By Courtney Suciu
With the publication of Discourse on Method1 and the famous maxim, “Cogito ergo sum” (“I think therefore I am.”), Rene Descartes profoundly shaped the future of scientific inquiry, mathematical reasoning and our understanding of consciousness. Additionally, inspired by early modern inventors, Descartes wrote copiously on the subject of bodies, minds and machines – exploring concepts and theories which, as we’ll explore, continue to shape the development of 21st-century technologies including artificial intelligence.
In his Discourse, Descartes pondered at length about machines and mechanization which was a burgeoning topic in the 17th century, an age of prolific invention. The steam turbine, pendulum clock, air pump, adding machine, barometer, refracting telescope and reflecting telescope were among the emerging technologies of the time. Amid such advancements, Descartes was inspired to explore their nature and their relationship to the way the human body functions.
Michael Wheeler, one of the authors of the book The Mechanical Mind in History2, categorized and described the three kinds of machines on which Descartes focused:
Type A: A material system that unfolds purely according to the laws of blind physical causation.
Type B: A material system that is a machine in the sense of A, but to which in addition certain norms of correct and incorrect functioning apply.
Type C: A material system that is a machine in the sense of B, but that is also either (1) a special-purpose system or (2) an integrated collection of special subsystems.
For Descartes, Wheeler explained, the whole universe, including the bodies of humans and animals, could be defined as A type machines, acting in response to physical causes. An example of a B type machine might be a clock which is judged on its ability to correctly tell time – if it fails to do so, an explanation can be determined to find out why. In the same way, the systems of the human body, such as the circulatory system, functions properly as it pumps blood throughout the body. When it ceases to function correctly, a diagnosis can be made to explain why.
The type C machine, Wheeler said, is represented in Descartes’s theory of the nervous system, of which in the “fine-grained details” the philosopher was wrong but in the abstract, his ideas correspond to electrical and biochemical transmission systems recognized by modern neuroscientists. The key to this type of machine, said Wheeler, is Descartes’s envisioning of “the existence of a locally acting bodily process through which the unaided machine can, in principle, continually modify itself, so that its future responses to incoming stimuli are partially determined by its past interactions with its environment.”
In other words, a C type machine is one capable of automated learning and adaptation.
Descartes conceived of the mind as distinct and separate from the body, which characterized the philosopher as a dualist. According to Descartes, the mind could not be a machine because it is “immaterial,” while a machine is essentially a material system. Basically, for Descartes (and other dualists), the body serves as an automated vessel for the mind.
But, what do we mean by “mind”? In a nutshell, according to Cartesian thought, it is the parts of the human that cannot be automated. If we think of the human body as a B type machine, it’s easy to point to the specialized subsystems that work correctly or incorrectly according to norms and standards – the circulatory system, the digestive system, the nervous system, etc.
But the human body can also be considered a C type machine, Wheeler explained, capable of automated learning and adaptation through exposure to new stimuli and from experience. C type machine functioning might include qualities like learning to walk and talk, sensory reception and even the development of human attachment and other emotional needs.
At first thought, we might consider these C type functions of a psychological nature related to the mind, but as we now know, and as Descartes was starting to explore, much of our learning and emotional experience is based in the body’s specialized subsystems. For example, we learn to avoid hot surfaces through sensory stimulus; our ability to form emotional attachments to our parents and children is tied to the endocrine system.
These C type functions are also true of animals, and potentially, Descartes knew, of man-made machines which could be capable of such automated learning and modification. But if this is true, what makes us “human”?
According to Descartes: consciousness.
Descartes argued consciousness is not automated by the body’s specialized subsystems. Consciousness, which we define here as the unique human ability to deliberate and reason – and say to ourselves, “I think, therefore, I am” – comes from the mind.
Wheeler explained a popular scholarly view of Descartes’ body-mind distinction: the ability to reason results from the many complex overlapping, subsystems continuously at work together in the body. While the mind is not part of the body, it cannot exist without the body.
This distinction is fundamental to Cartesian duality and is key to long-held doubts about the possibility of creating artificial intelligence.
How can a non-human reasoning machine exist without a mind? Is it possible to create an artificial reasoning mind?
Crudely summarized, here is how Wheeler explained the limitations of machine “thinking”: Machines function to perform a specific task. Just as the circulatory system works to specifically pump blood throughout the body, a robot can be commanded to successfully perform a specific task based on a specific set of directions and data.
However, when the robot is given contradictory directions or new data or stimuli which alters the original command, it will fail to function because it doesn’t have the ability to reason.
Because, in our Cartesian understanding of the bodily machine, the human body consists of a multitude of interrelated subsystems that can receive and respond to unlimited stimuli, including contradictory directions and new data. Then the human mind considers which information is relevant, contemplates the consequences of possible outcomes and takes appropriate action.
Would it be possible for a machine to also function this way?
Well, consider an example provided by Wheeler: A robot is tasked with retrieving its power source from a room which also contains a bomb.
The robot knows the power source is resting on a wagon, so it decides (quite reasonably it seems) to drag that wagon out of the room. Unfortunately, the bomb is on the wagon too. The result is a carnage of nuts, bolts, wires and circuit boards.
The robot failed because it is limited in its ability to consider the consequence of moving the wagon – it simply was not designed to “think” that way. Wheeler described what would happen if the robot was replaced by a more sophisticated one “that operates by checking for every side effect of every plan that it constructs. This robot, too, is unsuccessful, simply because it never gets to perform an action. It just sits there and ruminates…there are just too many side effects to consider.” [Personal note – who hasn’t been this robot?]
To succeed, such a robot mind would need the plasticity to determine, in a “dynamically changing world” of infinite stimuli, which information is relevant to its intended function, how possible outcomes will be affected by such information and determine how to proceed – in other words, it would need to be able to reason.
From Wheeler’s example, it’s clear why duplicating human intelligence is such a daunting endeavor. And the challenges are compounded when we start to think about what kind of a “body” would be required of such a reasoning machine.
Even if an artificial mind could be equipped with the kind of flexibility described above, it would still be constricted by the limitations of its bodily machine, as Brian Duffy and his co-authors explored in the article “The Future of Reasoning Machines: Mind and Body.”3
For Duffy et al, the artificial mind is referred to as an “agent,” and, as in the Cartesian model, they demonstrate how, while such a mind is separate from its body, it still requires a body – a platform or device – to function successfully. The body must be free from restrictions, they argue, which might hinder the artificial mind’s ability to dynamically gather and reasonably process data.
The problem is that “all platforms are not created equal,” Duffy et al wrote. They have “varying memory, processing power, bandwidth, display characteristics. Consequently, these agents have to be able to adapt to [these] different conditions” of their bodies, such as anticipating and responding to a shortage of memory or a slow processing speed, while still performing their original command.
While these concerns reflect some of the dilemmas presented in Cartesian notions of bodies, minds and machines, Duffy and his co-authors also addressed issues that go well beyond what Descartes would have likely anticipated. How could Descartes have imagined the pervasiveness of machinery in our society, and the intimate ways we interact with them?
For one thing, we live in a radically more image-conscious era than the 17th century, and as Duffy et al pointed out, “The influence of the appearance and the voice/speech of an entity on people's judgements of another's intelligence have been demonstrated in experimentation.”
Duffy and his co-authors make the case that the agent’s “body” must not only empower the functionality of an artificial mind, but also “maintain the agent's identity in the eyes of the user.” “The next step” in the development of artificial intelligence, they write, “is to look at how we will interact with these systems, how we will interact and understand these machines.”
Just as research shows “the more attractive a person, the more it facilitates others to rate the person as having higher intelligence,” the more attractive we find a machine, the more we are likely to find it trustworthy and be willing to engage with it. Of course, this isn’t an argument for building a “prettier” robot, but an assertion of the importance of aesthetics in the design of platforms and devices.
For further research
Learn more about how the influential writings of Descartes – and other pioneers of the scientific revolution – have shaped the future of physics, space exploration, medicine and technology.
Central to this archive’s Collection 14 are publications relating to the Cartesian revolution, along with hundreds of items that demonstrate the remarkable medical, scientific and intellectual advancements of the time. Encompassing more than 1,100 printed items and nearly 490,000 pages of content, highlights include works by Descartes, such as editions of Les Principes de la philosophie, copies of Les Passions de l'âme, the Discours de la methode and a volume of his letters. Supporting these are a host of items which shed light on Descartes’s wider philosophical milieu and his reception. A leading advocate of Descartes was Nicolas Malebranche (1638-1715). His two-volume De la recherche de la verité (1674-5) appears in Paris and Brussels editions. Other followers include Claude Ameline (1629-1706), and Pierre Cally (1630-1709), whose Universae philosophiae institutio appears in its original Jean Cavelier 1695 Caen edition.
Bodley, A. M. (2015). The Android and Our Cyborg selves: What Androids Will Teach Us About Being (Post)Human
Loghry, J. B. (2013). The Recreation of Consciousness: Artificial intelligence and Human Individuation (Order No. 3605083).
McGettigan, T. (2017). Artificial Intelligence: Is Watson the Real Thing? IUP Journal of Information Technology, 13(2), 44-69.
Obodiac, E. (2012). Transgenics of the Citizen. Postmodern Culture, 22(3).
Weigmann, K. (2012). Does Intelligence Require a Body? The Growing Discipline of Embodied Cognition Suggests That to Understand the World, We Must Experience the World. EMBO Reports, 13(12), 1066-9.
Wheeler, M. (2010). Plastic Machines: Behavioural Diversity and the Turing test. Kybernetes, 39(3), 466-480.
Baker, G., Morris, K., & Baker, G. (1995). Descartes' Dualism.
Descartes, R. (1990). Meditations On First Philosophy/ meditationes de prima philosophia : A bilingual edition.
Descartes, R. (1998). Descartes: The World and Other Writings.
Kolak, D., Hirstein, W., & Mandik, P. (2006). Cognitive Science: An Introduction to Mind and Brain.
Pfrehm, J. (2018). Technolingualism: The Mind and the Machine.
Wilkinson, R., & Wilkinson, R. (2000). Minds and Bodies : An Introduction with Reading.
Courtney Suciu is ProQuest’s lead blog writer. Her loves include libraries, literacy and researching extraordinary stories related to the arts and humanities. She has a Master’s Degree in English literature and a background in teaching, journalism and marketing. Follow her @QuirkySuciu