Published in Haworthiad 13:119 (1999). 

At last a reasonably rational review which is written with some insight and understanding.  It makes points which lead to communication and hence discussion. The then Editor of Haworthiad in relaying his feelings and that of his readers has criticised me many times for what is said to be my intolerance of the views of others. I want to set this record straight. I am intolerant of anything which stultifies and paralyses the understanding of Haworthia and distorts communication about it. I had written to Paul Forster before I heard he was preparing this review, and then I was thrown into a bit of a quandary. I do have enormous respect for people and it is not my wish to hurt anyone’s feelings. Paul is no doubt at all an excellent botanist and a human being of a high order. Unfortunately, I do not think his review is particularly good as it is couched in the paradigm of the orthodox and the conventional. I hoped for more, as I sense the complacence and almost smug security of the professional herbarium botanist who’s highly descriptive revisions will be held in awe by, and never ever tested in the crucible of, popular interest. There are many very positive things in the review, and in concentrating on the negative I really feel that progress is possible.

Firstly, I stated that herbarium space was a problem. It IS a problem. Whether Paul thought there was space or not is irrelevant. For many years random additions to the herbaria have been frowned on. Also for many years there was resistance to discarding the meaningless collections that occupied so much space in the cupboards. Paul should know how a herbarium is arranged, and that an addition to the full cupboard of a fern genus has a ripple affect all the way to the end of the collection. Space in the fern cupboard does not mean space in the Haworthia cupboard. The Bolus herbarium will only accept material regarded as necessary for bona fide research and the same is true for Compton Herbarium (NBG), where researchers are genuinely concerned about ‘their’ own space. Three of the only collectors I know (and who were persuaded and encouraged by me) to deposit specimens at all are Howard Gie, Emil Heunis and J.D. Venter. They were discouraged from doing so because of the real pressure for space that existed at the Compton and Bolus Herbaria – and still does. Gone are the days when collectors were welcomed with open arms and seen as a source of valuable new material. It is only in 1996, that space for Haworthia was created in the new herbarium building at Kirstenbosch. This is not because of a spacious new building which squeezed Haworthia into possibly less space than it had before, but because the curators allowed the transfer of about four boxes of G.G. Smith duplicates to Pretoria. I hate to think what problems these have created for Pretoria when I reflect on the lack of space I saw there in 1996, and available in terms of the total arrangement of the herbarium. I contributed many specimens to NBG and I share their concern that it is just not possible to provide the space necessary to reflect the diversity of the flora of the Cape. I must have driven them to distraction with my collections of Oxalis and Asparagus, which I think are packed in boxes in some remote and isolated room somewhere in the user unfriendly building. It is also not just a question of space.  The management of new material from the collector is a mammoth task. Specimens must be accessioned, treated, mounted and labelled. Those I submitted more than two years ago, are still not available for me to see again!

Why does Paul cite DMC3625 Kubusie Drift, near Kingwilliamstown if there is no specimen?  Because it is in fact meaningless and he intends this as an abject lesson to collectors who want a classification, but not the trouble required to support one. He later states that if there are no specimens (of intergrades) then they do not exist from a scientific data point of view. I would say, please leave out the red‑herring word ‘scientific’. It is beginning to convey to me the ultimate in insensibility. There are these intergrades and I really puzzled about how to cite them. I even described taxa to accommodate them e.g. H. decipiens var. minor, and H. gracilis var. viridis. Paul also writes of intergrades which he knows exists because he has seen them – but he also has made no specimens? His casual statement is thus good enough while mine is not? He does have the grace to say that new specimens have further blurred the distinctions between previously recognised taxa but he does not say which ones. Did he make specimens to support this statement? Paul does miss an important criticism ‑ the incorrect citation of specimens e.g. under H. cooperi var. pilifera, where Brigadoon is cited as Grahamstown and also King Williams Town; Peninsula is cited as Grahamstown, when it should have been King Williams Town. These are important errors because knowing this geography would perhaps have allowed one to connect the Kubusie Drift collection with other collections from nearby sites. Thus DMC3625 (if it existed) could in terms of my revision be only H. cooperi var. pilifera or H. cymbiformis var. setulifera. Also I find that I have several times erroneously cited the same specimens under different names. This arises from duplicates seen at different times and place, and ones which confused me often because of their intermediacy. However, this is what I hope the next revision will sort out ‑ thorough field work and assignment of each collection, in a more considered way, to each taxon. Part of my own frustration is that persons are fiddling about with nomenclature and guesswork when it is things like this that need attention and questioning. A herbarium specimen does not give the required information and ‘feel’ and sometimes a single dried specimen, and my memory, is all I had. Let me also say in mitigation – I have never been employed to do work on Haworthia and have never, in employ, been encouraged to do so. I do not have full-time access to the herbarium, and have never had. This is a serious problem when the herbarium record is the core of revisionary work and I might have made fewer mistakes had it been easier for me. I would also make fewer mistakes if half my mind was not held back so by the negative history of the literature on Haworthia.

There is also a serious and common problem with DMC numbers. This particular collector is well known to me and we have exchanged views on collection done without a permit, in numbers which no permit would concede, in a way which has no scientific principles at its base, and which does not ensure a herbarium record. 3600 meaningless numbers! It just is a fact of life that collectors do not make specimens and it plagues me more than it does him. In the normal course one should not expect them to do so, but for one he gets out into the field in the way that DMC does, it is another matter.

A bigger problem with herbarium material is that any haworthiophile values the plants he/she gets. To reduce them to squashed dried specimens is painful in the extreme. Apart from the pain of separation, it requires really tedious effort, more to make the specimens recognisable, and then the equipment and space to do so. Permit requirements, and a conservation ethic, severely limit what one can collect (my permit allows three offsets, and my ethic none!). As the plants require some observation, these are generally taken into cultivation with attempts to propagate and distribute material as well as produce a specimen from the product. The dried specimens seldom display the variation of the parent population and one usually finds a vegetatively propagated clone on the sheet. Miss Winsome Barker (past Curator of the Compton Herbarium) encouraged the use of colour photography to add to the usefulness of herbarium sheets. If there is any strength in my classification at all, it is because of what is in the herbarium like this. It disturbs me that so much of what I have seen, is not adequately documented in this way. Most of the herbarium records are of single specimens. While my permit allows three, I generally like to see at least six grown in cultivation to get an idea of what any population may be like.  I have grown plants from field collected seed, and to record some of this variability in terms of essential herbarium record is a monumental job – an impossibility. But I have tried to do it and am still trying. I cannot add an apology for the fact that I like the plants living rather than dead.

Paul says I have all this ‘information’ about vegetation and habitat ‑ indeed I do. So much so, that I cannot fool myself into thinking that one can make specific statements about habitat preferences. His comment makes me think back to the comment of Gordon Rowley’s about the soil pH data given in Col. Scott’s book. Was this pH data really any good at all, and if it was, did it have any practical value whatsoever?  (The story of the mima-like mounds of the winter rainfall Cape would be a useful reference in this respect.)  It should be clear to the expert who knows something of the South African flora and geography that no statement he can make will be clear to the novice, and that quite probably it will not be clear to any other expert either. I am still wrestling with this problem fantasy that vegetation can be classified with the same facility that discrete species can. I believe this is one of the great misconceptions and myths of botany, which like so many things in science, is never as simple as it seems. Paul is said to be a cycad researcher. I doubt if he could explain the distribution and habitats of the E. Cape cycads in terms of the recorded and described vegetation of the area, or in terms of their classification. This is a considerably easier task than it will be for Haworthia. Paul should know that Haworthia is just one of many genera which are virtually unresolvable in terms of classic taxonomy, or describable by vegetation and habitat. I have named some before and I can add to the list with Hermannia, Rhus, Tetragonia, Lycium, Pteronia, Zygophyllum as examples I am familiar with. Vegetation study will be more meaningful when we have good classifications for these genera. I have recently seen a Haworthia change facies over a distance of 50 meters, because of local habitat differences. The only differences I can record is the fact that one plant form was on exposed vertical slabs of rock, and the other in sloping terrain with the same rock substrate, but with more soil and grass. One would need to be a geologist, geomorphologist, and soil scientist, in addition to be able to identify and quantify the associated plant species to do what Paul idealises. Perhaps this is the prerogative of a reviewer. Nothing is impossible and it should have been done.

Paul makes the excellent point that the only person who can say if this new classification is any good is I. So I say “Thank you, I actually think, and know, that it is very good indeed”! This should be true for the author of any revision. I am sorry to say that in my experience, many revisions which have been accepted and published by “scientific” journals (or as academic theses) have proved to be pretty useless. I should add that this is not only for new material not seen by the authors, and despite meeting all the requirements Paul has for my ‘synopsis’. He goes on to say that “While it may be easy (probably not) for Bruce to attempt to fit new and morphologically distinctive collections into his system, it is well nigh impossible for anyone else for the lack of data presented”. The unnecessary words ‘probably not’ points to the wrong place. The subgenera in Haworthia are morphologically distinctive and that is practically where it stops. My complaints have been directed at the fact that persons unlimited are creating new names and fiddling with old ones because they are so unaware of this problem. My ‘data’ includes geographic information and no one should even attempt to fit a grossly ‘new and morphologically distinctive’ collection into my system without familiarity with that system. Nobody without knowledge of the herbarium collections can know what is new and distinctive. If they do not know this they should not be fiddling with Haworthia. The problem is not caused by my presentation and its shortcomings – it IS a fact of life for many more genera than just Haworthia. Writers are playing with MY classification, that have not seen the material in the herbarium and neither do they have the knowledge of those specimens in the field. They may have seen material that I have not seen, but still the responsibility is on them to first know what forms the body of an existing classification and its predictive ness, before embarking on a new mission to repeat the exercise.

Paul skates about a bit with the questions of descriptions and measurements. He complains that dimensions (‘size dimensions’!) are not provided. In an age when people do not read, and look at pictures instead, I decided that this kind of data could as well be deduced – at the level that it is practical and needed – from the many illustrations. Of course a lot of importance is attached to these details Paul calls for, because they give the impression of quantification and hence ‘materialistic ‘science’. Intellectuals have to have something to keep them busy or create the pretensions of business. It would have made the book a lot thicker (and a lot more expensive). I also know very well that with all that detail my credibility in the academic environment would be assured. Of course living organisms are measurable and there is also a statistical science (biometry) that goes with it. This science depends on randomness and replication, deals in mostly normal data and probabilities. It is my knowledge of this science which supports my understanding of Haworthia, and part of the reason why my work is sound. Nobody is going to get anywhere in Haworthia neither with three specimens, nor very far with the specific mensuration that Paul calls for. I do know the group. Contrary to Paul’s statement (sorry #, Paul), it is because I know the group (and many others) that I also know that this variability cannot be easily rationalised and summarised for meaningful usage. Haworthia is not like, say, Hoya where a new species (see Marsdenia as a new genus) can be recognised on the basis of a single flowering specimen. In a collection, often the decision on whether one or two clones are involved is to look and see if their flowers are different! The variability of the plants is what it is, and I do not understand why readers seem to think that there must be some character state lurking somewhere that is going to provide an ultimate solution. It is all well and good saying, as Paul does, that every time he plants seed he is struck by the differences. I challenge him to document this frivolous observation and to explain to me what exactly has struck him and what it’s more general value is. This approach of his is the same as the weaknesses he rightly perceives in my revision. How do we get out of the condition? In biometry (the statistics of biological systems?) we were taught that you do not attempt to seek significance when it is obvious what the variability in the sample tells your common sense. It is a laughing issue in agricultural science that researchers seek significance in statistical test after statistical test until, on the basis of the very probability they are testing, they find the significance they seek. Botany has apparently not reached this stage yet.

What I have said about voluminous and detailed descriptions is that they are in practise almost useless. I can say confidently that Paul’s comments about the woeful and incomparable descriptions (I think they are masterful in their brevity, contrary to Paul’s use of the word!) is a true statement about Haworthia. It cannot be wished away or blamed on me. I would really like to give him just the currently deposited collections of what appear to be two species viz. H. cymbiformis and H. cooperi. He must then write a description (aided by as much living material as he wants) which allows the reader to identify them with any degree of certainty at all, and in so doing also exclude elements such as H. gracilis, H. bolusii and H. aristata. If his then comprehensive description adds anything to the synopses of my revision, I will indeed be surprised. I know what goes on in taxonomic accounts which Paul so reverently cites, and I am not impressed by the vast amount of detail ‑ look how it has been used in cladistic studies in the Aloeaceae at genus level! If Paul knew Haworthia well enough to write this review, he would have also been able to say that Esterhuizen and Battista’s taxon (H. mucronata var. rooibergensis) bore no relation to reality at all. Of what value to Paul (and Col. Scott’s book would have been adequate for this particular exercise) was the detail these authors supplied? How could he possibly have used it if I had supplied the statistically true comparable values for just H. mucronata alone? I had already written a response to Esterhuizen and Battista which will be published elsewhere, and this response does deal again with this issue of detail. I must say that I am familiar enough with Halda’s work to suggest a blacklisting by the International Association of Plant Taxonomists. We could both be there together!

Yes, I am aware that I have not given, say, the difference in flowering time between bayeri and emelyae. It is only about six weeks and I think the actual flowering dates vary from east to west. However, it is a lot more complicated than that and I do not know what I can truthfully say. G.G. Smith recorded flowering times of many collections for several years and these varied by as much as 5 months. In H. magnifica var. splendens, the west of Albertinia populations flower in November, those east, in September and perhaps again in November. In the field I have observed a population of H. cooperi var. gordoniana (which has affinities with cymbiformis var. transiens! – Jeffries Bay MBB6792) which in one year flowered six weeks later than in the preceding year. Flowering time is as erratic and variable as the populations are. A good record should also be bounded by statistical limits. I separated maraisii and mirabilis on flowering time (Apr./May vs Feb./Mar), but it does not always work. The McGregor complex may flower from Nov. to May. In cultivation there are nearly always odd plants flowering out of synchrony and my field work has not been that of the professional who can set goals to observe phenological events. Yes, I did indeed do a lot of work on the flowers, and I cannot use it because I could find so little pattern. Like in the asclepiads, differences in the flowers within populations often exceeded the differences between populations. Again, underlying Paul’s complaint is the fact that there is no Holy Grail that is going to unlock the conundrum of Haworthia.

No, the book without pictures is not user friendly. It was not expected to be and I said as much. If the publisher could have accommodated another 450 (is it 450 or more?) photographs and the book remain affordable, I am sure they would have done so. The question of intergrades is also thin‑ice stuff. Does the nomenclatural system allow these things? In my collection book I quite happily use the chemistry equilibrium equation to indicate how I would classify a collection and many of the herbarium collections are so indicated. The object of my classification was to try and identify divisions rather than obscure them, which is what most classifications do. My next exercise (which I wish was a competent somebody else’s) is to show that intergrades are not just two-way, but perhaps three- or even four-way.

As to H. pumila, many (may I repeat, many) professional plant taxonomists have entered the debate and indeed there is no vacuum here. It is a space filled with vacillation and doubt, conflicting opinion and retroactive wisdom. Nomenclature is not my strong point, and perhaps for the very reason that any good practicable workable solution can be contested and upturned (and will be so contested and upturned) for no other reason that there is a juristic element which caters for it. That the International Association of Plant Taxonomists has not resolved the particular issue of H. pumila is their problem. I find it extraordinary that Paul does not see the irony of the situation. IF I have made a mistake, the following is what I will concede – the author citation of H. pumila should read (L.)Scott, and not (Duv.)Scott or (L.)Bayer. I deliberately took the onus on myself for any flack that my continued use of the name would draw, but I am very happy to give the credit back to a gentleman and a friend. This name pumila is the first epithet for Haworthia in the Linnaean system and if the ICBN does not allow its use I will happily continue to condemn that august body, all its pettifogging recommendations, and the Committee that empowers it or is empowered by it. Paul is just another of a very long list of professional taxonomists who can blithely say “I am not sure what the outcome should be at this point in time”. When will this momentous decision be taken and by whom?  He cannot thus be any more comfortable with nomenclature than I am, given that he is a professional. Paul is only going to be sure what the correct solution is when he is so informed by the high priests of nomenclature. So much for science. The fact that a committee has to resolve these things is, in my opinion, an indictment of the entire process. It is not just any committee of qualified taxonomists he can do this wondrous work as I painfully learned from a group discussion with four Doctored intellectuals. As for the “serious” oversight concerning the rule (Prof. Cronquist asked if nomenclaturists knew what they were doing, and I think not) of rank priority, forget it. I do not like the rule (or much else about the ICBN) and I do not see the necessity for its inflexibility. There seem to be more instability nurtured by the code than precluded by it. I prefer the names I have used and I also think they are more meaningful. Let the bandits do what they want to. It is perfectly clear what my names were derived from and what they mean. I think carping about the “rules” is just a nit-picking exercise derived from a failed historical attempt to rigidly control nomenclature. It does, and will continue to cause, more problems than it solves. As for “bandits”?  Who stole the familiar name Astroloba muricata away from Groen and Roberts-Reinecke and substituted a new name corrugata under the guise of the strongest words associated with the Code i.e. to avoid confusion? This is what I refer to as trivial botany and I hope the IAPT will mint a medallion for this kind of thing.

A fact is that there is not one of my names (or Scott’s) for which there is any doubt as to origin, usage, or typification – except that it was not done in the style and letter-correct formality of the ICBN. The fact that Breuer and Metzing could do what they did (and add to this, all the words that the former author generated for the non-existent enigma of H. arachnoidea) is further adverse commentary on the ICBN.

Paul does dangle the predictable carrot…”It can be expected that with further exploration…new data… may radically change classification”. What does he think happened between 1982 and 1996 when the book was drafted?  What is more important is what has happened since 1996? I have been almost horrified by the threatened return to the taxonomy of the early 1940’s, and in this respect Paul was kind enough to mention “the changing of a few names” by Breuer. The dearth of any signs of intelligent life in new ventures into Haworthia compelled me back into the arena by default. I have done some very thorough recent exploration and can say that my classification is even better than I had hoped it would be. Some very interesting new collections (and to date there are well over 400 new numbers in my accession book) can all be explained in terms of my classification if they can be explained at all. This is why I know my classification is so good. I hope to present the information somehow, in someway, as supplementary publications and the first has just appeared in Aloe 36:34 (1999).

There is a reference to refereeing of the book prior to publication. This is not fair to Peter Bruyns and I unfortunately cannot give a full explanation why I think this is so. We have worked together for more than 25 years. While his contribution was useful, this usefulness was not very forthcoming and it is not a door I would beg at. His big contribution to the debate about Haworthia is his forthright comment “Lump them!” It may well be the sensible thing to do and I have many times considered it seriously. Regrettably the opinion is based on an extraordinary ignorance of Haworthia, despite our long association.

Finally let me say this. For the full eighteen years that I was Head of the Karoo Botanic Garden (before, and even after) I sought the taxonomist Messiah that would lead us out of the unhappiness or confusion of Haworthia. Nobody has ever even nibbled at the bait.  Botanists just do not want to mess with “difficult” groups, and I have seen them really mess with simpler problems. See what Prof. E.A.C.L.E. Schelpe said and did about Gasteria, or about publishing a really marvellous work on Astroloba by Pandora Roberts. If I have presented a revision of Haworthia, it is simply because there has not been anybody more competent, and this may be an unpleasant truth. I have never wished to say “Trust me, I am the expert”. I consider it most unfortunate that I am, and I would rather be doing something else. After reading Paul’s review I can confidently say that my classification is a lot better than Paul is guessing. As I digest and work over these latest collections of mine my confidence grows. If good sense prevails there will be very little in the way of name change or circumscription and we will all be able to enjoy the plants a lot more. If somebody competent does appear who wishes to build on what I have done and take over where I want to leave off, they will have my unstinted blessing, assistance and gratitude. ♦