Illusion Research: An Infantile Disorder?

Some lucky contributor to Perception may be able to save themselves the purchase price of £162.50 by acquiring a review copy of a recent Compendium of Visual Illusions (Shapiro & Todorovic, 2016). I hope that their review will address the question ‘Is this a topic for a serious scientific publication?’ Exhibitions, books, conferences, websites, endless CVNet threads, special issues of journals (yes, fellow Editors of Perception, it is you I am addressing!) on the topic of Visual Illusions continue to proliferate. Why?

Lenin (1920) put down one group of his opponents with the pamphlet Left-wing Communism: An Infantile Disorder. Without instigating a Red Terror (just yet), I would like to suggest that the same scornful epithet can be applied to the study of illusions.

Let me make two things clear. First, I have no desire to enter the debate on how or whether illusory and non-illusory perception can be distinguished. Rogers (2014) and Morgan (1996), among others, have offered thoughtful arguments that such a distinction cannot and should not be made. I want to argue that, even if a watertight definition of illusion could be produced, it would still define a category that was deeply unhelpful for science. Second, I don’t want to dismiss any particular ‘visual illusion’ as a valuable subject for study. How could I, when stroboscopic motion, colour contrast, the Ames room, and comparative weight judgments can all be found under this heading? All these effects lead to questions, and some of them have led to answers, that are profound for our understanding of the processes of perception. But what have we gained by putting them in the same box, however elegantly we might phrase the definition? If we had a good mechanistic theory of the Muller-Lyer illusion (and perhaps it is revealing that over 120 years after Franz Carl Müller-Lyer, we don’t), would it be any help in deciding how colour contrast works?

I don’t adopt Lenin’s word ‘infantile’ simply to be provocative. Illusions appear in children’s science books, science museums, and gift shop fodder for good reasons. They have an immediate impact that doesn’t require any thought – pull out the ruler or mask off the surround – ‘Wow! they’re really the same! They do say you can’t believe your eyes!’ First-year Psychology students can be persuaded by the lecture slides that perception might be as interesting as the lectures on criminal psychology or racial attitudes (although not so many remain convinced, alas, when they get to the slides about metamerism or lateral inhibition).

Suppose instead of showing the apparently bent lines of the Hering illusion, your slide depicted a single straight line. You told your students ‘Nobody knows how your brain works out whether this line is straight or curved. What kind of process might do that? How could we try to find out?’ It would be a very rare first-year student who responded ‘Wow!’ Infantile material is absolutely essential as a developmental scaffold, and can give innocent pleasure even to the grown-up. We enjoy The Very Hungry Caterpillar with our two-year-olds. But we’d be a bit worried if they were still deep into it at fourteen, and would smirk at any University literature department where it was a set text.

So, by all means let’s use visual illusions in our scientific sweetshop window. But let’s not kid ourselves, or our students, that they provide a valid title for a textbook chapter, a conference, or a research programme. And let’s improve the quality of our thought, and our experiments, by motivating them through questions about how perception works, not by the first-year student’s ‘Wow’. If we want to understand how we assign surface colours, by all means let’s explore colour contrast. If we want to dissect the processes by which we can make accurate spatial judgments, maybe the classic geometrical illusions provide a source of data (although, in this case, I see rather little evidence). But these topics belong in the chapters on colour vision and spatial vision, rather than together. The ‘fact’ that things may not be what they seem is not a unifying principle that gets us anywhere.

My tone may seem unnecessarily polemical. I have an excuse, in that in the very first volume of this journal, I wrote a review of a book on illusions in which I made much the same arguments at greater length (Braddick, 1972). Needless to say, it hasn’t become a citation classic—‘sunk like a stone’ would be more accurate. Maybe by a mildly abusive title, I can at least enhance my h-index and Perception’s impact factor. ¹