Volume 3, Chapter 12:- Plant classification – a sociology of ignorance?

Introduction:     This essay is inspired by the chapter, of “Taxonomy and the sociology of botanical knowledge” written by Charles Craib in his book “Grass Aloes in the South African veld” (Umdaus, 2005).  That particular chapter was written purportedly to explain why he prefers the taxonomic revision of Reynolds (1950) to that of Glen and Hardy (2000).  Craib states that Reynolds provides “… a better reflection of the infrageneric taxonomy of the group”, and also better suited to “elucidation of processes concerning the autecology of the SA grass aloes”.

Craib maintains that… “The classification used by Glen and Hardy takes its place in the knowledge production process as a more abstract model than that proposed by Reynolds”.  However, Craib states that abstraction is a trend “…in the development of SA botanical knowledge and that such trends can be expected in the history of this development”.   He maintains that the discipline of a “sociology of knowledge” is essential for understanding how classification systems work as they do, and also useful in accounting for the regular revision of plant genera and changes in plant names.  I would like to ask if this argument is true for Haworthia and can it in any way explain the remarkable plurality of names that exist there and being added to exponentially.

In my opinion, Craib’s views are extremely relevant.  Not necessarily because they clarify the problem so much as demonstrate that there is indeed a problem.  Many of Craib’s assertions need to be questioned if the acquisition of knowledge is really the motive for writing and reading. I am truly seeking closure on the whole issue of why I ever wrote about Haworthia and still do.

Analysis:  What does the title “Taxonomy and the sociology of botanical knowledge” actually mean?  Taxonomy according to Collins Dictionary is… “the branch of biology concerned with the classification of organisms into groups based on similarities of structure, origin etc”.   Sociology is…”the study of the development, organization, functioning and classification of human societies”; and knowledge is…”information about a subject”.  So the title appears to mean that the object of the chapter is to enlarge on the way in which the classification of plants generally is related to the functioning of society with respect to the grass aloes and the communication and sharing of knowledge about them.

The impression I get from Craib’s chapter is that there are two available classifications which present him with something of a dilemma in deciding which one to use.  The particular problem in Aloe needs to be properly explained and understood if any contribution is to be made in that respect to the “knowledge production process” in botany generally, and this is indeed what Craib attempts to do.  Reynolds’ revision of the Southern African Aloe was published in 1950.  It was and still is a remarkable product by a non-botanist (Reynolds was an optician).  The publication of this work initiated a remarkable and quite extraordinary popular interest in these plants.  Aloes became a collector’s realm and Latin names were the key to the collectibles.  A vast new body of information became available and some of this was reflected in new names (species) in the formal literature.  Thus it was evident as early as 1970 that a new revision to unify and consolidate that information was necessary.  At that time there was also a drive in South African botany for a new Flora to list and synthesize all plant species.  In respect of Aloe, it was probably entirely fortuitous that Glen and Hardy were available and interested, to be assigned the task of revising the genus.

The essence of Craib’s thesis is that Glen and Hardy’s classification is a “more abstract model” than that of Reynolds’s and implies that the reasons are to be found in new technologies and processes which inspire the revisionary processes in botanical classification.  What concerns Craib is that such botanical revisions invariably lead to name changes and re-arrangements of species.  He is not alone in his contention that this process confounds the various segments of society outside and even within botany that use Latin names to generate, accumulate and share knowledge.  In Haworthia confusion is rife.

However, it appears to me that Craib’s chapter is only symptomatic of the problems that beset and confound the literature of Haworthia and similar popular plant groups.  The revisionary process in the hands of competent taxonomists is required to produce a clearer and better understanding of a genus and its complement of species.  If it does not do so and an abstraction results, we need to reflect on the people and the processes.  For science to triumph it is absolutely necessary for there to be organized and proper skepticism.  If this is lacking the whole process of progress is reversed and we come no closer to knowing anything. This is the problem I perceive in Haworthia .  It does not require a “discipline of sociology’ to explain botanical revisionary cycles as these are the natural product of exploration and the inquiring mind.  A discipline is needed to examine the increasing abstraction and theorizing of revisions and the rifts that are widening between users of names and the people who generate, change and argue about those names.  Not only this, but revisions can appear, as in Haworthia, that have no intellectual substance at all.  How does this happen?

Thus any real contribution to the knowledge of Aloe (and other plant genera) in society (Craib’s “sociology of botanical knowledge”) requires a more realistic appraisal of botanical science and the role of individual persons and the aptitudes and competence that they bring to the task.  It is people and personalia which trade in knowledge and it is the invidious fact that these are the factors that have to be evaluated and for which there is no provision in any code.   It is painful to have to express an opinion on the basis of personal merits.  In respect of Aloe, Glen is by observation, a theorist, a nomenclaturist and an historian, whereas Hardy was acknowledged worldwide as a collector, grower and field naturalist par excellence.  It is pure fantasy to suggest that a combination of these skills of two different people would necessarily result in a happy union.  A key issue, which Craib has not properly addressed, is ensconced in his use of the term “model”.  This is when he suggests that Reynolds generated one model for the classification of Aloe, and that Glen and Hardy generated another.  This is at the heart of the entire problem and what the sociology of botanical knowledge needs to address.  It is where the argument already begins to unravel, because Glen and Hardy used Reynolds’ revision in generating their own version of the SAME model.  It should not take any place in the “knowledge production process” unless it merits a place and substantially contributes more information.  The simple assertion that revisions become necessary once names have been assigned to the different entities in a genus is not correct.  Revisions become necessary when exploration and the publication of new information and names proliferate to the degree that a new statement is necessary to synthesize and re-organize all the available information.

If Glen and Hardy have now produced another classification, it is clearly not as a question of model that it can be queried, it is one of competence.  Have any changes, apart from obvious and necessary new additions, been based on tangible new evidence and new information to justify them?  Craib provides some instances where he obviously disagrees and it is quite certain that he is the best positioned and equipped to make any contribution towards the knowledge and truth of the matter.  If one takes the question of Aloe bowiea and its inclusion in a section Leptaloe along with other “grass” aloes, it is evident that Craib disagrees.  Rather than discredit and question this action he makes the cognitive choice that it simply does not serve his purpose.  But what is the truth of the matter.  Knowledge can essentially only be of something that is true.  Aloe bowiea was at one time the unispecies genus Chamaealoe africana.  Despite the argumentation and merit put forward by Smith and van Wyk(1993) it is not an absolute certainty that its transfer to the genus Aloe represents any fundamental truth.  It is unlikely that Glen and Hardy had any better insights and information by which they took a further step and put the species in the section of grass aloes.   Any “sociology of knowledge” has to acknowledge that while “species” may be real, genera and infrageneric groups may very often not be so.  Thus the transfer of Chamaealoe africana may be justified in several ways, but the driving force may have been derived primarily from the assessment… “This resulted in the unsatisfactory situation we have today where 27 generic names are available for some 450 species which could easily be accommodated in seven genera”.

The classification model used by Reynolds and that used by Glen and Hardy are depressingly and disappointingly exactly the same.  There is no difference.  The model is the two-dimensional dichotomously branching tree which is universally used to depict evolution and the relationships of species at the various levels of the classification hierarchy.  All Glen and Hardy have done is to add in the new names not known to Reynolds and to shuffle a few about on the basis of some new information (hopefully) or on the strength of their opinions.  It is an unfortunate reality for us that Reynolds is unavailable to express his opinion of this new product, and no doubt fortunate for Reynolds.

The question being asked in the botanical literature is if this existing old model is in fact adequate and is the binomial nomenclatural system valid.  The consensus seems to be that the system does work and besides that, new advances in scientific technology (molecular biology particularly) are in any case producing such dramatic new insights into the genetics and hence the evolution and phylogeny of species.  Hence any debate on the fundamentals of any model can be put aside for another time and place. The problem does not go away.  It can be categorically stated that Haworthia will never be satisfactorily understood and interpreted on the basis of the current model as used by either Reynolds or Glen and Hardy.  My contention is that this will be found to be true for many other plant genera.

This is a reflection on the model used by both Reynolds and Glen and Hardy in which families genera and species should be accommodated in a formal near symmetrical and easily managed hierarchy.  This is the “sociological” aspect which determines how names and classification should be understood.  The further examples that Craib cites where Glen and Hardy depart from Reynolds, has less to do with abstraction than it does with the competence and integrity of the taxonomic decisions.  Craib is in fact evidencing the better knowledge on which a classification can be based without committing or exposing himself to the scrutiny of the botanical community (professional and very amateur alike) and which can indeed be unpleasant, trivial and prejudiced.   The South African National Botanical Institute accepted the Glen and Hardy revision of Aloe, and Craib tacitly denounces it.

Craib raises one further issue which is extremely important.  This is that there are groups and individuals who have different needs in respect of classifications and that some of these groups require certainty in plant nomenclature and identification.  The answer is not to have different options in terms of classifications that can be used.  The answer is to generate one sound classification which is not abstract and which can be properly understood.

The basic unit of this classification is the species and this is what Latin binomials denote.   Do we ask how present society (the community that has a non-professional interest in plants) understands these names and their arrangement (taxonomy)?  Do we ask if taxonomists understand what it is they are doing and if what they are doing is meeting the needs of society?  Do we ask if taxonomists and society have and share a common definition of the “species”?  In the role that language plays in sociology, names are given to objects to facilitate communication and thus an association of names and objects is built.  The foundation of plant taxonomy is that single specimens serve as the primary objects (the types) to which any name is attached.  Unfortunately it happens that the association of those names is very often built around objects other than the primary specimen.  That is one source of confusion.  The other source of confusion is the association of the single prime specimen or type with others of the same kind.  There is no single common perception of what constitutes a single kind.  The “single kind’ is of course the “species”, and while definition and perceptions of this differ from one person to another, there can never be a true shared knowledge about how they are classified. If we consider consciousness and experience, we have to recognize that life is manifest in an array of living things at many different levels of organization from simple to complex.  We recognize these as different life forms and we use the word “species” to denote them.  It is remarkable that the system of plant names works at all given the fact that there is so much difference in the minds of people of the associations that exist between Latin name and its primary object.  A Latin binomial is not a common noun and a name for a single thing.   The root of the problem is that the association of a name with a type, and the association of that type with any living system of organisms, is confounding.    It is ridiculous to expect the Latin binomial system to be descriptive of all botanical variation.  It is equally ridiculous that interpretations of the meaning of the Latin binomial and its attachment to a hopelessly inadequate definition and diffuse usage of the word “species” remain a sociological problem.  Taxonomists, whatever the individual idiosyncrasies may be, are striving towards a classification which truly reflects evolutionary processes.  There cannot be alternatives, but we also need to realize that there is embedded knowledge that becomes associated with the use of a name.  It is absolutely incumbent on persons who generate name changes, and on user groups, to mutually and cognitively ensure that names and changes serve a common “knowledge” process.

It is just an obvious fact that in society, interest in plant names can range from the trifling collection of oddities to the deep wish of individuals to understand creation itself.  The ancient Greeks had a saying “Man know thyself”.   To many this instruction has no meaning, but there is a word in language which indicates that to a few it is important.  This is “teleology” – do natural phenomena have a predetermined purpose and are not just determined by mindless mechanical laws?  Can we start to consider this fundamental question if we do not properly understand those phenomena we observe?   People are engaged in taxonomy in a way which may only meet the needs of a segment of society, but which does not meet the needs of science nor can serve the interests of a cognitive thinking greater society.


My observation is that the knowledge of taxonomy within society is woefully inadequate.   Hence“sociology of knowledge” is unlikely to find any more science than art in classification.  It has been said that in botany the measure of a classification is that of the nomenclatural juristics rather than that of the practical usefulness or even scientific accuracy and relevance.  It seems unlikely that the botanical community will come any nearer to providing better practical classifications in the future than they have been able to do in the past.  Few botanists can match the application, energy and enthusiasm that a genuinely interested amateur brings to his chosen subject.  A revision can only be a reflection of the degree of sampling and the capacity of the observer to synthesize what he has seen and, axiomatically, the usage which can be obtained from that revision will be relative to the knowledge of the user.  It is evident therefore that the gauge of the usefulness and value of a revision will lie with the knowledgeable user.

Such users may be few and far between and Crain happens to be one of them.  If there were more like him botanical classification might in fact be found to be quite fraudulent.  An additional part of this problem is the excessive fragmentation of the literature and the huge volume of literature that is being produced as a by-product of the need of scientists to validate themselves.  This confounds the knowledge production process and non-professionals find themselves able to join the ranks of taxonomy without much fear of redress.

A new paradigm is needed by which the Latin binomial serves the needs of science and is appropriately understood and modified to also meet the needs of interest groups.  What a specialist group needs to do is to cognitively recognize the ideals of science, the need for proper and good information and a stable reliable structure of meaningful names of equal value by which they can communicate.  Societies, committee and journal editors need to similarly recognize their responsibilities in terms of the stated or unstated objectives of their activities.  This is surely to ensure that a screening process is in place by which the interests of their members and subscribers is protected and served.  Individuals need to be identified who genuinely have the credibility and skills to effectively and truthfully evaluate comment similar to Craib’s and my own, and identify and apply corrections where they are needed.  These in turn need to be fed back to, and respected by, competent scientists. Finally we need to find a less tortuous and circumlocutory way of expressing what needs to be said.

Acknowledgement:  I would like to express my appreciation to Charles Craib for his positive and constructive responses to my communication with him.


By Chareles Craib

Bruce Bayer has written two very useful, interesting and insightful essays with the above titles. One of them “Comments on the Sociology of Knowing Haworthia” refers in part to the chapter “Taxonomy and the Sociology of Botanical Knowledge” in my book “Grass Aloes in the South African Veld” (Umdaus Press 2005: 10 – 12).

The purpose of these comments is to look at the kinds of analysis which the Sociology of Knowledge can provide with specific reference to some aspects that Bayer has written about. The techniques used for analysis examine the knowledge production process. They do not provide alternative types of knowledge nor do they replace existing paradigms. Changes to existing knowledge or paradigms always originate from the knowledge producing communities themselves in this case people involved with the disciplines and sub-disciplines of botany.

Bayer correctly points out that Glen and Hardy use the same paradigm in their revision of Aloe (2000) as Reynolds (1969). The model however which Glen and Hardy use (meaning the constituent elements and concepts in the classification) is presented at a level of abstraction different from that used by Reynolds. Reynolds taxonomy was essentially rooted in extensive fieldwork with numerous justifications provided for his classifications, based on observable characteristics of plants. Glen and Hardy rarely, if ever in grass aloe, provide justification for the nomenclatural changes they make. I cite a number of examples of this in the book (2005: 33 – 34, 39 – 40, 50 – 52, 57 – 61, 75 – 76, 85 – 86, 88 – 90, 101 – 102 and 123 – 126). Essentially the abstraction does not represent any advance from the classification that Reynolds presented and is probably less accurate as it has little to do with aloes in the veld.

Techniques used in the Sociology of Knowledge judge this revision as another attempt to classify aloe with a number of shortcomings in the way in which the knowledge was articulated. Thus this revision does take its place in the knowledge production process concerning Aloe. The use and value of the revision must be determined by the botanical community itself. In the book I essentially used the techniques of criticism, provided for in the sociology of knowledge, to look at what existing classification provided the best model for an autecological study of grass aloes.

Revisions of plant genera occur for several reasons. As Bayer points out (2006 ….) a proliferation of plants names may necessitate a new synthesis. Other reasons are that analysts with different concepts of species, genus and family validly publish new species. All that is currently required for the naming of a new species is a description that has been validly published. The new species cannot be proved to exist empirically. This process will continue to have but a semblance of accuracy until such time as botanists are able to formulate a universally accepted code of what a validly published species concept comprises of. This is very significant since species are the primary units of understanding before grouping into families and genera. This problem has been extensively recognised and discussed by Bayer to some degree or another in most of his many publications.

The Sociology of Knowledge has one primary concern here and that is the characteristic that plant classifications probably represent an art rather than a science. This implies that unless paradigms change there will be any number of revisions of many plant genera in the future. Users of classifications will need to decide which one best suits their purposes. This will be based on principles such as the internal logic of the proposed system and whether or not it represents an improvement from a previous model for understanding the genus in question.

The sociology of knowledge focuses on published information. The competencies of botanists are important to the botanical community when plant genera are revised. However, from the point of view of the analyst of new knowledge, trends in knowledge production processes need to be examined, particularly those which might lend to paradigm changes.

It would be helpful if there was one universally accepted classification of a genus at a given point in time. In terms of the currently accepted methods for producing botanical knowledge this is unlikely to be achieved for two important reasons. The concept of species has no defined universally accepted boundaries and any newly described plant owes its existence to the fact that it was validly published. The huge amount of leeway that this provides will ensure a proliferation of names and provide room for various interpretations of what “species” are. If the concept of species is already abstract (it is the primary unit of understanding) then so much more so the higher level concepts of genus and family.

It is incumbent on the botanical community to change the methods for producing their own knowledge. This will require a paradigm change in the knowledge production process. In this respect perceived inadequacies of the status quo will be required to motivate botanists to change the way in which plants are classified.

Experience with plants in the field is and will probably remain essential for any plant revision. This is not always recognised as the formal point of departure for revisions and is likely to account for the abstractions that are published, which bear little resemblance to the physical reality. None of this is controlled by formal requirements and as a result it is legitimate to publish any kind of revision. The sociology of knowledge in this instance is only concerned with the consistency with which the internal components of the model are articulated at the level of abstraction they are presented.

There are several questions which will need to be asked and answered before botanists will be able to provide alternative or improved paradigms for the presentation of knowledge. One of the most significant is that empirical techniques of science which can be proved and demonstrated are used uncritically to elucidate concepts such as genus. The legitimacy for the combination of these two fundamentally different knowledge types needs to be theorised at an appropriate level before it is accepted as legitimate. Another problem is that there is probably insufficient motivation for botanists to involve themselves in these sorts of significant theoretical debates. A solution may be an expanded range of courses offered to botanists at tertiary level. More people would enter the discipline of botany with an enhanced range of skills for facing these sorts of intellectual challenges.



Bayer M.B. Haworthia Revisited.  Umdaus Press, Hatfield, Pretoria, 1999.

Thoughts on Haworthia. Spiderwalk Services, Durbanville, 2002.

Haworthia Update Vol.1. Umdaus Press, Hatfield, Pretoria.

Haworthia Update Vol. 2. Alsterworthia, Preston, UK. In ms.

Craib C. Grass Aloes in the South African Veld. Umdaus Press, Hatfield, Pretoria, 2005.

Glen H. and Hardy D. Aloaceae, Flora of Southern Africa Volume 5, National Botanical Institute, Pretoria, 2000.

Reynolds G. The Aloes of South Africa. A.A. Balkema, Cape Town , 1969.