The overriding theme of Mazis' book, as I understand it, is that, the more one looks for the boundaries between humans, animals, and machines, the harder they are to find, and the three form a unity that is somehow at once both three and one.  As he writes near the beginning:

It is uncanny, as we will see in case after case in this book… that when one proceeds to the core of what seems to be the animal, both the machine and the human emerge into the discussion.  When one proceeds to the core of the human, both the animal and the machine emerge. (Page 6)

Though he does not directly carry on to say so, the same is implied for the "core of the machine", and indeed, he gives examples through the book of where he sees the animal or the human in the machine.

If this is the correct reading, then for any of the three, one should be able to substitute them in _____ are _____ to get a statement that is, in some contexts at least, literal and meaningful.  That humans are animals I find biologically self-evident.  That we increasingly look for and find the human in the (non-human) animal I likewise have no problem with:  witness the increasing ease with which we attribute to other species rights that originally were reserved for humans (and even then often denied to many of them).  As we come to learn more about our neighbors on this planet, our definition of what it is to be human gets stretched, and a strictly biological definition of human will not suffice:  this, too, I find self-evident.

The idea of humans or animals as machines I have more difficulty with; it seems to raise shades of Cartesian thinking, and not the parts I might care to hold onto.  "Like the purring of the machine, operating smoothly, the quiet pulsing of the healthy human body reassures in a similar rhythm."  (Page 213)  Perhaps that is true, but is that enough to establish any blurring of boundaries between humans and machines, let alone kinship?

Sure enough, there are parallels between the characteristics of machines Mazis sees as definitive and tendencies in contemporary human culture.  So for example, he writes:  "Surely this is what makes the machine a 'mere object' — it has no relationship to its environment" (Page 49); and certainly, there are times that humans seem alienated from their environment.  (Non-human) animals he sees as being maximally and unavoidably continuous with their environment; their relation to machines comes from the "mechanical" rhythm of their lives and the dependability with which a certain stimulus will yield a certain response.  But it is a far cry from saying that there are useful comparisons to be drawn to saying that "humans are machines" or "animals are machines" in any but the most loosely metaphorical of senses.

Likewise I have difficulty with "machines are animals" or "machines are humans", even while acknowledging the power of a Kismet or a Cog to convince us, even perhaps on occasion to convince their designers, that they are emotionally alive, that their responses to our emotional expressions are themselves intentional.  Kismet and Cog strike me as updated, much more embodied versions of Eliza, using similar tricks that still withstand only so much scrutiny.

I am not always clear when Mazis is being literal or when he is being metaphorical – at least in part because he would probably deny a clear boundary between the two.  When he writes, "…Perhaps one day it will not seem strange to talk about the person that is the mountain in the distance or the person that is the lake outside our cabin" (Page 243), I assume the meaning is metaphorical.  But when he writes "…As persons, they [machines and animals] may be also be owed a direct voice in governmental decisions that affect them…", the intended reading (even with the qualifying phrases that surround the passage) seems quite literal.

Part of my difficulty with reading Mazis may well be a confusion with terms.  His definition of machine is a very broad one:  e.g., "In some sense, though we never put it this way, we accord to language the status of a person." (Page 249)  If even language can be a machine, then it's easier to see how and why animals and humans might be.  I may simply be trapped within an understanding of "machine" that Mazis would see as far too narrow.

Likewise there are important nuances to his use of "human" and "animal".  Although he occasionally writes of "human animals", most of the time Mazis writes of "humans and animals" rather than, as I might find more appropriate, e.g., "humans and non-human animals".  Mazis is inclined to reserve the term "human", in the end, for what I take to be a strictly biological definition (ironically), talking of granting not humanity but personhood to animals and machines.

There are many ideas in this book with which I deeply agree.  So, for example, Mazis talks of the heart of free will being not so much a choice of actions but a choice of perspectives. Likewise I agree with his view that emotions and rationality are not oppositional but that rationality arises from emotions.  I like his implication that what is ultimately human is what we recognize to be human.

I agree that apparently contradictory perspectives may both have something valid to tell us.  His approach is unabashedly pragmatist:  most questions of interest do not have one correct answer.  As he writes, "Rather than have to decide between these oppositional claims that one is true and the other false, it is more helpful to see the sense gained by considering both positions as revealing something important.  This approach leads to a lack of closure to our questions, especially the wider and the deeper they go, but yields interpretations that are more suggestive." (Page 13)  This is an idea that resonates deeply with my own research in theories of concepts.  It allows such powerful insights as the one that "the kind of responsiveness we consider evidence of some sort of agency is a simultaneous distance and insertion — an 'interrelation' — with the surround." (Page 22)  That is, either distance or insertion on its own does not yield agency.

I agree that the idea of humans as "a separate creation" from animals is unmaintainable and destructive, regardless of whether its motivation is religious or scientific.  I agree with his grounding premise that the boundaries between humans, animals, and machines (as perhaps between many if not most categories; there need be nothing special here about humans, animals, and machines) are impossible to fix with any absolute certainty; I am less convinced by his examples of where those boundaries get transgressed.  Though I am sympathetic to the points he wants to make by citing them, his reading of both Alan Turing and Thomas Nagel are completely different from mine, which makes me less willing to accept at face value the views he attributes to Merleau-Ponty, whom I have not read.

This is a book of powerful ideas.  If it does not always proceed by reasoned argument, it does provoke.  That, perhaps, is ultimately its strength, one that its author might well embrace.

© 2009 Joel Parthemore

Joel Parthemore is a third-year DPhil student studying theories of concepts at the University of Sussex in Brighton, UK. He is a member of the Philosophy of AI and Cognitive Science research group in the Department of Informatics. In his spare time he plays with Linux computer systems. You can find him online at http://www.parthemores.com/research/.

Keywords: cybernetics