Medieval Islamic Swords and Swordmaking
The main objective of this book is to render al-Kindi’s thesis “On Swords and their Kinds” into English. Translating a text that describes Muslim swordmaking techniques during medieval times is not an easy task. While there are several depictions of swords in medieval Muslim art and everyday objects (like coins and seals) and in literary works, there is precious little source material available on the technical aspects of weapons in general and swords in particular. There are several reasons for this. Among them is there is not much material written on the production of weaponry. Manufacturing techniques were passed down through the generations through apprenticeships and were closely guarded family secrets. As there were few techniques available to quantify and control key characteristics (such as high temperatures, cooling rates and concentrations of trace impurities), a great level of “feel” was required, and that came more often with experience than with formal study. The skilled workers making these weapons were by and large illiterate; hence written records would not be of much value to them. A lot of the written material is lost or destroyed as armed conflict (internal and external) swept across the Muslim lands.
Translating technical works poses an additional level of complexity. The scientific models we use today to understand our world are different from those used by medieval Muslim blacksmiths. Our challenge is to understand their explanations in terms of their models. Most technical texts assume a certain level of understanding, making our task more difficult as we do not have a clear understanding of the intricacies of their models. Sometimes colloquial terms are used as technical terms. One of the difficulties that arises is that technical terms generally have precise definitions that their colloquial counterpart does not capture. For example, the colloquial use of “drop” (as in “drop of water”) is not as precise as its technical equivalent in measuring dosage (where it is defined volumetrically as 1/76th of a teaspoon or roughly 65 microliters). If one is familiar with the colloquial usage of “drop”, one would have an idea of its technical usage. Other terms may not have such a close equivalent. For example “necking” in metallurgy refers to the reduction in cross sectional area of a member that is subjected to tensile loading. The same term has a very different meaning in colloquial usage.
Translating technical Arabic manuscripts poses its own unique challenges. The first is that diacritical marks are often omitted under the assumption that the reader is familiar with the language’s usage to deduce the intended meaning. While this would be an expedient convention for the original author’s contemporaries, it poses difficulties for later reader. In this case, the original manuscript is lost and we have later copies to work with. Later scribes may not have understood al-Kindi’s intensions and may have inadvertently introduced errors. Scribes are primarily interested in transmitting the author’s intended meaning and took liberties to alter the original text to conform to contemporary conventions. As the scribes may not have had a background in metalworking, they may not have been sensitive to the subtleties in the original work or the evolution in meaning of technical terms over time. Hence rendering a work like this on into English (or any language for that matter) is fraught with numerous challenges and is very likely to have its shortcomings.
Chapter 1 provides a brief introduction to al-Kindi’s life and his numerous works on a variety of topics including philosophy, mathematics, astronomy, music, and others. He also wrote on more “practical” topics like cooking, perfumes, jewels, dyes, apiculture, natural phenomena, and others. This chapter also discusses the manuscripts we have available on al-Kindi’s work on swords and swordmaking. Ibn al-Nadim wrote a biography on al-Kindi and mentions two works on swords. One is titled “the kinds of swords and iron” which is the focus of this book, and “that with which swords and iron are treated so that they are not broken or blunted”. The first is probably a thesis on production methods, while the second focuses on quenching techniques and is more chemical in nature and scope. As the original manuscripts are lost, Hoyland is forced to rely on later ones for this work. He mentions four such manuscripts (namely Leiden, Istanbul, Turin and the Chester Beatty Library in Dublin). While Leiden and Istanbul are close, and Hoyland correctly points out that they may well be copied from the same original, the other two manuscripts contain compositional and quenching sequences that could be from al-Kindi’s second thesis. Zaki rendered al-Kindi’s thesis into printed Arabic by trying to amalgamate both the Leiden and Istanbul texts and selecting readings that he though best represented the passage. Hoyland asserts that in the process Zaki produced a new text. Hoyland, on the other hand, chooses the Liden text as his basis for this translation. One may take issue with Hoyland’s assertion, however if the goal is to understand swordmaking during al-Kindi’s time, then Zaki’s and Hoyland’s contributions are valuable, and much additional work remains.
The next chapter is Hoyland’s version of al-Kindi’s text. On facing pages are Zaki’s printed Arabic version and Hoyland’s English interpretation of what he believes al-Kindi meant. It would not be fair to call Hoyland’s work a translation in the true sense of the word as it is not a translation of any of the manuscripts or of Zaki’ work. Hoyland’s rendition is his opinion of what he feels is as close as possible to al-Kindi’s original work. Whether one agrees with Hoyland’s interpretation is an open question, but he deserves the credit for the work he has done – especially for rendering a difficult text into English. For interested scholars, Hoyland provides Zaki’s Arabic text as well as reproductions of the four manuscripts so they may form their own opinions.
While Hoyland’s translation of Arabic is for the most part exemplary, technical terms require careful interpretation as these terms have different colloquial and technical counterparts in English. For example when discussing writing or putting images on Yemeni swords to hide their veins, Hoyland states in the footnotes “Though the sense is relatively clear here (cf. the previous note), the exact reading of these two words is uncertain. Zaki gives “it is ampleness and returns from the iron,” which is meaningless.” This could refer to the impurities (slag) from the casting process. Essentially, the blacksmith is trying to cover up areas where slag is visible and where the sword would be likely to split. In another area, when discussing Indian swords, Hoyland uses the term “clear.” In the footnote he adds “It is not clear how this word should be read; Zaki converts it into the indefinite relative particle ma, which is definitely wrong, but my “clear” (mubin) is only a very tentative suggestion.” When trying to understand the process, it seems that “clear” refers to the lighter grey (ferrite) nodules that are visible in crucible steel. In another area, Hoyland translates what could be interpreted as “forging” as “striking with sword.” When discussing Frankish swords, Hoyland translates “ancient” as “most well-formed.” The reader needs to realize that “ancient” here is a technical term and refers to the materials and manufacturing processes used to make the swords; hence arguably the term “ancient” would be more appropriate in its technical sense (as defined by al-Kindi) than “most well-formed.” There are other examples where metallurgical insight would have helped the reader.
Chapter 3 is Gilmore’s commentary on al-Kindi’s thesis. It is obvious that Gilmore has a through understanding of metal processing techniques. He brings his expertise to bear when discussing possible steps of producing swords, and the visual clues that betray their processing methods. An area which needed elaboration was exploring the impact of these processing techniques on the steel’s microstructure, and rendering it within the framework of contemporary metallurgy. While it is true that few swords from al-Kindi’s era have survived in modern times, that some tests may not be suitable for these specimens, and that corrosion and other similar phenomena would alter surface characteristics; having insight on what to look for would help additional work in this field. What is missing is a metallurgical based discussion that starts with the premise of what makes a good blade. To achieve this, blacksmiths have learned to work harden the steel and use chemistry and temperature to achieve their goal. They have discovered the necessity of forming austenite (so it dissolves carbon) and using the magic of quenching to destabilize austenite and form carbon-poor ferrite precipitate out carbon-rich cementite. The faster and further the steel is cooled, the more finely dispersed the cementite and the harder the steel.
Specifically, carbon content plays a significant role on the steel’s microstructure, and largely defines the characteristic patterning of the resulting blade. Small variations in the quantity of carbon would make the difference between hypoeutectoid (<0.8% C) and hypereutectoid (>0.8% C) steel. From Gilmore’s descriptions, it can be discerned that hypoeutectoid ingots are likely to produce ferrite-pearlite banding, which is dependent on impurities introduced by the iron or carbon sources. During solidification, pearlite would remain in the interdendritic region while ferrite would makeup most of the dendrite itself. Traces of manganese would promote banding and slower cooling rates would produce bigger bands. This technique is likely for producing Sham patterns. The Damascus pattern was probably produced with hypereutectoid steels and is due to the alignment of spherical cementite in the interdendritic zones. The cementite aligns because of the presence of impurity elements present in the interdendritic zone. Elements such as vanadium, molybdenum, chromium, niobium, and manganese promote alignment of cementite.
It is probable that cooling rate affects the appearance of the pattern. This is because the faster the ingot cools, the smaller the dendrites are, and therefore, the closer the interdendritic zones. The closer the interdendritic zones, the closer the aligned spheroidal cementites are, and therefore, the finer the final surface pattern. Therefore a blade forged from a slowly cooled ingot would have a coarser pattern than a blade from a quickly cooled ingot, assuming that the blades require a similar amount of forging. Forging does affect the final pattern. Therefore the fineness of the final pattern would depend on the cooling rate of the liquid steel and the number of forging cycles the blade undergoes. A slowly cooled ingot would make a course pattern or, if forged for a long period, a fine pattern, but a quickly cooled ingot could never make a course pattern. With the development of testing methods (such as SEM-EDX, XRD, and others), it can be argued that a metallurgical analysis like the one above, if included in the commentary, would help those scholars involved in analytical or replication work.
Chapter 4 is titled “Swords in Arabic Poetry” and is claimed as a translation of Schwarzlose’s work in German that discusses Arab swords in early Arabic poetry. Allen in the introduction provides a rational for including this translation, however if one is interested in the technical aspects of swordmaking (as the title of al-Kinid’s work would suggest), then this chapter would yield little in this area of study. Appendices 1 and 2 are commentaries on ibn Hayyan’s and al-Biruni’s thesis on iron respectively. While the Arabic text is not provided, these additions facilitate the comparison among all three works on medieval Islamic iron-working practices.
If a career warrior living in the Muslim lands during medieval times were investing in a sword, what would he be looking for? One would imagine that microstructure of the blade would be an important consideration. This the warrior could ascertain through the patterning on the blade. Quality of the forging and etching could be determined by the surface roughness of the blade. This he would do by running his hand along the flat. The “ring” (pitch and intensity) of the blade would give a qualitative indication of the level of internal defects (tears and voids) in the blade, and hence an indication of its strength in battle. He may verify that the tang and blade are riveted together as welding would cause differential heating of the blade, and would adversely affect its microstructure close to the tang. One would also imagine that he would want to know how maneuverable hi would be with his sword during battle. While weight is an important consideration, beyond that he would want to have some idea of what we now understand as the sword’s center of percussion, and radius of gyration. While sword designs suggest that the medieval Muslim blacksmiths had an intuitive understand of these dynamic characteristics, al-Kindi does not refer to them in his thesis. Why this “obvious” omission is a matter of speculation.
For a technical work such as this one, the Arabic itself needs interpretation so that more relevant translations of technical term may emerge. The interpretation needs to explain scientific works in technical terms, and as science has evolved, the interpretation needs to be presented within the framework of contemporary scientific thought. In this area, much work remains. Just by rendering al-Kindi’s, ibn Hayyan’s and al-Biruni’s works into English, Hoyland and Gilmore have made it accessible to a wider audience. Thus they have created an invaluable reference for scholars interested in Islamic iron and steel working in general, and swordmaking in particular.