Ducks and Rabbits: on AI and the joyful possibilities of university
‘Concepts constrain possibilities.’ This is one of the many insights drawn from our final year readings in anthropology this week, building upon the work of Sapir and Whorf back in the 1920s and ‘30s on the influence that language exerts over our perceptions of the world.
It is a truism that we encounter in several modules here at Trinity College Dublin Classics. In anthropology, for instance, as one of our final year students this year brilliantly demonstrated, we encounter it in (amongst other examples) Lisa Feldman Barrett’s work on emotion, wherein the puzzle of Serena Williams’ facial expression pictured below (Fig. 1) beautifully demonstrates (against Darwin, Paul Ekman, and their followers in Classics) how facial expressions do not index basic emotions, but rather require contexts for viewers correctly to interpret them. If I frame your perception of the image on the left below with the choice of ‘anger’ or ‘elation’, you will read Williams’ expression against the frame of those concepts -and statistically on facial expression alone, ‘anger’ will be your choice as Paul Ekman argued of faces like this one to make the claim that faces index basic emotions — but if our picture pans out to the wider view of Williams’ accompanying fist pump, you are likely to read ‘elation’, and this is of course the right interpretation. So contexts — ‘anger’, ‘fist pump’ — matter. ‘Concepts constrain possibilities.’
But here is a further and deceptively simple illustration — discussed in another of our final year modules, on comparative ethics and concepts of happiness — in an illustration the first variant of which was published in 1892, and many times in psychology since, but was perhaps made most famous by its use in Wittgenstein’s (1953) Philosophical Investigations and Gombrich’s Art and Illusion (1960) (Fig. 2).
My version is drawn in class so that it necessitates we activate it by revolving the paper like this:
Isn’t this the perfect Christmas surprise for your young nieces and nephews? I invite our group to reflect. In one fell swoop you can change a toddler’s world irredeemably and forever, for once the paper is turned never more will this squiggly duck be just a duck and not something more. And the larger philosophical point, as Wittgenstein argued, is that the world comes into being for us according to the concepts we bring to bear in our encounter with it, or here, as we instantaneously resolve the same collocation of lines as duck or rabbit according to the conceptual precedents of cartoon protagonists we have seen before ( ‘percepts ordered by precepts’ as Marshall Sahlins’ structuralist theory of history put it); the anthropological question is what then would they become if ‘seen-as’ by those unfamiliar with Donald or Watership Down? And, as Sahlins demonstrated with Captain Cook, something else, no doubt — a question for us to explore — because ‘concepts constrain possibilities.’
But in this example there is more to notice. As the paper is revolved (or our eye attends to a different aspect of the image) and one possible ‘seeing-as’ concedes to another, we experience something marvellous. The world as we first saw it gives way to something surprising and new, and what was once one image, instantly grasped, known and put away, now reveals itself to us to be uncertain and more (Wittgenstein rather nicely called this a ‘dawning’). And herein lies the magic of the moment. For as the singular and familiar becomes plural and undecidable (from Duck to Duck-Rabbit!), we experience not just the flush of unforeseen novelty, but also a small wonder: the transformative joy of possibility. (Didn’t you?)
At this present moment we in the university are awash with anxiety about novelty and what now is possible (see The Guardian 15th Dec). AI is the unreadable face of radical newness. And this is why we are terrified of it. If the value of the university lies in its grand collection of knowledge (its promise to capture the ‘whole’, universitas) organized and policed according to the canons of academic disciplines which set the boundaries of its legitimate forms and application, we ask, how can and should we withstand the emergence of AI, which in all these respects, promises the formerly unthinkable: the boundless consolidation and instant retrieval of something far greater than our authoritative collection: the internet’s ever-expanding universe (a word also from universitas) of knowledge (replete with its misinformation, and new ‘hallucinations’) unfettered and unconstrained, and worse, repackaged as the user’s own.
But what if its threat is itself a ‘hallucination’? For the concept of the university active here in our fear of the rival universitas of AI is born of just one construal of the Latin root of its name, the university as unifying custodian of what is known about the ‘whole of everything’ (universitas). But suppose instead that the true value of the university lies not in its anachronistic claim to such totalizing (read: ‘universalizing’, also from universitas) knowledge (its ambition the mastery of all domains of knowledge, each discipline contributing a part of the whole) or even in its claim to host and authorize the disciplinary regime (its purview, in each field, command over what is and should be known, and in what ways), but rather in the unique space it affords in its gathering together or ‘association’ of multiple lives (the other, medieval, sense of universitas), in and beyond its walls, to engender and ponder that remarkable transformative experience of the uncertain and the new? For the novelty of AI, after all, can only tell you what is already seen and known (even if, worryingly for examiners, not by whom). But the novelty of the university, by contrast, if anywhere, arises in the skilful direction it offers to all —in and beyond its disciplines and dogma—to attend with others to the joyful possibilities of the world ever on brink of dawning.
So should we in the university fear AI in quite the way we currently do? Or does our fear really belong closer to (the concepts by which we define our) home? For, when all is said and done, as anthropologists are want to observe, at work or at home, it is our concepts that constrain our possibilities.
Figure captions and sources
Figure 1: https://www.researchgate.net/publication/376746804/figure/fig1/AS:11431281213958961@1703254892456/Barrett-et-al-2011-287.png
Figure 2:
left: Rabbit and Duck.’ Fliegende Blätter, 23rd October 1892 (caption reads: ‘Which animals are most like each other?’), Wiki Commons Images; right: ‘Duck-Rabbit’ Wittgenstein, L. (2009 [1953]) Philosophical Investigations. Blackwell. (Trans G.E.M. Anscombe, P.M.S. Hacker, and Joachim Schulte). 204.
Ashley Clements
is Associate Professor in Greek literature and Philosophy & Head of Department of Classics, Trinity College Dublin. His most recent book is Humans, among Other Classical Animals (Postclassical Interventions). Oxford University Press (2021)