Your complimentary articles
You’ve read one of your four complimentary articles for this month.
You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please
Experimental Philosophy versus Natural Kind Essentialism
Mark Pinder puts Hilary Putnam’s essential philosophical theory to the test.
The stereotypical philosopher sits in an armchair carefully thinking through difficult conceptual problems. As with any stereotype, however, there are exceptions, and in recent years a number of self-styled experimental philosophers have expressed unhappiness with this armchair methodology. These philosophers have got up out of their armchairs, and are performing experiments and doing other empirical research to put a variety of philosophical views to the test.
This turn of events may be surprising. After all, philosophy is not science. Whereas we can experimentally verify scientific hypotheses, it does not at first seem possible to experimentally verify philosophical views. For example, whereas DNA has been used to verify the hypothesis that humans are closely related to chimpanzees, there is no empirical evidence that could be used to verify the utilitarian view that the morally right action is whichever choice leads to the most overall happiness, since we cannot observe morality or experimentally determine what is morally right. Likewise, while particle physicists have used the Large Hadron Collider at CERN to test their prediction of the existence of the Higgs boson, the philosopher cannot test the sceptical view that we don’t really know whether there is an external world. That view seems instead to depend principally on where one puts the bar for real knowledge, and so is not subject to empirical verification or falsification. At face value, philosophical views aren’t the sort of views that can be experimentally tested.
Nonetheless it turns out that empirical evidence is relevant to at least some philosophical views. In order to see this, let’s look at one example in depth. The example I will focus on is a philosophical view called Natural Kind Essentialism, which you may already know of in connection with the infamous ‘Twin Earth’ thought experiment. (More on that shortly.) After introducing the view, I will show you how a recent empirical study bears importantly on the issue – and so why not every philosopher can rest easily in their armchair.
Natural Kind Essentialism
Natural Kind Essentialism is, unsurprisingly, a view about natural kinds. But what are natural kinds?
There are various different kinds of things in nature. For example, objectively speaking, one molecule of water is the same kind of thing as another molecule of water, but it is a different kind of thing from a molecule of ethanol. This means that water and ethanol are examples of natural kinds. There are many other examples. Cats are an objectively and naturally different kind of thing from great white sharks; hearts are an objectively and naturally different kind of thing from kidneys; protons are an objectively and naturally different kind of thing from electrons; and so on. So all these things are examples of natural kinds.
If nature really is divided up into objectively different kinds of things, then we can try to understand what the differences are, or what makes one kind of thing different from another. Natural Kind Essentialism aims to do just that, by saying that natural kinds have essences, and then trying to discover what these essences are. For example, the essence of water is typically said to be its chemical composition, H2O, so that H2O is the only thing that counts as water. On this view, then, a molecule of ethanol is not water because ethanol does not have the right chemical composition. Similarly, the essence of the species ‘cat’ might be its genome; the essence of the heart might be the function to pump blood; and the essence of the proton might be its composition – two up-quarks and a down-quark. According to Natural Kind Essentialism, every natural kind has an essence.
Perhaps the most famous argument for Natural Kind Essentialism is Hilary Putnam’s ‘Twin Earth’ thought experiment, to be found in his essay ‘The Meaning of “Meaning”’ (1975). It goes like this. Imagine that, far across the universe, there’s a planet remarkably similar to Earth in many ways, which we will call ‘Twin Earth’. At first sight, Twin Earth is just like Earth: there are human-like people living in cities, who comb their hair in the mornings and go to work as accountants, builders, teachers, philosophers, etc.; there are similar varieties of trees and plants that cover the land; and there are vast oceans filled with a clear colourless liquid that, after desalinisation, quenches thirst. However, it turns out that there is no H2O on Twin Earth. Rather, the liquid that fills the oceans and comes out of taps has a very different chemical composition, for which we will use the formula ‘XYZ’.
Rough Sea © Steve Lillie 2017. Please visit www.stevelillie.biz
Now ask yourself: Is there any real water on Twin Earth? According to Putnam, the intuitive answer is ‘no’. He tells us that the meaning of the word ‘water’ is fixed by the chemical composition of whatever liquid we usually call ‘water’. And, as the liquid we usually call ‘water’ is H2O, Putnam concludes that ‘water’ means H2O. If Putnam is right here, then it straightforwardly follows that all water is H2O, and all H2O is water. It also follows that a molecule of XYZ, or of ethanol, or of anything else, won’t count as water, just because it doesn’t have the right chemical composition, however similar it might otherwise appear to water. So, as there’s no H2O on Twin Earth, there’s no water on Twin Earth. The conclusion is that water has an essence, namely its chemical composition, H2O. Putnam argues that there’s nothing special about water: we could run parallel arguments for ethanol, cats, hearts, protons, and other natural kinds, saying that for each kind there is a distinction between the appearance and the essence. And if we did run these arguments, then we would have an argument that each natural kind has an essence, that is, an argument for Natural Kind Essentialism.
Let’s see how empirical evidence is relevant here. Notice that in arguing for Natural Kind Essentialism, Putnam uses the intuitive premise that ‘water’ means H2O. But word meaning does not depend on Putnam’s intuitions. Rather, the meaning of a word depends upon how that word is conventionally used in a linguistic community. So, although Putnam would sit in his armchair arguing that the meaning of ‘water’ is fixed by its chemical composition, or that the meaning of ‘cat’ is fixed by its genome, etc., these are in fact empirical issues about how the words are actually used by speakers of English. And the empirical evidence suggests that Putnam is wrong.
Let’s consider an empirical study carried out by Jussi Jylkkä, Henry Railo and Jussi Haukioja, Psychological Essentialism and Semantic Externalism: Evidence for Externalism in Lay Speakers’ Language Use (2009). This aims to determine, in two stages, how the meaning of a word is fixed.
In the first stage of the study, the experimenters presented participants with ideas along the following lines:
“There is a yellowish, bitter-smelling, fragile mineral common in Serbia, that scientists believe to have chemical composition ABC. It is called ‘zircaum’. Furthermore, a deposit of a yellowish, bitter-smelling, fragile mineral, which scientists believe to have chemical composition ABC, has recently been found in Norway.”
The experimenters asked their participants to say whether they thought the recently-discovered mineral found in Norway was zircaum. The majority judged that it was zircaum.
The second stage is where it gets interesting. The experimenters told the participants that the scientists in the story had been wrong about the mineral in Norway. Contrary to what the scientists originally thought, the mineral actually had the chemical composition KLM. The participants were asked whether, reflecting upon this new information, their first judgement that the mineral was zircaum had been correct, incorrect, or correct in one sense and incorrect in another.
If Putnam is right about how words mean what they do, we should expect participants to say that their first judgement had been incorrect: they would think that ‘zircaum’ means ABC, while the mineral found in Norway has turned out to be KLM, and so the mineral found in Norway was not zircaum after all. The actual results, however, were as follows: 22% said that the original judgement was correct, 48% said that the judgement was incorrect, and 17% said the judgement was correct in one sense and incorrect in another. The other participants felt that they couldn’t say. So, although 48% of participants judged in accordance with Putnam’s intuition, the majority of them – the remaining 52% – did not share Putnam’s judgement.
How can we make sense of the data? I think that the best interpretation is that ‘zircaum’ is ambiguous: the word can have multiple meanings. Sometimes its meaning is fixed by its chemical structure, and sometimes by its surface properties – being a yellowish, bitter-smelling and fragile mineral. That is why some participants understand ‘zircaum’ to mean ABC, some participants understand ‘zircaum’ to mean ‘yellowish, bitter-smelling, fragile mineral, and some participants recognise that ‘zircaum’ can be understood in either way, and that we cannot objectively say that any one use of the word is better than any of the others.
Although there is no such thing as zircaum, the word ‘zircaum’ is just like ‘water’ and ‘cat’ in the relevant respect that they are all understood by the relevant language users to be words for natural kinds. So, just like ‘zircaum’, we can expect ‘water’ and ‘cat’ to be ambiguous. Sometimes people will use ‘water’ to talk about H2O, and at other times people will use ‘water’ to talk about any clear, ocean-filling, thirst-quenching liquid. Sometimes people will use ‘cat’ to talk about animals with the relevant genome, and at other times people will use ‘cat’ to talk about aloof, mouse-catching creatures of a certain type of appearance that people take as pets. And so on. The empirical evidence from this study, then, suggests that Putnam is wrong about the meaning of words for natural kinds. His premise that ‘water’ always means H2O is false.
Would the real zircaum please stand up?
To see what becomes of Putnam’s argument, we need to think about how it is affected by the discovery that sometimes ‘water’ means ‘clear, ocean-filling, thirst-quenching liquid’.
Often this happens precisely when people are discussing Twin Earth. For example, many people find it natural to describe Twin Earth by saying “The water in its oceans isn’t H2O, but XYZ.” And in flashes of imagination, it can be just as natural for them to include such embellishments as, “We can imagine that Twin Earth operates under totally different physical laws, and that the only H2O on Twin Earth isn’t water at all, but a rare mineral used in luxury shampoo.” But those statements wouldn’t make sense if ‘water’ was being used to mean H2O. Rather, here ‘water’ is being used to mean ‘clear, ocean-filling, thirst-quenching liquid’; which, when we’re talking about Twin Earth, is XYZ. Taking these uses of language into account, it is easy to see that some water isn’t H2O – namely, the water in the oceans on Twin Earth – and some H2O isn’t water – namely, the H2O used in shampoo on Twin Earth. So it can’t be the case that to be water is to be H2O, nor that H2O is the essence of water. The consequence is that we’re not going to be able to argue for Natural Kind Essentialism on the basis of what ‘water’ means. That is to say, in light of the empirical evidence we have been discussing, it seems that we should give up on Putnam’s argument for Natural Kind Essentialism, after all.
Does it follow that Natural Kind Essentialism is false? No. Regardless of the evidence we’ve been discussing, nature might nevertheless be divided up into objectively different kinds of things, each kind with a distinctive essence. For example, all H2O might have one essence (characterised by being H2O) and all XYZ another (characterised by being XYZ). What the empirical evidence has shown us is that Putnam’s argument for Natural Kind Essentialism, using what words mean, fails. If you think that Natural Kind Essentialism is true, you’re going to have to find a new way to argue for it.
The above discussion shows that empirical evidence can be relevant to philosophy, even to the abstract sort of views usually confined to armchair theorising. Natural Kind Essentialism posits essences in an attempt to understand the objective differences that we find in nature. But, according to the empirical evidence we’ve looked at, Putnam’s argument for Natural Kind Essentialism fails. Of course, a single study doesn’t settle the empirical facts, and so we would perhaps need to look at many more experiments or studies before finally consigning Putnam’s argument to the scrapheap.
This is just one example of how empirical evidence can be useful in philosophy. Philosophers are increasingly using experimentation as an objective basis to test their intuitions and arguments. Evidence has already been deployed in a wide range of philosophical debates, for example about consciousness, free will, moral responsibility, race, the mind-world relation, and knowledge, and the list is growing. It is fair to say that we can expect this trend to continue.
All of this suggests that, somewhat surprisingly, philosophy may not be so different from science after all. Philosophers are taking a more empirically-minded approach to their quest for knowledge, getting up and going out to perform experiments and studies. And so, if nothing else, you may just see a few more second-hand armchairs for sale over the next few years.
© Dr Mark Pinder 2017
Mark Pinder is a Visiting Professor at the University of Hertfordshire.