Práticas de Tradução Pt-Ing

03/10/2023

Theory of Translation

Translation transferre

Trans= across ferre= to carry or to bring

Transferre

To carry or to bring

(kasparek 1983)

Process Product

  • Translation is a communicative act of transport between languages.

  • It enables communication between members of different cultural communities

  • It bridges the gap between situations where differences in verbal and nonverbal behaviour, expectations, knowledge and perspective are such that there is not enough common ground for the sender and receiver to communicate effectively by themselves.

  • Translation is the art of making choices- choices of the right words, phrases, sentences, structures, etc.

Equivalence

1964

Eugene Nida

  • Formal: focus on the message itself both in form and content

e.g. if you have slang in the source, we don’t use the same word in the target, but you use another word in the local target source.

  • Dynamic: was not about the message but rather the effect that the message had

If you have a long sentence with 3 lines, you could theoretically change the two sentences, but the effect of message remains

1965

John Catford

Scottish linguist

-Level shifts: in the source and in the targe you need to change from one level to another level

e.g. I have just seen John- translate into Pt Acabei de ver o João

-Category shifts: maintain the same level but you just change the category

-Structure:

I drink water- Bebo água

- class: change the verb to a noun

Changing to shampoo verb in English in pt

  • Unit: cottage type of house, word that doesn’t exist

Intra-system shifts: sometimes you change words more than others

1995

Werner Koller

Swiss linguist

-Denotative: intend to refer to everything that was extra linguistic

- Conotative: intend to elect all the

Consider things like synonyms, register (informal/ formal)

Text-normative:

Pragmatic: easy going in the source is going to be the same in the target

Formal:

1992/2011

Mona Baker

Egyptian, British translation, studies academic

  • Word level: sometimes there is no equivalent.

  • Above word level: phrases or a stretch language

  • Grammatical: in one language we use a verb and sometimes we use a noun

  • Textual:

  • Pragmatic: it’s about the context itself

Semantics vs Pragmatics

Semantics: meaning of the word

Pragmatics. Issues about time, place

Everything related to background knowledge, intentionality, contextuality, acceptability social culture situation

Refers to how words are used in a practical sense

Will you crack open the door? I am getting hot

10/10/2023

Translation norms

Word-for-Word – Literal

VS

Meaning- for-meaning

Equivalence

Vs

Non-equivalence

IS THERE A ONE-TO-ONE RELATIONSHIP BETWEEN WORD AND MEANING??

Tenista is considered as professional and tennis player is considered as jogador de ténis.

“Words have blurred edges”

Non-equivalence

Common problems

Vs.

Strategies to deal with non-equivalence

Culture-specific concepts

e.g. airing cupboard:

Source language concept is not lexicalized in target language

Spicy:

Savoury: in English something can be savoury and not necessary be very salt.Can be everything but not desert

Source language word is semantically complex

Tomar café: in portuguese really means hang out with friends and chat

Vs

Tomar um café :

Source and target languages make different distinctions in meaning

Mackerel: in Portuguese is carapau

Target language lacks a superordinate

Facilities: the general term is missing

Target language lacks a specific term

Jump:

Plummet/plunge/ dive

Bounce/ leap/ vault/ spring

Differences in physical or interpersonal perspective:

Uncle in Chinese:

Differences in expressive meaning :

homosexuality in Arabic: in Arabia means sexual perversion

Differences in form

Affixes

Differences in frequency and purpose of using specific forms

ing form: use to express the future in Portuguese we don’t use

Use of loan words in the source text:

Chic

“false friends”

NON – EQUIVALENCE Common problems

vs

Strategies to deal with non-equivalence

Cultural substitution: Lanchar:

Use of more general word

To shampoo: in Portuguese we say lavar com

Use of more neutral/ less expressive word

To mumble:

Use of loan word or loan word plus explanation

Fazer um like no Facebook.

Tirar uma selfie.

17/10/2023

Paraphrasing using related words

Creamy= which resembles cream

Milky: contains or resemble milk

Paraphrasing using unrelated words

Affidavit: any official document that stated that I’m the one who did it.

You paraphrase using the words that are not the same.

Simple omission :

Illustration

Tagged teabags : use the picture to explain what is inside

Mona baker

Translation strategies :

7 procedures to build equivalence :

Borrowing from ST

Calquing: borrowing a term and then translating literally (sci-fi)- ficção-cíentifíca, levar uma knaifada, )

Literal translation: direct transfer

Transposition: replacing word from one class by another class (no smoking- proibido fumar)

Modulation: (até logo- see you later)

Equivalence: Finding an expression or idiom that can replace the source (isto é canja- it’s a piece of cake )

Adaption: a complete replacement of a cultural ( eng Shakespeare-pt- Camões,

Anthony Pym

Different terminology

Translation solutions:

(instead of techniques or strategies)

Norms in Translation

Perceptions:

  • Numerous alternative translation solution/strategies

  • Translator’s decision and responsibility

  • Sameness is not always possible

  • Translation problems (non-equivalence) are common

Non-equivalence: Grammatical differences, lexical non-equivalence, ambiguity, cultural

24/10/2023

Machine Translation / Artificial Intelligence

Vs

Machine Assisted Translation

Some facts

  • While MT systems sometimes produce howlers, there are many situations where the ability of MT systems to produce reliable, if less than perfect, translation at high speed is valuable.

  • MT/AI does not threaten translator jobs. The need for translation is vast and unlikely to diminish, and the limitations of current MT/ AI systems are still numerous.

  • There are still many open research problems to be solved before MT/ AI systems become close to the abilities of human translators.

  • Building these systems are arduous and time consuming jobs, involving the construction of grammars and very large monolingual and bilingual dictionaries. There are no magic solutions.

Cognition and MT/AI

  • Thought is the mechanical manipulation of abstract symbols, but thought is also embodied. It is influenced by the nature of the human body.

  • Even though the mind can be considered an abstract machine (manipulating symbols in the same way a computer does), thought is more than the mechanical/ artificial manipulation of symbols. Symbols only get their meaning via correspondences to things in the real world.

  • Thought is abstract but it does require a human body. Thought is not independent of the human body, the human perceptual system, and the human nervous system.

  • Our minds do not manipulate symbols the same way computers do. General cognitive mechanisms are essential to language.

  • Mathematical logic is insufficient to model human thought because thought is imaginative and goes beyond a simple correspondence to things in the real world. Will AI ever be imaginative enough?

What is machine translation?

“ a process consisting of the translation from one natural language into another performed by a computer system”

Dorr, Jordan, & Benoit (1999)

  • Rule-Based Machine Translation is when grammar and syntactic rules are analysed

(identifying the deep structures of a source language sentence), transferred (finding the corresponding deep structure in the target) and then generated ( creating a target language sentence from the deep structure).

  • Example-based Machine Translation proposes a comparison of sets of parallel sentences.

  • Neural Machine Translation is a system using mathematics to execute translations.

  • Statistical Machine Translation encompasses several different methods to produce a translated text without human intervention but relying heavily on statistics.

Carmo (2017)

ARTIFICIAL INTELLIGENCE TRANSLATION MODELS

AI uses multiple State of the Art (SOA) models: e.g., Deep Learning (DL), Natural Language Processing (NLP) Machine Learning (ML), Large Language Models (LLMs) = an extrapolation both ML-NLP models

It cannot be denied that AI is:

innovative;

relatively logical;

cohesive;

pertinent;

fluent;

Translating/Revising/ Post-editing

“revise is the new translating”

“using professional translators to revise and optimize, i.e. to post-edit translation produced by an MT system can lead to significant productivity gains”

“A number of my recent graduates have managed to find permanent in-house jobs- not as translators but as revisers of MT.”

“While MT fully replacing human translators seems unlikely human-machine interaction is certainly becoming a larger part of the work for many professionals translators”

31/10/2023

Machine Translation (MT)

Vs

Machine Assisted Translation (MAT)

MAT is normally taken to mean

The book, uses your own memory bank.

Star: