(Published in Haworthiad 18:3:102-106, 2004)

The systematics of Haworthia and the production of a practical and useable list of names that reflects a predictive classification for these plants, has kept me preoccupied for more than 40 years.  Initially I simply produced a list of names published in 1976 as “Haworthia Handbook”.  This was not a great work but there was a desperate need then for some kind of conspectus that established order where there was none.  I rewrote the Handbook published in 1982 as “The new Haworthia Handbook”.  This was followed by C.L.Scott’s  “The genus Haworthia, A taxonomic Revision” (1985), which effectively snatched defeat from the jaws of victory.  In 1996 I was asked to produce a synopsis of Haworthia for two projects, and Steven Hammer and Kobus Venter persuaded me to formerly revise the genus.  It should be explained that the greatest obstacle to such a work was the very difficult question of the validity and priority of names as required by an international code for plant nomenclature that sets out to standardize and lay down norms for this process of typification.  The problem in Haworthia is that the illustrations and specimens on which the process is based, were and are, so confounded.  Either there were no specimens at all, or the basis of the names existed only as barely identifiable illustrations.  An additional difficulty is that specimens from different species so often resemble each other, that unless locality data was available, there was often no certainty to what the name actually applied to.  Thus most of the available types were ambiguous and subject to alternative interpretations.  This is a self-evident truth.  The overriding consideration was to give meaning to the names in terms of biology and to give an objective reality to the different kinds of Haworthias that are the basic units of biological diversity i.e. species.  This latter objective seems very mundane and straightforward and it is extraordinary to now recognize what a minefield of discussion, dispute and argumentation is appearing in the literature concerning the subject.

With nomenclature being so confounded, I had in my Handbooks not attempted to formalize the application of names, but rather to cite specimens to make it quite clear on what my usage was based. The application of names to specimens preserved in herbaria is a base-line requirement in systematic botany, but consistently appears to be a difficult concept to persuade eager amateur botanists to adhere to.  At the same time I tried very conscientiously to get as close to a realistic appraisal of the sources of all the names with respect to the fundamental types required by the code.  With the decision to prepare a list of names for Urs Eggli’s Illustrated Lexicon of Succulent Plants, and almost simultaneously the decision to rewrite the Haworthia Handbooks, the formalization of names became inevitable.  Enter disaster.  I. Breuer in Germany had become inspired to undertake the same task.  He presented me with a manuscript of his  “The World of Haworthia – Vol.1” with a request for an introduction and intellectual sponsorship.  It was evident to me that he had no more capacity to do the work any more effectively than Col Scott or myself and there was no time to negotiate a solution.  I declined.  Breuer with Dr Detlev Metzing had already submitted a formal typification of Haworthia names to the botanical journal “Taxon” and the die was cast.  Of course it was a logical step for them and it is recognized in taxonomic botany that one can approach a revision in two ways.  The first is to establish the validity and priority of existing names and then proceed to the field to establish how they can be used.  The second is to proceed to the field and then only decide how best the available names can be assigned.  The advantage of the second method is that it grants one the grace of being able to better interpret ambiguous types and to generate some sense from poor original illustrations.  It also allows one the opportunity of fully appreciating the history of usage, past and present; and then properly formulating the way in which these names can be used with predictive value into the future.  All of this assumes that there is a solid understanding of biological diversity and the nature of the classification process with respect to “species”.  It was evident that Breuer and Metzing were not competent to undertake their mission with respect to either option and definitely not for the first.

In the introduction to my second Handbook, I wrote, “The complexity of what species are, is beyond the object of this book…”.  It is very difficult for me to explain my thoughts on the subject against the background of the extraordinary statements now being made in the literature with regard to species concepts.  The advent of molecular biology and the analysis of DNA have suddenly hit taxonomic botany as a revealed truth, and the vision of a true evolutionary history of plants is seen as an attainable goal.  Having been educated in the environment of animal science it never occurred to me that there was any other objective than this in classification.  Hence the ideal that DNA study now encourages viz. a phylogenetic species concept, is thus in my estimation simply a very damp squib.  The fact that botanists have never had a clear and unobstructed definition of the species which they have been identifying and naming so industriously for so long is only evident now from the convoluted discussions about the reality of the concept and the serious proposals to even abandon the binomial system of names.  The juxtaposition of the words phylogenetic, species and concept is not my construction.  It is one by botanists who have several other species concepts and yet no precise definition of the one word “species”.

My Revision “Haworthia Revisited”, drafted in 1996, was printed in 1999.  The text for Eggli’s Handbook was drafted earlier in 1996 and only published in 2002.  There are some minor discrepancies in these two works, but far the greater confusion has been the impact of the typifications by Breuer and Metzing and the subsequent publication by Breuer of a series of books and articles that add fuel to the fire.  Other authors and their editors seem to have taken every advantage of the existing confusion to make matters worse and fan the fire into a runaway inferno.  The inevitable result has been a relentless struggle on my part to maintain some kind of credibility on behalf of the plants I am interested in, and also for those people who I imagine present my reading public.  It can be likened to sitting at the control panel of a rocket fired wildly at the moon by a sightless scientist, and then firing directional rockets by remote control and a faulty keyboard, to get the act together and the rocket on course.

What has emerged now in one of the house journals of the National Botanical Institute of South Africa, Strelitzia 14, is an annotated checklist of the plants of Southern Africa with the section for Haworthia compiled by Ingo Breuer, E.M.E. Steyn, G.F. Smith, and N.L. Meyer.  There is no doubt very good reason why my participation was not sought considering that I have curated the NBI herbarium collection at the Compton Herbarium in Cape Town, effectively since 1970.  The most telling one is perhaps, facetiously, the assumption that I was senile or dead.  Surprisingly the list of names is remarkably true to my classification, in stark contrast to the surprising and odd opinions maintained by the principal author in his prior publications.  It is not clear quite when the list was drafted, but the book appeared in 2003.

Superficially the Strelitzia list seems to be very authoritative and certainly is a vast improvement to a similar earlier listing.  That was a ridiculous compound of both Col. Scott’s classification and my own.  The reason for that mess is simply explained in the miasma of a classification activity that had, or has no standard and by which any ambitious individual can assume or be granted credibility as an expert merely by publishing new names, irrespective of whether they reflect real taxa.  However, the new list does have problems and a large measure of incongruence.  My classification is founded, within the limits of my intellect, on a thorough effort to get as close as was reasonably possible to “how it could all have happened”.  Where the list seems to follow my classification, I must therefore accept it as sound.  However, the introduction of “new species” described primarily by Hayashi, makes a mockery of the intended all-encompassing and predictive value of my classification as a whole.  This is a wholly different set of problems added to those already generated by the Breuer and Metzing typifications and associated interpretations and applications.  I am just very grateful, as every other Haworthia enthusiast should also be, that Breuer has been seemingly constrained by his co-authors, to abandon many of his other radical and misconceived opinions or at least conceal them for the purposes of the publication.

The prime changes I would make to the list are:-

Haworthia pumila (H. maxima).  I did introduce the use the name maxima in the manuscript version for the Eggli publication and I seriously doubt if any author would have ventured to promote this name in the absence of myself as the “fall-guy”.  While the name pumila is inappropriate, it is the earliest name for this Haworthia and the late Dr Onno Weinands is credited with the formal typification of the name on the basis of the same illustration on which maxima is based.  Further debate on this issue that has already consumed so much paper and ink seems ridiculous.

The introduction of the name integra exposes the flawed nature of the whole list when it comes to the key problem of how these names are USED.  There is a massive interplay of what I have recognized as two distinct species.  These are H. arachnoidea and H. mucronata.  I have written about this problem at length in many places and I cannot go on and on presenting the same arguments over and over again.  The essence is that one has to recognize that “species” in Haworthia are not separate elements in the way in which we want to segment them for any purpose.  The type specimen that Breuer has triumphantly unearthed is a pickled one found in the Berlin herbarium.  It is quite evident to me that it has the darkened leaf-tips of plants that I know very well from Lemoenpoort near Barrydale, across to the area between Riversdale and Ladismith, which Dekenah patrolled so thoroughly.  These glabrous elements generated the type material for the original concept of H. venteri and the entire ramification that culminated in my decision (among others) to regard the name integra as ambiguous and to exclude it as insufficiently known.  Were I to reconsider this decision in the light of this new type, I would concede that it relates to H. arachnoidea var nigricans but as nothing more than another problem variant.   Here again, fussing over this minor problem when it is difficult enough to separate the major components in the puzzle is just incongruous.  I see no need to refute my variety rycroftiana or to recognize Esterhuizen’s standeri.  I make no pretensions that my system has any greater virtue than is possible in the face of the real nature of field variability and my capacity to sustain some kind of consistency across the genus as a whole.  The name H. mucronata var rooibergensis is similarly upheld in the list when I know very well that in this interplay of the two species aforementioned it is not possible to uphold any such small distinctions – what to say of the problem extended eastward to the “species” there.

One cannot pass this issue by without back reference to the uses of names.   There is no doubt that the way in which Breuer would use the name aristata would reflect the confusion generated by Col Scott when he had several species names such as mucronata, aristata, mclarenii and unicolor in different sections for the same one complex in the vicinity of Barrydale.  I was inspired to think that the name aristata could more effectively be used for a dubious group of populations in the H. cooperi complex of the eastern areas.  The list is in fact unsatisfactory because it excludes the changes that I was constrained to make when I wrote Haworthia Update in 2001 to describe and explain the results of my further exploration.  These authors have been confused over elementary things like the zygomorphic structure of the flowers in the minor genera of the family, the spiral arrangement of leaves in H viscosa, or the fact that the illustration of rossouwii cannot be equated with a depauperate specimen of serrata, and I have my doubts that they are able to apply the names they have listed, with any greater sensibility.  The last example is replicated many times in Breuer’s books where there is incongruence between type material and added images of other plants that are taken to represent the same entity.

What is now left is the really curious array of one-off names that are paraded as “species”.  They are in total conflict with my entire conceptual view of Haworthia and it is quite senseless to include my name anywhere in the list other than with the words “sensu Breuer et al” appended if this is the paradigm for which the list is required.  I use the name geraldii as the prime example.  If classification is the expression of a purely personal opinion, and many leading writers amazingly tout this nonsensical point of view, then of course one can generate new names in any way one chooses.  The fact is that I generated a system of names with a view to explaining a complex system that includes ALL haworthias and relates to my experience with many other genera, both plant and animal.  My system was intended to serve science and to efface my own predilections and opinions as far as it is possible to do so.  Geraldii is a name that owes its origin to Col. Scott and it relates to one single population where it is not even clear that the plants are non-vegetative in origin.  Similarly the odd agnis refers to a single population of the greater complex nortieri – a local ecotype.  All the names generated by Hayashi fall into this same category of names generated purely by personal attraction to a particular variant and for self edification (viz. H. hayashi Hayashi), which can actually quite easily be accommodated in a descriptive statement of a far more predictive system.  The hidden part of the floating iceberg, which this checklist is, is a barrage of new names being generated by Breuer, among others, before the print has cooled.  This kind of classification has eventually to sink from its own weight and will do so when any real and honest attempt is made to apply it to the great range of ecotypic variants and continuities that are such an obstacle to our sensibilities.  My few loyal friends suggest to me that I should just ignore these nomenclatural and descriptive incongruities and leave it to history in the belief that “truth will out”.  I appreciate the sense in this remark, but the reality is that I am in the present and would prefer to experience commonality of sense now too.  I plead that my system should stand in its entirety with no changes until these can be soundly supported by good technology, good statistics, and good appreciation and observation of the actual natural system.  A last requirement is that botanical classification finds a way through proper definition, purpose and communal agreement to arrive at a standard of names for all living systems.