This paper will give a general overview of the venture that is machine translation with particular focus on linguistic aspects. It will display history of MT and will deal with some of the major issues in the realisation of MT like the difficulty of translating prepositions or integrating semantics, as well as the importance of real world knowledge. To illustrate these difficulties with examples on a basic level, a practice test with a moderately complex translation engine provided by Google has been carried out and will be explained. Finally, I am going to introduce three of the largest and most powerful translation machines currently in use. I will also give a brief over-view of methods of MT. The aim of this paper is to show that the realisation of the primal idea of machine translation in its original sense, which was to perform translation without human intervention (except during the construction phase of the system), is still markedly far away at present and machines are still unlikely to take over the jobs of human translators.
Table of Contents
I. Introduction
I. 1.) What is Machine Translation?
I. 2.) Why Machine Translation Matters
I. 2. a) Social and Political Importance of MT
I. 2. b) Scientific Importance of MT
I. 2. c) Commercial Importance of MT
I. 2. d) Philosophical Importance of MT
II. The History of Machine Translation
II. 1.) The First Years of Translation Machines
II. 2.) A Pioneer: Warren Weaver, Founder of the Idea of MT
II. 3.) The Latter Years in MT
III. Machine Translation in Practice
III. 1.) MT Test: Google
III. 2.) To Avoid Mistakes
IV. Linguistic Aspects in MT
IV. 1.) Semantics
IV. 2.) Pragmatics
IV. 3.) Real World Knowledge
V. Computational Linguistics
V. 1.) Methods of MT
V. 2.) Commonly Acknowledged Translation Systems
V. 2. a) LOGOS
V. 2. b) METAL
V. 2. c) METEO
VI. Epilogue: On the Future of MT
Research Objectives and Core Themes
This paper explores the field of machine translation with a primary focus on the linguistic challenges inherent in automating the translation process. It examines the history of the discipline, evaluates current translation engines, and analyzes whether fully automated, high-quality translation without human intervention is a realistic prospect or remains a long-term scientific dream.
- Historical development of machine translation and key pioneers.
- Linguistic difficulties including semantics, pragmatics, and real-world knowledge.
- Practical testing and evaluation of modern translation software (e.g., Google).
- Methodologies such as direct translation, transfer approaches, and example-based systems.
- The future of human translators in an era of machine-aided translation.
Excerpt from the Book
IV. 2.) Pragmatics
Pragmatics is also difficult to integrate in the system of translation machines. As Levinson puts it, “pragmatics is the study of the relations between language and context that are basic to an account of language understanding” (2003:21). Here, the term language understanding is used to draw attention to the fact that understanding an utterance involves a great deal more that knowing the meanings of the words uttered and the grammatical relations between them. Furthermore, Arnold et al. state that a sentence had to be interpreted relative to the previous discourse and to the situation in which it is uttered.
The authors illustrate this with the following example (139):
The front cover should be closed.
The translation of this sentence will be affected by whether the hearer/reader will interpret the sentence as a command, as in ‘close the front cover’, or as a statement, which describes the state the cover is likely to be in.
Chapter Summaries
I. Introduction: Provides a definition of machine translation as a subfield of artificial intelligence and outlines its social, scientific, and commercial importance.
II. The History of Machine Translation: Traces the origins of translation machines from early 20th-century ideas to the influence of Warren Weaver and the later downturn caused by the ALPAC report.
III. Machine Translation in Practice: Explores the accessibility of modern translation engines via the internet and offers practical guidelines for pre-editing text to minimize machine errors.
IV. Linguistic Aspects in MT: Discusses the profound difficulties machines face regarding semantic analysis, pragmatic context, and the lack of general real-world knowledge.
V. Computational Linguistics: Details the methodologies used in machine translation, including transfer approaches and statistical methods, while reviewing systems like LOGOS, METAL, and METEO.
VI. Epilogue: On the Future of MT: Concludes that while automation is useful for specific tasks, total replacement of human translators is improbable due to the inherent complexity and creative nature of human language.
Keywords
Machine Translation, Computational Linguistics, Artificial Intelligence, Semantics, Pragmatics, Real World Knowledge, Translation Engines, Pre-editing, Post-editing, Human-aided Machine Translation, LOGOS, METAL, METEO, Syntax, Natural Language Processing
Frequently Asked Questions
What is the core focus of this research paper?
The paper provides a comprehensive overview of machine translation with a particular emphasis on the linguistic barriers that currently prevent systems from achieving high-quality results without human intervention.
What are the primary themes discussed in the work?
The work covers the history of translation machines, the technical methodologies behind various systems, the challenges of semantics and pragmatics, and the practical application of translation tools in modern contexts.
What is the main research question regarding machine translation?
The central question is whether the original goal of machine translation—performing translation completely without human intervention—is achievable or if human translators remain essential to the process.
Which scientific methods are analyzed in the paper?
The paper examines methodologies such as direct word-for-word translation, the transfer approach (analysis, transfer, and synthesis), example-based machine translation (EBMT), and statistic-based machine translation (SBMT).
What content does the main body of the paper address?
The main body moves from historical foundations to practical testing of tools like Google, followed by a deeper linguistic analysis of why machines struggle with idioms, context, and common sense.
Which keywords best characterize the research?
Key terms include Machine Translation, Computational Linguistics, Pragmatics, Semantics, Real World Knowledge, and Human-aided Machine Translation.
How does the paper differentiate between semantics and pragmatics in MT?
Semantics is described as the internal meaning and grammatical relations of words, whereas pragmatics involves interpreting a sentence relative to the specific context, discourse, and the intentions of the producer, which machines currently struggle to process.
What role does the ALPAC report play in the history of MT?
The 1966 ALPAC report is cited as a major turning point that depicted machine translation as generally unrealizable, causing a significant disruption in global research and funding for many years.
Why is "real-world knowledge" considered a boundary for AI?
The paper explains that machines lack the common-sense understanding of how the world functions, making it difficult for them to resolve ambiguities, such as identifying the correct referent of a pronoun in complex sentences.
What is the conclusion regarding the future of human translators?
The author concludes that human translators are not threatened with obsolescence; instead, machines are likely to automate tedious, repetitive tasks, allowing humans to focus on higher-level translation and system refinement.
- Citation du texte
- M.A. Alexander Täuschel (Auteur), 2006, Linguistic Aspects in Machine Translation, Munich, GRIN Verlag, https://www.grin.com/document/117455
