07.80.0019 Srestha F

Perpustakaan Unika
An Analysis of the Translation from English Sentences to

Indonesian Sentences by Using Google Translate
A THESIS
by:
Srestha Fem
Student Number: 07.80.0019
ENGLISH LETTERS STUDY PROGRAMME
FACULTY OF LETTERS
SOEGIJAPRANATA CATHOLIC UNIVERSITY
SEMARANG
2011
Perpustakaan Unika

A Thesis Presented as Partial Fulfilment of the Requirements to Obtain the
Sarjana Sastra Degree in the English Letters Study Programme
by:
Srestha Fem
ENGLISH LETTERS STUDY PROGRAMME

FACULTY OF LETTERS
SOEGIJAPRANATA CATHOLIC UNIVERSITY
SEMARANG
2011
Perpustakaan Unika
A THESIS ON

by
Srestha Fem
Approved by:
Drs. Y E Budiyana, MA
Major Sponsor
June __, 2011
Emilia Ninik Aydawati, SP. M.Hum

Co-Sponsor
June __, 2011
Perpustakaan Unika
A thesis defended in front of the board of Examiners on June 20, 2011

and declared acceptable.
BOARD OF EXAMINERS
Chair Person : Drs. Y E Budiyana, MA
........................
Secretary
: Emilia Ninik Aydawati, SP. M.Hum
........................
Member
: Heny Hartono, SS., M. Pd.
........................
Semarang, June __, 2011

Faculty of Letters
Soegijapranata Catholic University
Dean,
Heny Hartono, SS., M. Pd.

058.1.1998.221
Perpustakaan Unika
ACKNOWLEDGEMENT
My very first thanks will be dedicated for my Lord who resides in heaven.
Thank You for the guidance which you give me for these whole twenty one years
of my life. You have given me the most amazing life I could ever have by putting
me around these people in my family and my friends. You give me strength and
comfort during all of the hard times. I feel so lucky and blessed to have such
Almighty and Omnipotent God as You.
Second, I would like to give thanks to Drs. Y E Budiyana, MA and Emilia
Ninik Aydawati, SP. MHum and appreciate them for the incredibly wonderful
guidance and support. I am very conscious of my weakness in every subject,
especially grammar, and completing this thesis is one of my biggest
achievements. I really think that I will not be able to write this thesis without your
helps and advices. For that reason, I give you my deepest thanks.
The next is for my family A. Sandren and Pudjiwati, Thanks for giving
me a time to complete my thesis. I am very thankful for it. For my sister,
Saraswathi, Thank you for not bothering me too much when I made this thing,
it's really helpful"
The last, but not the least, I would like to thank all members of Mbzzttz
family: Andreas Wijaya, Alliance Tantosa, Agatha Astrid, Cicilia Kintariwati, F.A
Inez H, Julius Ardi, Marsha Marieanne, Marcelly, Ryan Sandjaya, and the other
members which are unlisted here. "I guess being friends with you guys is enough
to make me forget all of my problems, NB: just forgetting, not clearing. For
i
Perpustakaan Unika
'Madam' Angelika Riyandari, "Thanks for watching and listening to my disastrous

life from Australia " For James Eko Nugroho, Vincentius Marco, Ferdyanto
Handoyo, Christian 'Anzai' Yuwono and Ferdyanto Handoyo, of course with their
respective girlfriends, "Thank you for Sticking up with me since high school". For
Kedo Arya who accompanies and being accompanied for consultation time during
the making of this thesis "at last, right?" For the person I dare not say the name,
"Sorry I cannot fulfill my promise to you, well at least I only fall a step behind
you, and I'll catch up to you for sure, so wait for me". And for all the people
around me that I barely mention their names one by one, Thank you, I guess?
ii
Perpustakaan Unika
TABLE OF CONTENTS
ACKNOWLEGEMENT
...................................................................................i
TABLE OF CONTENTS
................................................................................iii
LIST OF APPENDIX
................................................................................vii
ABSTRACT .........................................................................................................x
ABSTRAK
.........................................................................................................xi
CHAPTER I INTRODUCTION
......................................................................1
1.1
Background of the Study
..........................................................1
1.2
Field of the Study
......................................................................2
1.3
Scope of the Study
......................................................................2
1.4
Problem Formulation ......................................................................2
1.5
Objective of the Study
..........................................................3
1.6
Significance of the Study
..........................................................3
1.7
Definition of Terms
......................................................................4
CHAPTER II REVIEW OF LITERATURE .........................................................5

2.1
Translation
..................................................................................5
2.1.1
Definition of Translation
..........................................................5
2.2
Computational Linguistics
..........................................................6
2.2.1
Trans Tool
2.2.1.1
Early Trans Tool
......................................................................8
2.2.1.2
Present Trans Tool
......................................................................9
2.2.1.3
Types of Trans Tool ....................................................................10
..................................................................................7
iii
Perpustakaan Unika
2.3.
Effective Translation. ....................................................................13
2.4
Sentences
................................................................................19
CHAPTER III METHOD OF DATA COLLECTION AND ANALYSIS ........21

3.1
Method of Data Collection
........................................................21
3.1.1
Material
................................................................................21
3.1.2
Instrument
................................................................................22
3.1.3
Procedure
................................................................................22
3.2
Method Data Analysis
3.2.1
The Parameters
3.2.2
The evaluation of Translation Result ............................................24
........................................................24
....................................................................24
CHAPTER IV DATA ANALYSIS ....................................................................26

4.1
ACADEMIC SENTENCES ........................................................26
4.1.1
Simple Sentences
4.1.1.1
Word Level Equivalence
4.1.1.2
Above Word Level Equivalence
4.1.1.3
Grammatical Equivalence
4.1.1.4
Textual Equivalence ....................................................................32
4.1.2
Compound Sentences ....................................................................33
4.1.2.1
4.1.2.2
4.1.2.3
4.1.3
Complex Sentences
4.1.3.1
....................................................................26
........................................................27
........................................................32
........................................................33
............................................38
........................................................42
....................................................................42
........................................................43
iv
............................................30
Perpustakaan Unika
4.1.3.2
4.1.3.3
4.1.4
Compound Complex Sentences
4.1.4.1
4.1.4.2
4.1.4.3
4.2
NON ACADEMIC SENTENCES
4.2.1
Simple Sentences
4.2.1.1
4.2.1.2
4.2.1.3
4.2.2
Compound Sentences ....................................................................65
4.2.2.1
4.2.2.2
4.2.2.3
4.2.3
Complex Sentences
4.2.3.1
4.2.3.2
4.2.3.3
4.2.4
Compound Complex Sentences
4.2.4.1
4.2.4.2
4.2.4.3
............................................51
........................................................51
............................................57
........................................................60
............................................61
....................................................................61
........................................................62
............................................62
........................................................64
........................................................65
............................................68
....................................................................70
........................................................71
............................................73
............................................77
........................................................78
............................................48
............................................80
........................................................85
Perpustakaan Unika
4.2.4.4
4.3
General Findings
....................................................................87
CHAPTER V CONCLUSION AND SUGGESTIONS
................................92
5.1
Conclusion
................................................................................92
5.2
Suggestions
................................................................................92
BIBLIOGRAPHY
............................................................................................94
APPENDICES
vi
Perpustakaan Unika
LIST OF APPENDIX
Appendix 1:
Table 4.1.1 Translation Mistake in Academic Simple Sentences
Appendix 2:
Table 4.1.1.1 Word-level Mistake in Academic Simple

sentences
Appendix 3:
Table 4.1.1.2 Above word-level Mistake in academic simple

sentences
Appendix 4:
Table 4.1.1.3 Grammatical Mistake in academic simple

sentences
Appendix 5:
Table 4.1.1.4 Textual Mistake in academic simple sentences
Appendix 6:
Table 4.1.2.1 Translation Mistake in Academic Compound

Sentences
Appendix 7:
Table 4.1.2.1 Word-level Mistake in academic compound

sentences
Appendix 8:
Table 4.1.2.2 Above word-level Mistake in academic compound

sentences
Appendix 9:
Table 4.1.2.3 Grammatical Mistake in academic compound

sentences
Appendix 10:
Table 4.1.3 Translation Mistake in Academic Complex

Sentences
Appendix 11:
Table 4.1.3.1 Word-level Mistake in academic complex

sentences
vii
Perpustakaan Unika
Appendix 12:
Table 4.1.3.2 Above word-level Mistake in academic complex

sentences
Appendix 13:
Table 4.1.3.3 Textual Mistake in academic complex sentences
Appendix 14:
Table 4.1.4 Translation Mistake in Academic Compound Complex Sentences
Appendix 15:
Table 4.1.4.1 Word-level Mistake in academic compoundcomplex sentences
Appendix 16:
Table
4.1.4.2
Above
word-level
Mistake
in
academic
compound-complex sentences
Appendix 17:
Table 4.1.4.3 Grammatical Mistake in academic compoundcomplex sentences
Appendix 18:
Table 4.2.1 Translation Mistake in Non- Academic Simple

Sentences
Appendix 19:
Table 4.2.1.1 Word-level Mistake in Non-academic simple

sentences
Appendix 20:
Table 4.2.1.2 Above word-level Mistake in Non-academic

simple sentences
Appendix 21:
Table 4.2.1.3 Grammatical Mistake in Non-academic simple

sentences
Appendix 22:
Table 4.2.2 Translation Mistake in Non- Academic Compound

Sentences
Appendix 23:
Table 4.2.2.1 Word-level Mistake in Non-academic compound

sentences
viii
Perpustakaan Unika
Appendix 24:

compound sentences
Appendix 25:
Table 4.2.2.3 Textual Mistake in Non-academic compound

sentences
Appendix 26:
Table 4.2.3 Translation Mistake in Non- Academic Complex

Sentences
Appendix 27:
Table 4.2.3.1 Word-level Mistake in Non-academic complex

sentences
Appendix 28:

complex sentences
Appendix 29:
Table 4.2.3.3 Textual Mistake in Non-academic complex

sentences
Appendix 30:
Table 4.2.4 Translation Mistake in Non- Academic Compound

Complex Sentences
Appendix 31:
Table 4.2.4.1 Word-level Mistake in Non-academic compound

complex sentences
Appendix 32:

compound complex
Appendix 33:
Table 4.2.4.3 Grammatical Mistake in Non-academic compound

complex
Appendix 34:
Table 4.2.4.4 Textual Mistake in Non-academic compound

complex
ix
Perpustakaan Unika
ABSTRACT
Trans Tool or Machine Translation (MT) has become commonly used
nowadays. This program provides a fast translation from one language into
another language. There are many kinds of Trans Tool, but the most commonly
used, and easily accessed is "Google Translate", an online Trans Tool. Though it
is very convenient to use, Trans Tool is not flawless. Therefore, in this study the
writer tries to analyze the translation result provided by "Google Translate" to find
out the problems with the translation result provided by the Trans Tool. The writer
tries to figure out the effectiveness of "Google Translate" in translating English
sentence into Indonesian sentence. This becomes problem, since the translation
result that seem to be correct at the first glance, might still contain a problem.
This study is conducted based on four parameters: word-level equivalence,
above word-level equivalence, grammatical equivalence, and textual equivalence.
As the result, "Google Translate" has a low effectiveness in translating English
sentence into Indonesian sentence, with the highest probability of 5 %, which is
only achieved when translating non academic simple sentence.
Keywords: Translation, Machine Translation (MT), Google Translate
Perpustakaan Unika
ABSTRAK
Trans Tool atau Machine Translation (MT) telah umum digunakan pada
masa sekarang ini. Program ini memberikan hasil terjemahan yang cepat dari satu
bahasa ke bahasa lain. Ada banyak jenis Trans Tool, tapi yang paling sering
digunakan dan paling mudah dijangkau adalah "Google Translate" yang mana
adalah Trans Tool online. Meskipun sangat nyaman digunakan, Trans Tool
tidaklah tanpa cacat. Maka, di studi ini penulis berusaha untuk menganalisa hasil
terjemahan yang diberikan oleh "Google Translate", untuk mengetahui apakah
masalah dari hasil terjemahan yang diberikan oleh Trans Tool. Penulis berusaha
untuk mengetahui seberapa efektif kah "Google Translate" dalam
menterjemahkan kalimat bahasa Inggris ke bahasa Indonesia. Hal ini menjadi
masalah, karena hasil terjemahan "Google Translate", meskipun terlihat benar,
masih mungkin memiliki masalah.
Studi ini didasarkan pada empat parameter: kesamaan tingkat kata,
kesamaaan di tingkat atas kata, kesamaan susunan bahasa dan kesamaan konteks.
Sebagai hasilnya, "Google Translate" memiliki efektifitas yang rendah dalam
menterjemahkan kalimat bahasa Inggris ke bahasa Indonesia, dengan
kemungkinan terbesar 5% yang hanya dicapai ketika menterjemahkan kalimat
sederhana yang tidak akademis.
Kata Kunci: Terjemahan, Terjemahan Mesin (MT), Google Translate
xi
1
Perpustakaan Unika
CHAPTER I
INTRODUCTION
1.1 BACKGROUND OF THE STUDY

Computer is a multifunctional device that can help us in many aspects. One of
the aspects is in translation. There is a program called Trans Tool or known better as
Machine Translation (MT). This Program is a program which is mainly used to
translate a text into another language according to the users will; it can be English to
Indonesian, French to Italian or other language in just a few minutes.
Trans Tool development has started since many years ago. According to
Yusuf (1994: 144) the first man to claim inventing this device is Troyansky, a pioneer
with Russian nationality. After being acknowledged of its advantages, the
development of this device is continued to increase the quality.
In using Trans Tool, the user will get advantages. One of which is that Trans
Tool translates faster than manual translation, and it covers up many languages. Thus,
the user is bound to complete the translation task in no time, and the translation
results are not limited to only a single language.
Despite its advantages, translating using Trans Tool has a certain risk. The
problem lies in the Trans Tool itself. Mostly, Trans Tool makes mistakes in
Translating an ambiguity, wrong context usage, Sometimes Trans Tool even
2
Perpustakaan Unika
translates words that do not even need to be translated, just as stated by Zaharin et. Al
(1994) (as cited in Isabelle, 2002):
The problem existing in fully automated MT systems was most take
the sentence as the basic unit for translation with little or no
reference to the environment (context) in which the sentence is used
This means that Trans Tool is completely blind when it comes to translating texts
which use a special kind of structure or grammar, context, and even ambiguity. These
mistakes commonly happen when the Trans Tools are given a task to translate
sentences.
Normally, after the flaws are detected, they will be immediately repaired, but
it cannot easily be done. Despite its flaws, sometimes the Trans Tool still manages to
translate the text correctly into the desirable target language. Based on the problems
above, this paper discusses about the effectiveness of Trans Tool. It discusses
whether it can translate the text on the sentence level correctly.
1.2 FIELD OF STUDY

This study is in the field of Linguistics
1.3 SCOPE OF STUDY

In this study, the writer focuses on machine translation as one of the studies of
Computational Linguistics. Furthermore the writer only focuses on Google
Translate Trans Tool. The translation which will be discussed in this paper is
English Indonesian translation, and on the level of sentences.
3
Perpustakaan Unika
1.4 PROBLEM FORMULATION

According to the theory of translation used as the main theory in this study,
the writer formulates the problem as follows:
How effective is 'Google Translate' able to translate the sentences from
English to Indonesian sentences?
1.5 OBJECTIVE OF STUDY

Considering the problem mentioned above, the writer formulates this studys
objective as:
To see how effective 'Google Translate' is able to translate the sentences from
English to Indonesian sentences.
1.6 SIGNIFICANCE OF STUDY

The writer expects that this study could be a feedback to the Google
Translate machine Translation system users, about the effectiveness of Google
Translate so that they can be more careful in using Google Translate. The writer
hopes there will be fewer English texts that are translated weirdly because translators
rely solely on the Machine Translation System.
In addition, the writer also hopes that the result of this study could contribute
something useful for both the lecturers and students in the Faculty of Letters in
Soegijaprananta Catholic University.
4
Perpustakaan Unika
1.7 DEFINITION OF TERMS
Google Translate:
Machine translation service which is provided by the

website http://www.google.com
Machine Translation: The term 'machine translation' (MT) refers to

computerized systems responsible for the production
of translations with or without human assistance.
(Hutchins, 2002)
Trans Tool:
Indonesian term of Machine Translation
5
Perpustakaan Unika
CHAPTER II
REVIEW OF LITERATURE
2.1. Translation
Translation has become an important issue since the world began to globalize.
Much information from other parts of the world is now easy to get, but the way
people interpret the information can be a serious issue. Wrong interpretation of the
information can lead to whole new meaning of the information, and of course it will
not give us the desirable result.
The act of translation has already begun since a long time ago. One of the
proofs is the Bible translation activity, and it became a key issue of translation until
the seventeenth century (Bassnett, 1991). The development of translation continues
until right now, the development is also accompanied by the development of many
rules suggested to give us a good translation result.
2.1.1. Definition of Translation

There are many definitions of translation; some of them are as follows:
1. It is rendering the meaning of text into another language in the
very way that the author intended the text. (Newmark, 1988,p. 5)
2. [] Involves the rendering of a source language (SL) text into the
target language (TL) so as to ensure that (1) the surface meaning of
the two will be approximately similar and (2) the structures of the
SL will be preserved as closely as possible but not so closely that
the TL structures will be seriously distorted. (Bassnett, 1991)
6
Perpustakaan Unika
From these two quotations, the writer concludes that Translation is the act of
rendering the Source Language (SL) into Target Language (TL), without changing
the original meaning that is intended by the author.
Translation, based on the process is divided into two kinds:
1. Manual Translation: Translation done manually by human.
2. Machine Translation: Translation done by machine (mostly a computer).
Manual Translation is mostly discussed in the field of Translation while Machine
Translation is discussed in the field of Computational Linguistics.
2.2 Computational Linguistics

This study is within the field of Computational Linguistics, just as stated by
OGrady et. al. (1996, p.665):
Computational Linguistics is a relative new discipline that lies in the
intersection of the fields of linguistics and computer science. It is but one
of many hybrid disciplines involving computers that require
computational expertise as well as a background in another field.
In Simpler terms, this field is the combination of Computer Science and Linguistics
which still carries the characteristics of each field.
Within this field there are also other fields, for example: Computational
Morphology, Computational Syntax, Computational Semantic, etc. In this study the
writer will discuss one of the fields within Computational Linguistics, which is
Machine Translation.
7
Perpustakaan Unika
2.2.1 Trans Tool

Trans Tool or known better as Machine Translation, is a program which is
used to translate a certain text from 1 language into another language. Trans Tool can
translate any language into any language, depending on the Trans Tool itself.
Basically the Trans Tool or Machine Translation work routine is just like what
OGrady (1996, p. 696) presents in the following:
Source Language Machine Translation System Target Language
The source language which the user wants to translate is typed in the Trans Tool or
Machine Translation System. After the text is typed, the user can select the command
to translate the text in the program. After the translate command is selected, the
program will begin to translate the text, the result will be displayed after a short
period of time. This work routine is the basic work routine that is used by any Trans
Tool, despite its type.
Bilingual Trans Tool is the oldest type of Trans Tool; Hutchins (1999, p.2)
states that:
For many years, the systems were based primarily on direct
translations via bilingual dictionaries, with relatively little detailed
analysis of syntactic structures.
This means that the bilingual Trans Tool in its time is very unreliable since it relies
on the bilingual dictionary to give translation result.
After several years of the development, the Trans Tool has undergone a lot of
improvement. One of them is the development of multilingual translation feature,
8
Perpustakaan Unika
One of the example called Systran system is provided by Hutchins (1999). He states
that the Systran system, which is originally designed for translation only from
Russian into English, is now available for a very large number of language pairs:
English from and into most European languages (French, German, Italian, Spanish,
Portuguese), Japanese, Korean, etc. The Trans Tool which has multilingual
translation features like this is called: Multilingual Trans Tool
2.2.1.1. Early Trans Tool

In the beginning of the development of Trans Tool, which according to
Hutchins (1995, p.1) is in the early 1950s, research was necessarily modest in its
aims. It was constrained by the limitations of hardware, in particular inadequate
memories and slow access to storage, and the unavailability of high-level Language
programming. What makes it worse is that it did not look for assistance from the
language experts. Syntax was a relatively neglected area of linguistic study and
semantics was ignored. Hutchins (1995) blames this problem to the behavior of the
leading scholars. It was not surprising that the first Trans Tool researchers turned to
crude dictionary based approaches, or in simpler terms; word for-word translation.
With such limitations, early researchers set out with modest aims. They knew that
whatever systems they could develop would produce low quality results, and
consequently they suggested the major involvement of human translators both for the
pre-editing of input texts and for the post-editing of the output, and they proposed the
9
Perpustakaan Unika
development of controlled languages and the restriction of systems to specific domains.

Hutchins (1995, p.1) states that:
Above all, they proposed that MT systems could progress by the cyclical
improvement of imperfect approaches, i.e. an application of the
engineering feedback mechanism with which they were familiar.
They in this quotation refers to the researchers at that time. Those researchers have
agreed despite its imperfectness, Trans Tool can still be improved. In this atmosphere
the first demonstration systems was developed, according to Hutchins (1995, p.1) the
first notable collaboration is the collaboration between IBM and the Georgetown
University in 1954
Hutchins (1995, p.1) states that: US agencies were encouraged to support a
large number of projects. This shows how the enthusiasm on Trans Tool is already
spread. Enthusiasm for Trans Tool spreads throughout the world, and in this period
from the mid 1950s to the mid 1960s many of the approaches which are still current
were first put forward
2.2.1.2 Present Trans Tool

Trans Tool has launched on a great significance of growtha historic shift in
the way is being used and a phenomenal increase in the number of people who rely
on it. People now have Trans Tool software that is viable, affordable, and can be
operated on personal computer. According to Vasconellos (2010)s research: there are
more than 500 vendors of MT software for the personal computer around the world,
10
Perpustakaan Unika
and among them they put out well over 1,000 products. One of the vendors,
Globalink, sells its extensive line of software in at least 6,000 stores in North
America alone, and at present Europe is its fastest-growing market.
The development of the personal computer with access to the Internet is
giving a great growth in the usage of Trans Tool. People now have on-line Trans
Tool, which can easily be accessed through websites. Vasconellos (2010, p.1) states
that:
The on-line phenomenon is changing our whole way of thinking about
Trans Tool. Together, these two developments-the abundance of lowcost Trans Tool in a form of CD or DVD, coupled with on-line Trans
Tool are turning the Trans Tool into an everyday commodity that is
within the reach of anyone with a late-model personal computer.
Vasconellos (2010) statement shows that Trans Tool is become easier to be accessed
by many people, and therefore become more popular. Vasconelloss (2010) statement
is also supported by Cribb (2000, p.2): As the volume increases, MT systems
become more useful and therefore more financially viable.
2.2.1.3. Types of Trans Tool

Based on its capability Trans Tool has 2 types: bilingual and multilingual, just
as stated by Hutchins (2002, p.1): Systems are designed either for two particular
languages (bilingual systems) or for more than a single pair of languages
(multilingual systems).
11
Perpustakaan Unika
1.Bilingual Trans Tool.

Bilingual Trans Tool, just like its name, is a Translation Device for bilingual
use. Bilingual use in this Trans Tool differs from the type of the Trans Tool, for
example: Toogle Text and Rekso Tool for Indonesian English.
This kind of Trans Tool has two types based on its program location: Online
Trans Tool and Offline Trans Tool. Online Bilingual Trans Tool is more dependable
than the offline one since its vocabulary database is always updated by the
administrator of the program, and of course it makes the user easier to translate the
new words, but people who want to use the Trans Tool need to connect to the internet
to use it.
Example of bilingual Trans Tool translating:
Before Translated
1 Saya ingin makan mie ayam
After Translated
1
I wanted to eat chicken

noodles
2 I wanted to eat chicken noodles 2
Saya mau makan daging

ayam mi
Table 1. Translation result ofToogle Text Trans Tool
The table shown above is a sample taken from 'Toogle Text' bilingual Trans Tool.
The Indonesian sentence Saya ingin makan mie ayam is translated without
problems to I wanted to eat chicken noodles. But when the process is turned around
the result is not matched correctly. Also, the Trans Tool did not use the correct tense
12
Perpustakaan Unika
form. The Indonesian sentence has no tenses form, while the translation result uses
past tense.
2.. Multilingual Trans Tool

Multilingual Trans Tool is a Trans Tool which has the capability to translate
text to not only one language to another language, but one language to many other
languages. Multilingual Trans Tool rarely has an offline version; it is different from
the Bilingual one which has the online and offline version.
Even though it seems that Multilingual Trans Tool is not as simple as the
Bilingual Trans Tool, Multilingual Trans Tool has its own advantages:
Example of Multilingual Trans Tool translation
Before Translated
1 Saya ingin makan mie ayam
After Translated
1
I wanted to eat chicken

noodles
2 I wanted to eat chicken noodles 2
Saya ingin makan mie

ayam
Table 2. Translation result of Google Translate Trans Tool
The table above shows the result of translation using Google Translate multilingual
Trans Tool. Both translations have similar result, which are different from the
Bilingual Trans Tool, but the use of wrong tense still happens.
13
Perpustakaan Unika
The writer uses Google Translate which is categorized as this (multilingual)

type to conduct this study, but the writer only uses the Trans Tool for bilingual
purpose only, that is English sentences Indonesian sentences.
Even though it seems that Google Translate is effective (referring to the
example shown above); sometimes this Trans Tool is not able to translate the text
correctly.
2.3. Effective Translation

There are many theories that can be used to help translate effectively. Baker
(1992) in her book thinks that an effective translation is a translation that fulfills the
parameters that she mentioned in her book.
The parameters consist of 6 criteria, which are already in hierarchical order.
According to Baker (1992, p.5), the organization of these criteria is based on a
straightforward principle: it starts with the simplest possible level and grows in
complexity by widening its focus on each criterion. Those criteria are as follows:
1. Equivalence at Word Level

This parameter is the simplest of all of the criteria. Though it is the simplest, it
is the most vital essence in translating. One problem that is most commonly happens
is like what is said by Baker (1992, p. 10): what does the translator do when there is
no word in the target language which expresses the same meaning as the source
language word? though translating word from source language to target language
14
Perpustakaan Unika
seems to be easy, it would cause us a big trouble when translator meet the problem
shown above. This case most likely appears when translator is translating a jargon or
slang.
Example: He has bad blood with her dia memiliki darah buruk dengannya
.
Btw, where did she go to? Btw, mana dia pergi?
The word(s) Bad Blood in the first example is translated as Darah Buruk, is
actually translated correctly, because Bad=> Buruk and Blood=> Darah.
But, in this case this translation is false, since Bad Blood here means Bad
Relation so it should be translated as Hubungan buruk. A different case happens
in the second example. The word Btw is not translated into other word. The word
Btw is a jargon commonly used in online chatting to replace the phrase By the
way which is translated into Omong-omong in Indonesian.
2. Equivalence above Word Level

This parameter does not imply to the equivalence at sentence level, it implies
mostly on the phrase level. Baker (1992, p. 46) states that:
It goes without saying that words rarely occur on their own; they almost
occur in the company of other words. But words are not strung together at
random in any language; there are always restrictions on the way they can
be combined to convey meaning.
Words that are combined can sometimes misleading to be translated. If translator
does not know the correct meaning of the combined words, the whole translation
result will be weird.
15
Perpustakaan Unika
Example: I want to kick his ass Saya ingin menendang pantatnya.

This Job is a piece of cake Pekerjaan ini adalah sepotong kue.
The Phrase kick his ass and a piece of cake is not an individual word, so it should
not translated word per word just like what happened above. The phrase Kick his
ass means to beat, while the phrase piece of cake means very easy. Therefore,
those phrases should be translated as Menghajarnya and Sangat Mudah.
3. Grammatical Equivalence
In translating, a translator also needs to take a caution in the grammar section.
In some cases, different grammatical order can have a different meaning for
sentences. Jakobson (1959) (as cited in Baker, 1992, p. 83) states that:
The grammatical pattern of a language (as opposed to its lexical stock)
determines those aspects of each experience that must be expressed in the
given language
The statement of Jakobson here is the key to solve most problems that mostly occur
in this part: Some languages have different grammatical order than the source
languages. This means that a translator must understand the grammatical order of the
target language before they begin to translate the text.
Example: I wanted to eat chicken noodle saya ingin makan mie ayam.
At a first glance this translation seems fine and correct, but there is one mistake. The
source text uses past form of the sentence, while the translation result does not
indicate any tense.
16
Perpustakaan Unika
4. Textual Equivalence: Thematic and Information Structures

In translating something, word arrangement is important. Please remember
that when translator translates texts, the translator should choose the word wisely so
that the information delivered correctly. This statement is also supported by Enkvist
(1978) (as cited by Baker, 1992, p. 117):
A sentence is not autonomous; it does not exist for its own sake but as part
of a situation and part of a text. And one of the most important functions
of information dynamics is precisely to link a sentence to its environment
in a manner which allows the information to flow through the text in
desired manner.
Example:
1. I love you Aku mencintaimu
2. I love birds Saya suka burung
The information is delivered correctly because the word love is translated

differently in different sentences. In English the word love to express both taken a
liking and Sexually attracted, but in Indonesian it is different. In Indonesian the
word Cinta is used to express sexually attracted, while the word suka is used to
express taken a liking.
5. Textual Equivalence: Cohesion

The connection of grammatical and lexical are important in a sentence, one
cannot stand without the others. Baker (1992, p.180) states that: Cohesion is the
network of lexical, grammatical, and other relations which provide link between
17
Perpustakaan Unika
various parts of a text. The cohesion device such as: reference, substitution, ellipsis,
conjunction, and lexical cohesion (Halliday and Hasan (1976) as cited by Baker,
1992, p. 180) arrangement plays an important role in helping readers to understand
the text.
Example:
1. Max is gone; this delighted his foe Max hilang, ini musuh senang
nya.
2. My knife is too dull; I must get a sharper one Pisau saya terlalu
tumpul, saya harus mendapatkan yang lebih tajam.
3. Here are thirteen cards, pick any Berikut adalah tiga belas kartu,
memilih apapun.
4. He punches me, and then flees Dia memukul aku, dan kemudian
melarikan diri.
5. Max phoned Bryan when he is working Max menelepon Bryan ketika
ia bekerja.
In the first example, the translation result is wrong. The word This, which indicates
reference to Max Gone, is translated into Ini, and this causes the translation result
became weird. In the second sentence the translation result is translated correctly. The
word one, which indicates a substitution, is translated to yang which also
indicates a substitution. The third sentences translation is a mess. There is no
indication that there is an ellipsis in the translation result. Ignoring the fault happened
18
Perpustakaan Unika
to the pick the word any would be better translated as yang mana saja or yang
manapun. There is nothing wrong with the fourth example; the conjunction Then
is translated correctly, and it still indicates a conjunction. The fifth example, which
indicates a lexical cohesion, is quite similar as reference. In this kind of text,
translator should provide enough information so that the people who are reading the
translation result are not confused, or translator needs to provide a lexical chain
which links the word to other word. The translation of the example is fine, but it does
not provide any link to the word He.
6. Pragmatic Equivalence
After dealing with the surface of the text, the next one is to deal with the
implementation of the text in the utterance, or more specifically coherence. Charolles
(1983) (as cited by Baker, 1992, p. 219) states that:
No text is inherently coherent or incoherent. In the end, it all depends on
the receiver, and on his ability to interpret the indications present in the
discourse so that, finally, he manages to understand it in a way which
seems coherent to him- in a way which corresponds with his idea of what
it is that makes a series of actions into an integrated whole
The quotation above does not mean that the translator does not have a part in this.
Translator part in this section is to arrange the words and sentences, with good
selection of diction in a way that it will be understand by the people who read the
translation results, since not all reader has the same extend of knowledge.
19
Perpustakaan Unika
Example: Syntactically speaking, this sentence is wrong. Sintaksis

berbicara, kalimat ini salah.
Ignoring the translation mistakes that happened in the translation result, the word
Syntactically may be understood by linguist, but not with the common people, so
the translator needs to change the word Sintaksis, or at least translator can give a
foot note about the definition of Sintaksis.
The difficulty of the effective translation differs from one sentence to another.
It depends on the sentence, and the sentence types.
2.4. SENTENCES
Sentence is the unity of several words which arranged grammatically. In more
technical term; A sentence is a full predication containing a subject plus a predicate
with a finite verb (Frank, 1972, p. 220).
According to Frank (1972), sentences are classified into two ways: by types
and by number of formal predication. Through the classification of the number of
formal predication there are four types of sentences, they are as follows:
Simple Sentences
Sentences which only has one predication, in the form of an independent
clause
Example: The man stole the Jewelry
20
Perpustakaan Unika
Compound Sentences
Sentences which have two or more full predications, in a form of independent
clauses
Example: The man stole the jewelry and hid it in his home
Complex Sentences
Sentences which have two or more full predication, one of the predication is
independent clause and the other one is dependent clause.
Example: The man who stole the jewelry hid it in his home
Compound-Complex Sentences
Sentences which contain two or more independent clauses and one or more
dependent clauses.
Example: The man stole the jewelry and he hid it in his home until he could
safely get out of town.
21
Perpustakaan Unika
CHAPTER III
METHOD OF DATA COLLECTION AND ANALYSIS
The type of this study is qualitative study because the aim of this study is to
know the effectiveness of Trans Tool able to translate English Sentence into
Indonesian sentence effectively, which is will give a new understanding in the
subject. It is just as stated by Creswell (2001): Qualitative research is descriptive in
that the researcher is interested in process, meaning, and understanding gained
through words or pictures. In this case, the writer dealt with a set of words in a form
of a sentence.
3.1. Method of Data Collection

3.1.1. Material
The writer chose sentences and translated using Google Translate. These
sentences , then, were analysed. The numbers of sentences used in this study are as
follows:
Sentences
Academic
Non Academic
Simple
10
10
Compound
10
10
Complex
10
10
22
Perpustakaan Unika
Compound Complex
10
10
Table 3. List of the sentence types and numbers
There are two genres of sentences that are used in the analysis; the academic
sentences and the non-academic sentences. The academic sentences are taken from
academic articles, while the non-academic sentences are taken from literary works
such as novels, and short stories.
The materials are taken based on the theory of purposive sampling, which is
as stated by Cohen et.al. (2007, pp. 114-115):
Researchers handpick the cases to be included in the sample on the basis
of their judgment of their typicality or possession of the particular
characteristic being sought. In this way, they build up a sample that is
satisfactory t their specific needs.
The writer chose up to ten sentences for each type of sentences in both genres.
However, if the data analysis results were not satisfying enough, the writer added
the number of sentences to be analyzed.
3.1.2. Instrument
The instrument which the writer used was a personal computer with an
internet access to use the Google Translate, since the Google Translate cannot be
installed into the computer. The writer also used documents as a material.
3.1.3. Procedure
In doing this study, there were some steps to collect the materials; they were
as follows:
23
Perpustakaan Unika
1. Selecting sentences to be translated.

Sentences were taken from academic articles entitled MODELING THE
SPREAD OF ANTHRAX IN BUILDINGS (Sextro et. al. 2002), Effective
Phrase Translation Extraction from Alignment Models (Venugopal et. al.
2003), Reflection on the History and Present State of Machine Translation"
(Hutchins, 1995), Machine Translation A Brief History (Hutchins, 2002), The
development and use of machine translation systems (Hutchins, 1999) and
Why Computers Do Not Translate Better" (Hutchins, 1991), and a literary
work entitled Resident Evil: Genesis (Decandido, 2004). The sentences
were picked based on the type and the quantity mentioned in previous section;
also the sentences were picked using the theory of purposive sampling. After
the sentences were chosen, the sentences were put on a list before the
sentences are translated.
2. Translating the sentences using the Google Translate.

After the original sentences were listed, they were translated into Indonesian
sentences using the Trans Tool, which in this case is Google Translate.
3. Analyzing the translation result of Google Translate.

The translation results provided by Google Translate were then analyzed to
determine whether the result was translated correctly or not.
24
Perpustakaan Unika
3.2. Method of Data Analysis

3.2.1 The Parameters
The parameters used in this study are based on the compilation of parameter by
many experts in translation field, which is compiled in by Baker (1992):
1.
Equivalence at Word Level
2.
Equivalence above Word Level
3.
4.
Textual Equivalence: Thematic and Information Structures
5.
Textual Equivalence: Cohesion
6.
Pragmatic Equivalence
Due to the limitation of the Google Translate or the Machine Translation to refer to
a certain reference or context, the writer did not use the fifth and sixth criteria.
3.2.2. The Evaluation of Translation Result

The translation result of the Google Translate analyzed using the
parameters compiled by Baker (1992). Every sentence was analyzed using every
parameter. After the sentences had been analyzed, the writer tried to find where
'Google Translate' made a mistake.
25
Perpustakaan Unika
The writer counted percentages of the sentences translated correctly. The

Percentage of sentences was formulated as follows:
Number of correctlytranslated sentences
X 100%
Total Number of Sentences
There are no fixed standard for translation. Newmark (1988) states that
"Ultimately standards are relative, however much one tries to base them on criteria
rather than norms". So, in order to score the effectiveness of "Google Translate" the
writer determined the effectiveness based on the standard below
1. Very High
: 100% -81%
2. High
: 80% -61%
3. Standard
: 60% -41%
4. Low
: 40% -21%
5. Very Low
: 20% - 0%
26
Perpustakaan Unika
CHAPTER IV
DATA ANALYSIS
In this chapter, the writer analysed the collected sentences. In the process, the
writer analyzed a number of data with the pre-determined number of sentences. To
analyze the data the writer used four parameters which are taken from Baker (1992)'s
theory: word-level equivalence, above word-level equivalence, grammatical equivalence,
textual equivalence.
4.1 Academic Sentences

The writer took every single sentences mentioned in this section from academic
journal, whose title already mentioned before. The sentences collected in this section
categorized as: Simple Sentences, Compound Sentences, Complex Sentences, and
Compound-Complex Sentences.
4.1.1 Simple Sentences

In this section the writer analyzed ten simple sentences taken from the article:
"MODELING THE SPREAD OF ANTHRAX IN BUILDINGS (Sextro et. al. 2002).
These sentences are analyzed using four parameters; Word-level equivalence, above
word-level equivalence, grammatical equivalence, and textual equivalence.
27
4.1.1.1 Word- Level Equivalence
Perpustakaan Unika
Among ten academic simple sentences, the writer found eight word level
equivalence translation mistakes. They are as follows:
1. `We examined the effects of two parameters Kami meneliti efek dari dua
parameter (S.1)
At the first glance this sentence is correctly translated, each word is translated
correctly, except for one word; the word 'effects' is translated as 'efek'. The word
'effects' is a plural word while the word 'efek' in Indonesian is a singular word.
Alwi et. al. (2003) states that eventhough usually the form of plural and singular
words are different in Indonesian, mostly the form differences are not used if it
has connection with certain verb. Eventhough the form of singular and plural
word is not applied, the sentence should, at least, show that the noun is not a
singular but plural, the translation result should be 'beberapa efek' or 'efek-efek'.
2. We established activity patterns for six people in the space Kami

mendirikan pola aktivitas selama enam orang di ruang (S.2)
A similar mistake to the previous translation also happens in this translation. The
word 'patterns' is translated to 'pola'. The word 'patterns' is a plural word while
'pola' is a singular word and the translation result does not show that the word
'pola' is a plural word. The next mistake is that there are some words that are not
translated correctly; 'for' and 'established'. The word 'for' should be translated as
'untuk' instead of 'selama', since the word 'for' in this sentence has the meaning of
28
Perpustakaan Unika
'intended to 1 ' not as duration'. The word 'established' should be translated as
'membuat' instead of 'mendirikan', since in Indonesian, the word 'mendirikan' is

ussualy used for building.
3. The activity rate, the building surface for exchange, and the zone for
resuspension, all may be scheduled within the model to vary over time
Tingkat aktivitas, permukaan bangunan untuk pertukaran, dan zona untuk
resuspension, semua dapat dijadwalkan dalam model bervariasi dari waktu ke
waktu. (S.3)
The first noticeable mistake is the word 'resuspension' that does not undergo any
translation at all; this word can be translated as 'resuspensi'. Another mistake is
the word 'to', located between the word 'model' and 'vary', is ommited.
4. The resulting system of coupled ordinary differential equations describes

mass conservation in the storage locations. Sistem yang dihasilkan
digabungkan persamaan diferensial biasa menjelaskan konservasi massa di
lokasi penyimpanan. (S.4)
The mistake in this translation is once again the plural noun is translated to
singular noun ('equations' 'persamaan' and 'locations' 'lokasi'), and the
sentence does not show that both of the words are plural words. To show that
these words are plural words, the word can be translated as 'persamaan-
CambridgeAdvancedLearner'sDictionaryThirdEdition(2008),p.557
29
Perpustakaan Unika
persamaan' and 'lokasi- lokasi'. Next, the translation of the word 'of' cannot be
found in the translation result.
5. The exchange and resuspension processes vary directly with the activity level
Proses pertukaran dan resuspension bervariasi langsung dengan tingkat
aktivitas (S.6)
The translation mistake of translating plural noun into singular noun also happens
in this translation. The word 'processes', which is a plural noun, is translated as
'proses' which is a singular noun and the sentence does not show that this word is
a plural word. To show that the word is a plural word, this word can be translated
as 'proses-proses' or 'banyak proses'. Another thing is that the word 'resuspension'
is not translated at all, which is similar to (S.3).
6. Deposition of airborne particles occurs uniformly on surfaces in both

categories Deposisi partikel udara terjadi secara seragam pada permukaan di
kedua kategori (S.8)
Another mistake of translating plural noun into singular noun happened in this
translation. It happened to the word 'particles' which translated as 'partikel' which
is a singular in Indonesian. To make the word has a plural meaning, the word can
be translated as 'partikel-partikel' or 'beberapa partikel'.
30
Perpustakaan Unika
7. We established a set of scenarios Kami mendirikan satu set skenario (S.9)
Similar to the previous translations, mistake in translating plural word into

singular word also happened in this translation. This time it happens on the word
'scenarios', which is translated to 'skenario'.
8. For the present paper we describe the results obtained from one scenario
Untuk tulisan ini kami akan menjelaskan hasil yang diperoleh dari satu skenario
(S.10)
The mistake is the word 'results' which is a plural word translated as 'hasil' which
is a singular word; furthermore, the translation result does not indicate that the
word 'results' is a plural word. To make the translation result indicates that the
word is a plural word the word can be translated as 'hasil-hasil'
Based on the data above, most of the mistakes that happen in the translation of simple
sentences are; English plural words are translated to singular words. The next mistake
found is that some words are not translated by 'Google Translate' (see S.3 and S.4). There
is also one case in which the word which should be translated is ommited (S.3).
4.1.1.2 Above Word -Level Equivalence

Among 10 academic simple sentences, the writer found two above word level
31
Perpustakaan Unika
1. The resulting system of coupled ordinary differential equations describes
mass conservation in the storage locations. Sistem yang dihasilkan

digabungkan persamaan diferensial biasa menjelaskan konservasi massa di
lokasi penyimpanan.(S.4)
The translation of the subject, which is 'The resulting system of coupled ordinary
differential equations' is strange, because there are two Indonesian verbs located
next to each other ('dihasilkan' and 'digabungkan'), not only that, the translation
result seems more like a word per word translation than a phrase translation. To
make it easier to be understood, the clause can be translated as 'Sistem yang
berasal dari persamaan-persamaan diferensial biasa yang dipasangkan' instead
of ' Sistem yang dihasilkan digabungkan persamaan diferensial biasa'.
2. Figure 2 shows the floor plan of the office suite and the location of the
anthrax source Gambar 2 menunjukkan denah office suite dan lokasi sumber
anthrax (S.5)
One thing that is obvious from this translation is that the phrase 'office suite' didn't
undergo any translation at all. For an option, the phrase can be translated as
'ruangan kantor'.
There are two different phrases in here; one is a long phrase and the other is a short one.
For the long phrase, it seems that rather being translated as a phrase, the phrase is
translated word per word and then is arranged. The short phrase is different; this phrase is
not translated at all.
32
Perpustakaan Unika
4.1.1.3 Gramatical Equivalence

Among 10 academic simple sentences, the writer found one grammatical
equivalence translation mistake. It is as follows:
1. We partitioned floor surfaces into tracked and untracked areas kita

dipartisi permukaan lantai ke daerah dilacak dan tidak terlacak (S.7)
The verb 'partitioned' is in the past form but 'Google Translate' could not translate
it in past tense, but it was translated as a passive form. If the translation is as it is,
it would give the meaning of the humans are being partitioned by the floor
surfaces.
Mistake of translating active past form into passive form and vice versa commonly
happens because of the similarities of past form and passive form, the suffix '-ed'.
4.1.1.4 Textual Equivalence

Among 10 academic simple sentences, the writer found one textual equivalence
translation mistake. It is as follows:
1. For the present paper we describe the results obtained from one scenario
Untuk tulisan ini kami akan menjelaskan hasil yang diperoleh dari satu skenario
(S.10)
33
Perpustakaan Unika
The word 'paper' can be translated as 'tulisan' but since this is for an academic
purposes it would be better if it were translated as 'makalah' instead of 'tulisan'.
The word 'tulisan' is one of the translation possibilities of the word 'paper', but the word
'tulisan' is not quite formal, especially because this translation is for academic purposes.
4.1.2 Compound Sentences

In this section the writer analyzed ten compound sentences taken from four
articles: MODELING THE SPREAD OF ANTHRAX IN BUILDINGS (Sextro et. al.
2002), Reflection on the History and Present State of Machine Translation" (Hutchins,
1995), Machine Translation A Brief History (Hutchins, 2002), and Why Computers Do
Not Translate Better" (Hutchins, 1991) These sentences will be analyzed using four
parameters; Word-level equivalence, above word-level equivalence, grammatical
equivalence, and textual equivalence.
There is one sentence that is translated correctly in this section, for more detail,
refer to table 4.1.2
4.1.2.1 Word Level Equivalence

Among 10 academic compound sentences, the writer found nine word level
34
Perpustakaan Unika
1. We modeled the spread of anthrax for a hypothetical office suite and

estimated the distribution of mass and resulting occupant exposures. Kami
model penyebaran anthrax untuk office suite hipotetis dan diperkirakan distribusi
massa dan eksposur penghuni yang dihasilkan. (CD.1)
The word 'modeled' should be translated as 'memodelkan', but 'Google Translate'
translated it as 'model'. The word' modeled' is a verb, but the word 'model' was
translated as a noun.
2. MT research had not looked at the economic use of existing less than
perfect systems, and it had disregarded the needs of translators for
computer-based aids. MT Penelitian tidak melihat penggunaan ekonomi yang
ada 'kurang sempurna' sistem, dan itu telah mengabaikan kebutuhan penerjemah
untuk alat bantu berbasis komputer. (CD.2)
There are two words that were translated and the result was word-level translation
mistake. They are 'economic' and 'systems'. The word 'economic' is an adjective,
but it becomes noun in the translation result. The word 'systems' is a plural word,
but it becomes a singular after translation process, without the indication that the
word is a plural word in the translation result. Another mistake is: there is a use of
capital letter which is incorrect, since it was not meant for name, title, nationality,
etc.
35
Perpustakaan Unika
3. Now it is recognised that MT systems should be designed ab initio for
controlled language, and a number of independent companies outside the

academic MT research community have been doing so in recent years (e.g.
Volmac). Sekarang diakui bahwa sistem MT harus dirancang initio ab untuk
bahasa dikontrol, dan sejumlah perusahaan independen di luar komunitas riset
MT akademik telah melakukannya dalam beberapa tahun terakhir (misalnya
Volmac). (CD.3)
The problem that occurs is that the word 'systems' which is a plural word becomes
a singular word after translation process, and the translation result does not
indicate that this word is a plural word. This is different to the word 'companies'
which is translated as plural word.
4. It is to be hoped that this research will both accelerate the development of

MT prototypes and improve the operational flexibility of commercial
systems. Perlu Diharapkan penelitian ini akan baik mempercepat
pengembangan prototipe MT dan meningkatkan fleksibilitas operasional sistem
komersial.(CD.4)
There is a word that needs to be omitted from the translation result, since these
words make the translation result become strange; this word is 'baik', which
istranslated from the word 'both'. There is also another mistake in this translation;
the word 'systems' and 'prototypes' which are plural words becomes singular in the
translation result, and it does not indicate that these words are plural.
36
Perpustakaan Unika
5. Logos systems were based on a direct translation approach, but later systems
are closer to a transfer design and incorporate sophisticated means for

recording and applying semantic features. Logos sistem didasarkan pada
pendekatan terjemahan langsung, tetapi kemudian sistem ini lebih dekat dengan
desain transfer dan memasukkan alat canggih untuk merekam dan menerapkan
fitur semantik. (CD.5)
The mistake of translating plural words into singular words also happens in this
translation; it happens to the word 'systems' and 'features', and similar as before,
the translation result does not indicate that both of these words are plural words.
6. Of course, the idea or dream of translating the spoken word automatically

was present from the beginning (Locke 1955), but it has remained a dream
until now Tentu saja, ide atau impian menerjemahkan kata yang diucapkan
secara otomatis hadir dari awal (Locke 1955), tetapi tetap mimpi sampai
sekarang (CD.6)
The problem in this translation is that there is an ommited word in the translation
result. The word 'it' in the second half of the sentence is ommited.
7. The rule-based approach was most obvious in the dominant transfer systems
(Ariane, Metal, SUSY, Mu and Eurotra), but it was at the basis of all the
various interlingua systems Pendekatan berbasis aturan paling jelas dalam
sistem transfer dominan (Ariane, Metal, Susy, Mu dan Eurotra), tetapi pada
dasar dari semua sistem berbagai interlingua (CD.7)
37
Perpustakaan Unika
There is word ommitance in this sentence; the translation of the word 'was' from
the first half of the sentence and the word 'it' from the second half of the sentence
is ommited from the translation result. The word 'systems' which is a plural word
also becomes a singular word after translation process, and the result does not
indicate that this word is a plural word.
8. Statistical methods were common in the earliest period of MT research, in

the 1960s (see sect.3 above), but the results had been generally disappointing
metode statistik yang umum pada awal periode penelitian MT, pada tahun
1960 (lihat sect.3 di atas), namun hasilnya telah mengecewakan umumnya (CD.8)
First mistake is caused by the wrong translation of the word 'were', which should
be translated as 'adalah'. The word 'were' is a verb in the sentence, but the word
'yang' is a conjunction in Indonesian. The translation of plural word into singular
word also happens in this translation. It happens to the word 'methods' and
'results', and also the translation result does not indicate that these word are a
plural word.
9. Furthermore, I shall not be describing any particular system of whatever

kind, past or present, nor shall I be giving details of methods of computation
analysis or processing Selain itu, saya tidak akan menjelaskan semua sistem
tertentu jenis, masa lalu atau sekarang apa pun, tidak akan saya akan
memberikan rincian metode analisis perhitungan atau pengolahan (CD.9)
38
Perpustakaan Unika
First mistake of this translation is the wrong translation of the word 'any', the
word 'any', which mainly has the meaning some, or even the smallest amount of 2 ,
become 'semua' which have the meaning of 'everything', so the translation of the
word should be 'manapun'. Next, the translation of the word 'nor' also unclear; the
result can be changed as 'tidak pula' to make it clearer. Translation of plural word
into singular word and the result does not indicate that the word is plural also
happens here, there are two words that experience this mistake; 'details' and
'methods.
The translation mistake in the word-level on compound sentences has more variant
compared to the simple sentences. Even though it has more variant, the most dominant
mistake is still the mistake of translating the plural word into singular word. It happens to
most of the sentences listed above.

Among ten academic compound sentences, the writer found six above word level
1. MT research had not looked at the economic use of existing less than
perfect systems, and it had disregarded the needs of translators for
computer-based aids. MT Penelitian tidak melihat penggunaan ekonomi yang
39
Perpustakaan Unika
ada 'kurang sempurna' sistem, dan itu telah mengabaikan kebutuhan penerjemah
untuk alat bantu berbasis komputer. (CD.2)
The mistake of this translation is the mistake of translating 'MT research' phrase.
If it was meant for word per word translation, the translation is correct, but since
this is for phrase translation, the order of the words is wrong. The correct order is
'Penelitian MT'. Similar mistake happens to the phrase 'existing "less than perfect"
systems', which should be translated as ' sistem-sistem 'kurang sempurna' yang
ada' instead of ' ada 'kurang sempurna' sistem'.
2. Now it is recognised that MT systems should be designed ab initio for

controlled language, and a number of independent companies outside the
academic MT research community have been doing so in recent years (e.g.
Volmac). Sekarang diakui bahwa sistem MT harus dirancang initio ab untuk
bahasa dikontrol, dan sejumlah perusahaan independen di luar komunitas riset
MT akademik telah melakukannya dalam beberapa tahun terakhir (misalnya
Volmac). (CD.3)
The noticeable phrase translation mistake is in 'independent companies outside
the academic MT research community' phrase. There is a wrong word order
translation; the correct translation should be 'sejumlah perusahaan independen di
luar komunitas riset akademik MT'' instead of 'sejumlah perusahaan independen
di luar komunitas riset MT akademik'. The other phrase which is incorrectly
translated is ' ab initio', which only undergoes word rearranging process.
40
Perpustakaan Unika
3. Logos systems were based on a direct translation approach, but later systems
are closer to a transfer design and incorporate sophisticated means for

recording and applying semantic features. Logos sistem didasarkan pada
pendekatan terjemahan langsung, tetapi kemudian sistem ini lebih dekat dengan
desain transfer dan memasukkan alat canggih untuk merekam dan menerapkan
fitur semantik. (CD.5)
There are two phrase translation mistakes in this translation. The phrase 'later
systems' has different meaning after being translated; the correct translation
should be 'sistem-sistem setelahnya' instead of ' kemudian sistem'. The phrase
'Logos systems' also has wrong word order in the translation result. The
translation result of the phrase should be 'Sistem- sistem Logos' instead of 'Logos
sistem'.
4. The rule-based approach was most obvious in the dominant transfer systems
(Ariane, Metal, SUSY, Mu and Eurotra), but it was at the basis of all the
various interlingua systems Pendekatan berbasis aturan paling jelas dalam
sistem transfer dominan (Ariane, Metal, Susy, Mu dan Eurotra), tetapi pada
dasar dari semua sistem berbagai interlingua (CD.7)
There is a mistake in the phrase translation of the 'various interlingua systems'
phrase; there is a wrong word order in the phrase. The translation result of the
phrase, which is ' sistem berbagai interlingua', should be translated as ' sistemsistem interlingua yang bermacam-macam' .
41
Perpustakaan Unika
5. Statistical methods were common in the earliest period of MT research, in

the 1960s (see sect.3 above), but the results had been generally disappointing
metode statistik yang umum pada awal periode penelitian MT, pada tahun
1960 (lihat sect.3 di atas), namun hasilnya telah mengecewakan umumnya (CD.8)
The phrase translation of 'generally disappointing' provides us with an ineffective
translation result. It would be better if the phrase were translated as ' pada
umumnya mengecewakan' instead of 'mengecewakan umumnya.
6. Furthermore, I shall not be describing any particular system of whatever

kind, past or present, nor shall I be giving details of methods of computation
analysis or processing Selain itu, saya tidak akan menjelaskan semua sistem
tertentu jenis, masa lalu atau sekarang apa pun, tidak akan saya akan
memberikan rincian metode analisis perhitungan atau pengolahan (CD.9)
The translation result of the phrase 'whatever kind' gives a different meaning from
the original sentence. The phrase 'whatever kind' should mean 'any kind', but if
the translation result is similar as translation result provided above, it would mean
'a specific kind', the translation result can be changed as 'jenis apapun' to convey
the correct information .
In this section, the problem most likely to occur is the wrong word order of the
translation result; the phrase is translated word per word and then arranged incorrectly.
There is special case in (CD.8), where there is a word that should be added to make the
translation result sounds better.
42
Perpustakaan Unika
4.1.2.3 Grammatical Equivalence

Among ten academic compound sentences, the writer found one grammatical
1. We modeled the spread of anthrax for a hypothetical office suite and

estimated the distribution of mass and resulting occupant exposures. Kami
model penyebaran anthrax untuk office suite hipotetis dan diperkirakan distribusi
massa dan eksposur penghuni yang dihasilkan. (CD.1)
Failure to distinguish between passive form and active past form is the problem in
this translation. The word 'estimated' in the original sentence is in active form, but
'Google Translate' perceives it as a passive form. Thus, it gives the wrong
translation result ('diperkirakan') The correct translation for this word should be
'memperkirakan'.
The problem in this section is similar to the grammatical equivalence problem in the
simple sentence translation. 'Google Translate' fails to differentiate between past active
form and passive form.
4.1.3 Complex Sentences

In this section the writer analyzed ten complex sentences taken from four articles:
MODELING THE SPREAD OF ANTHRAX IN BUILDINGS (Sextro et. al. 2002),
Effective Phrase Translation Extraction from Alignment Models (Venugopal et. al.
43
Perpustakaan Unika
2003), Machine Translation A Brief History (Hutchins, 2002), and Why Computers Do
Not Translate Better" (Hutchins, 1991) These sentences were analyzed using four
parameters; Word-level equivalence, above word-level equivalence, grammatical

4.1.3.1 Word Level Equivalence

Among ten academic complex sentences, the writer found ten word level
1. As a consequence, they are sometimes unable to distinguish between the

failings of particular systems and the problems which the best system would
have. Akibatnya, mereka kadang-kadang tidak dapat membedakan antara
kegagalan sistem tertentu dan masalah yang sistem yang terbaik akan (CX.1)
There are some cases where 'Google Translate' translates plural words to singular
words without the indication in the translation result that the words are plural
words in this sentence. These cases are found in the words; 'Systems', 'Failings',
and 'Problems'. The next mistake is the omittance of the word 'have' which is at
the end of the sentence.
2. These effects are also illustrated by the data in Table 2, which show a slight
decrease in the mass fraction for all tracked surfaces over time Efek ini
juga digambarkan oleh data dalam Tabel 2, yang menunjukkan sedikit penurunan
dalam fraksi massa untuk semua permukaan dilacak dari waktu ke waktu (CX.2)
44
Perpustakaan Unika
Another case of translating plural word into singular word happens here. The
word 'effects', which is a plural word, becomes a singular word after translation
without the indication that the word is a plural word, to indicate that the word is a
plural word, the word can be translated as 'beberapa efek' or 'efek-efek'.
3. For our simulations we postulated that the source is an envelope containing 1

gram of anthrax opened in Room 4. Untuk simulasi kami mendalilkan
bahwa sumbernya adalah sebuah amplop berisi 1 gram anthrax dibuka di Kamar
4 (CX.3)
The word 'simulations' which is a plural word becomes a singular word without
the indication that it is a plural word after the translation process. Also, please
notice that the word 'our' in the beginning of the sentence is ommited from the
translation result. Another thing is that between the word 'anthrax' and 'dibuka'
the conjuction 'yang' must be added, because the translation result becomes
inacurate if the conjuction is not added.
4. These simulations help identify critical information or processes that

contribute most to predicted outcomes or to uncertainties. Simulasi ini
membantu mengidentifikasi informasi penting atau proses yang paling
berkontribusi pada hasil-hasil diperkirakan atau ketidakpastian. (CX.4)
The mistake of translating plural word into singular word without the indication
that the word is a plural word also happens in this translation. It is found in the
word 'simulations', and 'processes'.
45
Perpustakaan Unika
5. Phrase translation pairs that are generated by the method described in this
paper are finally scored with estimates of translation probability, which can
be conditioned on the target language if necessary. Pasang terjemahan
Frasa yang dihasilkan oleh metode yang dijelaskan dalam makalah ini akhirnya
mencetak dengan perkiraan probabilitas terjemahan, yang dapat dikondisikan
pada bahasa target jika perlu (CX.5)
Translation mistake is found in the word 'pairs'. Not only translated to a singular
word when it is supposedly a plural word, the translation result, which is the word
'pasang', has a different meaning, the correct translation of this word should be
'pasangan-pasangan'. The current translation result would mean ' tide' if it were
translated back to English word. Wrong translation that gives another meaning is
also found in the word 'scored'. If the translation result is re-translated, it will
become 'print'. The correct translation of this word is 'dinilai', since the word
'scored' has the meaning of 'giving a mark' in the original sentence. Another
mistake is: there is a capital letter usage for the word 'Frasa' which is incorrect,
since it is not meant for name, title, nationality, etc.
6. Experiments were run with a language model (LM) built on a 20 million

word news source corpus using our in house decoder which performs a
monotone decoding without reordering. Percobaan dijalankan dengan
model bahasa (LM) dibangun di atas sumber berita 20 juta kata korpus
46
Perpustakaan Unika
menggunakan decoder kami di rumah yang melakukan decoding monoton tanpa

penataan kembali. (CX.6)
The word 'experiments' at the beginning of the sentence is plural word, but when
translated, it becomes singular word (eksperimen), and also without the indication
that the word is a plural word.
7. My intention in this paper is to provide some explanation for the difficulties

encountered by present computer systems which attempt to produce partial
or complete translations of texts from one natural language into another.
Tujuan saya dalam tulisan ini adalah untuk memberikan penjelasan atas kesulitan
yang dihadapi oleh sistem komputer ini yang mencoba untuk menghasilkan
terjemahan parsial atau lengkap teks dari satu bahasa alami ke lain. (CX.7)
The word 'difficulties', 'texts', 'translations', and 'systems' which are plural word
become singular word after translated. Next, the translation result of the word
'another' is ineffective. It would be better if the word is translated as 'bahasa lain'.
Wrong translation which gives a different meaning happens to the word 'present'.
If the result is re-translated, the result of re-translation will be 'this'; the word
should be translated as 'sekarang' since the word 'present' in this sentence has the
meaning of 'right now'.
8. My aim is to give an introduction, for those unfamiliar with machine

translation (MT), to the main phenomena which must be taken into
consideration Tujuan saya adalah untuk memberikan pengantar, bagi mereka
47
Perpustakaan Unika
yang tidak terbiasa dengan terjemahan mesin (MT), untuk fenomena utama yang
harus dipertimbangkan (CX.8)
If the translation of the word 'to', which is located between the words '(MT) and
'the main' is as it is (untuk), it would mean 'for' if the word is re-translated. The
translation result of this sentence should be 'kepada', so the translation conveys
the right information.
9. The aim is to highlight in the broadest terms those areas of translation which
are relatively easy for computerised handling and those areas which are
relatively difficult Tujuannya adalah untuk menyoroti dalam hal luas wilayah
terjemahan yang relatif mudah untuk menangani komputerisasi dan daerahdaerah yang relatif sulit (CX.9)
The word 'Areas' which is plural word is translated as singular word without the
indication that the word is a plural word in this translation. This word can be
translated as 'area-area' to indicate that the word is a plural word.
10. The basic idea is that rather than formulating changes between languages in
terms of transfer operations, translation takes place from and into an
interlingua, i.e. elements or structures which are in some sense neutral for
the languages concerned Ide dasarnya adalah bahwa daripada merumuskan
perubahan antara bahasa dalam hal pengoperasian transfer, terjemahan
berlangsung dari dan ke Interlingua sebuah, yaitu unsur-unsur atau struktur yang
dalam arti tertentu netral untuk bahasa yang bersangkutan (CX.10)
48
Perpustakaan Unika
The word 'Operations' and 'Structures' which are plural words are translated as
singular words without any indication that the translation result is plural words.
The word should be translated as 'proses-proses' and 'struktur-struktur' to indicate
that these words is a plural words.
The mistake of translating plural word into singular word is dominating in this section,
which is quite similar to the previous section. Another thing is that the translation which
gives different meaning from the original word occurs more frequently in this type of
sentence.

Among 10 academic complex sentences, the writer found three above word level
1. Phrase translation pairs that are generated by the method described in this
paper are finally scored with estimates of translation probability, which can
be conditioned on the target language if necessary. Pasang terjemahan
Frasa yang dihasilkan oleh metode yang dijelaskan dalam makalah ini akhirnya
mencetak dengan perkiraan probabilitas terjemahan, yang dapat dikondisikan
pada bahasa target jika perlu (CX.5)
Ignoring the wrong word translation to the word 'pairs' in the phrase 'phrase
translation pairs', the order of the phrase translation result itself is wrong. The
word in the phrase is translated word per word and for the word order, 'Google
49
Perpustakaan Unika
Translate' just switching the first and the final word possition. The translation
result should be ' Pasangan-pasangan terjemahan frasa' instead of ' Pasang
terjemahan Frasa'.
2. Experiments were run with a language model (LM) built on a 20 million

word news source corpus using our in house decoder which performs a
monotone decoding without reordering. Percobaan dijalankan dengan
model bahasa (LM) dibangun di atas sumber berita 20 juta kata korpus
menggunakan decoder kami di rumah yang melakukan monoton decoding tanpa
pemesanan kembali. (CX.6)
The phrase 'monotone decoding' is translated ineffectively. The word 'Monotone'
is translated while the word 'decoding' is not; also the word order is wrong in this
phrase translation result. The correct translation for this phrase is 'decoding yang
monoton' instead of 'monoton decoding'.
3. The basic idea is that rather than formulating changes between languages in
terms of transfer operations, translation takes place from and into an
interlingua, i.e. elements or structures which are in some sense neutral for
the languages concerned Ide dasarnya adalah bahwa daripada merumuskan
perubahan antara bahasa dalam hal pengoperasian transfer, terjemahan
berlangsung dari dan ke Interlingua sebuah, yaitu unsur-unsur atau struktur yang
dalam arti tertentu netral untuk bahasa yang bersangkutan (CX.10)
50
Perpustakaan Unika
The word order of the phrase translation result of 'an interlingua' phrase is in a
reverse order, it should be 'sebuah interlingua' instead of 'Interlingua sebuah'.
The word order problem also appears in this section, but mostly accompanied by the
word-per word translation. Another thing is, mostly for phrase translation, 'Google
Translate' only switching the order of the first and last word in the phrase (CX.5).

Among ten academic complex sentences, the writer found one textual equivalence
translation mistake. It is as follows:
1. My intention in this paper is to provide some explanation for the difficulties

encountered by present computer systems which attempt to produce partial
or complete translations of texts from one natural language into another.
Tujuan saya dalam tulisan ini adalah untuk memberikan penjelasan atas kesulitan
yang dihadapi oleh sistem komputer ini yang mencoba untuk menghasilkan
terjemahan parsial atau lengkap teks dari satu bahasa alami ke lain. (CX.7)
Similar case to (S.10), the word 'paper' can be translated as 'tulisan' but since this
is for an academic purpose, it would be better if it were translated as 'makalah',
which have more formality than 'tulisan'.
51
Perpustakaan Unika
Based on the data (CX.7), 'Google Translate' made the similar mistake with (S.10). It
almost seems that whenever the word 'paper is written in academicals context, it will
always be translated into 'tulisan'.
4.1.4.1 Compound Complex Sentence

In this section the writer analyzed ten compound-complex sentences taken from
three articles: Effective Phrase Translation Extraction from Alignment Models
(Venugopal et. al. 2003), Machine Translation A Brief History (Hutchins, 2002), and
"The development and use of machine translation systems" (Hutchins, 1999). These
sentences will be analyzed using four parameters; Word-level equivalence, above wordlevel equivalence, grammatical equivalence, and textual equivalence.
4.1.4.1 Word-Level Equivalence

Among ten academic compound complex sentences, the writer found 10 word
level equivalence translation mistakes. They are as follows:
1. Analyzing these models from a generative mindset, they all assume that the
atomic unit of lexical content is the word, and re-ordering effects are applied
above that level. Menganalisis model ini dari pola pikir generatif, mereka
semua menganggap bahwa unit atom konten leksikal adalah kata, dan efek
kembali memesan yang diterapkan di atas tingkat itu. (CDX.1)
52
Perpustakaan Unika
The word 'Models', and 'effects' which are plural words is translated as singular
word in this translation without any indication that the word is plural words. Next
mistake found in this translation is the word 'reordering' which is translated as
'kembali memesan'. In (CX.6) there is a similar case, but the word is translated
correctly as 'penataan kembali'.
2. The question is thus whether the manufacturers of these products prepared

to invest in the research required or whether researchers in the MT
community are going to address this particular need. Pertanyaannya
adalah apakah dengan demikian produsen produk tersebut siap untuk
berinvestasi dalam penelitian yang diperlukan atau apakah peneliti dalam
komunitas MT akan alamat ini kebutuhan tertentu. (CDX.2)
The word 'Manufacturers', and 'researchers' which are plural words is translated as
singular word without the indication that the words is plural words in this
translation. Next mistake is the word 'address' which is supposedly means as
'Formal to speak or write to someone 3 ', but is perceived by 'Google Translate' as
'location'. The translation of this word would be better if it were translated
'menyebutkan' rather than 'alamat'.
3. Japanese input demands considerable pre editing, but it is acceptable to

operators of Japanese word processors who have to interpret Japanese script
Masukan Jepang tuntutan pra editing yang cukup besar, tetapi diterima
53
Perpustakaan Unika
operator pengolah kata Jepang yang harus menginterpretasikan skrip Jepang

(CDX.3)
The word 'Operators' which is a plural word is translated as a singular word in this
translation. The second mistake is that the word 'Japanese' have meaning as a
'Language' not as a 'Nation'. Third mistake, the word 'demands' is a verb, but in
the translation result it becomes a noun
4. We then discuss the implications of adding phrasal translation pairs to the

decoding process, and present evaluation results that show significant
improvements when applying the described extraction technique Kami
kemudian membahas implikasi penambahan pasangan terjemahan frase untuk
proses decoding, dan hasil evaluasi ini yang menunjukkan perbaikan yang
signifikan ketika menerapkan teknik ekstraksi dijelaskan (CDX.4)
The word 'Implications ',' pairs', 'results' and 'improvements' which are a plural
words are translated as a singular words without the indication that the words are
plural words in this translation. Another thing is, the word 'improvement' means
'advancement' not 'reparation'. The translation result of this word should be
'peningkatan' rather than 'perbaikan'.
5. (Marcu, Wong, 2002) illustrate the effects of assuming that lexical

correspondence can only be modeled at the word level, and motivate a joint
probability model that explicitly generates phrase level lexical content across
both languages. (Marcu, Wong, 2002) menggambarkan efek dengan
54
Perpustakaan Unika
mengasumsikan bahwa korespondensi leksikal hanya dapat dimodelkan pada

tingkat kata, dan memotivasi model probabilitas gabungan yang secara eksplisit
menghasilkan tingkat frase leksikal konten di kedua bahasa. (CDX.5)
The word 'Effects' which is a plural word is translated as a singular word in this
translation without any indication that the original word is a plural word. The
word can be indicated as a plural word if it translated as 'efek-efek' or 'beberapa
efek' or 'banyak efek'.
6. However, it does mean that before an MT system can be installed, the user
must have already available a well-founded terminological database, with
authorised translation equivalents in the languages involved, or at least
must make a commitment to develop the required term bank Namun, itu
berarti bahwa sebelum sistem MT dapat diinstal, pengguna harus sudah tersedia
baik-didirikan database terminologis, dengan setara terjemahan resmi dalam
bahasa yang terlibat, atau - setidaknya - harus membuat komitmen untuk
mengembangkan istilah yang diperlukan bank (CDX.6)
The word 'Equivalents' which is a plural word is translated as a singular word in
this translation without any indication that the original word is a plural word. The
word can be indicated as a plural word if it translated as 'persamaan-persamaan'
or 'beberapa persamaan'.
7. MT systems are finding a natural role, since they can operate virtually or in
fact in real-time and on-line and there has been little objection to the
55
Perpustakaan Unika
inevitable poor quality Sistem MT adalah mencari 'alami' peran, karena

mereka dapat beroperasi hampir atau bahkan dalam real-time dan on-line dan
ada sedikit keberatan terhadap kualitas miskin tak terelakkan (CDX.7)
The word 'Systems' which is a plural word is translated as a singular word without
any indication that the original word is a plural word in this translation. The
translation of the word 'finding' has different meaning from the original word.
The word 'mencari' has the meaning of 'to search', while the word 'finding' in this
sentence has the meaning of 'discover', which is why the word should be
translated as 'menemukan'. Next mistake is the ommitance of the word 'virtually'
from the translation result.
8. Our method is significantly less computationally expensive than methods that

attempt to explicitly model phrase level interactions within alignment
models, and recovers well from noisy alignments at the sentence and corpus
level. Metode kami secara signifikan kurang komputasi mahal daripada
metode yang mencoba untuk secara eksplisit tingkat interaksi frase model dalam
model penyelarasan, dan pulih baik dari keberpihakan bising pada kalimat dan
tingkat corpus. (CDX.8)
The word 'methods', 'models', 'interactions', and 'alignments' which are a plural
word are translated as a singular word in this translation, without any indication
that the original word is a plural word.
56
Perpustakaan Unika
9. It is clear that the potential for MT on, via and for the Internet is now being
fully appreciated no company can afford to be left behind, and all the
major players have ambitious plans, e.g. Lernout & Hauspie (McLaughlin
and Schwall 1998), which has now acquired MT systems from Globalink,
Neocor and AppTek as well as the old METAL system (from GMS). Jelas
bahwa potensi untuk MT pada, melalui dan untuk internet kini sepenuhnya
dihargai - tidak ada perusahaan yang mampu ditinggalkan, dan semua pemain
utama memiliki rencana ambisius, misalnya Lernout & Hauspie (McLaughlin dan
Schwall 1998), yang kini diperoleh sistem MT dari Globalink, Neocor dan
AppTek serta sistem LOGAM lama (dari RUPS). (CDX.9)
The abbreviation 'GMS' which should not be translated is translated as a 'RUPS'.
10. The purpose, therefore, is not to describe the inherent limitations of machine
translation, but to give a rather simplified explanation of what may be
expected from systems at the present time Tujuannya, oleh karena itu,
bukan untuk menggambarkan keterbatasan terjemahan mesin, tetapi untuk
memberikan sebuah penjelesan yang agak sederhana dari apa yang dapat
diharapkan dari sistem pada saat ini (CDX.10)
The problem this time is in the Indonesian word 'penjelesan', which is the
translation of the word 'explanation'. This is happen when the writer translate
whole of the sentence at once. But when the writer translates the word
'explanation' as a word, the result becomes 'penjelasan'.
57
4.1.4.2 Above Word-Level Equivalence
Perpustakaan Unika
Among ten academic compound complex sentences, the writer found five above
word level equivalence translation mistakes. They are as follows:
1. Marcu, Wong, 2002) illustrate the effects of assuming that lexical

correspondence can only be modeled at the word level, and motivate a joint
probability model that explicitly generates phrase level lexical content across
both languages. (Marcu, Wong, 2002) menggambarkan efek dengan
mengasumsikan bahwa korespondensi leksikal hanya dapat dimodelkan pada
tingkat kata, dan memotivasi model probabilitas gabungan yang secara eksplisit
menghasilkan tingkat frase leksikal konten di kedua bahasa. (CDX.5)
The word order of the phrase translation result of 'lexical content' phrase is in
reverse order. The translation result should be 'konten leksikal' rather than 'leksikal
konten'.
2. However, it does mean that before an MT system can be installed, the user
must have already available a well-founded terminological database, with
authorised translation equivalents in the languages involved, or at least
must make a commitment to develop the required term bank Namun, itu
berarti bahwa sebelum sistem MT dapat diinstal, pengguna harus sudah tersedia
baik-didirikan database terminologis, dengan setara terjemahan resmi dalam
bahasa yang terlibat, atau - setidaknya - harus membuat komitmen untuk
mengembangkan istilah yang diperlukan bank (CDX.6)
58
Perpustakaan Unika
'Google Translate' translates the phrase 'well-founded terminological database' by

word per word translation. The translation result of this phrase should be
'database terminologis dengan dasar yang kuat' rather than ' baik-didirikan
database terminologis'. Similar case happen to the phrase 'authorised translation
equivalents'. The translation result should be ' terjemahan setara yang resmi '
rather than 'setara terjemahan resmi'. Different case happens to the clause
'required term bank'. The translation result of the clause would give the meaning
of 'the term required by the bank'. it should be translated as ' daftar istilah yang
diperlukan'
3. MT systems are finding a natural role, since they can operate virtually or in
fact in real-time and on-line and there has been little objection to the
inevitable poor quality Sistem MT adalah mencari 'alami' peran, karena
mereka dapat beroperasi hampir atau bahkan dalam real-time dan on-line dan
ada sedikit keberatan terhadap kualitas miskin tak terelakkan (CDX.7)
The word order of the phrase translation result of 'natural role' phrase is in reverse
order, it should be 'peran 'alami' rather than 'alami' peran'. Word order of the
phrase 'inevitable poor quality' has no problem, but the phrase 'poor quality',
which supposedly has the meaning of 'low quality' , has the meaning of 'no money
quality'. This is because 'Google Translate' think the word 'poor' has the meaning
of 'no money 4 ' rather than 'bad 5 '. The Translation should be ' kualitas rendah yang
tidak terelakan' rather than ' kualitas miskin tak terelakkan'.
Cambridge Advanced Learner's Dictionary Third Edition (2008), p.1098
59
Perpustakaan Unika
4. Our method is significantly less computationally expensive than methods that

attempt to explicitly model phrase level interactions within alignment
models, and recovers well from noisy alignments at the sentence and corpus
level. Metode kami secara signifikan kurang komputasi mahal daripada
metode yang mencoba untuk secara eksplisit tingkat interaksi frase model dalam
model penyelarasan, dan pulih baik dari keberpihakan bising pada kalimat dan
tingkat corpus. (CDX.8)
The phrase 'less computationally expensive' is not translated as a phrase by
'Google Translate'; it is translated as three words. The translation result should be '
lebih mudah secara komputasional' instead of ' kurang komputasi mahal'. Next
problem is the translation of the phrase 'noisy alignments'; the translation result is
clearly a word per word translation rather than a phrase translation. The phrase
should be translated as ' penyelarasan yang rumit' instead of ' keberpihakan
bising'. Next, There is a need to add an extra word to the translation result of the
phrase 'recovers well' to make the translation result better. The result should be
'pulih dengan baik' instead of just 'pulih baik'.
60
Perpustakaan Unika
The phrase 'can afford' becomes 'mampu' which means 'able', while it supposed to
means 'wanting', so the translation should be changed as ' tidak ada perusahaan
yang ingin ditinggalkan' instead of ' tidak ada perusahaan yang mampu
ditinggalkan'.

Among ten academic compound complex sentences, the writer found one
grammatical equivalence translation mistake. It is as follows:
61
Perpustakaan Unika
The verb 'acquired' is in the past participle but 'Google Translate' perceived it as a
passive form. If the translation result is as it is, it would give the meaning of the
humans are being owned by the MT systems.
The mistake is similar to the previous sections (see 4.1.1.3 and 4.1.2.3). 'Google
Translate' fails to differentiate between active and passive form, resulting in strange
translation result.
4.2 Non Academic Sentences

The writer took every single sentences mentioned in this section from non
academic source, which is Resident Evil: Genesis (Decandido, 2004). The sentences
collected in this section are also categorized into: Simple Sentences, Compound
Sentences, Complex Sentences, and Compound-Complex Sentences
4.2.1 Simple Sentences

In this section the writer analyzed ten simple sentences taken from the novel
'Resident Evil: Genesis' by Decandido (2004). These sentences will be analyzed using
four parameters; Word-level equivalence, above word-level equivalence, grammatical
There are four sentences that is translated correctly in this section, for more detail
please refer to the table 4.2.1
62
Perpustakaan Unika

Among 10 non academic simple sentences, the writer found two word level
1. It was what Umbrella paid him for Itulah yang Payung membayarnya untuk
(NS. 8)
The word 'Umbrella' functions as a name of something, which is why it should not
change after the translation process.
2. The helicopter spun through the darkness over Raccoon forest. Helikopter
berputar melalui kegelapan atas hutan Raccoon (NS. 9)
The mistake in this translation is the word 'over ' translated as 'atas'. The word
'atas' has the meaning of 'top', while the word 'over' has the meaning of 'above'
Different from the previous sections, there is no problem regarding plural or singular
sentences here; instead, there is a new problem. 'Google Translate' fails to identify which
word that should be translated, and which word which should not be translated.

Among ten non academic simple sentences, the writer found three above word
63
Perpustakaan Unika
1. His car was at the far end of the station lot Mobilnya berada di ujung stasiun
banyak (NS. 4)
'Station lot' here means 'the parking space in the station'. The translation result of
the phrase is clearly a word per word translation. The word ' lot' has a multiple
meaning. In this sentence this word should have the meaning of 'land 6 ' not 'large
ammount 7 ', so the translation result should be 'tempat parkir stasiun' instead of
'stasiun banyak'.
2. He'd be drenched before he got halfway across Ia akan basah kuyup

sebelum ia sampai setengah di (NS. 5)
The phrase 'halfway across' becomes strange after the translation process. The
result seems to be unfinished, this is because the phrase is translated word per
word and the word order is wrong. The translation result should be ' setengah
perjalanan' rather than ' setengah di'.
3. A ten-year-old girl named Teresa Hardy had been involved in a clinical trial
for a new pediatric heart medication, Valifin Seorang gadis sepuluh tahun
bernama Teresa Hardy telah terlibat dalam percobaan klinis untuk pengobatan
hati yang baru pediatrik, Valifin (NS. 6)
64
Perpustakaan Unika
'Google Translate' translates the phrase 'new pediatric heart medication' by word
per word translation. The translation result of this phrase should be 'pengobatan
pediatrik hati yang baru' rather than ' pengobatan hati yang baru pediatrik.
4. It was what Umbrella paid him for Itulah yang Payung membayarnya untuk
(NS. 8)
This sentence is translated with word per word translation, which is why the
translation result is strange. The result should be ' Itulah mengapa Umbrella
membayarnya' instead of ' Itulah yang Payung membayarnya untuk'
There are no new problems appear in this section. The problem is similar to the previous
sections; the phrase is not translated as a phrase, but it is translated as a multiple word.

Among ten non academic simple sentences, the writer found one grammatical
1. The small pin marked him as a regular commuter Pin kecil ditandai
sebagai seorang komuter biasa (NS. 2)
Failure to distinguish between passive form and past form is the problem in this
translation. The word 'marked' in the original sentence is in active form, but
65
Perpustakaan Unika
'Google Translate' perceive it as passive form, thus it gives the wrong translation
result.
There is nothing new from the mistake found in this section, the problem found in this
section is similar to the previous grammatical equivalence sections.
4.2.2 Compound Sentences

In this section the writer analyzed ten compound sentences taken from the novel
4.2.2.1 Word-level Equivalence

Among ten non academic compound sentences, the writer found seven word level
1. Then she heard the tick of metal, the strangled last gasp of the engine, and
her own thundering heart, and realized that they were down Lalu ia
mendengar centang logam, jantung terakhir terkesiap mesin, dan gemuruh
mencekik sendiri, dan menyadari bahwa mereka turun (NCD.1)
The word 'tick' means 'making a ticking sound', but 'Google translate' translates it
as 'centang' which means 'giving mark'. The translation result should be ' Lalu ia
66
Perpustakaan Unika
mendengar bunyi berdenting logam' instead of ' Lalu ia mendengar centang

logam'
2. We weren't struck directly, but it was close enough Kami tidak menyerang
secara langsung, tapi cukup dekat (NCD.4)
The word 'struck' means 'getting hit', but 'Google translate' translates it as
'menyerang' which means 'attacking'. To make it better, the translation result
should be changed as 'terkena' instead of 'menyerang'.
3. They were delivering him to his execution, but they were following orders; it
wasn't personal mereka memberikan dia untuk eksekusi, namun mereka
mengikuti perintah, bukan pribadi (NCD.5)
The word 'delivering' means ' to escort', but 'Google translate' translates it as
'memberikan' which means 'giving', So the translation result should be
'mengantarkan' instead of 'memberikan'.
4. Umbrella's headquarters had moved off to Europe some time ago, but they
still owned practically the entire town. Payung markas sudah pindah ke
Eropa beberapa waktu lalu, tetapi mereka masih memiliki hampir seluruh kota.
(NCD.7)
The problem in this sentence is similar to the problem in (NS.8) which also
happens to the word 'Umbrella'.
67
Perpustakaan Unika
5. Edward shot her an encouraging smile, and Sully leaned across the crowded
cabin to reassuringly pat her leg Edward shot tersenyum mendorong, dan
Sully bersandar di kabin ramai untuk meyakinkan menepuk kakinya (NCD.8)
The word 'shot' is not translated at all. Another thing is, the word' accross' is
ommited from the translation result. Next, the word 'smile' at the beginning of the
sentence is a noun, but it becomes a verb after the translation process.
6. There were a few lights on, but they were faint Ada beberapa lampu, tapi
mereka pingsan (NCD.9)
The word 'light' should be translated as 'cahaya', since the word means 'brightness'
in this sentence , but Google translate translates it as 'lampu' which has the
meaning of 'lamp'. A similar case also happens to the word 'faint', which means
'blur'. This word should be translated to 'redup', but 'Google Translate' translates it
as 'pingsan' which means ' unconscious'.
7. They'd be nice about it, but she'd have to endure weeks, maybe months of
gentle ribbing Mereka akan sangat baik tentang hal itu, tapi dia harus
bertahan minggu, mungkin bulan ejekan lembut (NCD.10)
The mistake is the word 'weeks' and 'months' which are plural worda are
translated as 'minggu' and 'bulan' which are singular words. These words should
be translated as 'berminggu-minggu' and 'berbulan-bulan' rather than 'minggu' and
'bulan'. The word 'endure' should be translated as 'menahan' instead of 'bertahan',
since it is an active verb.
68
Perpustakaan Unika
Majority of the mistakes happen in this section is giving a translation result which has a
different meaning from the original words. The mistake of translating plural word into
singular word also appears in this section.
4.2.2.2 Above Word-level Equivalence

Among ten non academic compound sentences, the writer found four above word
1. Then she heard the tick of metal, the strangled last gasp of the engine, and
her own thundering heart, and realized that they were down Lalu ia
mendengar centang logam, jantung terakhir terkesiap mesin, dan gemuruh
mencekik sendiri, dan menyadari bahwa mereka turun (NCD.1)
The clause 'the strangled last gasp of the engine' should be translated as 'bunyi
mesin yang tercekik', but 'Google Translate' translates it as ' jantung terakhir
terkesiap mesin' which is wrong. The translation result of the clause indicates that
it was translated word per word not as a clause. This problem also happens to the
clause ' her own thundering heart' which should be translated as 'hatinya yang
bergemuruh' but instead it is translated as 'gemuruh mencekik sendiri'
2. She'd survive, but would likely spend the rest of her life on dialysis Dia
bertahan hidup, tapi kemungkinan akan menghabiskan sisa hidupnya di dialisis
(NCD.3)
69
Perpustakaan Unika
The problem in this translation is in the clause 'she'd survive', which is better
translated as ' Ia akan bertahan hidup' rather than 'Dia bertahan hidup'. There is
something omitted from the translation result. The result indicates that the word
('d) is not translated, in fact it is even omitted.
3. The first train car was empty of life, but something very bad had occurred
there, no question. Mobil itu kosong kereta pertama kehidupan, tetapi sesuatu
yang sangat buruk telah terjadi di sana, pertanyaan no. (NCD.6)
4. The translation result of the clause ' The first train car was empty of life' is
ineffective. The clause only translated word per word. The translation result
should be ' Gerbong kereta yang pertama kosong'. The phrase 'no question' also
is translated ineffectively; one of the words is translated and the other is not. Also,
the phrase 'no question' here is an expression which means 'very obvious' or 'clear'
5. Umbrella's headquarters had moved off to Europe some time ago, but they
still owned practically the entire town. Payung markas sudah pindah ke
Eropa beberapa waktu lalu, tetapi mereka masih memiliki hampir seluruh kota.
(NCD.7)
Ignoring the wrong translation of the word 'Umbrella', the word order of the
phrase 'Umbrella's Headquarters' itself is in reverse order. It should be 'Markas
Umbrella'. It seems that 'Google Translate' is unable to differ which word is the
modifier, and which word is the noun.
70
Perpustakaan Unika
There is nothing new from the mistake listed above, but there is one thing that is clear;
the longer the phrase or clause, the bigger the risk of wrong translation is occurring.

Among ten non academic compound sentences, the writer found one textual
1. They were delivering him to his execution, but they were following orders; it
wasn't personal mereka memberikan dia untuk eksekusi, namun mereka
mengikuti perintah, bukan pribadi (NCD.5)
The word 'personal' can be translated as 'pribadi', but in this sentence the word
'personal' should mean as ' personal matter', so the translation result should be
changed as 'masalah pribadi'.
There is nothing new in this section. 'Google Translate' still has problem to give the
translation result which is suited to the context of the text.
4.2.3 Complex Sentences

In this section the writer analyzed ten complex sentences taken from the novel
equivalence, and textual equivalence
71
Perpustakaan Unika
There is one sentence that is correctly translated in this section, for more detail
4.2.3.1 Word-level Equivalence

Among ten non academic complex sentences, the writer found four word level
1. Rebecca frowned, pushing at the sliding door, which was refusing to slide
Rebecca mengerutkan kening, mendorong di pintu geser, yang menolak untuk
slide (NCX.1)
The word 'Slide' which is at the back of the sentence is not translated. The result
should be 'digeser'. It is translated as a passive word, since syntactically it is
impossible for door to do an action.
2. The key you saw was for the dining car, which we've already sampled
Kunci yang Anda lihat adalah untuk mobil makan, yang kita telah sampel
(NCX.3)
The word 'sampled' at the back of the sentence is translated as 'sampel', which
means 'specimen'. The word 'sampled' does not mean as 'specimen', so the
translation result should be 'datangi' since the word 'sampled' means 'ever being
there'.
72
Perpustakaan Unika
3. Unless Billy and Rebecca were entirely moronic, which they'd already
proven they were not, they would figure out how to trigger the structure's
rotation Kecuali Billy dan Rebecca sepenuhnya tolol, yang mereka sudah
terbukti mereka tidak, mereka akan mencari cara untuk memicu rotasi struktur's
(NCX.6)
The word "structure's" become 'stuktur's after the translation process, when it
should be translated as 'stuktur tersebut'. The ('s) in the word "structure's" is an

adjective possessive which should not also become ('s) in the translation result.
4. Enrico asked the pilot, who was tapping at controls and flipping switches
Enrico tanya pilot, yang sadap di kontrol dan flipping switch (NCX.10)
The word 'asked' which means 'to give a question' after the translation process has
the correct meaning from the original sentence, but not clear enough. To make it
clearer the translation result should be 'bertanya pada'. The word 'flipping' is not
translated at all, for this word, we can translate it as 'membolak- balik', since the
suffix (-ing) describing a continuous action. The word 'switches', which is should
be translated as 'tombol-tombol' is not translated, and it also transferred from
plural word into singular word.
There is a new case where the word which supposed to be translated is not translated at
all. Not only that, the word which is supposedly a plural word now is translated as a
singular word (see NCX.10).
73
4.2.3.2 Above Word-level Equivalence
Perpustakaan Unika
Among ten non academic complex sentences, the writer found six above word
1. The key you saw was for the dining car, which we've already sampled
Kunci yang Anda lihat adalah untuk mobil makan, yang kita telah sampel
(NCX.3)
The phrase 'dining car' is translated word per word. This phrase is translated as
'mobil makan' when it should be translated as 'gerbong makan' The phrase which
means 'The body of train that is used for eating purpose' after the translation
process means 'a car that is used for eating'.
2. He opened his eyes, blinking, assessing his surroundings as quickly as his muddled
mind could manage, which wasn't very quickly at all Dia membuka mata,
berkedip, menilai lingkungannya secepat kacau pikiran bisa mengelola, yang
tidak sangat cepat sekali (NCX.4)
The phrase ' his muddled mind' word order is in reverse order, it should be
translated as 'pikiranya yang kacau'. Moreover the phrase 'his muddled mind' has
an adjective possessive in the original sentence, but in the translation result it does
not have any word or phrase that resemble as the adjective possessive.
3. He turned and walked across the room, telling himself it would be a waste of
ammo, which was true Dia berbalik dan berjalan melintasi ruangan,
mengatakan dirinya akan membuang-buang amunisi, yang benar (NCX.5)
74
Perpustakaan Unika
The translation of the clause 'telling himself it would be a waste of ammo' and the
phrase 'which was true' is a word per word translation. The clause ''telling himself
it would be a waste of ammo ', which should be translated as 'mengatakan pada
dirinya sendiri bahwa itu akan menghabis- habiskan amunisi' has the meaning of '
he is telling himself that something will only wasting the ammunition' but the
translation result which was provided by 'Google Translate' (mengatakan dirinya
akan membuang-buang amunisi) means 'telling his own self he will wasting
ammunition without something that triggers it'. The phrase 'yang benar', in the
original sentence refer to the word 'waste of ammo' but the translation result is not
refer to anything, the translation result should be 'yang mana hal itu benar'.
4. Not the most inviting environment, but Billy didn't see any zombies or leech
people, which made it a hell of a lot better than the last place Bukan
lingkungan yang paling mengundang, tetapi Billy tidak melihat zombie atau
orang lintah, yang membuatnya menjadi neraka jauh lebih baik daripada tempat
terakhir (NCX.7)
The phrase 'a hell of a lot better ', which should be translated as ' jauh lebih baik',
undergoes word per word translation. This phrase which means 'far better than' in
the original sentence, but now has the meaning of 'Hell is becoming a better
place'.
5. There was a big metal door on the south wall, set back into a recessed
doorway, which appeared to be the only other way out Ada sebuah pintu
75
Perpustakaan Unika
besi besar di dinding selatan, mengatur kembali menjadi pintu tersembunyi, yang
tampaknya menjadi satu-satunya cara lain di luar (NCX.8)
The clause 'set back into a recessed doorway' is only translated word per word, the
translation result also does not undergo any word rearranging process. It would be
better if it were translated as ' terletak di ambang pintu yang tersembunyi'.

Another similar mistake is the phrase 'other way out' which should be translated
as ' jalan keluar yang lain' instead of ' cara lain di luar'.
6. What little light there was came from a few stray shafts of sunlight coming
down from the motile ceiling, which meant they were only a single floor from
the surface Apa ada sedikit cahaya berasal dari sebuah poros tersesat
beberapa sinar matahari turun dari langit-langit motil, yang berarti mereka
hanya lantai yang tunggal dari permukaan (NCX.9)
The first half of the sentence is unclear. The writer can not determine what is the
meaning of the translation result, but there is one thing that is clear; the translation
happen in this part is only a word per word translation. It would be better if it
were translated as ' Beberapa berkas sinar cahaya matahari datang dari langitlangit motile'. The phrase 'a single floor' in the second half of the sentence, which
means 'one floor', after the translation process means 'a floor', it would be better if
the phrase were translated as ' satu lantai dari permukaan' instead of ' lantai yang
tunggal dari permukaan'.
76
Perpustakaan Unika
There are some cases that a mistake happens not to a phrase, but it happens to a clause.
The most severe case happens to (NCX.9). Half of the sentence is unclear, it almost
impossible to determine its meaning after the translation process.

Among ten non academic complex sentences, the writer found three textual
1. Unless Billy and Rebecca were entirely moronic, which they'd already
proven they were not, they would figure out how to trigger the structure's
rotation Kecuali Billy dan Rebecca sepenuhnya tolol, yang mereka sudah
terbukti mereka tidak, mereka akan mencari cara untuk memicu rotasi struktur's
(NCX.6)
The word 'proven' which should has the meaning of 'giving a proof by
themselves', now has a slight different meaning, the translation result should be
changed as 'buktikan'. The word 'figure' can be translated as 'mencari' but in this
context the result that is more effective is a word that has the meaning of 'know',
so it should be translated as 'menemukan'.
2. Not the most inviting environment, but Billy didn't see any zombies or leech
people, which made it a hell of a lot better than the last place Bukan
lingkungan yang paling mengundang, tetapi Billy tidak melihat zombie atau
77
Perpustakaan Unika
orang lintah, yang membuatnya menjadi neraka jauh lebih baik daripada tempat
terakhir (NCX.7)
the phrase 'leech people' can be translated as 'orang lintah', but the word 'people'
is not quite fit as the translation result, so it should be changed as 'manusia,' which
is the synonym of the word 'orang'.
3. Enrico asked the pilot, who was tapping at controls and flipping switches
Enrico tanya pilot, yang sadap di kontrol dan flipping switch (NCX.10)
The word 'tapping' can be translated as 'sadap' but in this context the result that is
wanted is a word that has the meaning of 'making a tap sound', the word can be
translated as ' mengetuk-ketuk', which has a simmilar meaning with the original
sentence.
Mostly the textual mistake happens in word class, but in (NCX. 7) it happens to a phrase,
this proves that textual mistake is not only happening to word.
4.2.4 Compound Complex Sentences

In this section the writer analyzed ten compound-complex sentences taken from
the novel 'Resident Evil: Genesis' by Decandido (2004). These sentences will be analyzed
using four parameters; Word-level equivalence, above word-level equivalence,
grammatical equivalence, and textual equivalence
78
Perpustakaan Unika
There is one sentence that is correctly translated in this section, for more detail

Among ten non academic compound complex sentences, the writer found four
word level equivalence translation mistakes. They are as follows:
1. As it turned out, the drug did exactly what it was supposed to do - but it also
caused renal failure, and in Teresa Hardy's case, the damage had been
severe. Ternyata, obat itu persis apa yang seharusnya dilakukan - tetapi juga
menyebabkan gagal ginjal, dan dalam kasus Teresa Hardy, kerusakan sudah
parah. (NCDX.3)
The word 'did', located between the word 'drug' and 'exactly', is omitted from the
translation result, this caused the translation result become wrong. If the word is
not omitted, the translation should become ' obat itu melakukan apa yang harus
dilakukanya' instead of ' Ternyata, obat itu persis apa yang seharusnya
dilakukan'.
2. There was no need to talk about what would happen when they reached the
train ... and though the channel was scrambled, it was best not to say more
than was necessary, anyway. Tidak perlu bicara tentang apa yang akan
terjadi ketika mereka mencapai kereta api ... dan meskipun channel itu bergegas,
itu baik untuk tidak mengatakan lebih daripada yang diperlukan, anyway.
(NCDX.6)
79
Perpustakaan Unika
The word 'anyway', which located at the back of the sentence, is not translated.
The translation of the clause 'it was best not to say more than was necessary,
anyway' should be 'lagipula lebih baik untuk tidak mengatakan lebih dari
secukupnya', instead of ' itu baik untuk tidak mengatakan lebih daripada yang
diperlukan, anyway.' Another thing is the word 'anyway' is a form of expression.
3. Lord knew there were enough people who had reason to sabotage Umbrella,
but it wasn't easy to obtain even a low-level clearance pass for any of the
Raccoon labs Tuhan tahu ada cukup banyak orang yang punya alasan untuk
sabotase Payung, tapi itu tidak mudah untuk mendapatkan bahkan lulus tingkat
rendah clearance untuk salah satu laboratorium Raccoon (NCDX.7)
The word 'Umbrella' functions as a name of something, which is why it should not
change after the translation process. This case is similar to (NS.8). The word
'sabotage' is a verb, but in the result it becomes a noun (sabotase), the translation
result should be 'mensabotase' which is a verb.
4. One of the cuffs was off, the right one, bashed open when the jeep had gone
over, but unless he wanted to be wearing a jangly and rather incriminating
bracelet, he had to get the other one off Salah satu borgol itu lepas, yang
tepat, bashed terbuka ketika jip sudah berakhir, tetapi jika ia ingin mengenakan
jangly dan agak memberatkan gelang, ia harus mendapatkan yang lain dari
(NCDX.9)
80
Perpustakaan Unika
The word 'bashed' and 'jangly' are not translated at all. The word 'bashed' can be
translated as 'terhantam', while the word 'jangly' can be translated as

'bergemerincing'
The case happen in (NS.8) is repeated again in this section. This show that 'Google
Translate' has difficulty in selecting which word which must be translated and which one
should not.

Among 10 non academic compound complex sentences, the writer found eight
above word level equivalence translation mistakes, they are as follows:
1. It was highly improbable that one of the virus carriers would just happen to
blunder their way to one of Raccoon's other labs, but it was possible. Hal
ini sangat tidak mungkin bahwa salah satu pembawa virus hanya akan terjadi
pada kesalahan cara mereka ke salah satu laboratorium lain Raccoon, tetapi hal
itu mungkin (NCDX.1)
The phrase 'would just happen' become strange after translation. It will be better if
the translation result is 'begitu saja' instead of 'hanya akan terjadi'. It is clear that
the result provided by 'Google Translate' is only a word per word translation, not a
phrase translation. The phrase 'blunder their way', which should be translated as
'bersusah payah', also experience the similar case.
81
Perpustakaan Unika
2. Billy was gone, lost to the fast-moving sewer beneath the cable car platforms,
which had swept him and that monster in this direction, but he was alive,
and she was going to find him Billy sudah pergi, hilang ke saluran
pembuangan yang bergerak cepat di bawah platform kabel mobil, yang telah
menyapu dia dan rakasa bahwa dalam arah ini, tapi dia masih hidup, dan ia akan
menemukan dia (NCDX.2)
The phrase 'cable car' clearly is translated word per word. The phrase which
means 'a vehicle which hangs from and is moved by a cable 8 ' now means 'a cable
of a car', so the translation result should be changed as 'kereta gantung' instead of
'kabel mobil. The translation result of the phrase 'that monster' also is translated
word per word. Not only that, the word 'monster' is also translated incorrectly, it
will be better if the phrase is translated as 'monster itu' instead of 'rakasa bahwa'.
3. Enrico stepped to her side, and within seconds, five more beams of light had
jerked toward the gleam of metal she'd seen, illuminating what was, in fact, a
narrow dirt road - and an overturned jeep Enrico melangkah ke sisinya,
dan dalam beberapa detik, lima lagi berkas cahaya telah menyentakkan terhadap
sinar logam dia melihat, menerangi apa yang, sebenarnya, jalan tanah sempit dan sebuah jip terbalik (NCDX.4)
The clause 'she'd seen' is not translated completely. The ('d) in the word 'she'd' is
omitted from the translation result. If the word is not omitted, the result will be '
yang telah ia lihat' instead of 'dia melihat'.
82
Perpustakaan Unika
4. She'd been so focused on the coming storm that for a crazed split second, she
thought it was thunder, even as the helicopter tipped wildly and dropped, a
terrible rising, clattering whine filling the cabin, the floor vibrating beneath
her boots Dia begitu terfokus pada badai yang datang untuk split gila kedua,
ia pikir itu guntur, bahkan ketika helikopter itu miring liar dan menjatuhkan,
yang merengek mengerikan naik, derap mengisi kabin, lantai bergetar di bawah
sepatu botnya (NCDX.5)
The phrase 'Crazed split second' which should be translated as 'sekilas' is only
undergoing word per word translation, and become 'split gila kedua' after the
translation process. For the phrase 'crazed split second', it seems that 'Google
Translate' only recognize 'Crazed split' as the phrase, while the word 'second' is
not part of the phrase. The phrase 'a terrible rising' which is only consist of three
words after the translation process become consist of four words with a strange
meaning, it will be better if the translation result is changed as 'keributan yang
sangat buruk' rather than ' yang merengek mengerikan naik'.
5. There was no need to talk about what would happen when they reached the
train ... and though the channel was scrambled, it was best not to say more
than was necessary, anyway. Tidak perlu bicara tentang apa yang akan
terjadi ketika mereka mencapai kereta api ... dan meskipun channel itu bergegas,
itu baik untuk tidak mengatakan lebih daripada yang diperlukan, anyway.
(NCDX.6)
83
Perpustakaan Unika
Both of 'the channel was scrambled' and 'it was best' are translated word per word,
when it should be translated as a clause. As the result, the translation result
becomes wrong. The result will be better if the clause 'the channel was scrambled'
is translated as 'sinyalnya diacak' and the clause 'it was best' is translated as 'lebih
baik'.
6. Lord knew there were enough people who had reason to sabotage Umbrella,
but it wasn't easy to obtain even a low-level clearance pass for any of the
Raccoon labs Tuhan tahu ada cukup banyak orang yang punya alasan untuk
sabotase Payung, tapi itu tidak mudah untuk mendapatkan bahkan lulus tingkat
rendah clearance untuk salah satu laboratorium Raccoon (NCDX.7)
There is a similarity of the mistake in this translation with (CDX.5) The phrase'
low-level clearance pass' is translated word per word, and for the word order,
'Google Translate' only recognize 'low-level ' as the phrase, while the word
'clearance' is treated as an another word. For a better result, the translation result
can be changed as 'akses tingkat rendah' instead of ' lulus tingkat rendah
clearance'. The phrase ' any of the Raccoon labs', which should be translated as '
laboratorium-laboratorium Racoon manapun' instead of 'salah satu laboratorium
Raccoon', has a different meaning after the translation result. In the original
sentence, the phrase has the meaning of 'every labs of Racoon', but after the
translation process it became 'one of the labs of Racoon'.
84
Perpustakaan Unika
7. It was his life's work, but he'd destroy it before he'd let it be taken away
Itu adalah pekerjaan hidupnya, tapi ia menghancurkannya sebelum dia biarlah

diambil (NCDX.8)
Similar case to (CDX.4), the ('d) from 'he'd' is omitted from the translation result.
The result will be 'ia akan menghancurkanya' if the word ('d) is not omitted. The
phrase ' let it be taken away' is also wrong, the translation result should be
'membiarkanya diambil' rather than 'biarlah diambil'. The result provided by
'Google Translate' is not referring to any specific object, even though in the
original sentence this phrase refers to 'His life work'.
8. One of the cuffs was off, the right one, bashed open when the jeep had gone
over, but unless he wanted to be wearing a jangly and rather incriminating
bracelet, he had to get the other one off Salah satu borgol itu lepas, yang
tepat, bashed terbuka ketika jip sudah berakhir, tetapi jika ia ingin mengenakan
jangly dan agak memberatkan gelang, ia harus mendapatkan yang lain dari
(NCDX.9)
The phrase 'gone over' and ' a jangly and rather incriminating bracelet' is
translated word per word, also there is no word rearranging process for the two
phrases. The phrase 'gone over' should be translated as 'terbalik' instead of sudah
berakhir' and the phrase 'a jangly and rather incriminating bracelet' will be better
translated as ' gelang yang bergemerincing dan memberatkan itu' rather than ' a
jangly and rather incriminating bracelet'. The clause 'get the other one off' also
experiences the similar case. The translation result should be 'melepaskan sisi
85
Perpustakaan Unika
satunya, instead of ' ia harus mendapatkan yang lain dari'. The translation result
of this clause which is provided by 'Google Translate' also has a different meaning
from the original sentence. The original sentence means 'to take off' while the
translation result means 'to own'.
There is something new in this section. It seems that 'Google Translate' often recognize a
phrase is only consist of two word, while the rest is another word which is not a part form
the phrase. Consequently, 'Google Translate' only rearrange the word it recognize as the
phrase, and the other word will be put after the words that is recognized as the phrase.

Among ten non academic compound complex sentences, the writer found one
grammatical equivalence translation mistake. It is as follows:
1. Enrico stepped to her side, and within seconds, five more beams of light had
jerked toward the gleam of metal she'd seen, illuminating what was, in fact, a
narrow dirt road - and an overturned jeep Enrico melangkah ke sisinya,
dan dalam beberapa detik, lima lagi berkas cahaya telah menyentakkan terhadap
sinar logam dia melihat, menerangi apa yang, sebenarnya, jalan tanah sempit dan sebuah jip terbalik (NCDX.4)
86
Perpustakaan Unika
The verb 'jerked' is in the passive form but 'Google Translate' perceived it as an
active form; the correct translation should be 'disentakkan' instead of
'menyentakkan'.
'Google Translate' translates a passive verb into an active verb. It seems that 'Google
Translate' really have a problem in translating a word which have the suffix (-ed).

Among ten non academic compound complex sentences, the writer found one
textual equivalence translation mistake. It is as follows:
1. Billy was gone, lost to the fast-moving sewer beneath the cable car platforms,
which had swept him and that monster in this direction, but he was alive,
and she was going to find him Billy sudah pergi, hilang ke saluran
pembuangan yang bergerak cepat di bawah platform kabel mobil, yang telah
menyapu dia dan rakasa bahwa dalam arah ini, tapi dia masih hidup, dan ia akan
menemukan dia (NCDX.2)
The word 'gone' which means 'missing' or 'disappearing', now has a slight
different meaning if the result is as it is (pergi), The translation result should be
'hilang' which has similar meaning as the original sentence instead of 'pergi'
which means 'went'.
87
Perpustakaan Unika
There is nothing new in this section. 'Google Translate' still has problem to give the
translation result which is suited to the context of the text.
4.3 General Findings

After conducting this study, the writer found that the highest success rate of
translation is in the Non academic sentence type. The other type of sentences has been
translated ineffectively and the success rate is close to zero percent. The table below is
the complete list of the translation success rate:
correctly
Sentence Type
sentences
Rate
translated
Academic Simple Sentence
10
0%
Academic Compound Sentence
10
1.25%
Academic Complex Sentence
10
0%
Academic Compound Complex Sentence
10
0%
Non Academic Simple Sentence
10
5%
Non Academic Compound Sentence
10
0%
Non Academic Complex Sentence
10
1.25%
Non Academic Compound Complex Sentence
10
1.25%
Overall
80
8.75%
Table 4.3 Translation success rate of 'Google Translate'
Out of ten Academic Simple Sentences, Academic Complex Sentences, Academic

compound - complex, and Non Academic Compound Sentences which used in this study,
there is none which is correctly translated (see appendix 1, 10, 14, and 22). Therefore
88
Perpustakaan Unika
these types of sentences have 0% successful translation percentage. While Academic

Compound Sentences, Non Academic Complex Sentences, and Non Academic
compound - Complex Sentences have one out of the ten sentences correctly translated
(see appendix 6, 26, and 30), so (1/80) X 100% = 1.25%. The highest successful
translation percentage is achieved by Non Academic Simple Sentences. It has four
correctly translated sentences, out of ten material sentences. Therefore (4/80) X 100% =
5%. Overall, out of eighty sentences which used as the material, " Google Translate" only
able to translate seven sentences correctly, while the other sentences has problem in the
translation, so (7/80) X 100%
= 8.75%. Therefore, the writer assumes that the
effectiveness of "Google translate" is very low.

The writer also found the mistakes made by 'Google Translate' in each parameter
that is used for analyzing the sentences; they are as follows:
4.3.1
1.
Word Equivalence:
Translating Plural word into singular word
There are many cases of the plural word, which isasupposedly translated
as plural word, is translated as singular word. This mistake happens to: [S1, S2,
S3, S4, S6, S8, S9, S10, CD2, CD3, CD4, CD5, CD7, CD8, CD9, CX1, CX2,
CX3, CX4, CX5, CX6, CX7, CX9, CX10, CDX1, CDX2, CDX3, CDX4, CDX5,
CDX6, CDX7, CDX8, NCD10, NCX10]
89
2.
Omitted word
Perpustakaan Unika
The word which supposed to be translated is omitted from the translation

result. Omittance of any word from the original sentence mostly makes the
translation result has different meaning. This mistake happens to: [S3, CD6, CD7,
CX1, CX3, CDX7, NCD8, NCDX3]
3.
Translating one functional word to different functional word

There are three different kinds of mistake, translating adjective as a noun
[CD2], translating verb as a conjunction [CD8], and translating verb as a noun
[CDX3, NCD8, NCDX7]
4.
Unable to choose which word should be translated, which one should not.
The word which should be translated is not translated in the translation
result. The word still exists in the translation result, but it retained its original
form. There is an exception where the word cannot be translated, but it mostly
occurs on proper noun. This mistake happens to: [S3, S6, NCD8, NCX6, NCX10,
NCDX6, NCDX9]. There are also two cases where the word which should not be
translated is translated: [NS8, NCD7]
5.
The translation result has different meaning after translation

This is the most basic form of the mistake in translation. If the translated
word has a different meaning from the original sentence, the information which
should be delivered will be wrong. This mistake happens to: [S2, CX5, CX7,
90
Perpustakaan Unika
CX8, CDX2, CDX4, CDX7, CDX9, CDX10, NS9, NCD1, NCD4, NCD5, NCD9,
NCD10, NCX3]
4.3.2
1.
Above Word Equivalence

Unable to choose which word should be translated, which one should not.
The word which should be translated is not translated in the translation
result. The word still exists in the translation result, but it retained its original
form. This mistake happens to: [S5, CD3]
2.
Using word per word translation

Phrase and clause should be translated as a phrase and clause, not as a set
of word. Translating phrase or clause by word per word translation would cause
the phrase or clause has a strange meaning in the translation result. This mistake
happens to: [S4, CD2, CD3, CD5, CD7, CD8, CD9, CX5, CX6, CX10, CDX5,
CXD6, CDX7, CDX8, CDX9, NS4, NS5, NS8, NCD1, NCD6, NCD7, NCX3,
NCX4, NCX5, NCX7, NCX8, NCX9, NCDX1, NCDX2, NCDX5, NCDX6,
NCDX7, NCDX8, NCDX9]
3.
Wrong word order

In translating phrase or clause, word order is one of the main aspects to be
considered. Wrong word order will give a different meaning to the phrase or
clause. This mistake happen to: [S4, CD2, CD3,CD5, CD7, CD8, CX5, CX10,
CDX5, CDX6, CDX7, CDX8, NS4, NS5, NS6, NCD1, NCD6, NCD7, NCX4,
91
Perpustakaan Unika
NCX5, NCX7, NCX8, NCX9, NCDX1, NCDX2, NCDX5, NCDX6, NCDX7,

NCDX8, NCDX9]
4.
Part of the phrase or clause is omitted and not translated

There are some cases that the phrase is translated, but a word which is the
part of the phrase or clause is omitted or not translated. This makes the phrase or
clause become strange in the translation result. The mistake of omitting the part of
phrase or clause happens to: [NCD3, NCDX8], and the mistake of not translating
the part of the phrase or clause happens to: [NCX6]
4.3.3
1.
Gramatical Equivalence
Translating passive form into active form and vice versa
This mistake mostly happen to the word with the suffix '-ed'. Both second
of active verb and passive verb use this suffix, and that is the main problem why
this mistake happens. This mistake happens to: [S7, CD1, CDX9, NS2, NCDX4]
4.3.4
1.
Textual Equivalence:
It cannot match the translation result with the context
Aside from correctly translating a word, phrase, or clause, translation
should also correctly match to the context, in order to convey the right
information. There are some mistakes where the translation result does not match
with the context of the sentence. These cases happen to: [S10, CX7, NCD5,
NCD10, NCX6, NCX7, NCX10, NCDX2]
92
CHAPTER V
Perpustakaan Unika
CONCLUSION AND SUGGESTION
5.1 Conclusion
After analyzing and interpreting the data, the writer comes to the conclusion that
the effectiveness of 'Google Translate' in translating English sentences to Indonesian
sentences is very low (8.75%) . Most of the sentences, which are used as the data, are
translated incorrectly. Based on chapter 4 (table 4.3), the highest percentage of correct
translation is at Non Academic simple sentences (5%), while most of the other sentences
percentage are very low ( 1.25%). The mistakes found in the translation result mostly
are mistakes at minor detail, so the translation mistakes are not too serious. But, the
writer also finds a case where the mistake is very serious. Moreover, the writer finds that
there are some cases where the 'Google Translate' often makes mistake. Each parameter
has at least one case in which 'Google Translate' often makes mistake.
5.2 Suggestion
From the result of the study, the writer suggests to the readers to use "Google
Translate" only for translating non academic simple sentences, since it has the highest
percentage of correct translation among the eight types of sentences. The writer also
suggests not to use the 'Google Translate' as a primary translation tool, but as a
translation aid. Because the mistakes that 'Google Translate' makes are not too serious, it
can still be used, but it must be accompanied by Human Translator, who should edit the
post-translation result.
93
Perpustakaan Unika
For further studies, the writer suggests two main points that can be used as
consideration:
1. For Qualitative Study, the writer suggests the use of another theory of translation,
in order to make a comparison with this study.
2. For Quantitative Study, the writer suggests that the study is done on the mistakes
that the writer found in this study.
94
REFERENCES
Perpustakaan Unika
Alwi, H., Dardjowidjojo S., Lapoliwa H., and Moeliono A.M. (2003). Tata
Bahasa Baku Bahasa Indonesia. Jakarta: Balai Pustaka
Baker, M. (1992).In Other Words a Coursebook on Translation.London:
Routledge. 10, 46, 82, 180
Bassnett,S. (1991). Translation Studies Revised Edition. London: Routledge
Cohen.L, Manion, L., and Morrison, K.(2007).Research Method in Education 6th
Edition. London: Routledge
Cribb, V. M. (2000, autumn). Machine Translation: The alternative for 21st
century?. TESOL QUARTERLY, 34 (3)
Creswell, J. W.(2001).Research Design Second Edition. London: Sage
Publications.
Frank, M. (1972).Modern English: A Practical Reference Guide. New Jersey:
Prentice Hall International.
Hutchins, J. (1995). Reflection on the History and Present State of Machine
Translation. Retrieved April 16, 2010 from http:// www.citeseerx.com .
Hutchins, J (1999). The development and use of machine translation systems
and computer-based translation tools, 2-13. Retrieved April 13,
2009 from http:// www.citeseerx.com
Hutchins, J (2002). Machine Translation: A Brief History. Retrieved September 4,
2010 from http://aymara.org/biblio/mtranslation.pdf
Isabelle Ee Hsia, T.,and Tang Enya K (2004). A Discourse of Cognitive Paradigm
on the Accuracy of Machine Translation. 6. Retrieved April 15, 2009 from
http:// www.citeseerx.com
Klavans, J. in OGrady,W., Michael D. and Francis K. (Eds.). (1996).
Contemporary Linguistics: An Introduction. London: Longman.
Newmark, P. (1988). A Text Book of Translation. Hertfordshire: Prentice Hall
International.
Vasconellos,M, and L.Chris Miller (2010). Recent Trends in Machine
Translation. 1. Retrieved April 16, 2010. from http:// www.citeseerx.com
Walter, E., Kate W., Melissa G. (Eds.). (2008). Cambridge Advanced Learner's
Dictionary Third Edition. Cambridge University Press: Cambridge.
Yusuf, S.(1994).Teori Terjemah: Pengantar ke Arah pendekatan Linguistik
dan Sosiolinguistik. Bandung: CV. Mandar Maju.
Perpustakaan Unika
APPENDIX
Appendix 1: Table 4.1.1Translation Mistake in Academic Simple Sentences

No
SOURCE
TRANSLATED
S.1
We examined the effects of

two parameters
Kami meneliti efek dari dua

parameter
S.2
We established activity
patterns for six people in the
space
Kami mendirikan pola

aktivitas selama enam orang
di ruang
The activity rate, the

building surface for
exchange, and the zone for
resuspension, all may be
scheduled within the model
to vary over time
Tingkat aktivitas, permukaan

bangunan untuk pertukaran,
dan zona untuk resuspension,
semua dapat dijadwalkan
dalam model bervariasi dari
waktu ke waktu.
The resulting system of

coupled ordinary
differential equations
describes mass conservation
in the storage locations.
Sistem yang dihasilkan

digabungkan persamaan
diferensial biasa menjelaskan
konservasi massa di lokasi
penyimpanan.
Figure 2 shows the floor

plan of the office suite and
the location of the anthrax
source
Gambar 2 menunjukkan
denah office suite dan lokasi
sumber anthrax
The exchange and

resuspension processes vary
directly with the activity
level
Proses pertukaran dan

resuspension bervariasi
langsung dengan tingkat
aktivitas
we partitioned floor
surfaces into tracked and
untracked areas
kita dipartisi permukaan

lantai ke daerah dilacak dan
tidak terlacak
Deposition of airborne
particles occurs uniformly
on surfaces in both
categories
Deposisi partikel udara

terjadi secara seragam pada
permukaan di kedua kategori
S.9
We established a set of
scenarios
Kami mendirikan satu set

scenario
S.10
For the present paper we

describe the results obtained
from one scenario
Untuk tulisan ini kami akan

menjelaskan hasil yang
diperoleh dari satu skenario
S.3
S.4
S.5
S.6
S.7
S.8
Word
A.Word
Grammar
Textual
Perpustakaan Unika
Appendix 2: Table 4.1.1.1 Mistake in academic simple sentences word-level

No
SOURCE
TRANSLATED
Mistake point(s)
S.1
We examined the effects of

two parameters
Kami meneliti efek dari dua

parameter
1.
Effects
S.2
We established activity
patterns for six people in the
space
Kami mendirikan pola

aktivitas selama enam orang
di ruang
1.
2.
3.
Patterns
Establish
For
S.3
The activity rate, the

building surface for
exchange, and the zone for
resuspension, all may be
scheduled within the model
to vary over time
Tingkat aktivitas, permukaan

bangunan untuk pertukaran,
dan zona untuk resuspension,
semua dapat dijadwalkan
dalam model bervariasi dari
waktu ke waktu.
1.
2.
Resuspension
To
S.4

coupled ordinary differential
equations describes mass
conservation in the storage
locations.

penyimpanan.
1.
2.
Equations
locations
S.6
The exchange and

resuspension processes vary
directly with the activity
level
Proses pertukaran dan

resuspension bervariasi
langsung dengan tingkat
aktivitas
1.
2.
Processes
Resuspension
S.8
Deposition of airborne
particles occurs uniformly
on surfaces in both
categories
Deposisi partikel udara terjadi

secara seragam pada
permukaan di kedua kategori
1.
Particles
S.9
We established a set of
scenarios
Kami mendirikan satu set

scenario
1.
Scenarios
S.10

from one scenario

1.
Results
Appendix 3: Table 4.1.1.2 Mistake in academic simple sentences above wordlevel

No
S.4
SOURCE
coupled ordinary differential
equations describes mass
conservation in the storage
locations.
TRANSLATED
penyimpanan.
S.5
Figure 2 shows the floor

plan of the office suite and
Gambar 2 menunjukkan
denah office suite dan lokasi
Mistake point(s)
1. coupled ordinary differential
equations
1.
Office Suite
the location of the anthrax

source
Perpustakaan Unika
sumber anthrax
Appendix 4: Table 4.1.1.3 gramatical Mistake in academic simple sentences

No
S.7
SOURCE
we partitioned floor surfaces
into tracked and untracked
areas
TRANSLATED
kita dipartisi permukaan lantai
ke daerah dilacak dan tidak
terlacak
Mistake point(s)
1. Partitioned
Appendix 5: Table 4.1.1.4 Textual Mistake in academic simple sentences

No
S.10
SOURCE
from one scenario
TRANSLATED
Mistake Point(s)
1. Paper
Appendix 6: Table 4.1.2.1 Translation Mistake in Academic Compound

Sentences
No
CD.1
CD.2
CD.3
CD.4
SOURCE
We modeled the spread of
anthrax for a hypothetical
office suite and estimated
the distribution of mass
and resulting occupant
exposures.
MT research had not
looked at the economic use
of existing less than
perfect systems, and it had
disregarded the needs of
translators for computerbased aids.
Now it is recognised that
MT systems should be
designed ab initio for
controlled language, and a
number of independent
companies outside the
academic MT research
community have been
doing so in recent years
(e.g. Volmac).
It is to be hoped that this
research will both
accelerate the development
of MT prototypes and
improve the operational
flexibility of commercial
systems.
TRANSLATED
Kami model penyebaran
anthrax untuk office suite
hipotetis dan diperkirakan
distribusi massa dan
eksposur penghuni yang
dihasilkan.
MT Penelitian tidak melihat
penggunaan ekonomi yang
ada 'kurang sempurna'
sistem, dan itu telah
mengabaikan kebutuhan
penerjemah untuk alat bantu
berbasis komputer.
Sekarang diakui bahwa
sistem MT harus dirancang
initio ab untuk bahasa
dikontrol, dan sejumlah
perusahaan independen di
luar komunitas riset MT
akademik telah
melakukannya dalam
beberapa tahun terakhir
(misalnya Volmac).
Perlu Diharapkan penelitian
ini akan baik mempercepat
pengembangan prototipe
MT dan meningkatkan
fleksibilitas operasional
sistem komersial.
Word
A.Word
Grammar
Textual
CD.5
CD.6
CD.7
CD.8
CD.9
CD.10
Logos systems were based

on a direct translation
approach, but later systems
are closer to a transfer
design and incorporate
sophisticated means for
recording and applying
semantic features.
Of course, the idea or
dream of translating the
spoken word automatically
was present from the
beginning (Locke 1955),
but it has remained a
dream until now
The rule-based approach
was most obvious in the
dominant transfer systems
(Ariane, Metal, SUSY, Mu
and Eurotra), but it was at
the basis of all the various
interlingua systems
Statistical methods were
common in the earliest
period of MT research, in
the 1960s (see sect.3
above), but the results had
been generally
disappointing
Furthermore, I shall not be
describing any particular
system of whatever kind,
past or present, nor shall I
be giving details of
methods of computation
analysis or processing
The Georgetown group
was the largest in the USA
and there were
considerable differences of
viewpoint among its
members
Logos sistem didasarkan

pada pendekatan terjemahan
langsung, tetapi kemudian
sistem ini lebih dekat
dengan desain transfer dan
memasukkan alat canggih
untuk merekam dan
menerapkan fitur semantik.
Tentu saja, ide atau impian
menerjemahkan kata yang
diucapkan secara otomatis
hadir dari awal (Locke
1955), tetapi tetap mimpi
sampai sekarang
Pendekatan berbasis aturan
paling jelas dalam sistem
transfer dominan (Ariane,
Metal, Susy, Mu dan
Eurotra), tetapi pada dasar
dari semua sistem berbagai
Interlingua
Metode statistik yang umum
pada awal periode penelitian
MT, pada tahun 1960 (lihat
sect.3 di atas), namun
hasilnya telah
mengecewakan umumnya
Selain itu, saya tidak akan
menjelaskan semua sistem
tertentu jenis, masa lalu atau
sekarang apa pun, tidak akan
saya akan memberikan
rincian metode analisis
perhitungan atau pengolahan
Kelompok Georgetown
adalah yang terbesar di
Amerika Serikat dan ada
perbedaan besar sudut
pandang antara para
anggotanya
Perpustakaan Unika
Appendix 7: Table 4.1.2.1 W ord-level Mistake in academic compound

sentences
No
CD.1
CD.2
SOURCE
the distribution of mass and
resulting occupant
exposures.
MT research had not looked
at the economic use of
existing less than perfect
systems, and it had
TRANSLATED
distribusi massa dan eksposur
penghuni yang dihasilkan.
ada 'kurang sempurna' sistem,
dan itu telah mengabaikan
kebutuhan penerjemah untuk
Mistake point(s)
1. modeled
1.
systems
CD.3
CD.4
CD.5
CD.6
CD.7
CD.8
CD.9
Perpustakaan Unika

community have been doing
so in recent years (e.g.
Volmac).
It is to be hoped that this
research will both accelerate
the development of MT
prototypes and improve the
operational flexibility of
commercial systems.
alat bantu berbasis komputer.

akademik telah
melakukannya dalam
(misalnya Volmac).
Perlu Diharapkan penelitian
ini akan baik mempercepat
pengembangan prototipe MT
dan meningkatkan
fleksibilitas operasional
sistem komersial.
1.
Systems
1.
2.
Prototypes
Systems

semantic features.
Of course, the idea or dream
of translating the spoken
word automatically was
present from the beginning
(Locke 1955), but it has
remained a dream until now
Logos sistem didasarkan pada

pendekatan terjemahan
sistem ini lebih dekat dengan
desain transfer dan
untuk merekam dan
Tentu saja, ide atau impian
menerjemahkan kata yang
diucapkan secara otomatis
hadir dari awal (Locke 1955),
tetapi tetap mimpi sampai
sekarang
1.
2.
Systems
features
1.
it

interlingua systems
the 1960s (see sect.3 above),
but the results had been
generally disappointing
be giving details of methods
of computation analysis or
processing

Metal, Susy, Mu dan
Interlingua
hasilnya telah mengecewakan
umumnya
1.
systems
1.
2.
Methods
Results
1.
2.
3.
4.
Any
Nor
Details
Methods
Perpustakaan Unika
Appendix 8: Table 4.1.2.2 Above word-level Mistake in academic compound

sentences
No
CD.2
CD.3
CD.5
CD.7
CD.8
CD.9
SOURCE
MT research had not looked
at the economic use of
existing less than perfect
systems, and it had
community have been doing
so in recent years (e.g.
Volmac).
TRANSLATED
ada 'kurang sempurna'
sistem, dan itu telah
mengabaikan kebutuhan
penerjemah untuk alat bantu
berbasis komputer.
akademik telah
melakukannya dalam
(misalnya Volmac).

semantic features.
interlingua systems
the 1960s (see sect.3
above), but the results had
been generally
disappointing
be giving details of methods
of computation analysis or
processing
Mistake point(s)
1. MT Research
2. existing less than perfect
systems
1.
2.
Ab initio
independent companies outside
the academic MT research
community
Logos sistem didasarkan

pada pendekatan terjemahan
sistem ini lebih dekat dengan
desain transfer dan
untuk merekam dan
Metal, Susy, Mu dan
Interlingua
hasilnya telah
mengecewakan umumnya
1.
Later systems
1.
various Interlingua systems
1.
Generally disappointing

1.
Whatever kind
Perpustakaan Unika
Appendix 9: Table 4.1.2.3 Grammatical Mistake in academic compound

sentences
No
CD.1
SOURCE
the distribution of mass and
resulting occupant
exposures.
TRANSLATED
distribusi massa dan
eksposur penghuni yang
dihasilkan.
Mistake point(s)
1. Estimated
Appendix 10: Table 4.1.3 Translation Mistake in Academic Complex Sentences

No
CX.1
CX.2
CX.3
CX.4
CX.5
CX.6
SOURCE
As a consequence, they are
sometimes unable to
distinguish between the
failings of particular
systems and the problems
which the best system
would have.
These effects are also
illustrated by the data in
Table 2, which show a
slight decrease in the mass
fraction for all tracked
surfaces over time
For our simulations we
postulated that the source
is an envelope containing 1
gram of anthrax opened in
Room 4.
These simulations help
identify critical
information or processes
that contribute most to
predicted outcomes or to
uncertainties.
Phrase translation pairs
that are generated by the
method described in this
paper are finally scored
with estimates of
translation probability,
which can be conditioned
on the target language if
necessary.
Experiments were run with
a language model (LM)
built on a 20 million word
news source corpus using
our in house decoder
which performs a
monotone decoding
without reordering.
TRANSLATED
Akibatnya, mereka kadangkadang tidak dapat
membedakan antara
kegagalan sistem tertentu
dan masalah yang sistem
yang terbaik akan
Efek ini juga digambarkan
oleh data dalam Tabel 2,
yang menunjukkan sedikit
penurunan dalam fraksi
massa untuk semua
permukaan dilacak dari
waktu ke waktu
Untuk simulasi kami
mendalilkan bahwa
sumbernya adalah sebuah
amplop berisi 1 gram
anthrax dibuka di Kamar 4
Simulasi ini membantu
mengidentifikasi informasi
penting atau proses yang
paling berkontribusi pada
hasil-hasil diperkirakan atau
ketidakpastian.
Pasang terjemahan Frasa
yang dihasilkan oleh metode
yang dijelaskan dalam
makalah ini akhirnya
mencetak dengan perkiraan
probabilitas terjemahan,
yang dapat dikondisikan
pada bahasa target jika perlu
Percobaan dijalankan
dengan model bahasa (LM)
dibangun di atas sumber
berita 20 juta kata korpus
menggunakan decoder kami
di rumah yang melakukan
decoding monoton tanpa
penataan kembali.
Word
A.Word
Grammar
Textual
CX.7
CX.8
CX.9
CX.10
My intention in this paper

is to provide some
explanation for the
difficulties encountered by
present computer systems
which attempt to produce
partial or complete
translations of texts from
one natural language into
another.
Tujuan saya dalam tulisan

ini adalah untuk
memberikan penjelasan atas
kesulitan yang dihadapi oleh
sistem komputer ini yang
mencoba untuk
menghasilkan terjemahan
parsial atau lengkap teks
dari satu bahasa alami ke
lain.
My aim is to give an
introduction, for those
unfamiliar with machine
translation (MT), to the
main phenomena which
must be taken into
consideration
The aim is to highlight in
the broadest terms those
areas of translation which
are relatively easy for
computerised handling and
those areas which are
relatively difficult
The basic idea is that
rather than formulating
changes between
languages in terms of
transfer operations,
translation takes place
from and into an
interlingua, i.e. elements or
structures which are in
some sense neutral for the
languages concerned
Tujuan saya adalah untuk

memberikan pengantar, bagi
mereka yang tidak terbiasa
dengan terjemahan mesin
(MT), untuk fenomena
utama yang harus
dipertimbangkan
Tujuannya adalah untuk
menyoroti dalam hal luas
wilayah terjemahan yang
relatif mudah untuk
menangani komputerisasi
dan daerah-daerah yang
relatif sulit
Ide dasarnya adalah bahwa
daripada merumuskan
perubahan antara bahasa
dalam hal pengoperasian
transfer, terjemahan
berlangsung dari dan ke
Interlingua sebuah, yaitu
unsur-unsur atau struktur
yang dalam arti tertentu
netral untuk bahasa yang
bersangkutan
Perpustakaan Unika
Appendix 11: Table 4.1.3.1 Word-level Mistake in academic complex sentences

No
CX.1
CX.2
CX.3
CX.4
SOURCE
As a consequence, they are
sometimes unable to distinguish
between the failings of particular
systems and the problems which
the best system would have.
These effects are also illustrated
by the data in Table 2, which show
a slight decrease in the mass
fraction for all tracked surfaces
over time
For our simulations we postulated
that the source is an envelope
containing 1 gram of anthrax
opened in Room 4.
These simulations help identify
critical information or processes
that contribute most to predicted
TRANSLATED
Akibatnya, mereka kadangkadang tidak dapat membedakan
antara kegagalan sistem tertentu
dan masalah yang sistem yang
terbaik akan
Efek ini juga digambarkan oleh
data dalam Tabel 2, yang
menunjukkan sedikit penurunan
dalam fraksi massa untuk semua
permukaan dilacak dari waktu ke
waktu
Untuk simulasi kami mendalilkan
bahwa sumbernya adalah sebuah
amplop berisi 1 gram anthrax
dibuka di Kamar 4
Simulasi ini membantu
mengidentifikasi informasi
penting atau proses yang paling
Mistake Point(s)
1. Systems
2. Failings
3. Problems
4. Have
1.
Effects
1.
2.
Simulations
Our
1.
2.
Simulations
Processes
outcomes or to uncertainties.
CX.5
CX.6
Phrase translation pairs that are

generated by the method described
in this paper are finally scored
with estimates of translation
probability, which can be
conditioned on the target language
if necessary.
Experiments were run with a
language model (LM) built on a 20
million word news source corpus
using our in house decoder which
performs a monotone decoding
without reordering.
CX.7
My intention in this paper is to

provide some explanation for the
difficulties encountered by present
computer systems which attempt
to produce partial or complete
translations of texts from one
natural language into another.
CX.8
My aim is to give an introduction,

for those unfamiliar with machine
translation (MT), to the main
phenomena which must be taken
into consideration
CX.9
The aim is to highlight in the

broadest terms those areas of
translation which are relatively
easy for computerised handling
and those areas which are
relatively difficult
The basic idea is that rather than
formulating changes between
languages in terms of transfer
operations, translation takes place
from and into an interlingua, i.e.
elements or structures which are in
languages concerned
CX.10
berkontribusi pada hasil-hasil

diperkirakan atau ketidakpastian.
Pasang terjemahan Frasa yang
dihasilkan oleh metode yang
dijelaskan dalam makalah ini
akhirnya mencetak dengan
perkiraan probabilitas terjemahan,
yang dapat dikondisikan pada
bahasa target jika perlu
Percobaan dijalankan dengan
model bahasa (LM) dibangun di
atas sumber berita 20 juta kata
korpus menggunakan decoder
kami di rumah yang melakukan
decoding monoton tanpa penataan
kembali.
Tujuan saya dalam tulisan ini
adalah untuk memberikan
penjelasan atas kesulitan yang
dihadapi oleh sistem komputer ini
yang mencoba untuk
menghasilkan terjemahan parsial
atau lengkap teks dari satu bahasa
alami ke lain.
Tujuan saya adalah untuk
memberikan pengantar, bagi
mereka yang tidak terbiasa dengan
terjemahan mesin (MT), untuk
fenomena utama yang harus
dipertimbangkan
Tujuannya adalah untuk
menyoroti dalam hal luas wilayah
terjemahan yang relatif mudah
untuk menangani komputerisasi
dan daerah-daerah yang relatif
sulit
daripada merumuskan perubahan
antara bahasa dalam hal
pengoperasian transfer,
terjemahan berlangsung dari dan
ke Interlingua sebuah, yaitu unsurunsur atau struktur yang dalam
arti tertentu netral untuk bahasa
yang bersangkutan
Perpustakaan Unika
1.
2.
Pairs
Scores
1.
Experiments
1.
2.
3.
4.
5.
6.
Difficulties
Texts
Translations
systems
another
present
1.
to
1.
areas
1.
2.
Operations
Structures
Appendix 12: Table 4.1.3.2 Above word-level Mistake in academic complex

sentences
No
CX.5
SOURCE
Phrase translation pairs that are
generated by the method described
in this paper are finally scored
with estimates of translation
probability, which can be
conditioned on the target language
if necessary.
TRANSLATED
Pasang terjemahan Frasa yang
dihasilkan oleh metode yang
dijelaskan dalam makalah ini
akhirnya mencetak dengan
perkiraan probabilitas terjemahan,
yang dapat dikondisikan pada
bahasa target jika perlu
Mistake Point(s)
1. Prase translation
pairs
Perpustakaan Unika
CX.6
Experiments were run with a

language model (LM) built on a 20
million word news source corpus
using our in house decoder which
performs a monotone decoding
without reordering.
Percobaan dijalankan dengan

model bahasa (LM) dibangun di
atas sumber berita 20 juta kata
korpus menggunakan decoder
kami di rumah yang melakukan
decoding monoton tanpa penataan
kembali.
1.
monotone
decoding
CX.10
The basic idea is that rather than

formulating changes between
languages in terms of transfer
operations, translation takes place
from and into an interlingua, i.e.
elements or structures which are in
languages concerned

daripada merumuskan perubahan
antara bahasa dalam hal
pengoperasian transfer,
terjemahan berlangsung dari dan
ke Interlingua sebuah, yaitu unsurunsur atau struktur yang dalam
arti tertentu netral untuk bahasa
yang bersangkutan
1.
an interlingua
Appendix 13: Table 4.1.3.3 Textual Mistake in academic complex sentences

No
CX.7
SOURCE
My intention in this paper is to
provide some explanation for the
difficulties encountered by present
computer systems which attempt to
produce partial or complete
translations of texts from one
natural language into another.
TRANSLATED
Tujuan saya dalam tulisan ini
adalah untuk memberikan
penjelasan atas kesulitan yang
dihadapi oleh sistem komputer ini
yang mencoba untuk menghasilkan
terjemahan parsial atau lengkap
teks dari satu bahasa alami ke lain.
Mistake Point(s)
1. paper
Appendix 14: Table 4.1.4 Translation Mistake in Academic Compound Complex Sentences
No
CDX.1
CDX.2
CDX.3
SOURCE
Analyzing these models
from a generative
mindset, they all assume
that the atomic unit of
lexical content is the
word, and re-ordering
effects are applied
above that level.
The question is thus
whether the
manufacturers of these
products prepared to
invest in the research
required or whether
researchers in the MT
community are going to
address this particular
need.
Japanese input demands
considerable pre
editing, but it is
acceptable to operators
TRANSLATED
Menganalisis model ini
dari pola pikir generatif,
mereka semua
menganggap bahwa unit
atom konten leksikal
adalah kata, dan efek
kembali memesan yang
diterapkan di atas tingkat
itu.
Pertanyaannya adalah
apakah dengan demikian
produsen produk tersebut
siap untuk berinvestasi
dalam penelitian yang
diperlukan atau apakah
peneliti dalam komunitas
MT akan alamat ini
kebutuhan tertentu.
Masukan Jepang tuntutan
pra editing yang cukup
besar, tetapi diterima
operator pengolah kata
Word A.Word Grammar Textual
CDX.4
CDX.5
CDX.6
CDX.7
CDX.8
of Japanese word
processors who have to
interpret Japanese script
We then discuss the
implications of adding
phrasal translation pairs
to the decoding process,
and present evaluation
results that show
significant
improvements when
applying the described
extraction technique
(Marcu, Wong, 2002)
illustrate the effects of
assuming that lexical
correspondence can
only be modeled at the
word level, and
motivate a joint
probability model that
explicitly generates
phrase level lexical
content across both
languages
However, it does mean
that before an MT
system can be installed,
the user must have
already available a wellfounded terminological
database, with
authorised translation
equivalents in the
languages involved, or
at least must make a
commitment to develop
the required term bank
MT systems are finding
a natural role, since
they can operate
virtually or in fact in
real-time and on-line
and there has been little
objection to the
inevitable poor quality
Our method is
significantly less
computationally
expensive than methods
that attempt to explicitly
model phrase level
interactions within
alignment models, and
recovers well from
noisy alignments at the
sentence and corpus
level.
Jepang yang harus

menginterpretasikan skrip
Jepang
Kami kemudian
membahas implikasi
penambahan pasangan
terjemahan frase untuk
proses decoding, dan
hasil evaluasi ini yang
menunjukkan perbaikan
yang signifikan ketika
menerapkan teknik
ekstraksi dijelaskan
(Marcu, Wong, 2002)
menggambarkan efek
dengan mengasumsikan
bahwa korespondensi
leksikal hanya dapat
dimodelkan pada tingkat
kata, dan memotivasi
model probabilitas
gabungan yang secara
eksplisit menghasilkan
tingkat frase leksikal
konten di kedua bahasa.
Namun, itu berarti bahwa
sebelum sistem MT dapat
diinstal, pengguna harus
sudah tersedia baikdidirikan database
terminologis, dengan
setara terjemahan resmi
dalam bahasa yang
terlibat, atau - setidaknya
- harus membuat
komitmen untuk
mengembangkan istilah
yang diperlukan bank
Sistem MT adalah
mencari 'alami' peran,
karena mereka dapat
beroperasi hampir atau
bahkan dalam real-time
dan on-line dan ada
sedikit keberatan terhadap
kualitas miskin tak
terelakkan
Metode kami secara
signifikan kurang
komputasi mahal
daripada metode yang
mencoba untuk secara
eksplisit tingkat interaksi
frase model dalam model
penyelarasan, dan pulih
baik dari keberpihakan
bising pada kalimat dan
tingkat corpus.
Perpustakaan Unika
CDX.9
CDX.10
It is clear that the

potential for MT on, via
and for the Internet is
now being fully
appreciated no
company can afford to
be left behind, and all
the major players have
ambitious plans, e.g.
Lernout & Hauspie
(McLaughlin and
Schwall 1998), which
has now acquired MT
systems from
Globalink, Neocor and
AppTek as well as the
old METAL system
(from GMS).
We then discuss the
implications of adding
phrasal translation pairs
to the decoding process,
and present evaluation
results that show
significant
improvements when
applying the described
Jelas bahwa potensi untuk

MT pada, melalui dan
untuk internet kini
sepenuhnya dihargai tidak ada perusahaan
yang mampu
ditinggalkan, dan semua
pemain utama memiliki
rencana ambisius,
misalnya Lernout &
Hauspie (McLaughlin dan
Schwall 1998), yang kini
diperoleh sistem MT dari
Globalink, Neocor dan
AppTek serta sistem
LOGAM lama (dari
RUPS).
Kami kemudian
membahas implikasi
penambahan pasangan
terjemahan frase untuk
proses decoding, dan
hasil evaluasi ini yang
menunjukkan perbaikan
menerapkan teknik
ekstraksi dijelaskan
Perpustakaan Unika
Appendix 15: Table 4.1.4.1 Word-level Mistake in academic compoundcomplex sentences

No
CDX.1
CDX.2
CDX.3
CDX.4
SOURCE
Analyzing these models from a
generative mindset, they all
assume that the atomic unit of
lexical content is the word, and
re-ordering effects are applied
above that level.
The question is thus whether the
manufacturers of these products
prepared to invest in the research
required or whether researchers in
the MT community are going to
address this particular need.
Japanese input demands

considerable pre editing, but it is
acceptable to operators of
Japanese word processors who
have to interpret Japanese script
We then discuss the implications
of adding phrasal translation pairs
to the decoding process, and
present evaluation results that
show significant improvements
when applying the described
TRANSLATED
Menganalisis model ini dari pola
pikir generatif, mereka semua
menganggap bahwa unit atom
konten leksikal adalah kata, dan
efek kembali memesan yang
diterapkan di atas tingkat itu.
Pertanyaannya adalah apakah
dengan demikian produsen
produk tersebut siap untuk
berinvestasi dalam penelitian
yang diperlukan atau apakah
peneliti dalam komunitas MT
akan alamat ini kebutuhan
tertentu.
Masukan Jepang tuntutan pra
editing yang cukup besar, tetapi
diterima operator pengolah kata
Jepang yang harus
menginterpretasikan skrip Jepang
Kami kemudian membahas
implikasi penambahan pasangan
terjemahan frase untuk proses
decoding, dan hasil evaluasi ini
yang menunjukkan perbaikan
Mistake Point(s)
1. Models
2. Effects
3. reordering
4. 'yang'
1.
2.
3.
Manufacturers
Researchers
Address
1.
Operators
1.
2.
3.
4.
5.
Implications
Pairs
Results
Improvements
Present
CDX.5
(Marcu, Wong, 2002) illustrate

the effects of assuming that
lexical correspondence can only
be modeled at the word level, and
motivate a joint probability model
that explicitly generates phrase
level lexical content across both
languages
CDX.6
However, it does mean that

before an MT system can be
installed, the user must have
already available a well-founded
terminological database, with
authorised translation equivalents
in the languages involved, or at
least must make a commitment
to develop the required term bank
CDX.7
MT systems are finding a

natural role, since they can
operate virtually or in fact in realtime and on-line and there has
been little objection to the
CDX.8
Our method is significantly less

computationally expensive than
methods that attempt to explicitly
model phrase level interactions
within alignment models, and
recovers well from noisy
alignments at the sentence and
corpus level.
It is clear that the potential for
MT on, via and for the Internet is
now being fully appreciated no
company can afford to be left
behind, and all the major players
have ambitious plans, e.g.
Lernout & Hauspie (McLaughlin
and Schwall 1998), which has
now acquired MT systems from
Globalink, Neocor and AppTek
as well as the old METAL system
(from GMS).
CDX.9
CDX.10
The purpose, therefore, is not to

describe the inherent limitations
of machine translation, but to
give a rather simplified
explanation of what may be
expected from systems at the
present time
menerapkan teknik ekstraksi

dijelaskan
(Marcu, Wong, 2002)
menggambarkan efek dengan
mengasumsikan bahwa
korespondensi leksikal hanya
dapat dimodelkan pada tingkat
kata, dan memotivasi model
probabilitas gabungan yang
secara eksplisit menghasilkan
tingkat frase leksikal konten di
kedua bahasa.
diinstal, pengguna harus sudah
tersedia baik-didirikan database
terminologis, dengan setara
terjemahan resmi dalam bahasa
yang terlibat, atau - setidaknya harus membuat komitmen untuk
mengembangkan istilah yang
diperlukan bank
Sistem MT adalah mencari
'alami' peran, karena mereka
dapat beroperasi hampir atau
bahkan dalam real-time dan online dan ada sedikit keberatan
terhadap kualitas miskin tak
terelakkan
Metode kami secara signifikan
kurang komputasi mahal daripada
metode yang mencoba untuk
secara eksplisit tingkat interaksi
penyelarasan, dan pulih baik dari
keberpihakan bising pada kalimat
dan tingkat corpus.
Jelas bahwa potensi untuk MT
pada, melalui dan untuk internet
kini sepenuhnya dihargai - tidak
ada perusahaan yang mampu
ditinggalkan, dan semua pemain
utama memiliki rencana
ambisius, misalnya Lernout &
Globalink, Neocor dan AppTek
serta sistem LOGAM lama (dari
RUPS).
Tujuannya, oleh karena itu,
bukan untuk menggambarkan
keterbatasan terjemahan mesin,
tetapi untuk memberikan sebuah
penjelesan yang agak sederhana
dari apa yang dapat diharapkan
dari sistem pada saat ini
Perpustakaan Unika
1.
Effects
1.
Equivalents
1.
2.
3.
Virtually
Systems
Finding
1.
2.
3.
4.
Methods
Models
Interactions
Alignments
1.
2.
Afford
GMS
1.
'penjelesan'
Perpustakaan Unika
Appendix 16: Table 4.1.4.2 Above word-level Mistake in academic compoundcomplex sentences
No
CDX.5
SOURCE
(Marcu, Wong, 2002) illustrate
the effects of assuming that
lexical correspondence can only
be modeled at the word level, and
motivate a joint probability model
that explicitly generates phrase
level lexical content across both
languages
CDX.6
However, it does mean that

before an MT system can be
installed, the user must have
already available a well-founded
terminological database, with
authorised translation equivalents
in the languages involved, or at
least must make a commitment
to develop the required term bank
CDX.7
MT systems are finding a

natural role, since they can
operate virtually or in fact in realtime and on-line and there has
been little objection to the
CDX.8
Our method is significantly less

computationally expensive than
methods that attempt to explicitly
model phrase level interactions
within alignment models, and
recovers well from noisy
alignments at the sentence and
corpus level.
(from GMS).
CDX.9
TRANSLATED
(Marcu, Wong, 2002)
menggambarkan efek dengan
mengasumsikan bahwa
korespondensi leksikal hanya
dapat dimodelkan pada tingkat
kata, dan memotivasi model
probabilitas gabungan yang
secara eksplisit menghasilkan
tingkat frase leksikal konten di
kedua bahasa.
diinstal, pengguna harus sudah
tersedia baik-didirikan database
terminologis, dengan setara
terjemahan resmi dalam bahasa
yang terlibat, atau - setidaknya harus membuat komitmen untuk
mengembangkan istilah yang
diperlukan bank
Sistem MT adalah mencari
'alami' peran, karena mereka
dapat beroperasi hampir atau
bahkan dalam real-time dan online dan ada sedikit keberatan
terhadap kualitas miskin tak
terelakkan
Metode kami secara signifikan
kurang komputasi mahal daripada
metode yang mencoba untuk
secara eksplisit tingkat interaksi
penyelarasan, dan pulih baik dari
keberpihakan bising pada kalimat
dan tingkat corpus.
RUPS).
Mistake Point(s)
1. Lexical Content
1.
2.
3.
well-founded
authorised
translation
equivalents
required term bank
1.
2.
'natural' role
inevitable poor
quality
1.
2.
3.
less
computationaly
expensive
noisy allignment
recovers well
1.
can afford
Perpustakaan Unika
Appendix 17: Table 4.1.4.3 Grammatical Mistake in academic compoundcomplex sentences

No
CDX.9
SOURCE
(from GMS).
TRANSLATED
RUPS).
Mistake Point(s)
1. acquired
Appendix 18: Table 4.2.1 Translation Mistake in Non- Academic Simple

Sentences
No
NS.1
NS.2
NS.3
NS.4
NS.5
NS.6
NS.7
NS.8
NS.9
NS.10
SOURCE
Bill Nyberg rifled
through the Hardy file
The small pin marked
him as a regular
commuter
Nyberg nodded back
His car was at the far
end of the station lot
He'd be drenched before
he got halfway across
A ten-year-old girl
named Teresa Hardy
had been involved in a
clinical trial for a new
pediatric heart
medication, Valifin
The case had to be
settled quickly
It was what Umbrella
paid him for
The helicopter spun
through the darkness
over Raccoon forest.
This wasn't a training
exercise
TRANSLATED
Bill Nyberg mengadukaduk isi file Hardy
Pin kecil ditandai sebagai
seorang komuter biasa
Nyberg mengangguk
kembali
Mobilnya berada di ujung
stasiun banyak
Ia akan basah kuyup
sebelum ia sampai
setengah di
Seorang gadis sepuluh
tahun bernama Teresa
Hardy telah terlibat dalam
percobaan klinis untuk
pengobatan hati yang
baru pediatrik, Valifin
Kasus ini harus
diselesaikan dengan cepat
Itulah yang Payung
membayarnya untuk
Helikopter berputar
melalui kegelapan atas
hutan Raccoon.
Ini bukan latihan
X
X
X
X
Perpustakaan Unika
Appendix 19: Table 4.2.1.1 Word-level Mistake in Non-academic simple

sentences
No
NS.8
NS.9
SOURCE
It was what Umbrella paid
him for
The helicopter spun
through the darkness over
Raccoon forest.
TRANSLATED
Itulah yang Payung
membayarnya untuk
Helikopter berputar melalui
kegelapan atas hutan
Raccoon.
Mistake Point(s)
1.
Umbrella
1.
Over
Appendix 20: Table 4.2.1.2 Above word-level Mistake in Non-academic simple

sentences
No
NS.4
NS.5
Before Translation
His car was at the far end
of the station lot
He'd be drenched before
he got halfway across
NS.6
A ten-year-old girl named

Teresa Hardy had been
involved in a clinical trial
for a new pediatric heart
medication, Valifin
NS.8
It was what Umbrella paid

him for
After Translation
Mobilnya berada di ujung
stasiun banyak
Ia akan basah kuyup
sebelum ia sampai
setengah di
Seorang gadis sepuluh
tahun bernama Teresa
Hardy telah terlibat dalam
percobaan klinis untuk
pengobatan hati yang baru
pediatrik, Valifin
Itulah yang Payung
membayarnya untuk
Mistake Point(s)
1.
station lot
1.
halfway across
1.
new pedriatic heart medication
1.
Appendix 21: Table 4.2.1.3 Grammatical Mistake in Non-academic simple

sentences
No
NS.2
SOURCE
The small pin marked him
as a regular commuter
TRANSLATED
Pin kecil ditandai sebagai
seorang komuter biasa
Mistake Point(s)
1.
Marked
Appendix 22: Table 4.2.2 Translation Mistake in Non- Academic Compound

Sentences
No
NCD.1
NCD.2
SOURCE
Then she heard the tick
of metal, the strangled
last gasp of the engine,
and her own thundering
heart, and realized that
they were down
He'd already reviewed
the material a number of
times, but he wanted to
be on top of every detail
TRANSLATED
Lalu ia mendengar
centang logam, jantung
terakhir terkesiap mesin,
dan gemuruh mencekik
sendiri, dan menyadari
bahwa mereka turun
Dia sudah meninjau
materi beberapa kali,
tetapi ia ingin berada di
atas setiap detail
NCD.3
NCD.4
NCD.5
NCD.6
NCD.7
NCD.8
NCD.9
NCD.10
She'd survive, but would

likely spend the rest of
her life on dialysis
We weren't struck
directly, but it was close
enough
they were delivering
him to his execution, but
they were following
orders; it wasn't
personal
The first train car was
empty of life, but
something very bad had
occurred there, no
question.
Umbrella's headquarters
had moved off to
Europe some time ago,
but they still owned
practically the entire
town.
Edward shot her an
encouraging smile, and
Sully leaned across the
crowded cabin to
reassuringly pat her leg
There were a few lights
on, but they were faint
They'd be nice about it,
but she'd have to endure
weeks, maybe months of
gentle ribbing
Dia bertahan hidup, tapi

kemungkinan akan
menghabiskan sisa
hidupnya di dialisis
Kami tidak menyerang
secara langsung, tapi
cukup dekat
mereka memberikan dia
untuk eksekusi, namun
mereka mengikuti
perintah, bukan pribadi
Mobil itu kosong kereta
pertama kehidupan, tetapi
sesuatu yang sangat
buruk telah terjadi di
sana, pertanyaan no.
Payung markas sudah
pindah ke Eropa beberapa
waktu lalu, tetapi mereka
masih memiliki hampir
seluruh kota.
Edward shot tersenyum
mendorong, dan Sully
bersandar di kabin ramai
untuk meyakinkan
menepuk kakinya
Ada beberapa lampu, tapi
mereka pingsan
Mereka akan sangat baik
tentang hal itu, tapi dia
harus bertahan minggu,
mungkin bulan ejekan
lembut
Perpustakaan Unika
Appendix 23: Table 4.2.2.1 Word-level Mistake in Non-academic compound

sentences
No
NCD.1
NCD.4
NCD.5
NCD.7
SOURCE
Then she heard the tick of
metal, the strangled last
gasp of the engine, and
her own thundering heart,
and realized that they
were down
We weren't struck
directly, but it was close
enough
they were delivering him
to his execution, but they
were following orders; it
wasn't personal
had moved off to Europe
some time ago, but they
TRANSLATED
Lalu ia mendengar centang
logam, jantung terakhir
terkesiap mesin, dan
gemuruh mencekik sendiri,
dan menyadari bahwa
mereka turun
Kami tidak menyerang
secara langsung, tapi cukup
dekat
mereka mengikuti perintah,
bukan pribadi
Payung markas sudah
Mistake Point(s)
1.
tick
1.
struck
1.
delivering
1.
Umbrella
NCD.8
NCD.9
NCD.10
still owned practically the

entire town.
Edward shot her an
encouraging smile, and
Sully leaned across the
crowded cabin to
reassuringly pat her leg
There were a few lights
on, but they were faint
gentle ribbing

seluruh kota.
Edward shot tersenyum
mendorong, dan Sully
bersandar di kabin ramai
untuk meyakinkan
menepuk kakinya
Ada beberapa lampu, tapi
mereka pingsan
lembut
Perpustakaan Unika
1.
2.
3.
shot
smile
accross
1.
2.
1.
2.
3.
lights
faint
weeks
months
endure
Appendix 24: Table 4.2.2.2 Above word-level Mistake in Non-academic

compound sentences
No
NCD.1
NCD.3
NCD.6
NCD.7
SOURCE
Then she heard the tick of
metal, the strangled last
gasp of the engine, and
her own thundering heart,
and realized that they
were down
She'd survive, but would
likely spend the rest of her
life on dialysis
The first train car was
empty of life, but
something very bad had
occurred there, no
question.
had moved off to Europe
some time ago, but they
still owned practically the
entire town.
TRANSLATED
Lalu ia mendengar centang
logam, jantung terakhir
terkesiap mesin, dan
gemuruh mencekik sendiri,
dan menyadari bahwa
mereka turun
Dia bertahan hidup, tapi
kemungkinan akan
menghabiskan sisa
hidupnya di dialisis
Mobil itu kosong kereta
pertama kehidupan, tetapi
sesuatu yang sangat buruk
telah terjadi di sana,
pertanyaan no.
Payung markas sudah
seluruh kota.
Mistake Point(s)
1.
2.
the strangled last gasp of the

engine
her own thundering heart
1.
She'd survive
1.
2.
The first train car was empty of

life
no question.
1.
Appendix 25: Table 4.2.2.3 Textual Mistake in Non-academic compound

sentences
No
NCD.5
NCD.10
SOURCE
they were delivering him
to his execution, but they
were following orders; it
wasn't personal
gentle ribbing
TRANSLATED
mereka mengikuti perintah,
bukan pribadi
lembut
Mistake Point(s)
1.
personal
1.
endure
Perpustakaan Unika
Appendix 26: Table 4.2.3 Translation Mistake in Non- Academic Complex

Sentences
No
NCX.1
SOURCE
Rebecca frowned, pushing
at the sliding door, which
was refusing to slide
NCX.2
The place felt cold and

dead, like a tomb which
was not a bad analogy at
all
NCX.3
The key you saw was

for the dining car, which
we've already sampled
NCX.4
NCX.5
NCX.6
NCX.7
NCX.8
NCX.9
He opened his eyes,

blinking, assessing his
surroundings as quickly as
his muddled mind could
manage, which wasn't very
quickly at all
He turned and walked
across the room, telling
himself it would be a waste
of ammo, which was true
Unless Billy and
Rebecca were entirely
moronic, which they'd
already proven they
were not, they would
figure out how to trigger
the structure's rotation
Not the most inviting
environment, but Billy
didn't see any zombies
or leech people, which
made it a hell of a lot
better than the last place
There was a big metal

door on the south wall,
set back into a recessed
doorway, which
appeared to be the only
other way out
What little light there
was came from a few
stray shafts of sunlight
coming down from the
motile ceiling, which
meant they were only a
TRANSLATED
Rebecca mengerutkan
kening, mendorong di
pintu geser, yang
menolak untuk slide
Tempat itu terasa dingin
dan mati, seperti sebuah
makam yang bukan
analogi buruk sama sekali
Kunci yang Anda lihat
adalah untuk mobil
makan, yang kita telah
sampel
Dia membuka mata,
berkedip, menilai
lingkungannya secepat
kacau pikiran bisa
mengelola, yang tidak
sangat cepat sekali
Dia berbalik dan berjalan
melintasi ruangan,
mengatakan dirinya akan
membuang-buang
amunisi, yang benar
Kecuali Billy dan
Rebecca sepenuhnya
tolol, yang mereka sudah
terbukti mereka tidak,
mereka akan mencari
cara untuk memicu rotasi
struktur's
Bukan lingkungan yang
paling mengundang,
tetapi Billy tidak melihat
zombie atau orang lintah,
yang membuatnya
menjadi neraka jauh lebih
baik daripada tempat
terakhir
Ada sebuah pintu besi
besar di dinding selatan,
mengatur kembali
menjadi pintu
tersembunyi, yang
tampaknya menjadi satusatunya cara lain di luar
Apa ada sedikit cahaya
berasal dari sebuah poros
tersesat beberapa sinar
matahari turun dari
langit-langit motil, yang
berarti mereka hanya

X
NCX.10
single floor from the

surface
Enrico asked the pilot,
who was tapping at
controls and flipping
switches
lantai yang tunggal dari

permukaan
Enrico tanya pilot, yang
sadap di kontrol dan
flipping switch
Perpustakaan Unika
Appendix 27: Table 4.2.3.1 Word-level Mistake in Non-academic complex

sentences
No
NCX.1
SOURCE
Rebecca frowned, pushing
at the sliding door, which
was refusing to slide
NCX.3
The key you saw was for

the dining car, which
Unless Billy and Rebecca
were entirely moronic,
which they'd already
proven they were not,
they would figure out how
to trigger the structure's
rotation
who was tapping at
switches
NCX.6
NCX.10
TRANSLATED
Rebecca mengerutkan
kening, mendorong di pintu
geser, yang menolak untuk
slide
adalah untuk mobil makan,
yang kita telah sampel
Kecuali Billy dan Rebecca
sepenuhnya tolol, yang
mereka sudah terbukti
mereka tidak, mereka akan
mencari cara untuk
memicu rotasi struktur's
flipping switch
Mistake Point(s)
1.
slide
1.
sampled
1.
structure's
1.
2.
3.
Asked
Flipping
switches

complex sentences
No
NCX.3
NCX.4
NCX.5
NCX.7
SOURCE
The key you saw was for
the dining car, which
He opened his eyes,
blinking, assessing his
surroundings as quickly as
his muddled mind could
manage, which wasn't very
quickly at all
He turned and walked
across the room, telling
himself it would be a waste
of ammo, which was true
didn't see any zombies or
leech people, which made
it a hell of a lot better than
TRANSLATED
adalah untuk mobil makan,
yang kita telah sampel
Dia membuka mata,
berkedip, menilai
lingkungannya secepat
kacau pikiran bisa
mengelola, yang tidak
sangat cepat sekali
Dia berbalik dan berjalan
melintasi ruangan,
mengatakan dirinya akan
membuang-buang amunisi,
yang benar
paling mengundang, tetapi
Billy tidak melihat zombie
atau orang lintah, yang
membuatnya menjadi
Mistake Point(s)
1.
Dining car
1.
2.
muddled mind
which wasn't very quickly at
all
telling himself it would be a

waste of ammo
2. which was true
1.
1.
a hell of a lot better
the last place

NCX.8
NCX.9
There was a big metal

door on the south wall, set
back into a recessed
doorway, which appeared
to be the only other way
out
What little light there was
came from a few stray
shafts of sunlight coming
down from the motile
ceiling, which meant they
were only a single floor
from the surface
neraka jauh lebih baik

daripada tempat terakhir
Ada sebuah pintu besi
besar di dinding selatan,
mengatur kembali menjadi
pintu tersembunyi, yang
tampaknya menjadi satusatunya cara lain di luar
Apa ada sedikit cahaya
berasal dari sebuah poros
tersesat beberapa sinar
matahari turun dari langitlangit motil, yang berarti
mereka hanya lantai yang
tunggal dari permukaan
Perpustakaan Unika
1.
2.
set back into a recessed doorway

other way out
1.
What little light there was came

from a few stray shafts of
sunlight coming down from the
motile ceiling
a single floor
2.
Appendix 29: Table 4.2.3.3 Textual Mistake in Non-academic complex

sentences
No
NCX.6
NCX.7
NCX.10
SOURCE
Unless Billy and Rebecca
were entirely moronic,
which they'd already
proven they were not,
they would figure out how
to trigger the structure's
rotation
didn't see any zombies or
leech people, which made
it a hell of a lot better than
the last place
who was tapping at
switches
TRANSLATED
Kecuali Billy dan Rebecca
sepenuhnya tolol, yang
mereka sudah terbukti
mereka tidak, mereka akan
mencari cara untuk
memicu rotasi struktur's
paling mengundang, tetapi
Billy tidak melihat zombie
atau orang lintah, yang
membuatnya menjadi
neraka jauh lebih baik
daripada tempat terakhir
flipping switch
Mistake Point(s)
1.
2.
Proven
figure
1.
leech people
1.
tapping
Appendix 30: Table 4.2.4 Translation Mistake in Non- Academic Compound

Complex Sentences
No
NCDX.1
NCDX.2
SOURCE
It was highly
improbable that one of
the virus carriers
would just happen to
blunder their way to
one of Raccoon's other
labs, but it was
possible.
Billy was gone, lost to
the fast-moving sewer
beneath the cable car
TRANSLATED
Hal ini sangat tidak
mungkin bahwa salah
satu pembawa virus
hanya akan terjadi pada
kesalahan cara mereka
ke salah satu
laboratorium lain
Raccoon, tetapi hal itu
mungkin
Billy sudah pergi, hilang
ke saluran pembuangan
yang bergerak cepat di
NCDX.3
NCDX.4
NCDX.5
NCDX.6
NCDX.7
NCDX.8
platforms, which had

swept him and that
monster in this
direction, but he was
alive, and she was
going to find him
As it turned out, the
drug did exactly what
it was supposed to do but it also caused renal
failure, and in Teresa
Hardy's case, the
damage had been
severe.
Enrico stepped to her
side, and within
seconds, five more
beams of light had
jerked toward the
gleam of metal she'd
seen, illuminating what
was, in fact, a narrow
dirt road - and an
overturned jeep
She'd been so focused
on the coming storm
that for a crazed split
second, she thought it
was thunder, even as
the helicopter tipped
wildly and dropped, a
terrible rising,
clattering whine filling
the cabin, the floor
vibrating beneath her
boots
There was no need to
talk about what would
happen when they
reached the train ... and
though the channel was
scrambled, it was best
not to say more than
was necessary,
anyway.
Lord knew there were
enough people who
had reason to sabotage
Umbrella, but it wasn't
easy to obtain even a
low-level clearance
pass for any of the
Raccoon labs
It was his life's work,

but he'd destroy it
before he'd let it be
taken away
bawah platform kabel

mobil, yang telah
menyapu dia dan rakasa
bahwa dalam arah ini,
tapi dia masih hidup, dan
ia akan menemukan dia
Ternyata, obat itu persis
apa yang seharusnya
dilakukan - tetapi juga
menyebabkan gagal
ginjal, dan dalam kasus
Teresa Hardy, kerusakan
sudah parah.
Enrico melangkah ke
sisinya, dan dalam
beberapa detik, lima lagi
berkas cahaya telah
menyentakkan terhadap
sinar logam dia melihat,
menerangi apa yang,
sebenarnya, jalan tanah
sempit - dan sebuah jip
terbalik
Dia begitu terfokus pada
badai yang datang untuk
split gila kedua, ia pikir
itu guntur, bahkan ketika
helikopter itu miring liar
dan menjatuhkan, yang
merengek mengerikan
naik, derap mengisi
kabin, lantai bergetar di
bawah sepatu botnya
Tidak perlu bicara

tentang apa yang akan
terjadi ketika mereka
mencapai kereta api ...
dan meskipun channel
itu bergegas, itu baik
untuk tidak mengatakan
lebih daripada yang
diperlukan, anyway.
Tuhan tahu ada cukup
banyak orang yang
punya alasan untuk
sabotase Payung, tapi itu
tidak mudah untuk
mendapatkan bahkan
lulus tingkat rendah
clearance untuk salah
satu laboratorium
Raccoon
Itu adalah pekerjaan
hidupnya, tapi ia
menghancurkannya
sebelum dia biarlah
diambil
Perpustakaan Unika
NCDX.9
NCDX.10
One of the cuffs was

off, the right one,
bashed open when the
jeep had gone over, but
unless he wanted to be
wearing a jangly and
rather incriminating
bracelet, he had to get
the other one off
He didn't know about
Wesker, but he was
starting to think that it
might be time to get
out of Raccoon.
Salah satu borgol itu

lepas, yang tepat, bashed
terbuka ketika jip sudah
berakhir, tetapi jika ia
ingin mengenakan jangly
dan agak memberatkan
gelang, ia harus
mendapatkan yang lain
dari
Dia tidak tahu tentang
wesker, tapi ia mulai
berpikir bahwa mungkin
ini saatnya untuk keluar
dari Raccoon.
Perpustakaan Unika
Appendix 31: Table 4.2.4.1 Word- level Mistake in Non-academic compound

complex sentences
No
NCDX.3
NCDX.6
SOURCE
As it turned out, the drug
did exactly what it was
supposed to do - but it
also caused renal failure,
and in Teresa Hardy's
case, the damage had
been severe.
happen when they
not to say more than was
necessary, anyway.
NCDX.7

enough people who had
reason to sabotage
low-level clearance pass
for any of the Raccoon
labs
NCDX.9
One of the cuffs was off,

the right one, bashed
open when the jeep had
gone over, but unless he
wanted to be wearing a
jangly and rather
incriminating bracelet, he
had to get the other one
off
TRANSLATED
Ternyata, obat itu persis
apa yang seharusnya
dilakukan - tetapi juga
menyebabkan gagal ginjal,
dan dalam kasus Teresa
Hardy, kerusakan sudah
parah.
Tidak perlu bicara tentang
apa yang akan terjadi
ketika mereka mencapai
kereta api ... dan meskipun
channel itu bergegas, itu
baik untuk tidak
mengatakan lebih
daripada yang diperlukan,
anyway.
banyak orang yang punya
alasan untuk sabotase
Payung, tapi itu tidak
mudah untuk
mendapatkan bahkan lulus
tingkat rendah clearance
untuk salah satu
laboratorium Raccoon
gelang, ia harus
dari
Mistake Point(s)
1.
did
1.
anyway
1.
2.
umbrella
sabotage
1.
2.
bashed
jangly
Perpustakaan Unika

compound complex
No
NCDX.1
SOURCE
It was highly improbable
that one of the virus
carriers would just
happen to blunder their
way to one of Raccoon's
other labs, but it was
possible.
NCDX.2

swept him and that
monster in this direction,
but he was alive, and she
was going to find him
NCDX.4

side, and within seconds,
five more beams of light
had jerked toward the
was, in fact, a narrow dirt
road - and an overturned
jeep
NCDX.5
She'd been so focused on

the coming storm that for
a crazed split second, she
thought it was thunder,
even as the helicopter
tipped wildly and
dropped, a terrible rising,
clattering whine filling
the cabin, the floor
vibrating beneath her
boots
happen when they
not to say more than was
necessary, anyway.
NCDX.6
NCDX.7

enough people who had
reason to sabotage
low-level clearance pass
TRANSLATED
Hal ini sangat tidak
mungkin bahwa salah satu
pembawa virus hanya
akan terjadi pada
kesalahan cara mereka ke
salah satu laboratorium
lain Raccoon, tetapi hal itu
mungkin
mobil, yang telah
bahwa dalam arah ini, tapi
dia masih hidup, dan ia
akan menemukan dia
Enrico melangkah ke
sisinya, dan dalam
berkas cahaya telah
menerangi apa yang,
terbalik
Dia begitu terfokus pada
badai yang datang untuk
split gila kedua, ia pikir
itu guntur, bahkan ketika
helikopter itu miring liar
dan menjatuhkan, yang
merengek mengerikan
naik, derap mengisi kabin,
lantai bergetar di bawah
sepatu botnya
Tidak perlu bicara tentang
apa yang akan terjadi
ketika mereka mencapai
kereta api ... dan meskipun
channel itu bergegas, itu
baik untuk tidak
mengatakan lebih
daripada yang diperlukan,
anyway.
banyak orang yang punya
alasan untuk sabotase
Payung, tapi itu tidak
mudah untuk
mendapatkan bahkan lulus
Mistake Point(s)
1.
2.
would just happen

blunder their way
1.
2.
cable car
that monster
1.
She'd seen
1.
2.
crazed split second

a terrible rising
1.
2.
the channel was scrambled

it was best
1.
2.
low-level clearance
any of the Racoon labs
for any of the Raccoon

labs
NCDX.8
It was his life's work, but

he'd destroy it before
he'd let it be taken away
NCDX.9
One of the cuffs was off,

the right one, bashed
open when the jeep had
gone over, but unless he
wanted to be wearing a
jangly and rather
incriminating bracelet, he
had to get the other one
off
tingkat rendah clearance

untuk salah satu
laboratorium Raccoon
Itu adalah pekerjaan
hidupnya, tapi ia
menghancurkannya
sebelum dia biarlah
diambil
gelang, ia harus
dari
Perpustakaan Unika
1.
2.
he'd
let it be taken away
1.
2.
3.
gone over
incriminating bracelet
get the other one off
Appendix 33: Table 4.2.4.3 Grammatical Mistake in Non-academic compound

complex
No
NCDX.4
SOURCE
side, and within seconds,
five more beams of light
had jerked toward the
was, in fact, a narrow dirt
road - and an overturned
jeep
TRANSLATED
Enrico melangkah ke
sisinya, dan dalam
berkas cahaya telah
menerangi apa yang,
terbalik
Mistake Point(s)
1.
Jerked
Appendix 34: Table 4.2.4.4 Textual Mistake in Non-academic compound

complex
No
NCDX.2
SOURCE
swept him and that
monster in this direction,
but he was alive, and she
was going to find him
TRANSLATED
mobil, yang telah
bahwa dalam arah ini, tapi
dia masih hidup, dan ia
akan menemukan dia
Mistake Point(s)
1.
2.
cable car
that monster
List of Abbreviations:
S
CD
CX
CDX
NS
NCD
NCX
NCDX
: Academic Simple Sentences

: Academic Compound Sentences
: Academic Complex Sentences
: Academic Compound Complex Sentences
: Non-Academic Simple Sentences
: Non-Academic Compound Sentences
: Non-Academic Complex Sentences
: Non-Academic Compound Complex Sentences
Perpustakaan Unika

07.80.0019 Srestha F

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

07.80.0019 Srestha F

Загружено:

Авторское право:

Доступные форматы

Perpustakaan Unika

An Analysis of the Translation from English Sentences to

An Analysis of the Translation from English Sentences to

ENGLISH LETTERS STUDY PROGRAMME

An Analysis of the Translation from English Sentences to

June __, 2011

Emilia Ninik Aydawati, SP. M.Hum

June __, 2011

A thesis defended in front of the board of Examiners on June 20, 2011

Chair Person : Drs. Y E Budiyana, MA

: Emilia Ninik Aydawati, SP. M.Hum

: Heny Hartono, SS., M. Pd.

Semarang, June __, 2011

Heny Hartono, SS., M. Pd.

'Madam' Angelika Riyandari, "Thanks for watching and listening to my disastrous

Background of the Study

Field of the Study

Scope of the Study

Problem Formulation ......................................................................2

Objective of the Study

Significance of the Study

CHAPTER II REVIEW OF LITERATURE .........................................................5

Early Trans Tool

Present Trans Tool

Types of Trans Tool ....................................................................10

Effective Translation. ....................................................................13

CHAPTER III METHOD OF DATA COLLECTION AND ANALYSIS ........21

Method of Data Collection

Method Data Analysis

The evaluation of Translation Result ............................................24

CHAPTER IV DATA ANALYSIS ....................................................................26

ACADEMIC SENTENCES ........................................................26

Word Level Equivalence

Above Word Level Equivalence

Textual Equivalence ....................................................................32

Compound Sentences ....................................................................33

Word Level Equivalence

Above Word Level Equivalence

Word Level Equivalence

Above Word Level Equivalence

Textual Equivalence ....................................................................50

Compound Complex Sentences

Word Level Equivalence

Above Word Level Equivalence

NON ACADEMIC SENTENCES

Word Level Equivalence

Above Word Level Equivalence

Compound Sentences ....................................................................65

Word Level Equivalence

Above Word Level Equivalence

Textual Equivalence ....................................................................70

Word Level Equivalence

Above Word Level Equivalence

Textual Equivalence ....................................................................76

Compound Complex Sentences

Word Level Equivalence

Above Word Level Equivalence

Textual Equivalence ....................................................................86

CHAPTER V CONCLUSION AND SUGGESTIONS

Table 4.1.1 Translation Mistake in Academic Simple Sentences

Table 4.1.1.1 Word-level Mistake in Academic Simple

Table 4.1.1.2 Above word-level Mistake in academic simple

Table 4.1.1.3 Grammatical Mistake in academic simple