instagram translate’s and human translation’s … · 2018. 2. 15. · instagram translate'sand...

INSTAGRAM TRANSLATE’S AND HUMAN

TRANSLATION’S PERFORMANCE IN TRANSLATING THE

CAPTIONS IN @BASUKIBTP INSTAGRAM ACCOUNT

AN UNDERGRADUATE THESIS

Presented as Partial Fulfillment of the Requirements

for the Degree of Sarjana Sastra

in English Letters

By

STEFANI VERONIKA

Student Number: 134214137

ENGLISH LETTERS STUDY PROGRAM

DEPARTMENT OF ENGLISH LETTERS

FACULTY OF LETTERS

SANATA DHARMA UNIVERSITY

YOGYAKARTA

2018

PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI

ii




AN UNDERGRADUATE THESIS

Presented as Partial Fulfillment of the Requirements

for the Degree of Sarjana Sastra

in English Letters

By

STEFANI VERONIKA

Student Number: 134214137

ENGLISH LETTERS STUDY PROGRAM

DEPARTMENT OF ENGLISH LETTERS

FACULTY OF LETTERS

SANATA DHARMA UNIVERSITY

YOGYAKARTA

2018


INSTAGRAM TRANSLATE'S AND HUMANTRANSLATION'S PERFORMANCE IN TRANSLATING THE

CAPTIONS IN @BASlJKIBTP INSTAGRAM ACCOUNT

. Harris Hennansyah Setiajid, M.Hum.Advisor

Anna Isti'anah, S.Pd.. M.Hum.Co-Advisor

III

November 30, 2017

December 15,2017


v

LEMBAR PERNYATAAN PERSETUJUAN PUBLIKASI KARYA ILMIAH

UNTUK KEPENTINGAN AKADEMIS

Yang bertanda tangan di bawah ini, saya mahasiswa Universitas Sanata Dharma

Nama : Stefani Veronika

Nomor Mahasiswa : 134214137

Demi pengembangan ilmu pengetahuan, saya memberikan kepada Perpustakaan

Universitas Sanata Dharma karya ilmiah saya yang berjudul




Beserta perangkat yang diperlukan (bila ada). Dengan demikian saya memberikan

kepada Perpustakaan Universitas Sanata Dharma hak untuk menyimpan,

mengalihkan dalam bentuk media lain, mengelolanya dalam bentuk pangkalan

data, mendistribusikan secara terbatas, dan mempublikasikannya di internet atau

media lain untuk kepentingan akademis tanpa perlu meminta ijin kepada saya

maupun memberikan royalti kepada saya selama tetap mencantumkan nama saya

sebagai penulis.

Demikian pernyataan ini saya buat dengan sebenarnya.

Dibuat di Yogyakarta

Pada tanggal 29 November 2017

Yang menyatakan,

Stefani Veronika


vi

STATEMENT OF ORIGINALITY

I certify that this undergraduate thesis contains no material which has been

previously submitted for the award of any other degree at any university, and that,

to the best of my knowledge, this undergraduate thesis contains no material

previously written by any other person except where due reference is made in the

text of the undergraduate thesis.

Yogyakarta, November 29, 2017

Stefani Veronika


vii

STAY FOOLISH.

-Steve Paul Jobs-

STAY HUNGRY.


viii

TO YOU, MA&PA


ix

ACKNOWLEDGEMENTS

First thing first, I would like to send my sincere gratitude to my best

friend, Jesus Christ for His blessing so that now I can write down my

acknowledgements joyfully.

Needless to say, for sure I always cannot list everyone’s names who

helped me not only finish my thesis but also give me such a meaningful

experience during my study in Universitas Sanata Dharma. Some names,

however, are mentioned in this page, just because I really cannot express how

thankful I am to be supported directly by them.

Obviously, to my mother Tan Li Li and my father Shi Siang Siu. Thank

you for everything, everything and everything. My cool thesis advisor, Harris

Hermansyah Setiajid, M.Hum for always helping me from the very beginning

of this research conducted every time, like every time I needed help. Special

thanks goes to my thesis co-advisor, Arina Isti’anah, S.Pd., M.Hum for

providing suggestions and insight to improve my thesis.

My Cendani III troops, Sesilia Gisela Serat, S.S and Novita Sari, S.S, who

always stand by my side, 24/7. For my friend, Theresia Esti M. and my always-

be-there-friends, Mareta Anggita, B.IHM (hons) and Hendrike F. Cherry

Sumarauw, S.T. I also would like to send my huge thanks to my friends in Public

Relations of USD, JLTC, and also Kelas D Masih Skripsi’s Group.

Stefani Veronika


x

TABLE OF CONTENTS

TITLE PAGE ....................................................................................................ii

APPROVAL PAGE ..........................................................................................iii

ACCEPTANCE PAGE ....................................................................................iv

LEMBAR PERNYATAAN

PERSETUJUAN PUBLIKASI KARYA ILMIAH ............................................v

STATEMENT OF ORIGINALITY ................................................................vi

MOTTO PAGE .................................................................................................vii

DEDICATION PAGE .......................................................................................viii

ACKNOWLEDGEMENTS ..............................................................................ix

TABLE OF CONTENTS ..................................................................................x

LIST OF ABBREVIATIONS ..........................................................................xii

LIST OF TABLES ............................................................................................xiii

ABSTRACT .......................................................................................................xiv

ABSTRAK ...........................................................................................................xv

CHAPTER I: INTRODUCTION ....................................................................1

A. Background of the Study ....................................................................1 B. Problem Formulation ..........................................................................4 C. Objectives of the Study ......................................................................4 D. Definition of Terms ............................................................................4

CHAPTER II: REVIEW OF LITERATURE ................................................6

A. Review of Related Studies ..................................................................6 B. Review of Related Theories ...............................................................10 C. Theoretical Framework.......................................................................16

CHAPTER III: METHODOLOGY ................................................................17

A. Area of Research ...............................................................................17 B. Object of the Study .............................................................................18 C. Method of Study .................................................................................18 D. Research Procedure ............................................................................19

1. Types of Data ................................................................................19

2. Data Collection ..............................................................................19

3. Population and Sample ..................................................................21

4. Data Analysis .................................................................................22

CHAPTER IV: ANALYSIS RESULTS AND DISSCUSSION .....................25

A. The Errors Found in @basukibtp Instagram Captions ......................25

1. The Translation Errors by Instagram Translate Found in

@basukibtp Instagram Captions ..................................................26

a. Omitted Concept .................................................................26

b. Added Concept ....................................................................29

c. Untranslated Concept...........................................................32

d. Mistranslated Concept .........................................................34


xi

e. Substituted Concept .............................................................38

2. The Errors Found in @basukibtp Instagram Captions by Human

Translator ..................................................................................41

a. Omitted Concept ................................................................41

b. Added Concept ...................................................................43

c. Mistranslated Concept ........................................................44

d. Substituted Concept ...........................................................46

e. Explicitated Concept .........................................................47

B. The Performance of Instagram Translate and Human Translator in

Translating @basukibtp Captions .....................................................48

1. The Performance of Instagram Translate and Human Translator

in Omitted Concept......................................................................50


in Added Concept ........................................................................51


in Untranslated Concept ..............................................................52


in Mistranslated Concept .............................................................54


in Substituted Concept .................................................................55


in Explicitated Concept ...............................................................57

CHAPTER V: CONCLUSION ........................................................................59

BIBLIOGRAPHY .............................................................................................62

APPENDICES ...................................................................................................64

Appendix 1: All Data Translated by IgT ................................................64

Appendix 2: All Data Translated by HT ................................................67

Appendix 3: Description of Errors .........................................................71


xii

LIST OF ABBREVIATIONS

ST Source Text

SL Source Language

TT Target Text

TL Target Language

MT Machine Translate

IgT Instagram Translate

HT Human Translator

No. Number of datum


xiii

LIST OF TABLES

No. Table page

1. Table 1 Example of the Data Coding 21

2. Table 2 Description of error in Instagram Translate 23

3. Table 3 Description of error in Human Translator 24

4. Table 4 Omitted Concept by Instagram Translate 26

5. Table 5 Added Concept by Instagram Translate 30

6. Table 6 Untranslated Concept by Instagram Translate 33

7. Table 7 Mistranslated Concept by Instagram Translate 35

8. Table 8 Substituted Concept by Instagram Translate 39

9. Table 9 Omitted Concept by Human Translator 41

10. Table 10 Added Concept by Human Translator 43

11. Table 11 Mistranslated Concept by Human Translator 44

12. Table 12 Substituted Concept of Human Translator 46

13. Table 13 Explicitated Concept by Human Translator 47

14. Table 14 Total Errors by IgT and HT 49

15. Table 15 Total Errors in Omitted Concept done by IgT and HT 50

16. Table 16 Total Errors in Added Concept done by IgT and HT 51

17. Table 17 Total Errors in Untranslated Concept done by IgT and HT 53

18. Table 18 Total Errors in Mistranslated Concept done by IgT and HT 54

19. Table 19 Total Errors in Substituted Concept done by IgT and HT 55

20. Table 20 Total Errors in Explicitated Concept done by IgT and HT 57


xiv

ABSTRACT

VERONIKA, STEFANI. Instagram Translate’s and Human Translation’s

Performance in Translating the Captions in @Basukibtp Instagram Account.

Yogyakarta: Department of English Letters, Faculty of Letters, Sanata Dharma

University, 2017.

The paramount aspect of translating is when one meaning in source

language is well-conveyed in the target language. Translating can be done not

only by a professional translator but also machines translate. In 2016, a social

media called Instagram started providing a translation feature for its users to

translate captions. However, the researcher finds errors related to the meaning

conveyed from source texts to target texts, Indonesian to English. The errors are

found in Instagram accounts which owned by an Indonesia politician, Basuki

Tjahaja Purnama or commonly known as Ahok. Based on that phenomenon, the

researcher would like to investigate further about translation result done by

Instagram Translate and compare it to a human translator.

In this research, there are two problems to solve. The first one is to find

error exists in the result of translation done by Instagram Translate and human

translator. The second one is to find out the different performances of each

translation which might be reflected through the errors made.

In order to deal with the problems, this research applied library and

explicatory methods. Furthermore, the qualitative research method was also

applied to demonstrate the findings.

In the final result, the researcher finds the total error found in Instagram

Translate is 54, which 15 errors in Omitted Concept, 6 errors in Added, 4 errors in

Untranslated, 24 errors in Mistranslated, and 5 errors in Substituted Concept.

Meanwhile, human translator makes 8 errors. 1 error in Added and Explicitated

Concept, and 2 errors in Omitted, Mistranslated, and Substituted Concept.

Generally, it can be said that HT makes better performance than IgT in translating

captions.


xv

ABSTRAK

VERONIKA, STEFANI. Instagram Translate’s and Human Translation’s

Performance in Translating the Captions in @Basukibtp Instagram Account. Yogyakarta: Program Studi Sastra Inggris, Fakultas Sastra, Universitas Sanata

Dharma, 2017.

Aspek yang paling penting dalam penerjemahan adalah arti dari Bahasa

sumber ke Bahasa target tersampaikan dengan baik. Penerjemahan Bahasa tidak

hanya dilakukan oleh penerjemah profesional saja, tetapi juga dapat dilakukan

oleh mesin penerjemah. Pada 2016, sebuah sosial media yang dikenal dengan

Instagram memfasilitasi penggunanya dengan memberikan fitur terjemahan pada

kolom takarir gambar. Namun, peneliti menemukan kesalahan dalam

penyampaian makna dari bahasa Indonesia sebagai teks sumber ke teks target

dalam bahasa Inggris. Kesalahan yang dimaksud, ditemukan pada akun Instagram

milik seorang politikus Indonesia yang bernama Basuki Tjahaja Purnama atau

biasa dikenal sebagai Ahok. Berdasarkan fenomena tersebut, peneliti berkeinginan

untuk meneliti lebih lanjut hasil terjemahan yang dilakukan oleh Instagram

Translate dan membandingkannya dengan hasil terjemahan penerjemah

profesional.

Terdapat dua rumusan masalah yang menjadi fokus penelitian ini. Masalah

pertama adalah menemukan kesalahan-kesalahan pada produk terjemahan

Instagram Translate dan penerjemah profesional. Masalah kedua adalah melihat

kinerja yang berbeda dari masing-masing terjemahan yang dapat tercermin

melalui kesalahan-kesalahan yang dibuat.

Metode yang digunakan dalam penelitian ini adalah studi pustaka dan

metode explicatory. Selanjutnya, metode kualitatif juga digunakan untuk

menjabarkan hasil penelitian.

Hasil penelitian ini menunjukkan kesalahan-kesalahan yang ditemukan.

Instagram Translate membuat sebanyak 54 kesalahan yang diantaranya adalah 15

kesalahan pada konsep omitted, 6 kesalahan pada konsep added, 4 kesalahan pada

konsep untranslated, 24 kesalahan pada konsep mistranslated, dan 5 kesalahan

pada konsep substituted. Berbeda dengan penerjemah profesional yang hanya

membuat 8 total kesalahan. Diantaranya 1 kesalahan pada konsep added dan

explicitated, dan 2 kesalahan pada konsep omitted, mistranslated dan substituted.

Secara garis besar, dapat dikatakan bahwa terjemahan yang dilakukan oleh

penerjemah profesional menunjukkan kinerja yang lebih baik daripada Instagram

Translate dalam menerjemahkan takarir gambar.


1

CHAPTER I

INTRODUCTION

A. Background of the Study

Language as a means of communication plays an important role in delivering

messages. People use language in many fields such as religion, education,

technology, etc. However, according to Anderson cited in (ethnologue.com), there

are more than 6800 distinct languages in the world. The fact that a few people can

speak other languages besides their mother tongues is evident. This case leads to a

development of technology that provides various features to help its users deliver

and receive messages.

In this modern era, the use of technology gives significant helps for its users,

including in language matter. The fact that technology is consumed by a lot of

people makes it inseparable from the use of social media, nonetheless. Through

technology, the users can provide and receive messages through their own social

media accounts.

Social media have their revelatory role in delivering messages. In social

media, people (users) can create and follow more than one account. Hence, the

messages can be found in many accounts, whether the official or non-official ones.

There are many politicians, artists, educators as well as economists who share

messages through their own social media. One of the most popular social media is

Instagram. Instagram provides its one core feature, which is to share pictures,

besides videos, captions, likes, comments, Instagram Live and Instagram Story.


http://www.ethnologue.com/

2

Instagram is one of social media which has global users. As cited from

(Instagram-press.com), the Instagram users in 2017 are 800 million people around

the world. It makes a translation feature of Instagram Translate (IgT) very useful.

Certainly, among other languages around the world, Indonesian language is one to

be translated.

This undergraduate thesis discusses an official account which is owned by

Indonesia politician, Basuki Tjahaja Purnama or commonly known as Ahok. The

account, which has more than two millions followers, uploads posts containing his

social and work life. Ahok’s Instagram account is chosen due to the fact that he was

one of the governors who attracts Indonesia citizen’s attention the most in these

past years. In addition, he is one of Indonesia politician who delivers message and

information about issues in Indonesia through his own social media, Instagram.

One of the captions in his posts on June 22 is, Selamat ulang tahun Jakarta,

semoga terwujud Jakarta Baru yang modern, tertata rapi, manusiawi dan

melayani.#JKT489 which is translated by IgT into “Happy birthday, Jakarta,

hopefully come to Jakarta the new modern, neat, inhumane and serve. #JKT489”.

From the translation, it is seen that the source text meaning undergoes change (ST)

in target text (TL). In Indonesian, the word manusiawi has a positive meaning but

then it is translated into exactly the opposite meaning, “inhumane”.

From this case, it is interesting to closely examine how IgT works, and what

errors might be found. According to Shivali, IgT started showing its translation tool

in June 2016 (2016). Therefore, its relative novelty as a machine translation (MT)

compared to other established MTs, such as Google Translate and Bing Translator,


3

is worthy of analysis related to what extent this new MT is able to render the ST

into TT successfully.

The undergraduate thesis focuses on the translation result done by IgT.

According to Newmark, translation itself is, “a craft consisting in the attempt to

replace a written message and/or statement in one language by the same message

and/or statement in another language (1981:7). In other words, translation is a

process of replacing source text to target text by substituting the language demanded

without changing the intended message. In order to analyze the translation result,

the researcher would like to compare IgT to the manual translation which is done

by human translator.

The manual translation is produced by a professional and certified translator

who is the Secretary of Himpunan Penerjemah Indonesia (Indonesian Translators’

Association), working in Translexi, a professional translator services, Yogyakarta.

The fact that this research has its focus on the both translation results, the

categorization of translation mistakes or errors is needed. In addition, the

explanation and discussion to show the different results they make are also required.

As for the researcher decides to conduct this research, it is expected that this

undergraduate thesis would help the readers, especially students in Universitas

Sanata Dharma, see the different performances of human translator and Instagram

Translate in delivering message from Indonesian to English captions. The

researcher also anticipates for the Instagram Translate and also human translator to

develop their translation in accordance with the errors found by the researcher.


4

B. Problem Formulation

There are some problems in this undergraduate thesis which would be

analyzed. They are formulated into two major problems as follows:

1. What errors are found in the English translation of @basukibtp Instagram

account done by Instagram Translate and human translator?

2. How do Instagram Translate and human translator perform in translating the

captions in @basukibt Instagram account?

C. Objectives of the Study

This research aims to find and analyze the translation errors made by both

Instagram Translate and human translator in translating Instagram’s captions. The

errors are used as a source to measure translation performances and to show the

difference result in translating captions.

D. Definition of Terms

In order to have the same perceptions and terminologies used in this study,

the researcher defines the following terms:

Translation, as stated by Bell, “is the replacement of a representation of a

text in one language by a representation of an equivalent text in a second language”

(1991:6).

Machine Translation, according to Hutchin, “is the now traditional and

standard name for computerized systems responsible for the production of

translations from one natural language into another, with or without human


5

assistance”. There are many machine translations available today, to mention some

are Google Translate, Bing Translator, Sederet.com, and relatively new Instagram

Translate (1992: 3).

Error Analysis in Machine Translation is “the identification and

classification of individual errors in a machine translated text. It is also a means to

access machine translation output in qualitative terms, which can be used as a basis

for generation of error profiles for different system” (Stymne and Ahrenberg, 2012:

1).

Instagram is “a relatively new form of communication where users can

easily share their updates. It is a popular photo (video) capturing and sharing mobile

application, with more than 150 million of registered users since its launch in

October 2010” (Hu, Manikonda and Kambhampati, 2013: 2). Instagram recently

released its machine translation, popularly called Instagram Translate to help its

various users understand each other despite the language barrier.

Caption, as defined by Oxford Advance Learner’s Dictionary, is “words that

are printed underneath a picture, CARTOON,etc. that explain or describe it” (2010:

209).


6

CHAPTER II

REVIEW OF LITERATURE

In this chapter, the researcher aims to show and elaborate the researches on

similar topic done by other researchers. This present research complies with

Kurnianto’s and Febriana’s theses and also two journals written by Li, Graesser and

Cai and also Saffari, Sajjadi and Mohammadi. Each of them is reviewed to find the

similarity and difference discussion in order to avoid topic duplication. Some

theories applied are also reviewed in order to set a solid ground on which this

research has the focus on.

A. Review of Related Studies

1. Kurnianto’s thesis “ Google Translate Assessment With Error Analysis:

An Attempt To Reduce Error”

This undergraduate thesis discusses the machine translate assessment done by

Google Translate in three kinds of texts. There are an iPad user guide, a National

Geographic article and Ownership Agreement document.

This research focuses on two problems. The first is to find what errors exist

in the result of translation and the second is to suggest effort to reduce all errors

found in the first problem. In analyzing the errors, the researcher uses Koponen’s

first error category which divides the errors into six subclasses. After all errors are

categorized, the researcher uses Farrús’s suggestion in three methods to attempt

reducing errors.


7

This thesis shows that in the final result there are 206 errors in the three texts.

The suggestions by the researcher shows the errors in three texts are reduced. The

first method, typing in isolated form, can reduce almost 50% errors for omitted,

untranslated, and mistranslated concept. The second method, text edition, reduces

errors related to orthographic errors and literal translation. And the last method,

combination of all method, can reduce all categories.

This present undergraduate thesis develops Kurnianto’s thesis and answer

questions not elaborated in his thesis. Although it uses similar errors category, this

present undergraduate thesis does not stop in measuring machine result. It continues

in giving a comparison to a manual translation which is done by human or a

professional translator.

2. Febriana’s thesis “The Translation Performance of Sederet.com And

Google Translate: A Comparative Study With Error Analysis”

In this undergraduate thesis, Febriana assesses two machine translates which

are Sederet.com and Google Translate to translate three short stories namely Dads

Blessings, The Grasshopper and the Ant, and The Princess and the Pea. Similar to

Kurnianto’s thesis, this thesis also uses Koponen’s error category in analyzing the

three short stories.

The errors found in the three texts translated by both MTs are classified into

six subcategories. They are Omitted Concept, Added Concept, Untranslated

Concept, Mistranslated Concept, Substituted Concept and Explicitated Concept.

This research finds out 387 errors made by MTs and 69.5% errors are in the

Mistranslated Concept category which is followed by Omitted, Added and


8

Untranslated Concept. In this thesis, the errors categorized are then used to measure

the MTs performance in translating. Overall, by counting the errors, this research

tells that Sederet.com performs better than Google Translate.

As this undergraduate thesis shares the similar error category by Koponen to

the present undergraduate thesis, the difference is also found. The difference lies on

machine translation which is compared. In this present undergraduate thesis, the

machine translation analyzed is Instagram Translate and instead of comparing it to

another MT, the researcher chooses human translation as the comparison.

3. Li, Graesser and Cai Journal “Comparison of Google Translation with

Human Translation”.

In this journal, Graesser and Zhiqiang Cai investigate the accuracy of Google

Chinese-to-English translation from the perspective of formality and cohesion with

two comparison. They are Google translation with human expert translation and

Google translation with Chinese source language. The text used is a collection of

289 spoken and written texts excerpts from the Selected Works of Mao Zedong in

both Chinese and English version.

The texts taken are analyzed by automated text analysis tools. They are

Chinese and English Linguistic Inquiry and Word Count (LICW) and the Chinese

and English Coh-Metrix (a computational tool). The result of this study shows that

correlations on formality and cohesion showed Google English translation is highly

correlated with both human English translation and the original Chinese text.

The present study shares similarity to the study conducted by Graesser and

Zhiqiang Cai which is to compare a machine translate and human translator.


9

However, besides the object, the machine that is compared show differences. The

object used by present researcher is captions from Instagram. In addition the

machine analyzed is not Google Translate but Instagram Translate. The languages

analyzed are also differ, the present research uses Indonesian and English while

Graesser and Zhiqiang Cai use Chinese and English. This present research develops

the previous research by conducting analyses on the pattern and giving possibility

reasons (of errors) to evaluate the translation being compared.

4. Saffari, Sajjadi and Mohammadi “Evaluation of Machine Translation

(Google Translate vs. Bing Translator) from English into Persian across

Academic Fields”

In this study, the researchers investigate English-to-Persian accuracy of

machine translation at lexical, semantic and syntactic levels by using Groves and

Mundt (2015) Model of error taxonomy. The texts selected are from 4 (four)

domains, namely law, literature, medicine and mass media. 60 texts from each

domain are taken. The texts translated by Google Translate, Bing Translate as well

as human translators are evaluated with respect to lexical, semantic and

grammatical accuracy.

The result of this study shows that the translation done by the four human

experts just served as the benchmark against which Google and Bing translation are

judged and scored. The machines translations are not yet as accurate as human

translation in translating texts from English into Persian translation.

The present study shares similarity of comparison, machine translate and

human translator. However, the machine compared are Google Translate and Bing


10

Translate while the present study only focuses on one machine, which is Instagram

Translate. Another difference lies on the number of human translator, instead of

having 4 human translators, the present study has one human translator only to

balance the comparison, one from machine and one from manual (human)

translation. The present researcher agrees with the accuracy being compared, which

is lexical. As for semantic and syntactic are not yet suitable for the Instagram

Translate which is a relatively new machine translate to focus on.

B. Review of Related Theories

1. Theories of Translation

There are many definitions about translation. Each expert has his or her

definition based on his or her own perspectives. However, the researcher would like

to see the relations among those experts’ definitions. The first definition is

according to Catford, who states “the replacement of textual material in one

language (SL) by equivalent textual material in another language (TL)” (1965: 20).

Here, according to Catford, translation concerns with the text material which

must be similar between the source text and target text. Even though Catford only

focuses on the text material which is hard for the translator to change the text

material while not to focus on the structure and meaning, Nida and Taber support

the definition as

Translating consists of reproducing in the receptor language the closest

natural equivalent of the source language message, first in terms of meaning

and secondly in terms of style (1974:14).


11

From the definition according to Nida and Taber, the translation should be

the closest natural equivalent in meaning of the source language message. It can be

detected by the target language readers that should have similar response to the text

as in the source language readers have.

The last definition is according to Bell. Bell states that translation is “the

expression in another language (or target language) of what has been expressed in

another, source language, preserving semantic and stylistic equivalences.”

(1991:5). Bell supports Nida and Taber’s definition in term of semantic or meaning

and stylistic or style in language. From those definitions, it can be concluded that

translation is a replacement of text material that concerns with meaning and also

style. Certainly the text material should bring the same meaning which indicated by

the same response from target text and source text’s readers.

2. Machine Translation

According to Hutchin, machine translation is

the now traditional and standard name for the computerized systems

responsible for the production of translation from one natural language into

another, with or without human assistance (1992: 3).

Hutchin also states that in the twentieth century, humanity’s oldest dream of

mechanization of translation has become a reality in which computer programs are

capable to translate a wide variety of texts from one language into another natural

language. Even though in reality, the translation is not always perfect, however

humans have already achieved the development of program in which MT “can

produce ‘raw’ translation of texts in relatively well-defined subject domains, which

can be revised to give good-quality translated texts”. (Hutchin, 1992 : 1).


12

According to Chéragui (2012: 165-166), Machine Translation has four

different types. They are Machine Translation for Watcher (MT-W), Machine

Translation for Revisers (MT-R), Machine Translation for Authors (MT-A), and

Machine Translation for Translators (MT-T). MT-T is usually incorporated into the

translation work stations and the PC based translation tools. It aims at helping

human translators do their job by providing online dictionaries, thesaurus and

translation memory.

In conclusion, machine translate is a machine that is used to translate texts

with or without human by providing dictionaries, thesaurus and translation

memory.

3. Human Translation

According to Hopkinson (2009: 1-2), the role of a translator is many-faceted.

He or she must hear the music of the original, and replay it for a new audience; a

good translation sings, and displays a rhythm that not only reflects the original text’s

origin but also beats to a new drum.

In other words, a translator is both the reader and writer. The translator must

transfer the meaning through message in a good way which reflects the original

message. Hopkinson adds that in translating a literary works, the translator’s job is

to “recreate this work of art sensitively and seamlessly in such a way that it is true

to the original, as well as being equally enchanting, poetic and perceptive”.

Therefore, a translation should have the same virtues as the original, and inspire the

same response in its readers.


13

4. Koponen’s Classification in Assessing Machine Translation

As stated by Hutchin (1992:2), that “a good translation which are produced

by both human and machine is a difficult concept to define precisely”, the error

category assessment is needed to help assessing translation quality. According to

Koponen (2010), translation quality assessment is important for both human and

machine to bring out interesting differences translation result. Furthermore, the

analysis should separate form and function, and focus on errors affecting the

accuracy of meaning (translation errors) over errors related to fluency only

(language errors) such as capitalization and punctuation.

Koponen defines errors category based on the tested text into two major

classes. They are namely relation between source and target concepts and relation

between concepts. The two major classes are then divided into 6 and 8 subclasses.

The mismatches are as follows:

a. Relation between source and target concepts (Koponen, 2010: 4-5)

i. Omitted concept happens when Source Text (ST) concept that is not

conveyed by the Target Text (TT). It means the concept that should appear

in TT is absent. For example, when the ST Jakarta Barat then the TT

translation shows Jakarta only, the concept of barat is omitted.

ii. Added concept exists when TT concept that is not present in the ST. This

category is exactly the opposite of Omitted Concept. TT adds new concept

that is absent in ST. For example, the ST Terima kasih, bapak Jokowi is

translated into “Thank you, Mr. Jokowi and wife”. From the TT, the

additional of the word “wife” is categorized as Added Concept.


14

iii. Untranslated concept is Source Language (SL) words that appear in TT. It

means that a word or some words from ST is not translated to the TT. For

example, Bersama Keluarga Besar Damkar is translated into “With

Damkar big family”. The word Damkar is categorized as Untranslated

Concept because it appears in TT as the SL.

iv. Mistranslated concept happens when a TT concept has the wrong meaning

for the context. In this category, the wrong meaning of a concept possibly

happens because the word has polysemous meaning. For instance, Dapat

laporan dari warga is translated into “Can report from citizen”. The word

dapat means both “to get” also “be able” (can).

v. Substituted concept happens when TT concept is not a direct lexical

equivalent for ST concept but can be considered a valid replacement for

the context. For example, the ST Peresmian RPTRA di Menteng is

translated to “Grand opening of RPTRA in Menteng”. The word

Peresmian is not a direct lexical equivalent to “Grand opening” but

“inauguration”. As for “grand opening” is a direct lexical meaning to

pembukaan perdana.

vi. Explicitated concept exists when TT concept explicitly states information

left implicit in ST without adding information. In other words, the concept

appears in TT does not change the meaning as a whole but giving the

information more explicit than the ST intended to. For example, the

addition of ohjelma or program to Norton Antivirus (Koponen, 2010:5).


15

b. Relation between concepts (Koponen, 2010: 6)

i. Omitted participant happens when ST relation not conveyed by the TT due

to an omitted head or dependent.

ii. Omitted relation is ST relation not conveyed by the TT due to morpho-

syntactic errors that prevent parsing the relation although both concepts

are present in the TT.

iii. Added participant is the TT relation not present in ST introducing an added

concept.

iv. Added relation happens when TT relation not present in ST arises due to

morpho-syntactic errors.

v. Mistaken participant exists when head or dependent of the relation

different in ST and TT, not same entity.

vi. Mistaken relation is when relation between two concepts different in ST

and TT, changed role.

vii. Substituted participant exists when head or dependent of the relation

different in ST and TT, same entity.

viii. Substituted relation is relation between two concepts different in ST and

TT, same semantic roles.

As it is categorized above, the relation between source and target concepts or

Individual Concept are only represented by content words such as noun, verb,

adjective and adverb. They can be unit larger than individual words, for example in

the case of compound nouns, names and idioms, while relation between concepts


16

or Relation, are expressed through function words, inflection and word order

(Koponen, 2010: 3)

C. Theoretical Framework

The definitions of translation, machine translation as well as human translator

are used as the basic understanding to the topic being discussed in this research.

After knowing the definitions, the theory proposed by Koponen on error analysis is

used to identify the errors done by Instagram Translate and also human translator.

It is done to answer the first problem formulated. In determining the errors, only

the first category, Individual Concept, is used. This category is represented by

content words, which are verb, adverb, adjective and noun. After that, the result in

the first problem is used to measure the total errors to answer the second problem.

The errors that have been categorized are looked deeper by the researcher to see the

different performance through the errors pattern.


17

CHAPTER III

METHODOLOGY

This chapter presents the methodology used in order to answer two problems

formulated. There are four parts discussed in this chapter, namely area of research,

object of the study, method of the study and research procedure. In the research

procedure, it discusses other four parts namely types of data, data collection,

population and sample, and data analysis.

A. Areas of Research

This study takes the translation of Instagram Translate and professional

translator. It focuses on the texts which are captions translated by Instagram

Translate and human who is a certified translator. It also evaluates the translation

from Indonesian language to English using translation quality assessment. There

are three approaches of translation quality assessment, one of them is target-

language oriented. In accordance with Williams and Chesterman “the relation at

stake is not with the source text but with the target language, equivalence is not a

central concept here. The idea is to measure the translation’s degree of naturalness”

(2002:8).

The researcher aims to find out the mistakes or errors in translating the

captions. By finding the errors, the researcher is able to compare the performances

of IgT and HT.


18

B. Object of the Study

This undergraduate thesis was to compare the translation of Instagram

Translate and human translator. Therefore, the focus was on the target text which

was English. The data collected as the object of the study were captions. They

consisted of word, phrase and sentence taken from an official and certified account

of @basukibtp captions in Instagram. The data were then translated by Instagram

Translate and a professional translator.

C. Method of the Study

This undergraduate thesis applied library and explicatory methods. The

library method “involves identifying and locating sources that provide factual

information or personal/ expert opinion on research question; necessary component

of every other research method at some point” (George, 2008:6). This method was

applied to gather the information and theories of translation, both human and

machine translation, and linguistics aspects in order to be able to analyze the data.

Meanwhile, in accordance with George, the explicatory method

entails a careful, close, and focused examination of a single major text, or of

evidence surrounding a single complex event, in attempt to understand one or

more aspect of it (2008:6).

This method was applied to examine the captions closely and carefully, as the

captions were the focus of this research. It was also done to indicate the errors and

to compare the errors found in the captions. Furthermore, the qualitative research

method was also applied. Qualitative “designates any research whose results are


19

captured in words, images, or non-numeric symbols” (George, 2008:7). It meant

that in analyzing the data, the researcher used descriptive explanation to

demonstrate the findings.

The data in this undergraduate thesis were primary data. In other words, the

data taken were not collected from the other studies or researchers. The data were

collected by the researcher from the caption in @basukibtp official account in

Instagram.

D. Research Procedure

1. Types of Data

The objective data collected in this undergraduate thesis were taken from a

ST and TT. The ST were captions in @basukibtp Instragram’s account, which were

collected from March until September 2016. The TT were the translation from 21

captions collected which had been translated by Instagram Translate and a

professional translator. As the source text, the captions collected consisted of 382

words which not all of them were in the form of sentences but a word. Meanwhile

for the TT, after the captions were translated, the amount of words were differ. The

captions translated by Instagram Translate consisted of 424 words and human

translator obtained more words which were 478.

2. Data Collection

In collecting the data for this undergraduate thesis, the researcher took several

steps in order to get completed and sufficient data. Firstly, the researcher chose a

verified account in Instagram. The fact that Instagram Translates’ result would be

analyzed, it was a must for the researcher to take texts from captions in Instagram.


20

In order to have a reliable caption, a verified account was selected. A verified

account, @basukibpt, showed that the owner follows the Community Guidelines.

Secondly, the captions collected were taken from March 2016 until

September 2016. They were collected from March 2016 since the See Translation

button in that account only started its availability from February. However, the

captions from February were not taken due to unavailability some captions to show

See Translation button. The fact that @basukibtp created the posting randomly, it

meant that the account owner had no regular schedule in uploading the posts.

Therefore, the researcher did not take all the caption created. The researcher decided

to collect the data in the first, middle and the end of the months in order to have the

same number of data taken in each month.

Thirdly, after all the caption collected, the researcher tapped the captions See

Translation button in Instagram to see the translated captions. Because this

undergraduate thesis aimed to compare human and machine translation result, the

next step done by the researcher was to ask a professional translator to translate all

those 21 captions. The person asked to translate was a professional translator who

had been professionally dealing with translation since 2010. She was working in an

international translation house. Afterwards, the translation done by human and

machine was put in the table. The table was arranged in such a way so that each

caption would be side by side with its translation. The table was displayed as

following.


https://help.instagram.com/477434105621119

21

Table 1 Example of the Data Coding

No. ST No. Instagram

Translate No.

Human

translation

1/ST/II/25

Disambangi

Ganjar

Pranowo

dan Ridwan

Kamil di

Balaikota

DKI

1/TT/II/25/IG

Disambangi

ganjar

pranowo

and ridwan

kamil in

central unit

1/TT/II/25/H

Got Visited

by Ganjar

Pranowo

and Ridwan

Kamil at

DKI’s City

Hall

The code can be read as follows

1 : The number of the caption

ST : Source Text

II : The month of the caption created, II means February

25 : The date of the caption created

TT : Target text

IG : The caption are translated by Instagram Translate

H : The caption are translated by human translator

3. Population and Sample

According to Hanlon and Larget, a population “is all the individuals or units

of interest” to a researcher. The population taken was a total of individual that has

certain characteristic in dealing with the researcher’s topic research. The data

population of this undergraduate thesis were all captions collected from February

2016 until September 2016. However, the researcher found that the caption created

by @basukibtp in the Instagram account did not share the same number of postings

every month. Therefore, sample was needed in this research.


22

In accordance with Hanlon and Larget, a sample “is a subset of the

individuals in a population”. Sample was selected from population to represent the

population as the larger set. In this research, the purposive sample was chosen. As

cited from (research-methodology.net), purposive sampling is “a sampling

technique in which researcher relies on his or her own judgment when choosing

members of population to participate in the study”.

All the captions collected had the same characteristic to the population.

Therefore, instead of taking all the captions, the researcher used the purposive

sampling in deciding captions to select. For the result, the researcher took 3 captions

each month, starting from the first, the middle to the end of the month. For the

caption picked as the first month, it should be the caption that was posted in the

very beginning in the month and for the end of the month caption should be posted

in the very late of the month. Meanwhile, for the middle month, the caption taken

should be posted exactly or nearly the date of 15. This sample was conducted to

avoid dissimilarity amount of caption collected from February until September.

When conducting the sample purposive, the researcher found some captions In

February did not show the See Translation button. Therefore, the sample was 21

captions, taken from March until September 2016.

4. Data Analysis

There were two steps used in order to answer problems formulated. First was

to find the error done by Instagram Translate and the professional translator. The

researcher used Koponen’s error categories focusing on individual concepts. The

error categories based on Koponen’s journal were consisted of Omitted Concept,


23

Added Concept, Untranslated Concept, Mistranslated Concept, Substituted

Concept and Explicitated Concept. After analyzing the data, every datum would be

grouped based on its categories.

As stated above, that the focus of this research was on the first category of

error analysis. It was Individual Concept. The consideration in choosing the first

category because Instagram Translate was a relatively new feature of translation

tool which just been launched in 2016. Meanwhile, the data collected and analysis

were also started at the same year. The researcher believes that it was not suitable

to expect such a new tool to perform immaculately. For the second category needed

more advanced translation result. Therefore, the category used was only the first

one since it was focus on the individual concept, how one concept was conveyed in

another language. In addition, the area to analyze was also limited, content words.

Data that had been collected were categorized and examined based on the

result of both translation. In order to find out the source of error, the analysis process

requires alternative translation so that the error can be compared. The alternative

given was needed to give an organized analysis of each error. However, the

alternative translation was not given side by side to the translation result done by

either IgT or HT. It would be given in the explanation. To give a clearer explanation,

the example was on the table below.

Table 2 Description of error in Instagram Translate

No. ST No. TT Type of

Errors

6/ST/I

V/29

Peresmian RPTRA di

menteng dalam, tebet…

6/TT/I

V/29/

IG

Grand opening in rptra

in menteng, tebet.

Omitted

Concept-noun


24

Table 3 Description of error in Human Translator


Errors

7/ST/

V/1

Kondangan tadi siang di

Cipinang Lontar …

7/TT/

V/1/H

Attending a wedding

reception in Cipinang

Lontar …

Omitted

Concept-adverb

In order to answer the second problem of this undergraduate thesis, the errors

found in the captions would be counted according to each category. The errors

found by Instagram Translate would be separated from the errors made by human

translator. After all errors were counted as each subcategory, the number would be

summed up as the total number of errors in both machine and human translator.

This number would lead the researcher draws a pattern of errors and help researcher

to consider the conclusion of this research.


25

CHAPTER IV

ANALYSIS RESULTS AND DISCUSSIONS

In this chapter, the discussion is presented as the result of data analysis. The

discussion covers the two problems formulated in the previous chapter. The first

one covers the errors in each category. There are two big classes of analysis result

from the first problem namely errors done by IgT and the latter is done by HT. Each

class has other categories error which are followed by table or diagram before its

elaboration. As for the following problem, the discussion is based on the first

problem result. It is done in order to look at the pattern of typical errors made by

different translation.

A. The Errors Found in @basukibtp Instagram Captions

The first discussion in this analysis is to find the errors in Instagram captions

translated by Instagram Translate. All captions translated into English are analyzed

and categorized according to Koponen’s category on error analysis. As explained

in the previous chapter, the individual concept error which has six subcategories is

the focus in this research. In order to see the number of each error subcategory,

table and diagram are provided before discussion. In this chapter, every error found

is not discussed. However, some examples are given to present each subcategory

error.


26

1. The Translation Errors by Instagram Translate Found in @basukibtp

Captions

The researcher finds out that errors made by IgT only occur in five out of six

categories. They are Omitted Concept, Added Concept, Untranslated Concept,

Mistranslated Concept, and Substituted Concept.

a. Omitted Concept

A translation error is categorized as omitted concept when “ST concept that

is not conveyed by the TT” (Koponen, 2010: 4). It means that the concept from ST

does not exist in TT. The absence of the concept has a huge possibility to cause the

information in TT incomplete and make the meaning in ST not totally delivered. In

the individual concept, the words analyzed are the content words (Koponen, 2010:

3). In this category, the errors are only found in three subcategories, namely Omitted

Concept of noun, verb, and adjective. IgT makes 15 out of 54 of errors found. The

most frequent error existed in Omitted Concept is the omission of noun which is

followed by adjectives and verb. The table below shows that the errors in Omitted

Concept consist of 11 nouns, 3 Adjectives, and 1 verb.

Table 4 Omitted Concept by Instagram Translate

11

1

3

00

2

4

6

8

10

12

Noun Verb Adjective Adverb


27

i. Omitted Noun

In Omitted Concept of noun, there are 11 errors found. The table below shows

the representative of Omitted Noun errors.


Error

11/ST/

VI/14

… dua cumi cumi datang

mengganggu …

11/TT/

VI/14/

IG

… two squid with

come bother …

Omitted

Concept-

plural noun

15/ST/

VII/19

… Ayo daftar magang di

Kantor Gubernur DKI …

15/TT

/VII/

19/IG

… Let's list of

internships in the

office of Jakarta …

Omitted

Concept-noun

The concept of plural noun as seen in the datum 11/ST/VI/14 is not conveyed

by IgT. In the TT, the noun cumi-cumi or squids is preceded by determiner two

which indicates plurality. However, IgT fails to translate one of the indications of

noun plurality by omitting the letter s in the noun “squid”. According to Oxford

Dictionary, the plural squid can be written as squid or squids. The researcher takes

the second spelling which is squids to examine this datum. The fact that a noun is

modified by determiner (number) makes the result of IgT’s translation irrelevant to

the noun mentioned. Therefore, the translation of dua cumi-cumi should be two

squids.

As for the datum 15/ST/VII/19, the noun Gubernur or governor is not

mentioned in the TT. As in ST, the word Gubernur has a function to specify the

“office” as its head. Even though the word Gubernur performs as a modifier in ST,

but IgT is considered failed to translate a word Gubernur as the word itself stands

as a noun. In ST, the caption tells about the Governor’s Office while IgT only


28

translates the “office” which leaves the information incomplete. The omission error

causes generalization for the TT readers since the office mentioned is not specific

yet. Therefore, the translation of Kantor Gubernur should be Governor’s office.

ii. Omitted Verb

In the Omitted Concept, there is only 1 error found in Omitted Verb. The table

below shows the only datum found as error.


Error 13/ST/

VII/5 Pemerintah

menetapkan 1 Syawal

1437 H jatuh pada

esok hari, …

13/

ST/VII

/5/IG

The Government

Assign 1 Shawwal

1437 h on tomorrow,

…

Omitted

Concept-

verb

In the datum 13/ST/VII/5, the verb jatuh is not conveyed in the TT. In the ST,

the verb expresses an action as the main verb to convey the core information in the

datum above. This error happened is not caused by the incompetent of IgT to find

the equivalence translation of word jatuh since IgT can translate the exact same

word correctly as “fall”. Therefore, the omission affects, of course, the ST intended

meaning. The result in TT brings incomplete information that make the inequality

exist in ST and TT. Therefore, this verb which plays as an action verb should not

be omitted but translated into “falls” or “is”.

iii. Omitted Adjective

In the Omitted Adjective, there are 3 errors found. However, there is only 1

datum is given below as a representative of 3 errors of Omitted Adjective.


29


Error

8/ST/V

/15

...Jakarta Barat

daerahnya kotor…

8/TT/

V/15/

IG

...Jakarta area

absolutely gross…

Omitted

Concept-

adjective

The concept of adjective Barat or west is not conveyed in TT. Meanwhile,

the function of the adjective is to modify the noun, Jakarta. It is supposed to describe

which Jakarta the ST refers to since Jakarta is divided to 5 regions, namely North,

West, South, East and Central Jakarta. By omitting the adjective, the noun becomes

more general than it is intended to be. The TT has lost the additional information of

the Jakarta referred to, which is the West one. This omission does not cause much

changes in the sentence, but the intended specific meaning from ST. Therefore, the

word Barat should not be omitted but translated into West or left untranslated

because it is a proper name.

b. Added Concept

Added Concept is an error that exists when “TT concept that is not present in

the ST” (Koponen, 210:4). In means that the TT adds information which is not

intended or presented by ST. As the result, IgT makes 6 errors of Added Concept.

The most occurred error is in addition of noun category and the rest categories just

have slightly different in number of errors.

Unlike the Omitted Concept category, Added Concept has less number of

errors and sub categories. However, the most frequent error existed is similar to the

Omitted Concept, which is noun category. It has 50% from 6 errors found. The table


30

below indicates that the Added Concept error found consists of 3 nouns, 2 adverbs,

and 1 adjective.

Table 5 Added Concept by Instagram Translate

i. Added Noun

The errors found in this concept are 3. The errors exist in datum 4/ST/IV/1,

7/ST/V/1, and 12/ST/VI/30. The table below shows one of the errors found.


Error

12/

ST/VI/

30

Amin, terima kasih atas

ucapan dan doanya ...

12/TT/

VI/ 30/

IG

Amen, thank you for

the birthday wishes …

Added

Concept-

noun

In this case, at a glance, the addition of the word “birthday” might be said not

totally wrong. This is due to the fact that the day when the posting is uploaded, that

is also one day after @basukibtp’s birthday. The additional of the word “birthday”

can be considered relevant since the wishes that refer to are actually birthday

wishes.

However, the word “birthday” or hari ulang tahun in the ST does not exist

and appears in the TT. In this case, by only looking at the caption or without the

picture, the addition of “birthday” is considered as an error. IgT fails to translate its

3

0

1

2

0

0.5

1

1.5

2

2.5

3

3.5



31

intended meaning from ST, which is “wishes” or doanya. The additional of

“birthday” has no relation to the word wishes because wishes alone is not the pair

of the word birthday, such as the words “not only” to its pair “but also”.

As the result, the word “birthday” is categorized as an error due to the analysis

focuses on the caption. There is an additional information that does not present in

the ST. Therefore, the translation of terima kasih atas ucapan dan doanya should

be “thank you for the greetings and wishes”.

ii. Added Adjective

There is only 1 error found in Added Adjective. The datum 18/ST/VIII/23 is

given to give a clearer comparison between message in ST and TT.


Error

18/ST/V

III/23

Pagi tadi bersama warga

di peresmian Ruang

Publik Terpadu Ramah

Anak (RPTRA) …

18/TT

/VIII/

23/IG

This morning with the

residents at the

national integrated

public space (child

friendly) …

Added

Concept-

adjective

The word “national” according to Oxford Dictionary, is either “owned or

controlled by government” or “connected with a particular nation”. However, in

the ST, that concept does not exist but appears in the TT. The addition of the word

“national” in TT creates meaning for the TT readers that the concept of national

plays a role as the modifier of RPTRA (Ruang Publik Terpadu Ramah Anak) or

Child Friendly Integrated Public Space in the caption. It adds the concept that the

RPTRA is owned/controlled by government. Meanwhile, the concept is absent in

ST. Consequently, the concept of adjective national should not be present in the TT.


32

iii. Added Adverb

There are 2 errors found in this concept. The first error is found in datum

8/ST/V/15 and the last one is in datum 11/ST/VI/14. The table below shows one of

the errors mentioned.


Error

8/ST/

V/15

Dapat laporan warga

kalau Kapuk, Jakarta

Barat daerahnya kotor …

(RPTRA) …

8/TT/

V/

15/IG

Can citizens report if

ceiba, Jakarta area

absolutely gross …

Added

Concept-

adverb

The word “absolutely” in the datum 8/ST/V/15 shows that the concept of

adverb is added in the TT. As seen in the ST, the adverb “absolutely” is not found

in ST. However, IgT adds the adverb to describe the adjective “gross” in the

translation result. According to Oxford Dictionary, absolutely is “used with

adjectives or verbs that express strong feelings or extreme qualities to mean

extremely”. The addition, in fact, gives unintended information as an emphasis to

the word gross. As for the TT readers, this addition delivers the stronger emphasis

compare to the ST intended. Moreover, the description of noun, Kapuk in West

Jakarta, becomes dissimilar.

c. Untranslated Concept

Deriving from its name, Untranslated Concept is an error that happens when

“SL words that appear in TT” (Koponen, 2010: 4). It means that the word from ST

is kept untranslated in the TT. Instagram Translate makes 4 errors in Untranslated

Concept, which 3 of them are in untranslated of noun and 1 error in untranslated

verb. The Untranslated Concept has less errors compare to the other categories.


33

From all the errors made by IgT, this concept makes 7.40% errors out of 54. The

table shows that there are 3 nouns and 1 verb found in Untranslated Concept error.

Table 6 Untranslated Concept by Instagram Translate

i. Untranslated Noun

In the Untranslated Concept, there are 3 errors found. The errors exist in data

1/ST/III/1, 3/ST/III/27, and 11/ST/VI/14. The table below shows the error

representative of this category.

No. ST No. TT Type of Error

1/ST/I

II/1

Bersama Keluarga

Besar Damkar

1/TT/I

II/1/IG

With big family

damkar

Untranslated

Concept-noun

In the datum 1/ST/III/1, the word Damkar appears as it is in TT. Damkar as

the shortening word from Pemadam Kebakaran or Firefighter is left untranslated

by IgT. As a blended word, IgT is considered failed to find the equivalence form in

TT. Hence, when it happens to translate the words as Pemadam Kebakaran, IgT

shows the better translation result. From this case, shortening words are not

recommended to make a good translation by IgT.

3

1

0 00

0.5

1

1.5

2

2.5

3

3.5



34

ii. Untranslated Verb

Similar to the category of Omitted Concept, there is only 1 error found in the

verb category. The Untranslated Verb is the word dikasih in datum 4/ST/IV/1. The

table bellows displays the error mentioned.


4/ST/I

V/1

Ekspresi ketika

mengharap dikasih

tempat duduk

4/TT/I

V/1/IG

Expression when hope

dikasih seats

Untranslated

Concept-verb

In ST, the word dikasih appears in the TT. IgT fails to translate the passive

voice in the datum 4/ST/IV/1. The verb should be translated into “is given” or “to

be given”. As there are two verbs in the datum above namely mengharap or

expecting as an active verb and dikasih as a passive verb, the error found could be

caused by the inability to translate the passive verb since the active one is translated.

IgT fails to see both verbs to be translated. For dikasih is another verb in datum

which is untranslated, it disables the TT readers to gain the information of the

caption as a whole. The translation should be “The expression when expecting to

be given a seat”.

d. Mistranslated Concept

Mistranslated Concept is an error when “a TT concept has the wrong meaning

for the context” (Koponen, 2010: 4). It means the TT gets the wrong concept from

the intended concept in ST. In this research, IgT translation gives the out of the

context information from what the ST intends to convey. There are 4 sub categories


35

of error exist in Mistranslated Concept, namely Mistranslated Concept of noun,

verb, adjective and adverb.

IgT makes 24 errors in Mistranslated Concept. The noun error sub category

still shows the highest number of errors. Then, it is followed by verb, adverb, and

adjective sub category as the least occurred errors. It is seen in the table below that

in Mistranslated Concept, there are 14 nouns, 5 verbs, 2 adjectives, and 3 adverbs

found as errors.

Table 7 Mistranslated Concept by Instagram Translate

i. Mistranslated Noun

The most occurred error is found in Mistranslated Concept. From 24 errors

found in Mistranslated Concept, 14 of them exist in Mistranslated Noun. The table

below shows the error as the representative of Mistranslated Noun.


3/ST/III/

27 … di daerah Koja …

3/TT/II

I/27/

IG

… in the area of tree … Mistranslated

Concept-noun

In the datum 3/ST/III/27 the noun Koja is mistranslated. The intended context

by ST should not be translated, due to the fact that Koja is a proper noun. However,

IgT mistranslates the concept into “tree”, which is completely not suitable for the

14

5

23

0

2

4

6

8

10

12

14

16



36

concept in the ST. By looking at the mistranslation of Koja as a sub district in

Jakarta into “tree”, IgT is considered failed to pick the proper meaning for a word

that has polysemous meaning. IgT picks the meaning of Koja as a tree for the reason

that Koja is also kind of leaf that is used to cook. In here, Koja refers to salam koja

or in Indonesia, it is also called daun kari (curry leaves). However, instead of

translating Koja as curry leaves, IgT translates the superordinate of it, which is tree.

This mistranslation, of course, creates the wrong information for the TT readers.

ii. Mistranslated Adverb

There are 3 errors found in the Mistranslated Adverb. The errors exist in data

number 2/ST/III/15, 20/ST/IX/17, as well as 21/ST/IX/30. The table below consists

of mistranslated concept in one of the adverbs found.


2/ST/II

I/15

… Sekarang bikin SIUP

dan TDP online dan

simultan …

2/TT/II

I/15/IG

… Now to make a trade

operation permit and the

company registration

number online and

simultaneous …

Mistranslated

Concept-adverb

IgT mistranslates the adverb simultan or “simultaneously” into

“simultaneous”. In the datum 2/ST/III/15, the concept of simultan or

“simultaneously” is supposed to be an adverb that modifies the verb, explaining

how to do a registration. In addition, there are two adverbs in the caption above

namely “online” and simultan. From this case, IgT fails to recognize the second

adverb and fails to translate the second adverb as it is intended in ST. Therefore,

the translation of adverb simultan should be translated into “simultaneously”.


37

iii. Mistranslated Verb

In Mistranslated Verb, there are 5 errors exist. The table below gives 2 data

to show the errors found.


8/ST/

V/15

Dapat laporan warga...

8/TT/

V/15/

IG

Can citizens report …

Mistranslated

Concept-verb

15/ST/

VII/19

… Ayo daftar magang di

Kantor Gubernur DKI …

15/TT/

VII/19

/IG

… Let's list of internships

in the office of Jakarta …

Mistranslated

Concept-verb

The mistranslated concept of verb in the datum 8/ST/V/15 exists in the word

dapat (mendapatkan) or “to receive” or “to get”. This happens because in Indonesia

language, dapat means both “to receive” and also “can” or “be able to”. Looking at

the translation result, IgT fails to differentiate which intended meaning the ST

would like to convey. Therefore, dapat is translated into “can” instead of “to

receive”.

As in datum 15/ST/VII/19, the mistranslated concept is also found. The word

daftar (mendaftar) or “to apply” is translated into the noun of daftar which is “list

of”. According to Kamus Besar Bahasa Indonesia, daftar is a noun and only stands

as a noun. However, in ST, it plays a role as a verb. From this phenomenon, the

mistranslation of verb is supported by the mistake made by ST. The word daftar

(noun) should be written as mendaftar (verb) if it has to stand as a verb. Therefore,

IgT as a machine translate, certainly fails to read the intended meaning.


38

iv. Mistranslated Adjective

The least occurred error in Mistranslated Concept in Mistranslated Adjective.

The errors found are 2. They exist in data number 10/ST/VI/1 and 21/ST/IX/30.

The table below shows one of them.


Error

21/ST/I

X/

30

… Mohon maaf

pelaksanaan MRT ini

membuat Jakarta tambah

macet …

21/

TT/

IX/30/I

G

… Sorry for the

implementation of this

make mrt jakarta add

crashed ...

Mistranslated

Concept-

adjective

Datum 21/ST/IX/30 shows that adjective macet or “jamming” is

mistranslated. The context presented in the ST is macet that refers to a condition

where vehicle cannot move or traffic jams in Jakarta. However, the result of

translation shows the message of having “crashed” which related to vehicle and its

accident. TT provides the wrong information of the intended concept, “jamming”.

Even though both macet and “crashed” are related to vehicle, still, an accident as

described in TT is absent in ST. Therefore, macet should be translated into “traffic

jams” so that the intended meaning of macet is conveyed.

e. Substituted Concept

An error is categorized as Substituted Concept when “TT concept is not a

direct lexical equivalent for the ST concept but can be considered a valid

replacement for the context” (Koponen, 2010: 4). It means that a translation is not

considered as a completely out of the intended concept but considered to share the

similarity from the ST. This error is not a fatal one since it has the valid replacement

concept as a whole. In this category, IgT makes 2 sub categories error which are

verb and noun. Unlike the other categories, Substituted Concept has more number


39

of errors in the verb rather than noun category. From all the category made by IgT,

Substituted Concept has 9.25% of errors. The table below shows that Substituted

Concept errors found are 2 nouns and 3 verbs.

Table 8 Substituted Concept by Instagram Translate

i. Substituted Verb

There are 21 data collected from Instagram’s captions and 3 of them have the

word kondangan which all are found as Substituted Verb error category. The table

below shows one of the words, kondangan, in the datum 8/ST/V/15.


8/ST/

V/15

… Hari ini pilih

kondangan ke daerah

Kapuk …

8/TT/V

/15/

IG

… Today choose

invitation to ceiba …

Explicitated

Concept-verb

One of the Substituted Verb exists in datum 8/ST/V/15. The verb kondangan

is translated into “invitation”. At first glance, the result might be categorized as

Mistranslated Concept. However, if it is analyzed deeper the result of IgT is not out

of the intended concept from ST. The context is not wrong but still related.

Moreover, the concept of “invitation” does not bring a new information in TT.

2

3

0

0.5

1

1.5

2

2.5

3

3.5

Noun Verb


40

Hence, the “invitation” and kondangan are regarded experiencing Substituted

Concept. According to KBBI V, kondangan is “pergi menghadiri undangan

perkawinan dan sebagainya (untuk mengucapkan selamat dan sebagainya” or

going to attend wedding invitations and so forth (to congratulate and so forth,

translated by researcher). Meanwhile in TT, the translation of kondangan is

“invitation” which has a more general meaning as it stands as a noun. However, it

still shares the information of inviting as mentioned in the word kondangan. The

fact that the translation is still a valid replacement due to the similarity of

information shared by both ST and TT, makes this error as Substituted Concept.

The translation of kondangan for the intended meaning in the datum above should

be “to attend a wedding reception”.

ii. Substituted Noun

Unlike the other categories which almost the most occurred error is in noun

category, in Substituted Concept, noun category shows the least occurred of errors.

They are 2 errors found in this category. The table below shows one of them.


15/ST/

VII/19

… di Kantor Gubernur

DKI ...

15/TT/

VII/19

/IG

… in the office of

Jakarta ...

Substituted

Concept-noun

The initialism of DKI or Daerah Khusus Ibukota is undergone Substituted

Concept of noun. DKI is supposed to be translated as either DKI which is left

untranslated or Special Capital Territory. However, in the datum above, IgT

translates the initialism into Jakarta. The fact is, in Indonesia, the initialism of DKI

only stands for Jakarta as the capital city of Indonesia. It means that DKI is kind of


41

“title” because this city has its special role for Indonesia. Therefore, DKI precedes

the noun Jakarta in order to show the city is distinct. This makes the form of DKI

always refer to Jakarta. Even though the concept of DKI is not a direct equivalent

for the concept of Jakarta, it is still considered as a valid replacement.

2. The Errors Found in @basukibtp Instagram Captions by Human

Translator

The errors found in Instagram Caption by Human Translator are classified

into 5 categories. The categories are based on Koponen’s theory on error analysis.

There are Omitted Concept, Added Concept, Mistranslated Concept, Substituted

Concept and Explicitated Concept.

Table 9 Omitted Concept by Human Translator

The table above shows errors made by HT, 1 in noun category and 1 in verb

category.

a. Omitted Concept

In the Omitted Concept, there are two sub categories found. There are omitted

adverb and verb. HT makes equal error of verb and adverb. Omitted Concept has

25% errors out of 8 errors made by a professional translator.

1 1

0 00

0.2

0.4

0.6

0.8

1

1.2

adverb Verb Adjective Adverb


42

i. Omitted Adverb

In the Omitted Adverb, there is only 1 error found. The error of this category

only exists in datum 7/ST/V/1. The table below shows the translation done by HT

is categorized as Omitted Adverb.


7/ST/

V/1

Kondangan tadi siang

di Cipinang Lontar …

7/TT/

V/1/H

Attending a wedding

reception in Cipinang

Lontar …

Omitted

Concept-adverb

In the data above, there is an adverb which is omitted, tadi siang or “this

afternoon”. The omission does not change the whole meaning as it is intended in

ST. However, the omitted adverb by HT makes the meaning in TT incomplete

because tadi siang stands as a complement to provide the information when the

event happens. This omission error might not cause confusion but incomplete and

different interpretation for TT readers. Therefore, the adverb tadi siang should not

be omitted but translated into “this afternoon”.

ii. Omitted Verb

It is similar to the Omitted Adverb, the error found is only 1. The table below

shows error exists in datum 4/ST/IV/1.


4/ST/I

V/1

Ekspresi ketika

mengharap dikasih

tempat duduk

4/TT/I

V/1/H

instagram translate’s and human translation’s … · 2018. 2. 15. · instagram translate'sand...

Documents