhere") is ever available.
Subjects making sense and encountering Umwelten, instead of interior minds perceiving alien exterior worlds. Finally, we bring in the most powerful tool in the armoury of this Enactive approach. Instead of hypothesising, and then reifying, a subject or mind who knows the world through sensory input, and acts on it as a form of output, we replace all that baggage with a single activity: sense-making. The whole business of sensing the ambient glucose concentration (standing here for a cluster of concepts around sensation and perception), of behaving one way or another, and of doing so in the service of the goals of a singular subject, and with that the arising of a lived perspective of that subject, all that can be subsumed under the term "sense-making", which seeks to characterise in the most general way possible the goings on of the animate. This vocabulary is not easily aligned with received psychological stories. [Table of Contents]
Sense-Making, alone and together
There are different ways to come at sense-making. We can think of it as dealing with the challenges of an environment, being affected by its solicitations and threats, navigating its landscape, and coping from moment to moment. As an activity, it is subject-directed, that is, we can only speak of sense-making if there is a subject for whom an Umwelt laden with significance arises. In that respect it is asymmetrical, arising from the activity of a subject. But the subject and the world that thereby arises form a mutually specifying pair. Remember, this is not a mindless, one-size-fits-all objective world of which we speak, but an Umwelt encountered by an organism as a function of its capacity to make distinctions and to act.
This cell is lonely, and lacks its own history. Now we come to the point at which this new vocabulary can be put to use in the contrast we wish to tease out between Mr Sparkles and the tumblebot. To do so, we need to augment the basic picture sketched above with two further considerations. Our cell in its petri dish is lacking two very important characteristics that are an inalienable part of every actual living being. It is asocial and it is ahistorical. We will look at each of these in turn.
All life forms are inextricably entangled. The asociality of the cell is quite unlike any real life form we know. Life arose on earth as a collective phenomenon, and in so doing, the activity of the living has gradually transformed the thin layer between bedrock and space, so that virtually every feature of the habitable zone of the planet has co-evolved with life, and could not be understood without life. The oxygen in the atmosphere, the soil of the ground, the weather patterns we observe, are all the result of interactions between living beings, an inert substratum providing raw material, and a constant energetic input from the sun. This is the foundation out of which every life form we know emerged, and as life developed, each distinct organism did so together with others, inextricably entwined with others, and inconceivable without others. This sociality applies among organisms of a similar kind: humans grow up in human societies, an ant is inconceivable without a colony, even a humble seed, one of perhaps many millions, is released into a world in which like seeds have found home and thrived. It applies too to the intertwining of radically different organisms. Humans are not viable without a vast amount of bacteria, plant roots die without fungi to work symbiotically with them, and every kind of life is bedded down in a vast and thoroughly interconnected network of other life forms. Life lives with life, among life, and is inseperable from life. Photosynthesizers aside, life also is built out of life, from the fungus on the rotten log to the killer whale plundering a seal colony. [Table of Contents]
The consequences for the picture of sense-making we have outlined are profound. When we view sense-making as the activity of an organism that maintains its being in a variable and dangerous world, then that world is, to a large part, made of other sense-making beings. Activities of the one influence, and are influenced by, activities of the other. In this manner our worlds become entwined, and we can then speak of "participatory sense-making." De Jaegher, H., & Di Paolo, E. (2007). Participatory sense-making. Phenomenology and the Cognitive Sciences, 6(4), 485-507. Two people engaged in a conversation illustrate this rather nicely, as their mutual activity, and reciprocal influence, gives rise to a temporary common locus, the common ground of the conversation, that provides a shared perspective on things. We sense-make together, which provides part of the reason we manage to coordinate our being so well that we come to inhabit a largely shared world.
Living beings share a long and intimate history. This extension to the basic picture we have sketched becomes vastly amplified when we now pull back, and situate ourselves within the 3 billion years or so of life on this planet. Any organism today is the result of co-evolution among all life forms, in which sociality and participatory sense-making have been at work the entire time. In that time, organisms have changed, and with each change in structure and capacity, comes a change in the world, the Umwelt, encountered. For each being, the distinctions between self and other, between my kind and other kinds, between prey and partner, has arisen on the back of countless acts of copying, continuous activity with others, so that any earthly organism today is embedded in the long history of life, and encounters a world that reflects that history. Each act of copying has passed on something remarkable, more than mere form or structure. It has passed on, possibly with minor variation, the means by which a world is brought forth through sense-making activity. The world encountered by our cartoon cell was simple, and unique to that cell. But real life forms arise, live, and die collectively, their sense-making activities thoroughly entwined. [Table of Contents]
What do we see?
Time to wrap up. Armed with all the above, can we finally say anything further about the contrast between Mr Sparkles and the tumblebot? Have we got two cents worth of commentary that we might contribute to the question of whether living beings are simply machines? I believe we do.
We have not discovered the secret of life, unfortunately. First we should probably reassure the reader that we are not claiming that there is some physical essence to be discovered that will differentiate artifact from organism. Physical measurement will not reveal any difference between the two. So far, so unsurprising. The cell of our little cartoonish rendition of life is likewise not distinguishable from an artifact. Taken on its own, against no background, it is a mechanism.
But we have articulated a new perspective. But we have been at pains to develop a story in which a physical account is not exhaustive. A physical account does not provide a lens through which we can identify, describe, or even talk about, behaviour. Behavour is goal directed, and the identification of goals implicates the observer. Goals do not simply exist. Like beauty (and ugliness), they demand a beholder, and any claim to their reality must perforce include the perspective of the beholder.
Mr Sparkles is "one of us". Now we can identify the difference we have been looking for. Like the cell of our story, the tumblebot is asocial and ahistorical. Mr Sparkles, for all his placid demeanour, is an integral element in the grand drama of life on earth. Here, we must recognize that we are speaking from our own perspective. We do not have a mythical view-from-nowhere. We too are part of the same story, and our capacity to see anything, including the goal-directed striving of others, has emerged over a very long time in which the twin capacities for discrimination and action were passed on, with minor variation, from generation to generation. The copying that lies at the heart of life does not only copy structure. Like the power to unlock doors that passes from key to key, reproduction passes on Umwelten, worlds that are met by subjects, and that are laden with significance.
And that might be as much as we can say. Let us recall Alan Turing's quandry. Intelligence was not something he could define. It might not be something anyone could provide a satisfactory definition for. But he wanted a good-enough-for-government-work solution that one could employ to see if a given artificial system could merit the accolade "intelligent" on the grounds that it had convinced, or fooled, one or more human observers. Our situation is not dissimilar, and we might recall that the tumblebot, when presented in just the right context with an appropriate back story, did indeed briefly fool a bunch of bright people. We might note now that the hypothetical encounter with any alien life form or machine left behind by a passing spaceship would present us with a completely different kind of quandry, as we would be seeing only half of the story. Without the equivalent of a biosphere, we would not expect any such being to be interpretable.
We are in and of life. We are of life, of the biosphere. We evolved within it, and it evolved along with us (for some suitably generous and temporally extended view of "us"). We recognize goals, see behaviour, because we are at home here. I will recognize your goals more easily than those of a chimpanzee, and a chimpanzee's more readily than those of a paramecium. Moreover, life forms are more densely entangled than a simple tree of structural descent might imply. I can read a dog's goals more readily than those of a cow, though both are approximately equally distant in the evolutionary tree. The social lives of elephants and cetaceans bear interesting and profound similarities to those of humans in many respects, yet there are many animals more closely related to us that have rather different local histories, and that appear behaviourally more distant.
A strict division between biological, cultural and historical facts cannot be sustained. And if we can adopt a sufficiently expansive sense of "us" that places us among all living beings, perhaps we can also resist the liberal urge to attempt to describe a fictional "humanity", as if humans were excluded from the remainder of life. We are not only larger than humanity, we are also smaller. The goals, the behaviour, I understand, I do so from a specific cultural and historical perspective. If you are not from Ireland, there will be behaviours I will see and understand, that you will not (and vice versa). Irrespective of my overt beliefs, my Umwelt is stamped with my country, the religious heritage of my neighbours, the idiosyncracies of my family, and the indelible effects of the peculiar path I have traced out in life. You and I will reach common understanding on many things, and we will both recognize a footballer kicking a ball as a specific kind of behaviour. But there will be subtleties, and perhaps quite obvious acts, that one or other of us does not understand, cannot parse, and cannot recognize. The story we have sketched out here does not decompose at all into distinct biological, historical and cultural narratives, each ranging over its own set of processes. The enactive approach, that integrates perspectives, and the cautious introduction of value into scientific narrative, allows us to avoid and evade those distinctions. [Table of Contents]
Genitum, Non Factum
The infant is at home. So what is the mewling infant doing when it sticks its tongue out at the other? If, as we have suggested, the world we meet directly, the Umwelt, is not simply "there", but is drawn in strokes given by our own constitution, lineage, and history, then it does not seem unreasonable to suggest that the new born infant is indeed engaging in an act of self-recogition, is responding in a very direct manner to an encounter that is not strange or alien, but that has been prepared for through countless acts of copying, each passing on the ability to make some distinctions rather than others. On this view, the infant is not born into a hostile and unknown world. True, it is not yet ready to act in any effective fashion. That will take time. But it is a living being born into the biosphere, the ground and home of life, and its first encounter is with a being of remarkably similar constitution to itself, or any other human being.
And we have more than one story to tell. Perhaps the reader feels a certain disappointment, akin to that felt when it becomes clear that Turing's test does not nail down the concept of intelligence in some objective and fixed fashion. We have not identified an objective characteristic of living forms that distinguishes them from machines. If we had, we would indeed be making claims at odds with what we know about biology, evolution, and more. But I think the path we have traced has offered a novel vantage point from which to view the question we have pursued with new eyes.
An enactive account takes a bit of work to put in place. The pieces from which the story is assembled take liberties with a few sacred cows of the contemporary landscape, but they are each grounded in concerted attempts to understand life and the living on their own terms. The enactive treatment of objectivity, for example, will be unfamiliar to many, but it is one among many constructivist accounts concerned with understanding the personal relation each individual must have with the world they encounter.
But we have need of new stories too. The more familiar treatment of a strict divide between the lived experience of sentient beings and the worlds that they inhabit has been developed in part through the progressive elimination of agency from our account of the cosmos. The mechanical universe of Newton and Gallileo, even that of Einstein, does not include any role for any agency. This is great for dealing with stars and atoms, but not so useful in understanding penguins, lichen, and clowns. That scientific picture serves many purposes, but it does not exhaust our need to understand and describe the world we live in, which is saturated with agency.
But I cannot help but notice that buried in the Christian Nicene creed is the intriguing claim that the embodied form of the Christian God was "begotten, not made", or genitum, non factum. I am not in the business of Christian apologetics, but I would suggest that from this vantage point, that this looks like a distinction that is of some consequence for us. And lest I be accused of religious partisanship, we might go back to Kali and the Hindu tradition. One of the most venerable phrases from the oldest of the Upanishads is "Tat tvam asi", or "Thou art that", which likewise seems to resonate well with the infant sticking its tongue out in recognition of its situation in a familiar world.