Medieval manuscript illustration of a goat and a person holding a disc, with gold circles in the background, surrounded by text in Latin script.

From Liber de natura rerum a 13th century encyclopedia by Dominican friar Thomas de Cantimpré. Courtesy the Bibliothèque de Valenciennes, France


The problem of erring animals

Three medieval thinkers struggled to explain how animals could make mistakes – and uncovered the nature of nonhuman minds

by Sam Alma + BIO

From Liber de natura rerum a 13th century encyclopedia by Dominican friar Thomas de Cantimpré. Courtesy the Bibliothèque de Valenciennes, France

You are standing on a boat that is drifting down a placid river. You watch the trees on the shore glide along. For a moment, it looks like the trees themselves are moving – not your boat. But this, of course, is mere appearance: the trees are still, and it is your boat that moves. This parallax effect was described by medieval philosophers, but it may be more familiar in another form: when you’re sitting on a train slowly rolling out of the station, it can seem like it is the stationary train next to yours that is departing instead.

A handful of 14th-century scholastic thinkers wondered how this parallax effect came about. What explains our perceptual error? Let’s call this the problem of erring. In finding a solution, the medieval philosophers had to take into account another observation: nonhuman animals err too. According to scholastic orthodoxy, human and nonhuman animals were alike in being animals. Even so, within this category, humans occupied a special place: they are the only type of animal that is endowed with an intellect, a rational soul. They are, in medieval parlance, rational animals.

What happens in the case of erring? Do nonhuman animals err? Together, these two questions posed a considerable philosophical challenge to William of Ockham (c1287-1347), Adam Wodeham (c1295-1358) and Gregory of Rimini (c1300-1358). The first question, as we will see, was answered that erring is an act of the intellect. To the second, the answer was a tentative yes, as empirical observations suggest. Together, these answers put the medieval philosophers in an awkward position: nonhuman animals carry out acts of the intellect, yet without having one. Call this the problem of nonhuman erring.

While centuries old, this debate reveals a question that is relevant for today’s debates on nonhuman minds. How much are we willing to give up in order to hold on to our basic assumptions? Nonhuman minds continue to be a source of discord. After all, we simply cannot infer with absolute certainty from the outside what is going on in the inside of someone else, let alone a nonhuman being.

The above three scholastics bend over backwards to fit the empirical findings of erring animals with the idea that only humans have an intellect. As such, the medieval debate on erring animals exposes the underlying structure of many disputes on nonhuman minds. The nonhuman mind perplexes them and mocks their basic ideas of how the world works – much like it challenges us nowadays.

William of Ockham was born around 1287 in a village in Surrey a day’s ride from London. In his early teens, he joined the Franciscan order and subsequently went to the London Greyfriars convent (as the Franciscan friars were also known). There he was taught grammar, philosophy and theology. Over the years, as he started to teach these topics himself, criticisms of Ockham mounted and he was even accused of heresy at the highest ecclesiastical level: the papal court in Avignon (in present-day France). In 1324, Ockham went to Avignon to defend himself. Then things got worse.

Ockham studied the writings of the pope, John XXII, and came to a remarkable conclusion: the pope himself was a heretic. This did not go down well in Avignon, and Ockham fled on the night of 26 May 1328, eventually arriving in Munich where he spent the rest of his life until his death in 1347. Despite his clash with the pope, Ockham proved to be one of the chief figures of medieval philosophy. He played a pivotal role in kickstarting a movement known as nominalism, whose followers favoured an austere ontology: not for nothing is the well-known parsimony principle called ‘Ockham’s Razor’.

Ockham’s discussion of the problem of erring was written in the early 1320s, before his Avignon venture, and displays his razor to its full effects. When you are standing on that boat riding down the river, Ockham observes that the illusion that the trees are moving is due to a mistaken judgment from your own movements. When you are moving and the trees are not, your perceptions are equivalent to the perceptions you would have if it was you who was standing still and the trees were moving. That is, if all you perceive is that the trees were two metres to your left, while a moment ago they were two metres to your right, it is an equally valid – but false – conclusion that the trees are moving. Let’s unpack this claim a bit.

The intellect either assents, dissents or doubts the proposition in question

The psychological process leading to the erroneous judgment of an illusion goes as follows. At the start, you perceive what is the case in the world around you. Standing on the boat, you see a tall green-brownish object over there. Medieval philosophers called these simple impressions intuitive cognitions. They are directly caused by present external objects, and for this reason they are not only simple – the basic building material of perception and cognition – but also reliable. That is, they cannot be false: if you intuitively apprehend a green-brownish object, it is the case that there is such an object.

Next, this raw sensory data is processed by the internal senses and the intellect. The so-called ‘common sense’ integrates different pieces of sensory information: the sound of rustle and the green-brownish colour are placed together as belonging to one object. Some of the simple apprehensions are abstracted into concepts. The rustling green-brownish object is a tree. Together, intuitive cognitions and concepts are bundled into propositions. A proposition is a mental sentence composed of mental ‘words’, namely, the concepts and intuitive cognitions. In our case, the following proposition results:

The tree is moving.

Finally, then, judgments come into play: the intellect either assents, dissents or doubts the proposition in question.

The assent to the proposition that The tree is moving is, as we know, erroneous. How come? Here Ockham’s theory is able to provide a neat solution. To start, you have a bunch of intuitive cognitions like the following:

(1) The tree is eight metres from you (at t1).
(2) The tree is four metres from you (at t2).
(3) The tree is two metres from you (at t3).

Now, if you have fallen for the illusion, you implicitly use something like the following rule of thumb:

(4) If the distance between some x and you changes over time, then x is moving.

From this, it follows that:

(5) The tree is moving.

However, step (4) is too quick: if the distance between an object and you changes, the only thing you can conclude is that something moves – whether it be you or the object. Therefore, one should not assent to the conclusion. It is a false inference, an erroneous judgment.

Erring runs on heavy mental machinery. Composing mental sentences and reaching a verdict are both acts of the intellect. In this way, Ockham is able to explain the illusion by appealing to something we all have: the rational soul. Nevertheless, this also presents a difficulty to Ockham, which he grudgingly points out: nonhuman animals err too. As his student Adam Wodeham says:

[A]lso to brutes, the trees appear to move. For when the ship moves towards the trees, they run away [as] if these [trees] are something frightening to them.

Yet if ‘brutes’ err, and erring is making a false judgment, and this in turn is an act of the intellect, it follows that nonhuman animals too have an intellect.

This goes straight against medieval orthodoxy. Many scholastic thinkers adhered to a form of (human) exceptionalism, as it is called nowadays. Following Aristotle, humans were distinct from all the other animals in virtue of being rational. Church authority had it that humans – and humans only – were created in the ‘image of God’. As such, they were the only ones to possess an immaterial and rational soul. The only animals, that is, with the capacity to think, form concepts and judge.

It has the absurd consequence that nonhuman animals never go wrong

Allowing for nonhuman intelligence runs the risk of throwing out both the Aristotelian theory and the Christian faith that formed the fundament of medieval philosophy – too big a cost. So Ockham’s solution poses three conflicting claims:

(a) Nonhuman animals do not have an intellect.
(b) Erring is an act of the intellect.
(c) Nonhuman animals err.

Take a step back and note the status of the different claims. Claim (a) is a background assumption; claim (b) is the theory that explains erring; and claim (c) is a piece of empirical observation. At least one of these claims has to go, but which one?

As I said, claim (a) is an unlikely candidate, being one of the cornerstones of medieval philosophy. Rejecting claim (b) is not much better for Ockham, as this would undermine his own solution to the problem of erring. Neither is claim (c) an obvious pick. One could reinterpret the evidence, but it has the absurd consequence that nonhuman animals never go wrong. The sanctioned background assumption, the cherished theory or the straightforward piece of evidence: Ockham wants to keep all of them.

In the end, Ockham chickens out. ‘For the sake of brevity,’ he says, ‘I refrain from explaining how this could happen’ in the case of ‘brute animals or those who lack reason.’ Luckily, Adam Wodeham and Gregory of Rimini are willing to take a closer look.

Like Ockham, Wodeham was a Franciscan friar. He was born near Southampton around 1295, going to London sometime in his 20s. Wodeham lived in the same convent as Ockham, who taught him until Ockham left to go fight the pope in Avignon. Wodeham assisted Ockham and edited his works; as such, he was a follower, although not uncritically. Wodeham later studied at Oxford, and afterwards taught across present-day England, where he died in 1358. Unfortunately, little else is known about Wodeham’s life (besides the fact that he went to Basel in 1339 to investigate some miracles; the obligations of medieval philosophers differed from today’s). While he may be less famous than Ockham, Wodeham’s teaching proved to be influential both in England and in Europe.

We already saw that Wodeham brought in the observation that a ‘brute’ runs away when sitting on a boat that approaches a tree on the shore. What exactly is the problem this poses? Since nonhuman animals have a sensory apparatus, they are susceptible to simple and reliable impressions of what is the case. That is, ‘brutes’ have intuitive cognitions – for instance of a green-brownish object – just like humans have them. These cognitions lead to certain action patterns: sometimes, Wodeham observes, an animal flees something, sometimes it pursues something. It is exactly this behaviour that suggests the capacity of judgment. After all, one only flees something that one judges to be bad; and one only pursues what one judges to be good.

Wodeham uses the following rule of thumb to guide the interpretation of ‘brute’ behaviour, which is an inference from animal behaviour to a mental state:

If x flees or pursues y, then x judges y (to be bad/good).

Much like a proposition – although Wodeham doesn’t call it that – the evaluative judgment is part of a mental complex or composition. This complex is composed of multiple elements:

intuitive cognition + evaluation

So, in the nonhuman illusion case that Wodeham described, the sum of the complex is something like this:

green-brownish object (intuitive cognition) + bad (evaluation) → flee

Both of these – mental composition and judgments – are marks of the intellect. So the behaviour of ‘brutes’ bespeaks a rational soul. As Wodeham himself concludes: ‘I do not see why they should not be called rational animals.

Some horses, for example, are scared of fireworks; others are not bothered at all

Having fleshed out the problem, Wodeham tries to refute it by denying the rule of thumb: while nonhuman animals often flee or pursue things, they do not judge. Rather, the flight or pursuit is an immediate reaction, an automatic response to an intuitive cognition. It is an instinct proper to the species of ‘brute’. Where it was:

intuitive cognition + evaluation → flight/pursuit

Wodeham cuts away the second half of the complex, the judgment:

intuitive cognition → flight/pursuit

Without the judgment, the complex has dissolved. No mark of the intellect is left in nonhuman behaviour. In other words, Wodeham reinterprets the empirical evidence in order to maintain both Ockham’s theory and the exceptionalism of human beings.

However, further observations challenge this view, as Wodeham is well aware. If nonhuman flight and pursuit is but an instinct, there should be no individual differences within a species. After all, this instinct would be hardwired in the sensory make-up of the animal in question. Yet there are differences. Some horses, for example, are scared of fireworks; others are not bothered at all. Some dogs flee the tree, others pursue it.

According to Wodeham, this is because of an individual’s memory or imagination. One dog may have had a bad experience with trees and therefore flees, while another dog may imagine a tree to be delicious food and runs towards it. While he stresses this is still a single and simple cognition, it now consists of the following sum:

intuitive cognition + memory/imagination → flight/pursuit

Is this not a complex, composed of different mental elements? If not, why not? The weaknesses of Wodeham’s strategy were spotted by Gregory of Rimini.

Gregory was known by his fellow medieval thinkers under the unsettling nickname ‘Torturer of Infants’, which was based on his views on predestination and the impossibility of the salvation of unbaptised children. He was born in Rimini (present-day Italy) around 1300. Unlike Ockham and Wodeham, he didn’t join the Franciscans, but the Augustinians, and he spent his life flitting between studying in Paris and teaching in Italy. At the end of his life, Gregory moved to Montpellier to briefly become head of the Augustinian Order. His work was printed into the 16th century and had an impact on a number of important figures, especially Martin Luther.

Gregory provides a handful of arguments to prove Wodeham wrong: nonhuman animals do have complex knowledge and make judgments. To start, Gregory stresses that nonhuman animal behaviour shows that it is not always instinctive:

This can be proven, first, because, as we see, a brute at some time apprehends some sensible thing, such as a bread, and moves towards it. At some other time, however, it apprehends the same [kind of] thing and does not move towards it.

As Wodeham already pointed out, there are intraspecific differences. One snow leopard, for example, reacts differently to the same stimulus as another snow leopard. Gregory’s point, however, is more refined: the very same individual facing the same stimulus can act differently at different times. Now, suppose with Wodeham and Ockham that nonhuman animals have only simple intuitive cognitions that never get assembled into larger mental compositions. If so, their actions are instinctive: the same simple cognitions result in the same type of action. Even if we grant Wodeham that there could be reactive immediacy in spite of differences between members of the same species, this immediacy implies that the same cognition assimilated by the same individual cognitive hardware should lead to the same behaviour. After all, what else could cause variation in reaction?

Sometimes, the bear finds out that the golden substance is not honey and walks away

However, there is intra-individual variation, as Gregory observes. We therefore need to assume larger mental complexes to explain this.

Not only do empirical observations suggest nonhuman mental composition, they also indicate the capacity of judgment. Gregory gives some examples:

[S]ometimes an animal that desires something sweet or something else with regard to taste moves towards a particular thing, the colour of which it has apprehended. This, however, would not happen, as it seems, if it would not judge that this thing is sweet.

What is happening here? Let’s make it a bear that sees a golden substance at a distance. This bear integrates different cognitions, from different cognitive faculties. On the one hand, there is the present intuitive cognition of a golden-coloured object. On the other, there is the past cognition of honey being golden and sweet, stored in the memory. The bear judges the yellow substances to be sweet on the basis of these two cognitions taken together (indeed, a mental complex). After all, Gregory adds, sometimes the bear finds out that the golden substance is not honey and walks away. It thought the substance was something different than it turned out to be.

As the empirical evidence piles up, Gregory concludes that nonhuman animals surely employ mental compositions and judgments. From this, however, Gregory does not conclude that nonhuman animals have an intellect. He dodges the bullet by concluding that judgments and mental compositions are not marks of the intellect. Rather, the boundary of the rational soul is drawn elsewhere, where it is not crossed by nonhuman animals. Human rationality has characteristics not shared by nonhuman animals, and it is on the basis of these that judgments ‘are called “rational” in the proper sense of the term’. Gregory adapts the theory of the intellect.

Human and nonhuman animals, it appears, have parallel rationalities. The ‘proper’ type of rationality involves general and abstract concepts, as well as the ability to compare the present, past and future. This response faces some worries. For one, do Gregory’s own examples not suggest any of the above characteristics in the nonhuman case? Does not the bear employ past experience to anticipate the future as well as act on a universal concept of honey? Gregory’s solution also faces a problem of relevance. Why would the use of general categories, abstract things and comparisons over time imply another, if not better, type of rationality? Pointing out a difference is not enough; Gregory should say why this matters. (Note this problem retroactively affects the initial marks of the intellect as well.)

So where does this leave us? Ockham came up with a clear-cut explanation of the erring. With it came another problem: nonhuman minds are rational too. Wodeham and Gregory each tried a different strategy to evade the problem of nonhuman erring. While they disagree about their strategies, Wodeham and Gregory both hang on to the divide between human and nonhuman animals.

In hindsight, it is easy to discard the whole endeavour as a misguided form of speciesism. The problem that all three shared was that they took for granted that humans occupied a unique and superior place because of the authority of the Church and Aristotle, in spite of the lack of evidence for this claim. This reply, however, would overly simplify the discussion, and overlook what is going on between these thinkers. When theory and observation on nonhuman minds do not match up, it can be far from clear from the outset what to address. Wodeham’s and Gregory’s strategies were, I think, the result of genuine puzzlement, not a piece of motivated reasoning from the comfort of the armchair. Wodeham and Gregory simply overlooked their speciesist assumption.

Even so, their reasoning is not reckless. It is guided by principles that mirror a tension between two guidelines in animal psychology phrased much later. Wodeham uses some sort of Morgan’s canon, a principle coined by the 19th-century British ethologist and psychologist C Lloyd Morgan: animal behaviour should be explained by the least complex psychological processes possible. (This fits very well with Ockham’s razor.) The flight or pursuit of nonhuman animals can be explained without referring to mental composition and judgments. Therefore, Wodeham argued, judgment and composition should not be ascribed to nonhuman animal minds.

Humans are exceptional in some rational abilities, just like giraffes are exceptional as regards their neck length

Gregory, to the contrary, is much more willing to ascribe these mental capacities. This attitude reflects one taken by David Hume some centuries later: if we ascribe a capacity to humans because they behave in a certain way, we should also ascribe it to nonhuman animals when they behave similarly. Otherwise, the standard would be arbitrary, or unreasonably high. Since ‘brutes’ are prone to the illusion just as humans are, they judge and compose just as humans do. Even so, in the end, there seem to be two types of rationality for Gregory, very much alike: one for humans, and another for the other animals.

Their two strategies forced novel ways of looking at nonhuman minds. What does being intelligent amount to, and how is erring related to this? How should we interpret observations? Can we incorporate individual differences in our theory? In tackling questions like these, the scathing and dogmatic speciesism that may seem to mark these medieval thinkers turns out to be a quite weak form of exceptionalism. Humans are exceptional in some of their rational abilities, just like giraffes are exceptional as regards their neck length. This does not mean, however, that no humans have necks or no giraffes are intelligent.

All this gives us a worthwhile insight. Nonhuman minds, ranging from dogs to whales to leafhoppers, confront our basic ideas and spur new ones. The study of the nonhuman mind (like any study) involves assumptions, theory and the interpretation of observations. It is not always clear how to disentangle these, or to pinpoint where, if so, it goes wrong. In a way, rationality, the intellect or any other mental capacity are for us to define. Gregory shows great philosophical insight when he notices this:

But if you still want to call these [animals] ‘rational’, too, you can; because words are arbitrary. And so, one could also deny any simple knowledge to them and conclude that others are rational, if one wants to call all beings that have cognition ‘rational’.

Even so, one thing is clear: nonhuman minds are baffling. They were to the medieval philosophers so distant from us, and they still are.