This paper is a set of explorations of topics related to classification, grouped round the work of Linnĉus, but with a quick look back at Aristotle, and leading to issues about mental disorders.
On a personal note, I might add that a (remote) origin of this paper is in the Faringdon Road in London, where, a long time ago, numerous vendors of second-hand books used to set up their barrows. I would go there as a school-boy and as a student, to browse and sometimes to buy. One of the items I bought, probably for a florin, was a late 18th century collection of nosologies, written in Latin. It was falling apart. From time to time over the years I have leafed through its apparently lapidary lists, with increasing curiosity.
ĥ1 There are two kinds of truth:
If I say “this is a goat”, I express a belief. The belief can be justified by giving acceptable evidence that the object in question meets some agreed criteria for something to be a goat. But we can also ask whether it is acceptable to employ the term “goat” at all. Aristotle thought, in the De Anima, that the question of truth arose in two different ways. First, there was what we might now call propositional truth: is it true that this is a goat ? But second, there was the question whether the term “goat” itself represents a true classification, i.e. whether it represents a justifiable way of dividing up the world. Some classifications may not be “true” or justifiable. They may contain an internal incoherence, or they may fail in other ways to function as a proper element in our mapping of the world.
ĥ2 A challenge: cannot the “truth” of classifications be analysed in terms of truth of beliefs?
Today, many philosophers and some scientists would hold that Aristotle’s second kind of truth is an illusion. Set theory is constructed extensionally: it permits any classification of items. The only questions arising are what the members of a given set (other than the null set) are, and which propositions about those members are true. In this framework, the question whether a classification is itself “true” cannot arise directly. An attempt to give an extensional account of such a notion of truth can be envisaged, in terms of how many well-entrenched propositions are true of all, most or many members of a given set, or of how many predicates occurring in such propositions are “projectible”. But I shall not make such an attempt here. For, despite the cleanliness of extensional logic, intensional structures are so deeply embedded in human thought about the world (and not only in the domain of the mind), that the ideal of a purely extensional language seems too unrealistic. (I shall not address the larger issues which are called up by these rather superficial remarks, such as whether we need a Fregean notion of “sense”.) Let us allow ourselves, at least for the time being, to retain the justifiability of classifications as a distinct question.
ĥ3 Democracy : an example
Take a more substantial example, to avoid a “one-sided diet of examples”. James Madison wrote, in connection with his advocacy of a democratic republic in America : “The people will have virtue and intelligence to select men of virtue and wisdom” This belief has been, and will continue to be disputed by those who believe that ordinary people cannot be relied upon to make good choices of this sort. Historical evidence can be adduced to support such a claim (Hitler was elected). On the other hand, similar evidence can be adduced to suggest that élites cannot be guaranteed to be, or to remain, virtuous and wise either. Whatever the rights and wrongs of such a disagreement, the presupposition of discussing it is that the parties share a sufficient common understanding of what virtue, intelligence and wisdom are, of what is meant by ‘the people”, of whether “men” includes women or people who do not own real estate or people of any ethnic origin, and of what constitutes “selection”. History reveals that these particular classifications are not unproblematic, either in thought or in practice. Before we can consider how we can justify our beliefs on such matters, we have to ask how we can justify our classifications. But how is this to be done?
Aristotle claimed that there are natural kinds, independent of human classificatory practice, and able to be discovered by empirical investigation. There are tigers, talipots and tourmaline. And there are jackals, jacaranda trees and jasper. Jackals are, in nature, different kinds of animals from tigers, the jacaranda is a different kind of tree from the talipot, and jasper is a different kind of semi-precious stone from tourmaline. These differences are real, and exist in nature independently of human classificatory practice. The task of enquiry is to investigate them, and to find out by appropriate empirical enquiry what the essential differences are. This will on occasion require a certain amount of ‘rectification of names’, when classifications represented by the words in a given language do not correspond to the real ones which we may discover. We shall learn, for instance, in the hackneyed example, not to classify whales as ‘fish’. The most remarkable early illustration of this approach is Aristotle’s History of Animals, which describes about 520 species.
Now Aristotle also claimed that artefacts have no “essence”, strictly speaking. Tables, tragedies, tunnels, or indeed the constitutions of city states, were not natural objects, and so should not be treated as falling into natural kinds. Consider Aristotle’s investigation of the form of dramatic presentation called ‘tragedy’ in his Poetics. Tragedy was about as recent for Aristotle as cinema is for us. It is well-known how he tried to give a sort of definition — we might better say a characterization — of this form of dramatic presentation, and one which had great influence in the much later development of European literature.
But Aristotle did not think that tragedy was a natural kind. While theoretical knowledge did have the empirical investigation of the differences between natural kinds as one of its aims, its goal being to establish the truth, productive knowledge was concerned with rules of production, like a recipe book. Nevertheless, the production of tragedies or tables, trifles or trains does not occur in a vacuum. These are not arbitrary objects, in spite of the considerable degrees of freedom which exist for creative work. In the case of tragedy, Aristotle thought that pity and fear are universally occurring human emotions, which, as it were, help create the psychological space for a certain kind of possible artefact, the tragedy: not a natural kind, nor yet immutable, yet a kind occurring in a specific connection with the natural and cultural world. In the case of social and psychological phenomena, no doubt, the interplay between the natural and the artefactual will create further complications.
ĥ1 The persistence of the Aristotelian view
In spite of the overthrow of Aristotelian teleology when “the mechanical philosophy” prevailed in modern Europe, Aristotle’s idea that some classifications have a direct natural basis has persisted. So while the Aristotelian view of teleological explanation has been largely abandoned in the ‘harder’ natural sciences, the Aristotelian view of classification has been more persistent.
ĥ2 Linnĉus: classification by genus and species
After Aristotle, perhaps the greatest classifier was the Swedish scientist Linnĉus, in the 18th century. Linnĉus is of course most well-known for his prodigious and epoch-making work in botanical classification, which rapidly became the standard for classification, and parts of which remain basic to today’s practice. Linnĉus initially thought that genera and species were indeed natural kinds, created distinct by God. But he came to the view that new species could come into existence over time. Nevertheless, varieties, i.e. subdivisions in a species marked by merely superficial differences, should be regarded as ‘Nature’s sport’. So we should concentrate on species and genera, and pay attention to nomenclature. He wrote in the preface to his Species Plantarum, “to acquire knowledge of these things properly, individual specimens should be covered by a distinct idea and a distinct name, without which the copiousness of things would necessarily overwhelm us, and, in the absence of a common language, all communication will cease”.
Here is a further theme. Even if there are natural kinds, there is also the pragmatic need to deal with them through language. Linnĉus wrote, in the foreword to his Journey in Västergötland “Without names, our knowledge of things would also perish”. So nomenclature was very important for him. He supplemented fuller descriptions of those characteristics which differentiated species within a genus by what he called trivial names in the margin. His general comment in the preface is that this enables us conveniently to distinguish any species of plant with a single species name. However, he adds, all sensible botanists, however distinguished, must never propose a trivial name without the basis of sufficient specific differentiĉ, “lest science should rush back into pristine barbarity”.
ĥ3 Linnĉus’ Sexual System
But there are many genera of plants (for Linnĉus, over a thousand), so the question arises how to group them in turn. Linnĉus’s new technique was to take the plants’ organs of reproduction, that is, pistils and stamens. He then classified the genera by the “sexual method”, which is based on the configuration and number of these organs. A modern textbook of systematics praises Linnĉus for this: “Linnĉus recognized that the floral parts were more useful than the vegetative organs in grouping like with like. He noted that vegetative parts were usually more variable ... On the other hand, floral parts usually were quite invariable under different environmental conditions. Therefore, a classification based on similarities of the floral parts reflected more accurately the phylogenetic relationships.” The author adds, however, that “for us today Linnĉus’ sexual system appears contrived”. The author does not give the details of how Linnĉus in fact characterized these groupings. Put your cursor here and then here to see the translation into English made by “a Botanical Society at Lichfield” and published in 1783 with thanks to “that great Master of the English tongue, Dr. Samuel Johnson, for his advice in the formation of the botanic language”.
ĥ4 “Injured Modesty”
Linnĉus’s sexual system quickly prevailed, though it was replaced within about two generations by more “natural” systems developed in France. But it was not universally well received. Johann Amman “doubted very much if any Botanist will follow his lewd method”, Siegesbeck said that “such loathsome harlotry” as several males to one female would never have been permitted by the Creator, and called it “a lascivious method”, and another commentator said that no text-book of botany should bring “the blush of injured modesty to the cheeks of the innocent fair” (click on “innocent fair” to see a letter of mine about this to the New Scientist).
ĥ5 “Strange Thoughts”
Thus we find in Linnĉus’s actual classificatory work in botany a remarkable combination of empiricism, pragmatism, formalism and fantasy, a mixture of the arbitrary and the necessary. It is basically Aristotelian in its view and treatment of genus and species, but it adds conventional nomenclature for convenience of identification (the binomial system of nomenclature which is still standard), and superimposes human marriage systems (real or imaginary) for convenience of finding your way about in the mass of material.
There are many questions that can be raised about Linnĉus’s “sexual system”. We could think, on the one hand, that his recognition of the crucial role of sexual reproduction in plants was an advance, but on the other, that his expression of this through a systematic analogy with human marriage customs was an indulgence. And if we wonder why he indulged himself in this way, it is interesting to note that he wrote: “The genitalia of plants we regard with delight, of animals with abomination, and of ourselves with strange thoughts”? It is worth lingering on this. We have a classification genitalia, i.e. the sexual organs. Let us assume that this is a justifiable and natural classification of features in the world around us, including, in spite of what may seem Linnĉus’s excesses, its application to the vegetable kingdom. And let us accept what Linnĉus reports, as a possible human reaction to these features. To contemplate the sexual organs of plants is a pleasure. To contemplate the sexual organs of animals tends to give rise to disgust. To contemplate our own sexual organs gives rise to strange thoughts. Assuming such reactions, but without assuming that they are, as it were, “correct”, let alone “true”(!), we may ask why they should occur. Perhaps it is that we do not perceive sexual reproduction in plants as involving desire. Instead, we can view it as an elegant, perhaps ingenious and beautiful mechanism. For ourselves, in spite of all that science can offer in a way comparable to what it can offer in the case of the vegetable kingdom, it is difficult merely to contemplate sexuality, since we are involved in it from the inside through our own body and desires. So it can give rise to “strange thoughts”. But some other animals are sufficiently different from us to preclude the inside view, yet sufficiently close to be quite close to home. So regarding the sexual organs of animals as of sexual interest to us is both possible and dangerous. This makes disgust possible. We now find a new topic in our discussion of classification. There is not only the question whether and how classifications or classificatory schemes can be justified, there is also the additional possibility that they might engage our thought structures and our emotions, at a quite deep level.
ĥ1 Uncertainties in the classification of mental disorders
I now turn from the classification of plants to the classification of mental dysfunction. Note that this itself is a more general category than “mental disorders” might seem to be. Indeed, Thomas Reid comments: “it were to be wished that we had ... a nosology of the human understanding”, going on to give a sort of epidemiology of errors of judgment based on Bacon. For Reid, errors typically arise from otherwise beneficial powers and proclivities. But he has nothing to say about “errors of judgment” which occur in what came to be thought of as pathological contexts.
Questions about madness and medicine were raised by Michel Foucault some thirty years ago. He tried to trace deep conceptual changes in the category madness itself which accompanied its coming to be treated as an object of clinical attention at a particular time and place in history.
More recently, Oliver Sacks has shown how some conditions of “mental abnormality”, well-described, have disappeared from history, only to reappear many years later. What can justify calling, or failing to call certain types of behaviour “Tourette’s syndrome”, for example?
ĥ2 Nosologies
The classification of mental disorders has been viewed as a subdivision of the classification of diseases in general. During the enlightenment, great attention was paid to this question. Cullen, for instance, wrote:
The art of discerning and distinguishing diseases may best be attained by an accurate and complete observation of their phenomena, as these occur in concourse and succession; & by a METHODICAL NOSOLOGY, or an arrangement of diseases according to their genera and species, established upon observation, abstracted from all reasoning. This arrangement we have attempted in another work.
Some examples of attempts at these tasks are those of Sauvages, Vogel, Sagar, MacBride, and Linnĉus. This veritable explosion of classificatory systems in the eighteenth century is itself noteworthy, and characteristic of the encyclopĉdic period.
Just one of Linnĉus’s eleven genera of diseases is “mental”. He lists the species of mental diseases, which are said to involve alienation of judgment, as ideal, imaginary, and pathetic. The full list, appears, to a modern eye, not so much a classification of sickness as a sickness of classification.
ĥ3 Questions
A central question here is whether the genus mental disease, or certain species of mental disease, can be viewed as in any sense being natural kinds.
ĥ1 Cognitivism
Linnĉus describes the whole genus of mental disorder as involving ‘alienation of judgment’, not only those forms which involve distortions of thought and imagination, but also those which involve distortions of the passions. The idea here seems to be that it is a defining characteristic of humans to have the ability to form beliefs and make judgments. But sometimes this ability undergoes a pathology, in which the judgments you make are no longer yours, or are no longer integrated, so that this (primarily cognitive) disorder can give rise to various further symptoms. Take Linnĉus’s sub-species of mental disorders which he called “imaginary”. People judge there to be a ringing sound when there is not, a visible object when there is not, and so on. Here is a defect of judgment, whatever its cause. And the defect is straightforwardly cognitive. But desires too are put under the same heading, in the sub-species “pathetic”. “Indecent desire for lovers” and “desire for country and relatives”, for instance, are given by Linnĉus as examples of alienation of judgment. What alienation of judgment can be thought to arise in these cases?
I shall do my best to offer an interpretation. Note, in the case of the indecent desire for lovers (erotomania), that it is distinguished by Linnĉus from an enormous desire for sex (satyriasis). Now, what is lust (a less clinical notion)? There are traditions, from Plato and the Buddha onwards, in which lust, like other desires, passions or emotions, is considered to be a sort of psychic force. It contains no judgment: judgment would come in simply as a means of satisfying the pre-existing brute desire, in line with the view of Hume that “Reason is the slave of the Passions”. From this perspective, there would be no deeper reason for distinguishing between erotomania and satyriasis. It would just be that people with what is vulgarly called “a strong sex drive” would make different instrumental judgements about how to satisfy it.
ĥ2 Satyriasis and Erotomania distinguished
Perhaps there is after all something interesting in Linnĉus’s distinction. If we grant, for the sake of argument, that these two conditions are pathologies, then we cannot view the role of reason as purely executive. Some mountaineers of my acquaintance take a similar view. They enjoy and value mountain climbing, but they think that those who simply want to increase the number of peaks which they have climbed miss the point of the activity. Peak-bagging, they consider, is a sort of perversion. A fetish has taken the place of the proper object of desire. Similarly, those who want to maximize, in number, their sexual liaisons, would show a defect, not primarily in desire, but in judgment. The ordinary desire for sexual congress has become distorted into a score-card. There is, again, a specific kind of defect of judgment. I am not asserting this analysis, but trying to offer a way of making Linnĉus’s classification comprehensible.
But what about satyriasis? Presumably, for Linnĉus, the satyr is a male who wants to have sex as much as possible, no matter which way. Sagar’s definition is that it is “a desire peculiar to males, of which the main symptom is an unrestrained and impudent desire for sex” (both authors distinguish this from nymphomania, which is “an unrestrainted, audacious and impudent desire for sex in women”). It seems that here too Linnĉus regards the condition of satyriasis as a pathology, and as a pathology of judgment. In other words, fetishism, for instance, as one channel for the satyr, would be taking something as an object for sexual interest which does not have the appropriate characteristics to be such an object. Here too, there would be a defect of judgment, but one which is significantly different from the case of erotomania.
ĥ3 Other views
For Linnĉus, mental diseases are treated as distinct entities, which have not only differentiĉ, but also symptoms. And we are to regard as separate what differentiates a disease in nature from what is merely a symptom of it. He wrote “symptoms are related to a Disease, as leaves and supports are to a plant”. (We have already seen how Linnĉus’s Sexual System for classifying plants gives central place to reproductive organs in plants, and only a secondary place to other features, such as foliage.) Now, other nosologists gave lists of disorders quite similar to that given by Linnĉus, though with some differences of terminology and definition. Sauvages, for instance, differentiates vesaniĉ as involving error in imagination, appetite or judgement. Sagar’s list is very similar to that of Sauvages, though he adds a fourth type of error: of memory. Vogel, for his part, has two distinct genera, one (number seven) is “hyperĉstheses”, involving excessive or perverse sensations, which is made to include vertigo and boulimia, for instance, and the other (number nine) is “paranoiĉ”, which includes melancholy, enthusiasm, stupidity and sleep-walking. There is therefore a good deal of variation in detail between these authors, as well as disagreement on the cognitivist hypothesis. Indeed, it is far from clear how far various of these conditions could really be identified as distinct, and as a disease. It is interesting to note that MacBride’s eighth order of diseases is “mental”, but that he gives no list at all. The editor of a compendium of nosologies including MacBride’s wrote:
The distinguished author inserted no list of the diseases which can be included under this title; and this, it seems, for the reason that he doubts whether the genera of mania and melancholy, commonly listed under this heading, actually differ in any way other than that in which a disease which is the same in nature manifests itself in different ways at different times.
ĥ1 Nostalgia : an affliction?
Let us turn to nostalgia (etymologically, shall we say, “aching for return”). This is interesting for us here in Hong Kong. There is the fact that many Hong Kong people feel a desire to return, or at least to make a pious visit, to the place of their ancestors. There is also the fact that there are quite a few people in Hong Kong, like myself, who grew up in other places and cultures, and some of whom have a desire or intention to return to them. The curious thing is that several of these 18th century authors, including Linnĉus, should have classified nostalgia as a disease. It may be worth repeating what Sagar wrote about nostalgia:
Nostalgia is a desire to see one’s parents, relatives and native soil which is so ardent, that those affected suffer deeply from this one passion; the interaction between body and soul is almost completely interrupted; all their senses and other appetites go to sleep; chronic fever consumes these miserable creatures, and leads to death if the opportunity of returning to the homeland and the relatives is denied; I once suffered from this at Krizevci in Croatia. I could not eat, I had obstipation, I was swollen with anasarca, insomniac and debile; but when I returned home, I got better without any medication.
It is also worth adding that nostalgia is included in a standard modern classification of diseases.
ĥ2 ‘Roots’
Now nostalgia, whether or not it may take a pathological form, is clearly related to the notions of “roots” and of “belonging”, which provide, in turn, the ground not only for nostalgia but also for loyalty, for nationalism and and indeed for racism.
There are many people in the world who believe that racial distinctions are important, a belief which is in some ways self-verifying, since, by acting on the belief that they are important, one tends to make them important in fact. Yet it is evident, for most human populations, that racial classifications are scientifically questionable. In other words, terms like, say, “Black, White and Yellow” in their actual application to humans do not denote groups of people with gene-pools which are systematically and sufficiently distinct enough to be biologically significant, so as to justify a distinction of race, rather than constituting a mere variety, as opposed, for instance, to a distinction of culture.
To take a couple of examples: in some respects (such as genetically determined dental characteristics) Southern Chinese populations have more in common genetically with other South-East Asian populations than with their fellow-Chinese in the North. Or again, the genetic predisposition to thalassĉmia (or sickle-cell anĉmia) is found in populations in Italy and Greece, in South China, and in American Blacks.
As for Caucasians and blacks and Jews, to take some examples at random, it is at least as obvious that none of them as a group will have a systematically distinctive gene-pool. No doubt, our groupings of people, such as those mentioned, vary in the extent to which there are underlying genetic differences, and where they do exist, the importance of distinctive genetic features in the genome as a whole will also vary. We still have good grounds to resist the kinds of racist research enterprises associated with the likes of Arthur Jensen in the USA. This is because we know that the classifications used in these efforts are doubly questionable. Even if “Black, White and Yellow” had a sound biological basis as a method of classifying humans into so-called “races”, there are clear historical reasons to assert that U.S. blacks, whites and yellows would not be representative samples of those putative races.
ĥ3 Stereotypes
Linnĉus himself could joke using stereotypes of other human beings, but there is no sign that he thought that they had a sound natural (as opposed to cultural) basis. Himself Swedish, wrote the following characterization of the Swede: “he eats like an Englishman, drinks like a German, dresses like a Frenchman, builds like an Italian, smokes like a Dutchman, takes snuff like a Spaniard, tipples vodka like a Russian.” These are cultural stereotypes. For Linnĉus, no doubt, some or all natural differences within the human species could be merely varietal, that is, “nature’s sport”.
My concluding remark is that we ought to consider the justifiability and viability of our classifications. We cannot afford to await the advent of a pure extensional language, even if that were possible. And in doing so, we shall have to take into account not only empirical and pragmatic considerations, but also the affective, what it is like to be the particular creatures that we are. This is not a plea for relativism: on the contrary. It is a plea to consider and take seriously and to see how to justify every factor which may play a direct or indirect role in our classificatory practices, something which I have tried to illustrate today by taking the work of one of the great classifiers in human history, warts and all.
In particular, if there is a viable distinction between natural kinds and kinds which are not natural, or are the mere sport of nature, two questions arise. First, how should we classify those which are not natural ? And second, how should we decide whether a kind is natural or not ? Is the classification of mental disorder something fabricated with our ways of living, or does it have a natural basis ? Or is the truth about this, as often, something much more complex and messy, in which empirical or natural, pragmatic or utilitarian, formal or nomenclatural, considerations can and will all play their part ? By answering these questions, we provide ourselves with a framework for understanding some of the ways in which the classification of mental disorder, from the eighteenth century onwards, has been so variable and refractory.
