People always invent new ideas in their minds, and therefore the way to express these ideas, either written or orally, needs to be invented. As the result, people started to invent new words, and at this time, neologisms are invented. However, neologisms, in some cases, can be used as a form to control ideas, a contradiction to the notion before. It is best illustrated in a novel called Nineteen Eighty-Four by George Orwell, the data source used in this research, which contains a new variant of language called Newspeak which contains new words. This research aims at investigating types of Newspeak neologisms found in the novel, the translation techniques used to translate them, and the techniques’ impact to the neologisms’ translation quality.

The research is a descriptive qualitative research, in which the researcher collected, arranged, and analysed data due to certain problems posed earlier, in order to be able to interpret the data as well as to draw conclusion. The data contained and found in the novel are in form of words and phrase-like words. By employing total sampling being kept pursuant to the criteria set by the researcher, 118 data containing neologisms were obtained. Besides, to obtain data about translation quality, the researcher employed questionnaires to three raters of translation quality. 

After conducting substantial analysis, the research findings show that there are seven types of neologisms employed by the translator, comprising of old words with new senses (5.08%), derived words (31.36%), abbreviations (19.49%), collocations (5.08%), phrasal words (31.36%), acronyms (1.69%), and pseudo-neologisms (5.93%). The researcher also notes that there are 17 translation techniques employed by the translator, comprising of adaptation (0.85%), pure borrowing (47.46%), naturalised borrowing (3.39%), discursive creation (4.24%), established equivalent (17.80%), literal translation (2.54%), adaptation and explication (0.85%), adaptation and literal translation (0.85%), explication and pure borrowing (2.54%), pure borrowing and explication (8.47%), pure borrowing and literal explication (1.69%), established equivalent and literal translation (2.54%), established equivalent and transposition (0.85%), particularisation and pure borrowing (0.85%), particularisation and naturalised borrowing (1.69%), particularisation and established equivalent (2.54%), and adaptation + literal translation + addition (0.85%).

The researcher also found that there are 102 accurate data (91.89%), four less accurate data (3.60%), and five inaccurate data (4.50%), with 2.83 average score. Furthermore, there are 24 acceptable data (21.62%), 24 less acceptable data (21.62%), and 63 not acceptable data (56.76%), with 1.69 average score. Moreover, there are 38 data with high readability (34.23%) and 73 data with medium readability (65.77%), with 2.36 average score.