Leiden Unified Transliteration
Hi folks! Recently, the ICE took place. What is the ICE? Itโs the largest Egyptological event that takes place every four to five years. This time Leiden was the venue, and about 800 Egyptologists attended. 15 of them met on August 8th to discuss transliteration. (It is a pity that Alexander Ilin-Tomich, Mark-Jan Nederhof and Serge Rosmorduc, some of whom were in Leiden, did not also participate. They could have strengthened the IT component. We will see below where the problem lies). At the end of the congress, the LUT, Leiden Unified Transliteration, was adopted. What does transliteration mean? Hieroglyphs - contrary to popular belief - are not pictographic writing, but like our alphabets, represent spoken language. Hieroglyphs (at least many of them) have phonetic values, so that they have the function โphonogramโ. To reproduce the phonetic value of a sign, transliteration is used. A ๐ has the phonetic value m, so the transliteration m is used. A ๐ also has the phonetic value m, so m is transliterated here as well. A transliteration is therefore not a representation of the graphic system. It does not matter which sign with the phonetic value m is used, the transliteration writes an m to the same extent. A new transliteration sign is needed if - shortened - another sound is present. For the Old Kingdom, the two graphemes ๐ and ๐ด represent two different phonemes, so - for the transliteration of the texts of the Old Kingdom - two transliteration signs are needed. In short, the transliteration reproduces the consonants written through the hieroglyphs without representing the peculiarities of the script with its determinatives, etc.
Egyptology has known the process of transliteration since the 19th century. What is so special about it that a major conference has to deal with it? Unfortunately, the transliteration of Egyptian is not uniform. Some write an s for the hieroglyph ๐ด, others a ล. Some write an แธณ for the hieroglyph ๐, some a q. This is very annoying in our digital age, because then you have four different options for the word ๐๐ค: qs, qล, แธณs, and แธณล. So if you search for this transliteration in a search engine of your choice, you have to enter four variants! In this respect, it would be nice if Egyptology had a standardized transliteration. This problem is compounded by the fact that Egyptology uses different signs for its transliteration. The Egyptological Alif appears as the correct ๊ฃ or as the incorrect ศ, which is actually a Middle English sign.
For a better understanding of the differences, we have created a small table here that shows the special signs of the transliteration in different projects:
| Project | ๐ฟ | ๐ | ๐ | ๐ |
|---|---|---|---|---|
| AED | ๊ฃ | j | y | q |
| AGรA | ศ | แป | y | q |
| IFAO Karnak Cachette | ศ | j | y | q |
| Karnak | ศ | j | y | q |
| ORAEC | ๊ฃ | j | y | q |
| PNM | ๊ฃ | j | jj | q |
| Ramsรจs | ๊ฃ | j | y | แธณ |
| TLA | ๊ฃ | j | y | q |
| Trismegistos | ศ | แป | y | q |
| TSL | A | j | y | q |
| VรgA | ศ | j | y | q |
| Wikisource | 3 | j | y | แธณ |
The table nicely shows the inconsistency, which is on two levels: First, different signs are used for a hieroglyph, i.e. y or jj or q or แธณ. Second, different Unicode signs are used for one sign, i.e., ๊ฃ, ศ, A, and 3 for ๐ฟ. The แป for ๐ is not even listed as a possibility for the Egyptological Yod at Rosmorducโs overview!
Letโs see what LUT proposes. LUT has a total of 26 transcription signs, which looks broadly traditional. It does not follow Rรถรlerโs system, which Schenkel introduced in Egyptology. Nor does it adopt Schneiderโs suggestions, which, if we survey it correctly, did not find favor either. (On those systems, see https://en.wikipedia.org/w/index.php?title=Transliteration_of_Ancient_Egyptian&oldid=1170366739) It is a little surprising that Joachim Quack, as convenor of the meeting that decided LUT, agreed to these transcription marks; for he uses Schenkelโs system exclusively. So, LUT says, no แนญ, ฤฬฃ and ฤ, but d, แธ and แนฏ! At the moment, we can only speculate about the motivations. Probably, LUT was meant to incorporate elements that many Egyptologists use. Schenkelโs system is not very well received, so LUT without Schenkelโs signs is more likely to gain acceptance. Perhaps the degree of diffusion is also the key reason for using q rather than แธณ. As we have seen above, most projects use q. There are possibly two other factors at play. First, especially for beginners, there is a risk of confusion between k and แธณ. Writing q avoids this problem. Second, on the computer, a q can be typed directly. A แธณ, on the other hand, is more difficult to generate. But these are only speculations concerning the motivation. Letโs go back to the signs! That a ลก is used for ๐ is not surprising. We limit ourselves to four cases: ๐ค, ๐ฟ, ๐ and ๐ญ, which we want to discuss.
๐ค
Peust, Egyptian Phonology, pp. 107 ff. discusses the possibility that Egyptian has seven different velars. Thus, in the Old Kingdom, the grapheme ๐ค could represent the phoneme k2w. Usually ๐ค is transliterated as g, so LUT also suggests it. One could consider choosing a separate sign for ๐ค. If you are wondering why we are even considering ๐ค: Well, LUT has its own transliteration sign for ๐ญ too!
๐ฟ
For the LUT there is a PDF where you can see the codes of the transliteration signs. Hooray! The LUT takes the correct ๊ฃ for the Egyptological Alif! This is an important statement. We saw the big mess at the top of the table. Hopefully the correct sign will prevail.
๐
Unlike ๐, the LUT does not suggest the j that most projects use. The LUT thus takes the typographically more difficult route, since the chosen sign is not as easy to produce as j. However, the real problem lies elsewhere and is comparable to the current situation with ๊ฃ. As Alexander Ilin-Tomich points out, completely different Unicode signs are used for the alternative sign to j. There are iฬ, iอ, and iา, which modify the normal i in three different ways. In addition to i as the base, the dotless ฤฑ is also used. So there are three more possibilities: ฤฑฬ, ฤฑอ, and ฤฑา. These are not just potential cases, but several of these variants are actually used and recommended, as Alexander Ilin-Tomich compiles. As of 2019, there is a separate sign in Unicode for the Egyptological Yod: ๊ฝ. Unfortunately, this sign is not yet implemented in many fonts, so again, the sign is not properly visible. The digital projects mentioned above that do not use j do not choose any of the previous seven possibilities, but แป. The LUT introduces another sign: In the PDF of transliteration signs, there is a แผฐ, which actually belongs to the Greek alphabet. Dear readers, have you lost the overview? We can well understand that. This is also our great concern. At the moment we already have the ๊ฃ catastrophe when different Unicode signs are used for the Egyptological Alif. There is a great danger that the rejection of the unambiguous j will create another transliteration disaster, namely many different variants for the ๐. It is not our place to criticize the rejection of the j. We simply point out that it unnecessarily creates problems in the digital world.
๐ญ
Surprisingly, ๐ญ gets its own transcription sign; i.e. there are three separate transcription signs for the three hieroglyphs ๐, ๐ and ๐ญ. This is very unusual. All the above projects have only two transliteration signs for these three hieroglyphs. Two signs are used because the two graphemes ๐ and ๐ represent two different phonemes in the rendering of Semitic words, cf. Albright, Vocalization, p. 34. So where does the third sign come from? Peust, Egyptian Phonology, pp. 49 f. wants to differentiate between the three signs in transliteration. He writes about the distribution of the hieroglyphs:
In Middle Egyptian all three signs indicated are in current use and variation among them becomes rare. It can be roughly stated that at this time ๐ is not restricted in distribution, whereas ๐ can appear at a morpheme boundary only, and ๐ญ is even more restricted in that it is used only morpheme-finally.
Thus, according to Peust, the distribution is done for morphemic and not for phonematic reasons. He writes further on the function:
At least as can be judged from Coptic, there seems to have no phonetic difference between ๐, ๐, and ๐ญ. The sounds rendered by all three signs are equally preserved as /j/ or lost by the same conditions in Coptic. It is possible that these signs never expressed a phonetic distinction but instead their functional difference basically was to indicate morphological boundaries or other features. It is also possible that they did originally express a phonetic distinction which was lost later.
So there is no evidence for a phonological difference. But this is exactly what is needed to postulate an independent transliteration sign! All that remains is the assertion that Middle Egyptian seldom exchanged the three signs. According to this, the three transliteration signs are exclusively peculiarities of the graphic system. But even this does not seem to be true. With our metasearch engine it is very easy to find spellings with ๐ญ. The data from the PNM project speak for themselves. (We stopped our survey after twenty hits, because we think the results are already clear).
| Name | Some spellings |
|---|---|
| jtjj | ๐๐๐๐, ๐๐๐๐, ๐๐๐ญ |
| jแนฏ๊ฃ.j | ๐๐ ท๐ฟ๐, ๐๐ ท๐ฟ๐ญ |
| nmt.j | ๐ฐ๐ ๐๐ญ, ๐๐ฐ๐ ๐๐, ๐๐ฐ๐ ๐๐ |
| rns.j | ๐๐๐ด๐, ๐๐๐ด๐, ๐๐๐ด๐ญ |
| แธซnt.j-แบt.jj-แธฅtp | ๐๐๐๐ก๐๐๐ต๐๐ช, ๐๐๐๐ก๐๐ญ๐ต๐๐ช, ๐๐ก๐๐๐ต๐๐ช |
| snb.tjโธsj | ๐ด๐๐๐๐ด๐ญ, ๐ด๐๐๐๐ด, ๐ด๐๐๐๐๐ด๐ |
| kjj | ๐ก๐, ๐ก๐ญ, ๐ก๐ |
The three hieroglyphs are interchangeable in the personal names! It seems to us a suitable topic for a final thesis to investigate the distribution of ๐, ๐, and ๐ญ. A working hypothesis would be that the regularity of the distribution depends on genre and register and is purely graphemic. Substantial evidence that a third phoneme is represented in ๐ญ, requiring a separate transliteration mark, does not seem to exist.
Therefore, in our opinion, a separate รฏ is not recommended. In addition, we would like to point out that adapting the data is very time-consuming, since it cannot be done automatically. Do you really want to go to all that effort just because of a questionable sign distribution without any phonemic reference? Sorry, no thanks, we cannot follow this suggestion.
Standardization
The above points have been lost in detail; in general it is a laudable view that Egyptology treats the same things in the same way, i.e. that one agrees on a uniform system. But we still see two general problems here:
Standardization prescribes the status quo. It is difficult to incorporate new findings into a standardized system. Transliteration signs already do not reflect the state of phonological research. ๐ง is considered by many to be an emphatic sound. Accordingly, a แนญ is a much better rendering than a d, which can lead those unfamiliar with the subject to make incorrect etymological equations.
We have described above the problem that to search for ๐๐ค one has to enter four variants into the search engine. But these are regular equivalents. Much worse are the cases where the Egyptological transliteration offers variants that cannot be handled in a regular way. The scribe is sลก or zแบ๊ฃ, the sealer แธซtm.tj or sแธ๊ฃw.tj, the sun god R๊ฅ or R๊ฅw, the head tp or dp, โto be illโ mr or mแธฅr, and the king nzw, nswt, or njswt. This is the real problem with transliteration searches in a search engine. You are dealing with an unpredictable variance.
So this time we conclude rather pessimistically. True, it would be nice to have a uniform system. We can certainly talk about the details mentioned above. But such variants as แธฅtp dj nswt, แธฅtp-แธ-nzw or แธฅtp แธiฬฏ njswt are the real problem.
This work is marked with CC0 1.0 Universal
