Translation, Lexicography and Computational Linguistics Questions and Answers

Answer: A

'Move' in computational linguistics is a last resort operation.

In the Minimalist Program, Move is triggered only when a feature has to be checked and no other way of satisfying it exists, so it applies as a last resort. In the options this appears with the misprint 'last report operation', which is the same wording.

Move is not described as costly in the sense of a free option. It is not optional, because it applies only when needed, and it is not a universal operation in the sense of applying freely. The key keeps the first option.

Answer: B

The transfer method assumes two intermediate stages.

In transfer-based machine translation the source text is analysed into an abstract source-language representation. A transfer module maps it to an abstract target-language representation, and the target text is then generated. These two representations are the two intermediate stages.

The interlingua method uses a single intermediate representation, and the direct method has none. Three or four intermediate stages are not part of the standard transfer design.

Answer: A

Both assertions are correct. Translation is the replacement of a text in one language by an equivalent text in a second language, which is a standard definition.

Translation is also described as a highly motivated, industrial activity supported by information technology, which has diversified in answer to the needs of this kind of communication. Both agree with the accepted view, so the options calling either wrong are rejected. The second assertion is a description of translation as a modern industry, in which tools such as translation memory and machine translation are used.

Answer: D

The correct matching is Eugene Nida and C. R. Taber with The theory and practice of translation, Basil Hatim and J. Munday with Translation: an advanced resource book, Peter Newmark with About translation, and Susan Bassnett and Andre Lefevere with Translation, history and culture. That is a-ii, b-iii, c-iv, d-i.

The other codes give these titles to the wrong authors. Each of these books is well known in translation studies, and the authors are best remembered by the title of their best known work.

Answer: C

The theory of translation based on a comparison of the structure of the source and target languages, and not of literary genres and style, is the linguistic theory. It uses the findings of linguistics to compare the two languages and work out equivalents.

The philological theory compares literary genres and style, the philosophical theory looks at the nature of language and meaning, and the text-type theory classifies texts by function, so none of them is built on a structural comparison.

Answer: A

Assertion (I) is false and assertion (II) is true. In a reverse dictionary the entries are ordered from the last letter, so words with the same ending stand together.

Urdu dictionaries in general are not reverse dictionaries; they are arranged in the usual order from the first letter. So only (II) is true, option (A). Where the arrangement is by the final letter, as in a reverse dictionary, it is useful for studying endings, rhymes and suffixes, which is not the purpose of an ordinary Urdu dictionary.

Answer: B

The correct matching is glossary with technical terms, dictionary with common vocabulary, thesaurus with related concepts, and electronic dictionary with digital. That is a-ii, b-iii, c-i, d-iv.

A glossary explains the technical terms of a field, a dictionary covers the common vocabulary, a thesaurus groups related concepts, and an electronic dictionary is a digital resource. The other codes assign these to the wrong kinds.

Answer: C

Statements a, b and d are correct for POS tagging. It breaks the sequential information in text into parts, marks up each word according to the grammatical class it belongs to, and simplifies the natural language for the computer into a form that is easier to recognise as patterns.

Statement c is not a description of it: putting hardware in computers for processing speech is not a task of tagging. Options with c are therefore wrong.

Answer: B

The correct matching is speech processing with speech recognition, information retrieval with stemming, parsing with the formal analysis of a sentence by computer, and OCR with pattern recognition. That is a-iii, b-iv, c-i, d-ii.

Stemming is used in information retrieval to reduce words to a common root, and optical character recognition is a task of pattern recognition. The other codes give these tasks to the wrong fields.

Answer: C

Speech synthesizers are used together with speech recognizers in artificial intelligence. A system that can both understand and produce speech, as in dialogue systems and voice assistants, combines the two.

Neurological disorders and articulation are subjects of the clinical and phonetic study of speech, not of this kind of pairing, and "none of the above" is wrong since artificial intelligence is the answer.