Cute as a function of familiarity

A while back, I read an article on the internet about how real was too real when it came to artificial intelligence, robotics and CGI. I remember the page had charts and a write up about it, but I can't remember what it was called or even if it was an actual scientific study or just a joke on xkcd*. Either way, the idea was that anytime you see a human-like representation of something that isn't actually human (anything from a robot to Gumby to Tom Hanks in Polar Express), as the realism increases, the emotional attachment increases to a point, then drops off. In other words, as it becomes more realistic, it reaches a peak cuteness, then just becomes creepy. It was meant to explain why you feel an emotional response to characters in Wall-E, but not the ones in Final Fantasy - The Spirits Within.

vs

The reason I got to thinking about all this was because I saw this picture of a cute baby:

When I see a picture like that, there is a definite emotional reaction - an internal "squee" of sorts. This might just be because I'm 6 months pregnant, but I think it's more than that. I suspect most people would have have an emotional reaction to this picture, although some might be more effusive than others. With this picture, my attention is especially drawn to the baby's eyes. This is because, when a person is born, their eyes are as big as they will ever be, even though their head starts out small. This results in baby's eyes being disproportionately large compared to the size of their faces. Supposedly this is part of the reason why people have that automatic emotional response to babies and tend to perceive them as "cute". From there, I got to wondering if you could plot a baby's "cuteness" as a function of how similar they look to a full grown adult similar to the way you can plot the emotional reaction to human-like representations as a function of their realism

If you were to plot such a graph, on one end you'd have babies that look the least like adults. This would include babies that have some sort of physical deformity, but I think it would also include newly born babies. I'm not talking about newborns that have been cleaned up and had cute hats places on their heads, I'm talking about babies moments after they're born. At first sight, there is no doubt a strong emotional reaction from the parents of this child, but I don't think an objective observer would judge these babies as "cute". They're sort of slimy and gross looking and would probably score low both on "cute" and "adult-like":
Further along on the graph would be babies that have grown a little and are starting to look more like adults. Take the baby pictured above, for instance. He's been cleaned up, can hold eye contact and probably smiles a lot. These babies exhibit many characteristics of an adult, but they're still disproportional in the way they look and awkward in the way they move. I think this puts them at the height of their cuteness and gives them the innate ability to elicit a strong emotional response. If you don't think so, just take another look at that baby up there. Can you think of anything you could change to make this baby cuter? I can't think of anything. Maybe a puppy in the background, but I think that just reinforces my point.

As for the far end of the "adult-like" spectrum, I can only speculate on what would cause a baby to score high in adult qualities. Perhaps one of those children with the rare disorder that causes them to age incredibly fast? I can't find a picture of a baby who has that to determine how early that becomes noticeable. The only real-life example I can think of is less specific, but if you've ever seen a baby that looks like an old man, you'll know what I'm talking about. I can't put my finger on exactly what gives a baby this look, but they just tend to be bald, have less baby fat and generally look like they're scowling:

VS

In closing, I'd just like you to ear in mind that any information seen in this post is not based on any scientific study, double blind study groups, or statistical analysis. Unfortunately, until I can either find or become a wealthy benefactor, this whole hypothesis will remain complete conjecture. In the mean time, I did make a graph to represent this jist of my theory and hopefully lend it some credibility:


* EDIT: Thanks to the help of an astute reader, the mystery has been solved. The original theory plots what is known as the "Uncanny Valley".