About Our Project

Research Question

At the beginning of the semester, we set out to explore translation differences between a well-known English text and its Japanese counterpart. We settles on Alice's Adventures in Wonderland, and asked: how are literary elements translated when their main purpose is to convey a sense of whimsy in a document?



Alice Translations in History

Alice’s Adventures in Wonderland was written by Lewis Carroll (a pseudonym created by Charles Lutwidge Dodgson) and was originally published in 1865. Carroll published three other Alice-related works: Through the Looking Glass and What Alice Found There (1871), Alice’s Adventure’s Under Ground (1886), and The Nursery ‘Alice’ (1890).

The first Japanese translation of one of Carroll’s works was of Through the Looking Glass published in 1899. Since this initial translation, there have been 1,271 editions of Alice’s Adventures in Wonderland and 321 editions of Through the Looking Glass translated into Japanese over the course of 119 years (it should be noted that the only other language with translations in the thousands is Spanish with 1,223 editions; the third most popular language, German, has less than half of this) (Lindseth 741). Titles often differ across translations, but the most commonly used title is Fushigi no Kuni no Arisu (「不思議の国のアリス」).

Alice’s Adventures in Wonderland continues to be translated, adapted, and used as inspiration for new works throughout Japan today. There is even a Lewis Carroll Society of Japan, which has two publications in both Japanese and English.

Project Background

The translation of Alice’s Adventures in Wonderland used for this project is titled Arisu wa Fushigi no Kuni de (「アリスはふしぎの国で」) and was translated by Yu Okubo (大久保ゆう訳) in 2014. There are several reasons why this edition was chosen. This edition, unlike many other translations, is out of copyright. One of the goals of this project is to allow the reader to view both the Japanese and English texts side-by-side, to truly be able to compare the original with its translation. The recent date of publication for this translation also made it appealing, as the language and referenced used would likely be more current, and any unknown features of the text would be easier to look up. Even the title of this work made it appealing; the deviation of the title from the accepted standard title used in the Japanese translation community indicated that there might be a larger amount of interesting linguistical features embedded within the text.

The texts used in this project can be found on the text pages, as well as on their respective original sources. The edition of the Japanese text can be found at Aozora Bunko, and the English text can be found at Project Gutenberg.

Translational Features

When translating a work from one language into another, differences between languages can cause difficulties in producing a target text from the original. Alice is well known for the difficulties it poses to translators.

Rhyme in particular is a feature that appears with some degree of frequency in English songs and poems, but that is almost nonexistent in Japanese. End-rhyme is a feature that is much easier to use in English than in Japanese due to word order; Japanese typically follows a subject-object-verb (SOV) pattern, whereas English word order is subject-verb-object (SVO). For end-rhyme to be present in Japanese poetry, one would have to rhyme verb endings and sentence-final particles, which vary based on several factors, including tense, formality, tone, and gender. Rhyme is also more common in poems that are metrical and contain parallelism (Brooks). English poems have strong focus on meter and parallelism, especially in couplets, whereas Japanese poems tend to embrace asymmetrical structures with few instances of strong stresses.

When translation occurs from Japanese into English, translators tend to not only change the asymmetrical structure and syllabic patterns of Japanese verse, but also that end-rhyme is placed in the English translations where it did not occur before. A Japanese tanka might be translated into a rhyming English quatrain (Brooks).

Idioms can be difficult to translate from one text to the next, as most languages do not have exact equivalents of the lexical properties of idiomatic phrases with the connotations. Idioms are usually translated in one of several ways: by using an idiom with a similar meaning (with either the same types of lexical items or different lexical items), by paraphrasing the meaning without an idiom, or by omitting the idiom completely (Takacs).

Similarly, Carroll’s use of word play makes for interesting translation techniques. Carroll’s use of puns can offer the same kinds of struggles as idioms do; translators must find a set of play on words that not only matches Carroll’s original meaning, but that also convey these puns within their own language (Lindseth). These problems are often tackled in a similar manner to idioms as well; some translators find ingenious ways to convey the similar meaning, while tweaking the play on words to make sense in their own language. Although most translators of Alice attempt to find a word play equivalent in their own texts, occasionally phrases or entire sections of chapter will be omitted due to the difficulty of translation.

Carroll’s use of phonetic and letter-based wordplay often requires translators to find similar sounds within their own languages. In cases of alliteration, consonance, and similar literary devices, it can sometimes be difficult for translators find a series of words with the same sound, especially if the attempt is to keep the same sound that Carroll used in his work.



Linguistic Methods

Two kinds of words were looked for within the texts; nonsense words and incorrect word usage. Nonsense words (“ns”) are words that were coined or made up by Carroll or Okubo. Incorrect word usage (“iu”) was used as an umbrella term for words that are accepted by the dictionary, but that were used in the improper context.

The translation of play on words (“pow”) was also looked for when analyzing the texts. The types of word play focused on were idioms and syntactic play, lexical play, and phonetic and phonemic play. Each kind of word play occurs under the same tag. Word play that was based on structure or typography were not marked; this kind of word play was not as common. The English and this edition of the Japanese are also very similar in structure, with the same number of paragraphs per chapter and same basic structure to poems and songs.

Within the text, play on words often occurred in a specific format; a character would make a statement that would be misinterpreted by another character, who would respond, creating a sense of verbal irony for the reader. These exchanges would often occur over multiple paragraphs, which could make tagging while maintaining well-formedness difficult. For ease of tagging play on words within the structure of the text, only the second part of the play on words phrase would be tagged. For example:

Onomatopoeia (“onomatopoeia”) was added later to the project. Japanese uses a larger quantity of onomatopoeia than English does, but onomatopoeia is still present within the English text.

Within the original XML files, distinctions were made between words and phrases when marking up the text; however, these distinctions have not been included in the XHTML. Both <word> and <phrase> tags contained an attribute that was allowed the same specifications. Because of this, and because the differences between “words” and “phrases” were determined to be insignificant for this project, the XHTML includes only the attribute specifications in a <span> without distinguishing whether the text is a word or a phrase.

Not all attribute specifications were included in the final markup on this site; originally the texts were also marked up for uncommon word usage (“uc”) and Japanese text that ‘matched’ a marked phrase in the English, but that contained no device itself (“match”). Uncommon usage was originally placed on the English words and phrases that had were considered out-of-date, but they were later left out for several reasons. The Japanese translation of Alice used in this project was published in 2014, and likely contains modern Japanese in most cases. The other reason lies in the proficiency of Japanese of the researchers; as neither Dana nor Emily are native Japanese speakers, it would have been difficult to know for certain which Japanese was out-of-date. For this reason, the Japanese text contains no “uc” tags. Match, however, was left out of the HTML when it was decided that the two texts did not to mirror each other in their entirety when rendered on the site.

Computational Methods

We were able to use a variety of languages and systems to fully annotate Alice and come out with the data we wanted. As mentioned above, we started out using XML, simply by tagging unique features in the text with their own tags. We used a Relax NG schema to validate our manuscripts and make sure they had been tagged and organized properly. Our website was built using the host obdurodon, the program Fetch, and XHTML. We also created several CSS stylesheets in order to properly configure our website's aesthetics.

As the semester went on, we began using XSLT to convert our XML files to XHTML that we could put on our website. XSLT also helped us to create visualizations of our data, made by turning our XML into SVG files.



Thank you to Dr. Birnbaum, our instructor, and Gabi Keane, our project advisor, for their advice, teaching, and support.

Thank you to the other students in our class for providing other perspectives to problems we encountered and helping us to look at things in a different way.

Thank you to Hiroyuki Good, for his help in locating crucial texts for this project.


About Us

Emily Byron is a freshman (Class of 2021) at the University of Pittsburgh, main campus, majoring in linguistics and Japanese.

Dana Kaufhold is also a member of the Class of 2021 at Pitt, and is majoring in linguistics and political science.

Neither of us have had prior coding experience, but we highly recommend this class to anyone majoring in Humanities or Social Sciences. You don't need experience - you just need enthusiasm.


Sources and Links


Brooks, E Bruce. “Nine Maxims on Translation.” Lectures: On Translation. 5 Dec. 2002, University of Massachusetts Amherst, Amherst, University of Massachusetts, www.umass.edu.
Carroll, Lewis. Alice's Adventures Under Ground. British Library, 26 Nov. 1864, www.bl.uk.
Lindseth, Jon A., and Alan Tannenbaum. Alice in a World of Wonderlands: the Translations of Lewis Carroll's Masterpiece. First ed., vol. 1 3, Oak Knoll Press, 2015.
Takacs, Csilla. “The Fascination of Translating Idioms.” Bulletin of the Transilvania Univ. of Braşov, no. 2, 2015, pp. 41–54. IV, www.ceeol.com.
Weaver, Warren. Alice in Many Tongues. Univ. of Wisconsin Pr., 1964.

Other Links

Our project's Github.
Our course's webpage.
Project Gutenberg, an online source for English texts that are out of copyright.
Aozora Bunko, an online source for Japanese texts that are out of copyright.