| Subject | : | What does it take to define an evolutionary path? |
| Author | : | Steve Price |
| Date | : | 09-03-2000 on 10:18 a.m. |
| sprice@hsc.vcu.edu Dear People, Daniel's Salon is basically directed toward trying to clarify the evolution - that is, the details of the historical development - of certain rug motifs. The approach he uses is to attempt to construct a classification - a taxonomy - for a particular subgroup of rugs. In another thread that has become almost unmanageable because of the number of side issues introduced and the general tone of some of the remarks, John Howe criticized the approach. He asserted that we can't work out the evolution without knowing the genetics (I assume that what he means by "genetics" is the mechanisms by which designs get transmitted from one person or culture to another). I pointed out that in biology and in linguistics, evolutionary details were worked out well enough to be persuasive to most (albeit not all) educated people in the complete absence of genetic information, while taxonomy provided the framework in each case. That is, when languages were arranged in a diagram based on similarities, a plausible historical underpinning could be posited to explain it. Likewise for biology. Those who don't believe that evolution occurred in biology posit that every species was created independently. At least in Rugdom, I don't think anybody doubts that some kind of evolution occurred. What, then, prevents us from doing the same sort of things with rugs? That is, why can't we generate a historical progression on the basis of the way we classify them? The answer is, not every classification system reflects the history. Let me illustrate with a subject about which I actually know something, biology. On the basis of certain criteria that aren't really relevant to us right now, we classify vertebrates (animals with backbones) into five groups: mammals, birds, reptiles, amphibians and fishes. Assuming that the most widely held notions about evolution are correct, these five groups are fairly large branches on the same limb of a tree-like diagram that represents the genealogy. That is, the taxonomy and the evolution are closely tied. We need to recognize that this classification is not the only one possible. It depends on the selection of appropriate criteria for classifying. For example, we could set up a classification system in which all aquatic organisms were in the same group, all desert organisms were in the same group, etc. That would give us a different taxonomy, and, most biologists believe, would not reflect the genealogical tree. By now you're probably wondering what this has to do with anything relevant to rugs. The answer is, that it suggests that if we are to use taxonomy to construct a genealogy, we will probably have to explore a lot of different taxonomic schemes and criteria before we get to the right one(s). The value of an exercise like the one Daniel has introduced here is that it is just such an exploration. Most of us believe that structural characteristics will be important to the system, and I think that is correct. Design similarities, especially in designs that are not trivially simple, are likely to also hold important genealogical information. Regards, Steve Price |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | Michael Wendorf |
| Date | : | 09-03-2000 on 11:18 a.m. |
| Dear All: If I follow Steve, an example of one taxonomy would be to group all the white ground rugs with memling guls into a single classification. However, as has been discussed, to do so would group rugs with different structures together. An alternative is to group the rugs together as we have done and then try to classify by other criteria such as whether the rugs have two ply or three ply warps. Once we understand how the members of the group relate and differ we can begin to try and relate them to weaving groups and origins. -Michael |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | Christoph+Huber |
| Date | : | 09-03-2000 on 12:28 p.m. |
| huber-ch@pilatusnet.ch Dear Steve While speaking of different taxonomic schemes you already have dealt with much of my uneasiness concerning genetics and taxonomy in the previous posts on the other thread. But "genetic" still sounds too linear to me to be the only way of thinking. We could for example try to speak about ornaments in the language of "behavioural science" which hasn't necessarily to do with genetic development. The "technical behaviour" seems to be Marla's main focus and/but there are many more aspects of "ornament behaviour", for example the horizontal vertical differentiation. Every "language" we chose has its own advantages and disadvantages and thus shapes our thoughts. It's therefore necessary to change the way of thinking from time to time to escape the inherent limitations of a single frame. From my point of view a rigid "genetic" way of looking at ornaments misses the point, that they aren't bound to the same kind of laws as biological beings. Look for example at the curled leaf and the ashik: In many carpets we can decide whether we see the former or the latter only out of the context (if we can do it at all): they merged in a certain sense. But despite this similarity I would still suggest two very different origins for the two ornaments. (This might be wrong, but nevertheless we can't neglect the possibility of mutual influence of ornaments from rather different branches of the "genealogical tree of motifs".) For the ethnic attribution of ornaments I see great problems. Virtually all carpets we have to compare were made after the arrival of the Turks in Western Asia. So it's very hard to decide which roots of ornaments are Turkish and which are Persian. East Asian influence during Timurid and Safavid times makes the whole thing even more difficult, especially as some groups of ornaments seem to have re-entered Persia after they first travelled to China from the West. So, I don't think that we can be really sure whether an ornament should be called Turkish, Persian or something else. And if we could agree on this point it would change the approach to the question of the attribution of ornaments to much smaller ethnic groups. The importance of the comparison of rug ornaments with Sassanian, Coptic, Bactrian Bronze Age (...)materials is in my eyes twofold: First it proposes lines of development which may can later be validated by supporting arguments and second (even more important!) it prevents us from to early labelling ornaments as "Turkish", "Persian", "Luristani", "Chinese" or whatever: it gives us the freedom we need to look at them from different points of view. Regards, Christoph |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | Bob Emry |
| Date | : | 09-04-2000 on 01:37 a.m. |
| emry@starpower.net Hello All: I see some real progress toward understanding the process of classifying, but too much reliance on the language of organic evolution still leaves the potential for confusion. Organic evolution (Darwin's idea) is based on the principle that traits (characteristics, features, whatever) are heritable-that they are transmitted from parents to offspring. It isn't necessary to understand the underlying genetic mechanism to understand how organic evolution works-it is sufficient to know that somehow through the process of reproduction, traits are passed from one generation to the next. The evolution of design in rugs must be limited to this aspect-inherited traits. A design element copied or modified from one in a preexisting rug is inherited, if you will. If a weaver makes a rug by copying another rug, then that rug has "inherited" its characteristics from a predecessor. However, the mechanism here is not reproduction, but the choice of the weaver, and this is where the analogy to organic evolution breaks down. For example, instead of simply copying a rug, the weaver might have chosen to make a rug that combines design elements from many different rugs, a process that has no analogy in organic evolution. One principle of biological taxonomy that would seem to be equally appropriate in classifying rugs is that the greater the degree of similarity the closer the relationship. I don't see how it will be possible to make one synthetic classification of rugs, using all characteristics, because groupings based on design similarity will often conflict with structural groups. Nevertheless, if we classify, we have to go as far as we can by grouping according to similarity. If the question is identifying ethnic provenance, then structural characteristics will probably be weighted more heavily. If the objective is interpreting evolution of design, then grouping by design will be important. As an example of design "relatedness" we can look at the four "Shahsavan" rugs that were illustrated in another thread: The Pap rug, the Thompson rug, the Rudnick fragment, and the rug I posted. These four rugs have such a high degree of similarity that they must be closely related (at least in design)-they not only share all of the same design elements, but the elements are all arranged to form the same overall design composition. With this complexity, it is simply impossible to imagine that these rugs could have been created independently by different weavers all selecting design elements randomly and arranging them randomly. There is "genealogical" information here-the overall design composition was "inherited" from a preexisting rug. I don't see how we can conclude otherwise, even if the structure might suggest different ethnic provenance. When I first read Wendel's "Coptic Connection" posting I intended to respond, but before I could do so, Marla's response made some of the same points I had intended to make. Since one of my points involved the "degree of similarity" principle, I'll make it here. Like some others who have replied, I see the Coptic "rosette" as depicting three dimensions, like an endless knot, and I perceive the other rosettes (Shahsavan, Beluch, Talish, etc) as being two dimensional, assembled from other common smaller elements, with no interlacing. I don't don't see any similarity here that I would consider significant in terms of design evolution. Similarly, with the Coptic medallion that Wendel showed next to the "Shahsavan" medallions-the Coptic medallion is octagonal overall, and the outline of its inner medallion is a stepped polygon with gabled top and bottom, but I see no other similarity in the details. To me these shapes are so general and ubiquitous that I'm doubtful that the similarity is significant. I can't disprove Wendel's hypothesis; I just don't find it very convincing without more evidence, and that must come from finding a transformation series showing how the Coptic medallion evolved to the 'Shahsavan." Best wishes, Bob Emry |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | Steve Price |
| Date | : | 09-04-2000 on 08:20 a.m. |
| sprice@hsc.vcu.edu Dear Bob, I must not have made myself clear. I'll take another crack at it. 1. My point about the "genetics" is exactly the same as yours - all we need to know is that designs could be transmitted somehow. The details of how it happens are not at all important. I mentioned this because John Howe's assertion was that trying to determine an evolutionary pathway is impossible without knowing the genetics, a position with which I disagree because in linguistics and in biology the evolution was reasonably well understood in the complete absence of understanding of method of transmission. 2. I didn't mean to imply that all the rules of evolution in biology (or linguistics) would apply to rug designs, only that to infer evolutionary pathways from taxonomy requires finding the approximately correct taxonomy or, more properly, the correct criteria for classification. I don't think it will be trivial to do this, and maybe it can't be done at all. On the other hand, if you don't go fishing you never catch fish. Regards, Steve Price |
| Subject | : | I goofed - correction |
| Author | : | Steve Price |
| Date | : | 09-05-2000 on 11:22 a.m. |
| sprice@hsc.vcu.edu Dear People, In denying that we need to know the "genetics" in order to explore possible evolutionary lines, my preceding message includes the phrase, "...all we need to know is that designs could be transmitted.." That has two possible meanings, one of which is ridiculous. It is more correct and unambiguous to say, "...all we need to know is that mechanisms for transmitting designs existed..." Sorry about that. Steve Price |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | Bob+Emry |
| Date | : | 09-05-2000 on 10:27 p.m. |
| emry@starpower.net Dear Steve, In your next-to-last posting, you say "I must not have made myself clear..." To the contrary; your discussion was very lucid, and agreed with what you wrote. I just reiterated (paraphrased) part of of it, so I could go on to make some other points: (1) that it's dangerous to push the organic evolution analogy too far; (2) that even though we don't have to know the "genetic" mechanism of design "inheritance" to understand design evolution, it seems that we can actually identify the mechanism as "weaver's choice." If a weaver copies designs from a preexisting textile, or selects from an established design repertoire, then the design is transmitted to another generation and becomes another step in the design evolution; if the weaver creates design elements de novo, it is not; (3) that the greater the degree of design similarity, the closer the "relatedness." (4) that is seems unlikely that a classification of rugs based on design will be congruent with one based on structure. Etcetera. Bob Emry |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | Steve Price |
| Date | : | 09-06-2000 on 08:44 a.m. |
| sprice@hsc.vcu.edu Dear Bob, I think we see things through fairly similar spectacles. Vis-a-vis structural versus design-based taxonomic systems, I suspect that both will be applicable at certain levels, and that recognition of this is implicit in much of what we actually do. Most people would agree that structure is more likely to be conservative than pattern, although we would have to define which elements of structure we're talking about to reach that agreement. If we're shooting for a taxonomy that reflects genealogy, for instance, structural matters like flatweave versus pile are likely to be relevant very early on, while things like number of plies in the warp are likely to be relevant in sorting out the relationship between more recent evolutionary events. Likewise for pattern. For instance, we probably agree that a textile with a field consisting of rows and columns of a gul, or alternating rows and columns of two guls, is almost surely Turkmen. Furthermore, since this pattern occurs frequently in the work of just about every Turkmen subgroup (the only exception I can think of is the Beshir group), we can reasonably infer that it predates the separation of the major tribes. That is, we can use this simple criterion as a tool for classification, and it works pretty well (I think) as a genealogical marker. Another might be, say, the kepse gul. For all practical purposes it occurs only in Yomud and late Ersari stuff, so we might reasonably infer that it was a Yomud invention that occurred after they separated from the Turkmen main groups, and that it migrated later to the Ersari. Again, reasonable inferences (with uncertainty built in, of course, but that's true of anything) based entirely on design. The trick to ultimate success, in my opinion, will be in the selection of appropriate criteria - structural, design, color, other things - for classification at the various taxonomic levels. Not likely to be a trivial undertaking, but little spots of success do suggest that it's worth pursuing. Steve Price |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | R. John Howe |
| Date | : | 09-07-2000 on 07:08 a.m. |
| Dear folks - It may well be as Steve has suggested that the "standard" (although I did not articulate it clearly enough for it to function as that) I have suggested is too tightly drawn and that it may be possible to delineate a less stringent but still useful set of requirements for saying what my philosopher friends might describe as some "meaningful" things about how designs progress and have progressed. (I can see now that I need to be very careful about importing such usages into our conversations here. Wittgenstein frequently makes reference to "non-sense" in a way that is entirely non-perjorative and refers only to the fact that the argument being made does not depend in any way on sensory data. Such is my use here of the term "meaningful.") I also think that Christoph Huber's wise suggestion that we utilize multiple perspectives in our analysis from time to time since the phenomena we are examining are complex and no single perspective is likely to permit us to look at them "full-faced," is both sophisicated and sound. I do think that there is also potential for confusion in such shifts in perspective, especially if they are not well-marked. The richness of the argument can be such that it overwhelms our ability to detect what is being "said." But I agree that we should not in our concern for precision, impose onto the likely richness of what we are attempting to examine, a set of standards that are too narrow and mechanical. I think I also sometimes mistake some of the objectives those who enjoy the examination of rug designs and patterns might be pursuing. I ofen project onto their discussions the notion that they are attempting to "demonstrate" something about the development of designs. It may well be that they are often, in part and importantly, merely savoring some aspect they have noticed and find interesting in and of itself, much as one might roll an agreeable wine about on one's tongue savoring its taste and complexity. I might not indulge much in this latter activity (although there are aspects of aesthetics applied to rugs that interest me in similar ways) but I do not denigrate it. I am interested in being able to tell when that is primarily what is intended. I think we are now positioned (perhaps with some attendant pain that might have been avoided) to do something close to what Steve suggests here: to describe however tentatively amongst ourselves some requirements that we would recommend that a sound demonstration of design development should attempt to meet. It may be that the order of thing we can come to will be a great deal more humble than our aspirations for it but it seems to me worth doing given the very obvious attraction the examination of rug designs hold not for just the members of this group but for rug students across the entire literature. Regards, R. John Howe |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | Michael Wendorf |
| Date | : | 09-07-2000 on 02:16 p.m. |
| Dear John: Try as I might, I still have no idea what you continue to write about imposing order or why the vague standards or requirements you refer to are or would be meaningful to our discussion, understanding or appreciation of rugs. I also do not understand what relevance or meaning Wittgenstein has to our discussion or why you always drop his name. In any event, would not a better translation of Wittgenstein be "non-sensory" rather than "non-sense?" We are making observations, trading observations, sketching out theories and debating and speculating on a big and complex picture from the fragmentary record available to us. Beyond that we are beginning to understand the underlying process of classification outlined by Bob, Steve and Christoph that is otherwise sort of intuitive. There are probably too many blanks, too many holes in the raw information available to go far beyond that. I think most of us recognize that even as we try to go as far as we can while also having fun. It seems to me that the order, standards and requirements you crave only tend to inihibit the discussion and have nothing to do with seeing and understanding rugs. Best personal regards, Michael |
| Subject | : | RE:What does it take to define an evolutionary path? |
| Author | : | R. John Howe |
| Date | : | 09-07-2000 on 03:17 p.m. |
| Hi Michael - Thank you for the continuing frankness of your comment. I thought Steve's heading of this thread invited not just examples of such analysis but some thoughts on "what it might take to define an evolutionary path." But I can see that this last contribution too does not seem useful to you. More, you apparently see it as "off topic," so to speak. Perhaps you are right that I have explained my thoughts on this point sufficiently or perhaps insufficiently but too extensively and I will, at least for the moment, stop here. I am sorry that you doubt my intentions in this discussion. Wittgenstein is often mentioned when one talks about conceptual analysis because he provides some of the most arresting examples and is a widely recognized reference point in labelling such matters. I have not examined my motivation deeply but I do not think all mentions of his name in such a context are necessarily instances of "name dropping" and I am sorry to have given you reason for doubting the authenticity of my contributions here. Regards, R. John Howe |