Poster: Translating a Korean Poem into English. A Case Study on Exploring the Connection between Phonetics and Emotions
|Title||Poster: Translating a Korean Poem into English. A Case Study on Exploring the Connection between Phonetics and Emotions|
|Year of Publication||2022|
|Conference Name||Dag van de Fonetiek 2022|
|Authors||Soehnlein, Mika Aya|
|Publisher||Nederlandse Vereniging voor Fonetische Wetenschappen|
|Conference Location||Utrecht, The Netherlands|
David Crystal showed in his book “Sounds Appealing” that there is at least some kind of symbolic meaning to vowels and consonants. He does not further elaborate on it, instead asking: “It doesn’t make sense to ask ‘What is the meaning of /i:/ or /u:/?’.” This is, however, the intention of this research. To explore the connection between emotions and phonetics. To see whether there is any connection at all, and if a potential connection could be universally applicable or if it is specific to each language and culture.
To find an answer to this, there were three steps needed. First, the phonetic inventory of both languages, Standard South Korean and General American, was analysed and compared to have basis on which to work on. In a second step, pre-existing research was used to investigate how emotions are represented in speech in order to see which phonemes seem to correlate with which emotions. In a third step, this was used in a practical application of translating a poem to see whether this actually holds true. The medium of poetry was chosen as it is a literary medium that was originally meant to be performed and therefore, relies more heavily on phonetics and phonology to communicate its message. The poem was translated twice. First, the traditional way and in a second translation, the research was used to match the emotional valence of source and target word as best as possible to translate the poem. Both translations were then compared with each other.
The pre-existing research clearly shows that there is a connection between emotions and phonetics and because some emotions are more similar to each other, the distinction between categories is not fully clear cut. On the basis of this pre-existing research, seven emotions were split into three bigger groups: Group 1 consisting of anger, surprise, joy, and happiness; Group 2 consisting of sadness, fear, and disgust; and a third group Neutral, comprising of a neutral state of being as well as other emotions for which there was not enough data available to clearly put them into either group. Comparing both translations, they do have equally good results concerning trying to match the emotional valence of the source. This begs the question why the first translation was already this close to the source. Is it in my specific case because of my knowledge of English or was the gut instinct right because there is an inherent emotional quality to phonemes that the brain latches onto subconsciously, this in turn then influencing the decision-making process? Right now, it is too time consuming to match the emotional valence of every word from the source and every potential word of the target language. However, if a software could be programmed that could in split seconds do the work for you, then it could certainly majorly help with preselecting useful words – making the whole translation process quicker and the quality better.