The Charge of Essentialism (a reply to David King)

Essentialism can be defined as the view that all things, or groups of things, possess essential features without which they would cease to be what they are. So, for example, all squares have four equal sides and four right angles. Without these fundamental attributes, a shape would not qualify as a square.


Essentialism has been around for more than two millennia, so inevitably there are several different versions on offer. Perhaps the most extreme version was developed by Plato who believed in an abstract realm of perfect "forms" of which the things of our world are merely imperfect copies. So, for example, all squares are approximations of an ideal square to which our only means of access is by way of ideas. Hundreds of years later, John Locke did not posit an immaterial realm of perfect forms but held instead that essences are in the mind, as ideas to which "things existing are found to agree, so they come to be ranked under that name."


Interestingly, it might be suggested that every form of essentialism must necessarily partake of some essential feature or features by which it might be identified as such. Just as all squares share essential features, so too perhaps, do all forms of essentialism. Nonetheless, as has been famously noted by Wittgenstein, many conceptual categories cannot be reduced to essential features, but share what Wittgenstein called "family resemblances." Different members of a family may share no single feature in common, yet several different features may be shared across two or more members. Games, Wittgenstein argued, are related in this way. As a conceptual category, there is no essential feature of all games, yet each game shares features in common with one or more other games. Accordingly, there can be no typical member of a family and no typical game. Therefore only a cross section of examples is likely to provide an indication of some, but not necessarily all, of the overlapping features of the group as a whole.

Wittgenstein's theory of family resemblances is often taken to be strongly anti-essentialist and is frequently cited in disputes over whether it is possible to determine necessary and sufficient conditions for certain concepts. It should be noted though, that family resemblance categories are not entirely boundless even if many conceptual boundaries are fuzzy. That we usually do not need to see a border between Scotland and England in order to know which country we are in does not mean that no border exists or that no boundary can be erected. Likewise, Wittgenstein is not suggesting that there are no linguistic rules at all or that it is pointless to explore the boundaries between one conceptual category and another. For example, if the concept of games had no boundaries, then it would be indistinguishable from the more general concept of "activities" which itself would be indistinguishable from the concept of "change" etc. Wittgenstein's central point is that we do not need to establish any boundaries or to consult any putative mental or ideal essences to know how to use a concept. But in the process of making this point, he is not saying that we cannot establish necessary and sufficient conditions for concepts under any circumstances. Nor is he saying that it is impossible to distinguish between different conceptual categories through conceptual analysis. My reasons for these important clarifications will become clear in a moment.
On several occasions, philosopher, blogger and Facebook discussion group moderator, David King, has labeled me a "fundamentalist", a "Platonist", a "language policeman" and, in response to my last blog post here on the subject of communication, a "rabid essentialist". Evidently, by characterising my position in these ways he hopes to discredit my view in as expedient as way as possible. King is well aware that my commitment to conceptual analysis owes a great deal to the work of Wittgenstein and since he is also aware that Wittgenstein is widely regarded as an anti essentialist, he sees it as "ironic" that my methods should be so at odds with what he takes to be Wittgenstein's position. I hope I have already made it clear that Wittgenstein's position on essentialism is all too easy to oversimplify. After all, it was Wittgenstein who wrote "Essence is expressed in grammar." Of this statement, the Stanford Encyclopedia of Philosophy explains:

The 'rules' of grammar are not mere technical instructions from on-high for correct usage; rather, they express the norms for meaningful language. Contrary to empirical statements, rules of grammar describe how we use words in order to both justify and criticize our particular utterances. But as opposed to grammar-book rules, they are not idealized as an external system to be conformed to. Moreover, they are not appealed to explicitly in any formulation, but are used in cases of philosophical perplexity to clarify where language misleads us into false illusions.

On at least two occasions King has asked whether I have conducted tests of native speakers or undertaken corpus analysis of people's every day speech interactions. In order to clarify where language misleads us into false illusions we do not need to consult native speakers to know that "communicating at someone" is not a strong candidate case of communication. Simply being an experienced native speaker—with all that that entails—is sufficient qualification to know that to communicate about someone is rarely, if ever, to communicate with them. And we can also readily observe that it is not entirely unintelligible—although it is somewhat strained—to say that we can "communicate to someone" or "communicate to ourselves" but we cannot do so without also communicating with someone or with ourselves. And to communicate with our hands, lips, mouth or voice is not to communicate to our hands, lips, mouth or voice. None of these points about the relation of the preposition "with" to the verb "communicating" require any corpus analysis any more than they require us to consult some inner essence, ideal, prototype or guide. Nor in fact do they require us to be experts in linguistics, although such credentials would probably help to convince some doubters. So, to suggest that science, linguistics or corpus analysis are necessary—essential even—to establishing anything about ordinary language is to have missed one of Wittgenstein's central points.

Even though it was clear that King's remarks were aimed more at undermining my arguments than engaging with them, it occurred to me that I could indeed test my conjecture in a way that might partially satisfy King's scrutiny. I Googled "communicating with", "communicating to", "communicating in" etc. and found, just as I had expected, that the preposition "with" is by far the most common usage. In fact it is sixteen times more common than its nearest rivals "to" and "in". With further research it might be possible to use Google to perform a statistical analysis of ordinary language locutions but in this case the results merely confirmed what should have been obvious from the outset. Nonetheless, for King, who seems unwilling to accept reasoned argument (philosophy) on its own merits, perhaps the possibility of providing some readily available statistical evidence might offer a certain utility.

My point in drawing attention to the relationship between the concept of "communication" and the preposition "with" has always been to emphasise that communication is a transaction, and like all transactions it usually occurs between two or more agents. We can communicate with ourselves of course—by making notes, keeping a diary or simply by talking to ourselves—but one group of locutions that are tellingly absent from Google are the following: "communicate in myself", "communicate inside myself" and "communicate in me". I take these as unequivocal evidence that the concept of inner communication is beyond the bounds of ordinary usage. King would dismiss these observations as a priori stipulations but such accusations do not change the fact that communication is a thoroughly public affair that can only ever occur between communicators or, at the very least, on the part of an individual already skilled in such transactions. There are no communicators within us, and without communicators there can be no communications.

So when theorists speak of inner communication, signals, codes, representations and information, they overstep the bounds of ordinary language. Communication is the mark of culture, not biology or physics. These theorists and researchers unwittingly expand the concept of communication to include causal influences, natural processes, stimuli, reflexes and autonomic mechanisms, none of which are communicative behaviours at all and simply do not qualify for inclusion in the category. Just because there are some family resemblances between one conceptual terrain and another does not mean that there is no boundary. Scotland is not a part of England.

When King accuses me of philosophical equivalent of nationalism, he fails to appreciate what I am trying to achieve. He argues: "You just assume without argument that neuroscientist are using the word 'communication' in a sloppy manner because they don't conform to your conventions. This is not an argument it is a statement of how you think they should speak." As I have repeatedly tried to make clear, the conventions I identify are not "my" conventions, they are the conventions of ordinary language. I do not berate anyone, but I am very critical of the misuse of ordinary language within the sciences and philosophy. But it should be noted that I am not critical of these misuses because of some quasi nationalistic commitment to ordinary language. I am critical of these misuses because of the devastating effect they have on our understanding of the difference between natural processes on the one hand and intentional actions on the other.

Intentional behaviours (of which communication is paradigmatic) require goals. What is the having of goals if not the having of abilities to represent the thing or things with which we are engaged? King consistently fails to address this question. If we cannot produce representations then we cannot communicate. This is why I am so insistent that communication is the turning point between biology and culture. If we bundle all natural processes into the same conceptual category as communication then there can be no clear distinction and many important insights will be obscured from view.

Evidently King regards all criticism of conceptual confusion as little more than an irksome punctiliousness on the part of a minority of "Wittgenstein fanatics". He does acknowledge that "Hacker and Bennett DEMONSTRATE [King's emphasis] some contradictory and confused uses of certain terms" but he is clearly unwilling to consider the possibility that the problems of conceptual confusion are much more far reaching.

The Concept of Communication

It may not seem to matter much how we each carve up our conceptual world. If you choose to apply a concept in one way and I choose to apply it in another, the potential for difficulty might seem to be of little consequence. Where, for example, is the conflict if, like the people of Lilliput, you choose to attack your conceptual boiled eggs from the narrow end and I, like the people of Blefuscu, choose to attack mine from the broad end?

If our efforts are limited to individual projects, then there is little likelihood of disagreement. We can agree to disagree. But if, on the contrary, we wish to cooperate, then the potential for confusion, loss, damage or harm can be very significant. Two examples can be used to illustrate this point.

In 1999, a $125 million NASA mission to send a probe into orbit around Mars, ditched into the Martian surface. Unbeknownst to NASA, one of the contractors had used metric units in their component instead of the imperial standard used by NASA.  

In 2003, the builders of a new bridge between Sweden and Germany discovered that the German side was more than half a metre higher than the Swedish side. The engineers were already aware that Germany and Sweden determine the height of sea level in different ways, but they had mistakenly reversed the correction, thus precisely doubling the 27cm difference rather than cancelling it out.

Problems like these are perhaps best regarded as conversion errors. The bridge engineers converted between two conceptual systems incorrectly and the rocket component engineers simply took it for granted that no conversion was necessary. When errors have glaring material consequences, it is often relatively straightforward to trace the source. But conversion errors need not be obvious and may survive over long periods due to a lack of appreciation of the significance and scope of the problem. I hope to show that we face exactly such a problem with the way that the concept of communication is understood and this has profound implications in all fields in which the concept is used.

The concept of communication can be understood in any of three incompatible ways. The first—what we might call "pervasive communication"—defines communication as the transfer of information ("differences that make a difference" as Bateson put it in 1972) between various entities. According to this view, all forms of life, and even their parts, communicate with one another. For example Baluška et al. (2009, p.123) write: “Roots are able to produce and to sense growth regulators, chemical messengers and metabolites that communicate to the whole plant the result of processing and integration of that information.” And Bais et al. (2004) claim that plants communicate with other organisms: “Increasing evidence suggests that root exudates might initiate and manipulate biological and physical interactions between roots and soil organisms, and thus play an active role in root-root and root-microbe communication.” Search Google for “bacterial communication” and you will find nearly 50 million results. Clearly this generalised notion of communication is extremely prevalent. Some theorists even speculate that there is communication of information at the quantum level. Whilst, at the other end of the spectrum, it is not at all uncommon for pheromones to be described as as a form of "nonverbal communication" or for body posture, eye movements etc. to be called "body language."

Most people recognise that language is a subordinate concept of which communication is the superordinate. This is why we distinguish between the verbal and the non-verbal; between language on the one hand and what are sometimes called "the mimetic arts". Nonetheless, it is possible to find cases where communication  is conceived as a "kind of language" (Argyle 1975, Hudin 2009). This sense of communication is what the British philosopher Gilbert Ryle (1949) would have called a "category mistake" (which, incidentally, I would argue, is a subordinate form of conversion error). Instead of conceiving of language as a sub-category of communication, communication is regarded as a sub-category of language. We find such category mistakes for example when people claim that language is a kind of picture (Wittgenstein 1922) or that a "picture is a model" (Ibid), that "models describe the world" (Daiper 2003) or that "Drawing is a mode of description" Ingold (2011). There is little illumination to be gained by explaining one form of representation in terms of another. Strictly speaking, pictures do not describe or model anything, language does not picture or model anything and models do not describe or picture anything. A cat is not a kind of dog. With this in mind, it is understandable that Wittgenstein later rejected his "picture theory" of language as misleading. And I suspect that this realisation may well have had a significant influence on his later important emphasis on conceptual analysis.

The final—and I think the most coherent—sense in which the concept of communication is commonly used can be found at the beginning of the first sentence of the current Wikipedia entry devoted to it: "Communication (from Latin commūnicāre, meaning 'to share') is the purposeful activity of information exchange between two or more participants..." In contrast with "pervasive communication", this more restricted sense emphasises that all communications are meant; that they are intentionally produced with a purpose, usually of eliciting a response on the part of another individual or individuals. This is why it is incoherent to conceive of communication as an act that can exist without an intention, goal or purpose. Bacteria do not intend anything, quantum particles do not pursue goals, roots do not deliberately influence microbes and people do not purposefully generate pheromones.

Communication is a purposeful activity because in principle it involves an anticipated outcome: a response on the part of another perceiver. We cannot communicate with trees because trees cannot communicate with us. Communication is a reciprocal affair. There is no such thing as communicating at something.

Embodiments of Mind

I have often mentioned on this blog that our actions are embodied. Sometimes I have also claimed that our thoughts are embodied in our actions, as I believe they are. There is nothing unusual about such turns of phrase. They are commonplace in ordinary language and serve very well to describe the relationship between our many skills and the ways these can be applied to the world.

In the last twenty years or so, a new breed of theorist has emerged, a breed who see themselves neither as immaterial minds pulling the strings of material bodies, nor as brains locked away in skulls, but rather as minds interwoven into the very fabric of the body. These Embodied Mind theorists come in several varieties, but their principal claim is that the brain is not the locus of mind, but rather the body as a whole. Whilst I think this view is a vast improvement on that of mind/body dualism, I think it is mistaken to suggest that the mind is dispersed throughout the body, or throughout anything for that matter.

To be a minded creature is to be in possession of a range of abilities, all of which can be performed and are thus communicable in principle. The possession of abilities is nothing like the possession of scars or dentures or artificial hips. Abilities are not bodily things we can point to or examine with a scanner, no matter how sophisticated or precise. They are not to be found in our bodies but in the skilful things we do, the actions we perform and the complex tasks we engage in and carry out.

To develop a new ability—to learn something—is to be capable of demonstrating it, and to be capable of demonstrating an ability is also to able to use it to envisage goals, to form expectations and to anticipate outcomes. Such predictive capacities are the embodiment of intelligence and are by no means limited to our fellow human beings. But what does it mean to use the word "embodiment" in this ordinary way and what light might this usage shed upon the theory of the Embodied Mind?

Firstly, it should be obvious that my thoughts of swimming are not embodied in my sitting down whilst thinking of swimming. Nor are my thoughts of dancing or of laughing or casting my eye over Rodin's "Le Penseur". My posture and furrowed brow might well be taken to be the embodiment of my thinking, but not my thoughts. I can only embody my thoughts if I enunciate or enact them. It makes no sense to say that I am the embodiment of my thinking. When I act, my thoughts are embodied in the act, not in my body. But when I do not act, my thoughts are not embodied in my inactivity.

When we say that Hitler was the embodiment of evil, we do not mean that evil took bodily form in Hitler (though we might be sorely tempted to think so). We mean that Hitler is a prime exemplar of everything we regard as evil: he was the personification of evil. In ordinary language, to embody something is to represent it, not to contain it or to instantiate it or to be it. A red thing does not embody redness. We reserve the concept of embodiment for symbols, not for things that actually instantiate the relevant property. When we say that an act was the embodiment of goodness, we mean that the act could be taken as a representative symbol of the concept of goodness, not that goodness has taken earthly form. Likewise, when we say that a person is the embodiment of innocence, this is an attribution, not an attribute that we should expect to find instantiated in their corporeal frame. You will not find any more truth in the body of someone who speaks truths than in the body of someone who speaks lies.

In a recent discussion with a neuroscientist and advocate of the Embodied Mind, he claimed that physiological sensations of guilt etc. are "100% instantiated in emotional somatic markers in the body." Putting aside the question of what emotional somatic markers might actually be, let's imagine that my young son unwittingly does something widely regarded as bad. If everyone agrees and tells him so, but he does not accept this, then according to our advocate of the Embodied Mind, my son's action cannot be bad because he will not have the requisite embodied markers of guilt. Something is very obviously wrong with this conception of wrongdoing because the criterion for the determination of guilt or innocence etc. is taken to be a personal possession (or lack of) rather than a publicly negotiated rule or convention. Our concepts are not private possessions. My son cannot choose to define his conceptual world by fiat, but must accept social codes that are instantiated in public practices, not somehow somatically inscribed in the bodies of individuals. Our abilities to use concepts are not instantiated in us. We are not the embodiments of our minds, our actions are.

Embodied Mind theories have emerged in opposition to various forms of Cartesian dualism. As commendable as this opposition is, it comes at a price, the price of coherence. To suggest that our minds are to be found distributed throughout our bodies is to suggest that our minds are somehow contained in the body in the way that we might mistakenly take the potential for my hand to dissolve in acid to be instantiated in my hand. Philosopher, Peter Hacker, identifies this as transcendentalism:

One tempting misconception is transcendentalism, i.e., the fallacious reification of powers, according to which they are conceived as occult entities mysteriously contained within the possessor of the power.

In their reification of the mind, embodied theories fail to recognise and to clarify that the mind is neither a thing nor a non-thing. Abilities are not to be found by examining the body, no matter how precisely. We are not the embodiments of our minds, our actions are.