Humanoids

As a child, I used to be very irritated by anthropomorphic robots. It felt like people weren’t taking the idea seriously when they attempted to depict an automaton in such a way that it was like Robbie The Robot from ‘Forbidden Planet’, and also that form did not follow function here. Why would a robot – an apparatus designed to do the work of a human – have to look like a human? In fact there are good reasons why that would be good, but before I go there, I ask you to consider this monstrosity:

For some reason, not hard to guess, someone decided to drape a woman across this machine’s arms and I expect it was quite successful. The other, main one, which springs to mind is this:

Nowadays I feel a kind of grudging respect for these designs, but even still, things like the obviously valve-inspired heads, also seen in ‘Fireball XL-5’ and elsewhere, are rather annoying. There seems to be neither a reason for the gyroscope-like “eyes” in the top illustration nor any way in which they could function as visual apparati if that’s what they’re supposed to do. One is left with the question of what the fictional designer was trying to “say” with these, in a fashion kind of ways. It doesn’t help that the appearance is so approximate. But on another level they’re fun, I suppose, but at the time I just found them irritating.

When I became a fan of ‘The Hitch-Hiker’s Guide To The Galaxy’, I took pains not to imagine Marvin as humanoid. It actually makes sense that he wouldn’t be because although H2G2 includes humanoid aliens, many of them aren’t, so why bother making “him” look like one? Sirius Cybernetics would be missing out on the non-humanoid market there. He was first represented as human visually in the 1981 TV series, but is implied to be an android in the radio, and in fact explicitly described as one, although he’s also called “paranoid”, which doesn’t actually make sense as he’s depressive rather than delusional. However, he’s bilaterally symmetrical, having a pain in all the diodes down his left side, and has two arms and ears:

I’ve worked out that if I stick my left arm in my right ear I can electrocute myself.

• Marvin, Fit the Tenth.

My default vision at the time of what a robot looked like was something like Hewey, Dewey and Louiey from the ‘Silent Running’ film, but with an air cushion instead of legs and metal rings to manipulate objects encircling the main body. If I could draw at all, I’d show you, but I can’t really.

In fact there is a good reason for robots operating in an environment shared with humans to be humanoid. A domestic robot, for example, would presumably be interacting with objects designed to be interacted with by humans, such as furniture, doors, windows and perhaps domestic machinery such as washing machines and vacuum cleaners, so the most logical design for such a machine is to make it pretty much the same shape and size as the majority of humans. There probably isn’t any external feature of a human body which is not somehow involved in this kind of interaction, and in any case the human musculoskeletal system and sense organ arrangements are quite well “designed”, so why not? In a way, we already have a basic design for a humanoid robot in the form of human anatomy, and if it ain’t broke don’t fix it, eh?

Other environments and functions might need different kinds of robot though. For instance, a remote-control device for bomb disposal like the one above is absolutely fine and having, for example, a walking robot to do the same, causing vibrations which might set the bomb off, is probably a really bad idea. In general, robots which move along solid surfaces have wheels rather than legs, and the main reasons for legs are to traverse highly irregular solid ground and because an organism is usually completely connected internally and therefore can’t have wheels. There are exceptions to this, such as flagella, whip-like organelles used for swimming which have gears, but on the whole animals need to have blood, muscles and nerves connected to locomotory organs, so they can’t easily have literal wheels.

A number of political and social issues arise with humanoid robots. You may have noticed that I’m not saying “‘droid” or “android” in this post. The reason for this is that the word “android” doesn’t mean “in the shape of a human” but “in the shape of a man”. An android pelvis is one which is shaped like the typical XY human pelvis, and can therefore be problematic for giving birth vaginally. The Greek equivalent to “humanoid” would be “anthropoid”, but this term is usually used biologically to refer to gorillas, orangutan, bonobos, chimps and humans rather than something mechanical, so that also is ruled out. Androids are assumed to be male. If a robot is “female”, it’s called a “gynoid”, and is likely to be a sex robot. The stereotypes of gender are much stronger and less questioned in the realm of robotics than the human world, and perhaps this is not surprising because robots are in a sense slaves, although that metaphor has now become rather tired.

There is a campaign against sex robots, founded oddly enough in Leicester. Its aims are to abolish porn robots of women and girls, prevent the manufacture of child robots for pædophiles, challenge the normalisation of pornbots (their term as far as I know), encourage a model of sexuality and sex as mutual, create a vision of technology where women and girls are valued and work across the political spectrum to value the dignity of women and girls. The issues are complex. Augustine used to argue that we only had duties towards non-human animals because we might otherwise transfer their abuse to humans, and there are cases of people who were cruel to cats and dogs when they were children who grew up to be serial killers as adults. I find that argument very suspect because, regardless of veganism, cruelty to other species was the norm among boys I knew as a child and they clearly did not all become serial killers as men. Applying this to sex robots, the issue seems to be similar, that the use of sex robots would lead to disrespect and abuse of women and children in the long run. The big problem with this view is what it ends up looking like if sex robots become sentient, because at that point it seems to become practically identical to genocide, and it’s also reminiscent of the argument that Cannabis is a gateway drug, which is clearly nonsense, at least where drugs are decriminalised. There is a story to be told here and I may do that one day.

Nonetheless, the Campaign Against Sex Robots does have a point. A gynoid right now is likely to be a sex object, and this is deeply problematic. However, there are examples of non-sexualised gynoids, particularly more recently. A related issue was mentioned on the BBC radio podcast ‘White Mischief’: there is a very strong tendency for androids to be White, and also white in colour. If androids in fiction were both servile and dark-“skinned”, that would communicate certain disturbing issues. The podcast also pointed out something I had never noticed before in ‘Blade Runner’: the unquestionably human population of Los Angeles is ethnically diverse, but all the androids and possible androids are White, because this allows them to pass. Whiteness constitutes inconspicuousness and neutrality, so all the escaped Nexus 6 androids are White, as are Deckard and Rachel. It was then suggested that rather than giving real androids humanoid colours either characteristic of a given ethnicity or associated with names associated with those, such as black, yellow or white, they should instead be made completely non-human in colour, such as saturated green or blue.

The history of the word “humanoid” is interesting. During the nineteenth Christian century, the word was applied by White people to indigenous people in South America and the like. In a somewhat related usage, the humans in the original ‘Planet Of The Apes’ films and TV series were referred to as humanoids. Early last century it referred to non-Homo sapiens hominids such as Australopithecines and Neanderthals. Nowadays the words used there, with narrowing scope, are “hominoid”, “hominid” and “hominin”. Current usage seems to apply most to aliens who look like us. That is, they’re bipedal, have arms and hands with opposable thumbs, binocular vision using forward-facing eyes in a separate head and horizontal mouths. They may or may not have hair. In fictional contexts, they probably turn up for two reasons. One is that if cinema, stage plays or TV programmes are involved, most actors are human (not all – dogs, cats and dolphins among many other species are also used as actors), so it’s impractical to use models or CGI for the most part unless animation is involved generally. Three-quarters of ‘Star Wars’ protagonists are human, for example. The other is that humanoids are easier to relate to and to write convincingly. This tends to condition people unconsciously to expect aliens to be humanoid. And there is an argument that they would be. Several separate ones in fact.

Suppose there are 600 million life-bearing planets in this Galaxy, and that 500 000 of them have intelligent civilisations originating on them – I should say “habitable worlds” incidentally because it’s easily possible that many of them would be moons rather than planets. In order to represent those numbers in binary form, twenty-nine and nineteen bits respectively are needed. Assuming, very simplistically, that evolution consists of a number of steps with binary outcomes and that there is always intelligent tool-using life at some point, this would mean that twenty-nine steps would be enough to ensure an unique path to intelligent life, and that nineteen steps were involved in producing each species of this type which currently exists. These assumptions should illustrate that the probability of humanoid life is very small. However, the outcome of evolution is not random, and many steps may have to be taken in a particular direction rather than another to lead to a successful species.

The evolutionary biologist Stephen Jay Gould once pointed out that at a certain time in the early Cambrian, the most successful phylum of animals was the priapulids, and that only one species of chordate (the phylum of which vertebrates form the majority of species) was known from that time. There was also a huge variety of other phyla most of which quickly died out. Looking at those representatives of animal life on this planet, most people would probably have bet on the priapulids as the most diverse and perhaps successful phylum later in the history of life. In fact, at the time of writing there were only eight known species of priapulid surviving and they constituted the very smallest animal phylum. There are now twenty-two known species of priapulid and sixty-five thousand known species of chordate, mainly vertebrates. The most diverse phylum of all is of course the arthropoda, with up to ten million species, mostly insects. I already went into why chordates might be rare even in a Universe where complex life is common. Vertebrates are unique in having a hard endoskeleton which facilitates movement. The other major phylum with such a skeleton is the sponges, and they use them to anchor themselves in place and they work as a kind of scaffolding. By contrast, animals with hard exoskeletons are very common, including for example the arthropoda and many molluscs. The specific combination found in vertebrates, including bilateral symmetry, gill slits, a spinal column, hard endoskeleton, a tail extending beyond the anus and discrete muscles, is unique. In fact each feature is quite rare except for bilateral symmetry, although some are found in other deuterostomes, the superphylum whereof we are members, distinguished by developing a mouth after an anus embryonically and having cells whose fate is not fixed early and which divide radially rather than in a spiral arrangement.

Given the success on this planet of other body plans, the arthropods seem more likely to resemble complex animal life elsewhere. Not only are they the majority of species here, but also they have features which have evolved independently several times such as possession of a hard articulated exoskeleton. However, just becausethe basic plan of the body is different, it doesn’t mean these very different animals wouldn’t be humanoid. If silicon-based life is possible, it could even go as far as having completely different biochemistry and still being humanoid if that’s dictated by evolutionary pressures in the right way.

There are some body shapes which seem to recur a lot. For instance, it used to be thought that there was a single phylum known as the polyzoa, which consisted of colonies of sedentary animals, but it turns out that there are two more or less unrelated phyla which happen to be somewhat similar: the entoprocta and ectoprocta, distinguished by the position of their ani inside and outside their feeding organs. Flower-like organisms are also a repeating theme all through the animal, plant and protist kingdoms, although it’s hard to imagine such an organism being intelligent. Why should it be? It does what it needs to do. However, the humanoid body form may not be repeated at all closely throughout the animal kingdom. I think of ants as very slightly humanoid, but it’s a stretch even with them. Nonetheless, it’s possible.

If there is an organised and peaceful galactic community out there, it’s possible that they would send delegates we could relate to, and if they have a wide choice of different life forms because, say, they have half a million such civilisations to select from, it seems plausible that they would choose the most human-like species available. This is a highly fanciful scenario of course, based on serious projection onto a blank canvas, but just maybe, if that situation exists, the first aliens we meet might be highly unrepresentative in that they look somewhat like us. For that to happen, it’s more likely that such aliens would be from a world which was particularly suitable for the evolution of humanoid life, and one major feature of such a world would be bipedal organisms. Bipeds have evolved separately on our planet, notably as dinosaurs and their subset, birds, and also as such forms as kangaroos, gerbils and the now extinct leptictids. All of these, however, have long tails balancing their bodies and none stand erect. An animal with a neck is also “strangleable”, i.e. it has a large number of vital supplies relatively unprotected stretching from the trunk into the head, and it might make more sense either for the brain to be buried deep inside the body or for it to have no neck and instead more eyes and ears arranged around the head, or of course both. The fact that bipedalism usually seems to involve some kind of tail also naturally requires a tail to exist, and the evolution of such an organ seems improbable as it’s only happened once and even then has some tendency to become reduced or disappear. Human erect posture is a combination of three trends, including a change in the position of the pelvis, hip joints and spine, and seems to be unique, on this planet anyway. But the appearance of erect bipedalism has various consequences for anatomy, such as foot-like feet, so to speak, and in the case of vertebrates a head which is placed in a particular position on top of the spine rather than in front of it, which changes the location of the foramen magnum, and there seem to be good reasons for erect bipedalism such as the ability to look out for potential predators and the freeing up of the forelimbs for tool use and carrying, which would otherwise have to be done with the mouth. That said, trunks or tentacular lips could do pretty well in this regard, and we’re talking bipedalism as well rather than tripedalism or something else because we’re assuming an animal is bilaterally symmetrical.

Hence I’m prepared to make a compromise here. Humanoid aliens could evolve as one of many suitable body forms for a tool-using intelligence, but are likely to be rare. If the Galactic Federation exists, I can see them sending humanoid emissaries to soften the blow caused by the shock that there is intelligent life elsewhere, but they would have to choose from a plethora of different types of life form to find the one or two other humanoid species in this Galaxy. And of course this all assumes that the massive array of filters between an initially lifeless world and the existence of a technological culture can be negotiated by the average biosphere.

But there is another possibility here which is fuel for paranoia and delusion, and would undoubtedly be within the capabilities of advanced alien technology: they might be able to produce beings indistinguishable from humans artificially in some way. They could possibly 3-D print them or genetically engineer them with identical genomes to our own, and perhaps implant memories so that even the beings themselves had no idea of their origins. I’m inclined to discount this possibility simply because it constitutes a mental health hazard, regardless of its plausibility, and in a way it’s unscientific because it would be in principle hard to falsify, since they would’ve made sure of that. It is, however, one solution to the Fermi Paradox: some of us are aliens but we don’t know it. Alternatively, maybe some of us are aliens and we do know it, or we’re like this:

I suppose there are a few things to say about this. Firstly, we don’t even know there is any life at all off this planet, barring the few tardigrades which accidentally ended up elsewhere and are currently dormant. Secondly, if anything did ultimately end up getting through all the obstacles and reached the stars, we can safely assume that it would have the technology to do this, and for all we know there are sleeper agents in the human population right now who don’t even know they are themselves. Thirdly, however, this is exactly the kind of thing someone who has lost contact with reality would think, and it plays into, for example, Capgras Delusion. It’s equally possible that this could reflect back on the person themselves and they could end up believing they are such an individual. But the question of motive does arise.

To conclude then, humanoids come up in at least three ways in our understanding of the Universe. One is in the form of robots, where one might question why they would need to be humanoid at all. Another is as aliens, which seem at first to be improbable, or perhaps inevitable, so I’ve chosen the middle way and concluded that if intelligent life is common in the Universe, humanoid aliens are rare but do occasionally evolve. Finally, there probably isn’t anything which would stop non-humanoid aliens from manufacturing humanoid replicas who don’t even know themselves that they aren’t human in origin, but there would have to be some motive for them to do that. Then again, they’d be aliens so maybe we can’t even understand those motives.