About — DocumentTranslationGrammarDictionary


In spring of 2015, I took a class about the works of J. R. R. Tolkien. For the final project, we had the option of writing an essay, or doing a creative project. I decided to make a language.

Tolkien had a love of languages and grammar, but I am more interested in typography and calligraphy, so I started by designing the script for my language. Originally, I wanted to type the document, so I limited myself to modifying an existing script that I could type. Early on, Georgian caught my eye. Just as Tolkien had a taste for Germanic languages, I liked Georgian because it was different enough from the Latin alphabet to be interesting, but still familiar. It is an alphabet system – unlike consonant-based writing sytems like Hebrew or Arabic, syllabaries like Japanese katakana or Cherokee, or ideographic systems like Chinese.

Historically, there have been three major alphabets in use with Georgian, Asomtavruli (ასომთავრული), Nuskhuri (ნუსხური), and Mkhedruli (მხედრული), which are roughly analogs of the Latin capitals, lowercase, and cursive scripts, respectively. However, Georgian today only uses the Mkhedruli script. In printed form, I think that the Nuskhuri script is the most pleasing, but handwritten Mkhedruli is even better. Of the 33 letters widely used in the alphabet, only four are on the baseline and at x height (as the Latin letters x, r, o, e, n, etc.). All of the others ascend (as h or b) or descend (as y or j) from the line. Some do both (as f in this font). These sweeping pen strokes leave a lot of room for stylish writing.

Over time, my writing style has diverged a bit from standard Georgian, especially in a few letters, and I have re-assigned some letter values to better fit the sounds of my language. Along the way, I also lost several letters either with shapes that I couldn’t reconcile with the rest, or that represented sounds that are different in Georgian but indistinguishable in my language. In actuality, the script is very different from Georgian, but the differences are probably no larger than the differences in handwriting and letter values of different languages that use the Latin alphabet, but fictionally, the script is unique.

Armed with a script to write with, I turned to writing a sample text to translate. This would help guide my word inventory and necessary grammatical constructions. A large inspiration came from the book A Canticle for Leibowitz by Walter M. Miller, Jr. Miller was a bomber pilot during WWII, and participated in the bombing of a monastery which housed enemy troops – I considered writing my midterm paper about him, but the themes are more relevant to the Cold War. The book is a post-apocolyptic story about a group of monks hundreds of years after nuclear war. The mission of their order is the preservation of knowledge from the old time, knowledge that is beyond their comprehension, but that they feel is important to preserve. One monk creates an illuminated copy of a circuit diagram for a missile guidance system, while electricity has just barely been re-discovered. The scribal origin of my script and the idea of monks working to preserve knowledge that might be lost influenced my story greatly.

Another large influence is the Old English account of the Battle of Maldon. It might even be said to be the same story, in the reductive way that Tolkien disputed in “On Fairy Stories.” The scenario is certainly similar, but the details give it a distinct difference.

Tolkien’s description of Beowulf in his essay really resonated with me. The idea of reading an old story that even when written down hearkened back to even older times is super interesting. My story uses that quite a bit. I tried to emulate the style of Anglo-Saxon translations in English, especially some Beowulf translations. The description is very sparse, and I used simple sentence structures and simple words. This was partly stylistic, but also helped me translate it. I tried to emulate simple Germanic sentence structure, and considered that my language would have extra cases “These things were told us by him,” rather than “told TO us.” With a dative case (which English doesn’t have), it could become “These things were told us him,” without ambiguity. The result, I think, is that my language, in this example, is far beyond Green Eggs and Ham, but not enough for writing a strong persuasive essay. Writing it in English first forced me to fill out the grammar rules to be at least as expressive as my (simplified) English.

I typed it up then ran a program to find all of the unique words in the story. There were about 200, but languages won’t have one-to-one correspondence between words. I went through the list manually, and built a kind of concordance to find all of the unique sense of each word, and whether I needed it in my language. I removed all structural words – conjunctions, prepositions, articles, etc., which probably not directly translatable. Other word only appeared in idiomatic expressions, which should be translated as a unit. These were phrases such as “yet to come,” “break bread,” “back toward,” “dark art,” and “at once.” I also de-duplicated synonyms killed/slain, strength/might. In this way I reduced my concordance by at least one third. By the time I was done, probably only half of those words were used, but I also had to invent more things.

When making up words, I mostly started with nouns then verbs. The process was mostly arbitrary – I just picked interesting sounding combinations. Sometimes I used a program I wrote to generate random combinations of consonants until I found something that sounded good, but often I could make connections between related words. Talk, tell, story, and the name of the language all have the same root. I didn’t consciously make connections with words in a real language, except a few cases (year - jari, know - kena, i - and), these, I supposed, are coincidental, not indicative of a relationship between the languages.

Once I had enough words, I started translation. I went clause-by-clause in the English, but sometimes what was one clause in English became several in my language or vice-versa. As I went, I made up and documented grammar rules. Unavoidably, much of the grammar is probably Germanic, even in ways I don’t know, because I have only studied English and German extensively, but I tried to use features from widely different languages. There is a large “conlang-ing” community online with many great resources and examples of grammar from other languages. My pronoun classes are borrowed from Quechua (spoken widely in the Andes region, and the origin of words such as llama and quinoa)– I liked the distinction between we (but not you), and we (all of us) when telling the stories of a culture to outsiders vs insiders.

Grammatical agency rather than gender seemed like an interesting concept which I borrowed from Navajo. My sample text doesn’t use it much, but I thought it would interesting when writing fairy-stories to personify nouns by giving them grammatical agency. So, the non-agent thunder becomes Thor when given agency.

The set of cases is the same as in German. The familiarity I have helped my immensely in avoiding mistakes in cases. There is some German inspiration in my verb phrases and prepositions as well, but they are probably as different from German as they are from English.

Finally, once I had finished my translation, I made the manuscript. I used parchment paper, soaked it for a bit in tea, wrote it down, then burned some of the edges. I didn’t have extra pages prepared, and doing a first draft in pencil wouldn’t work, because it didn’t erase well, so there are some errors that I left in, like when I would lose my place and start writing the wrong word, but I am really happy with how it turned out.

About — DocumentTranslationGrammarDictionary