Scarica The Cambridge Encyclopedia of the English Language - Riassunto e più Appunti in PDF di Inglese Commerciale solo su Docsity!
Varieties Of Discourse - Chapter 19 THE USE OF ENGLISH The way English is used can depend on many situations, in which features of spoken and written language appear in an apparently unlimited number of combinations and variations. In this case, it’s important to underline the difference between text and discourse: text is a product of the use of English, and it can be spoken or written, while the discourse is the way English is used in a social context. In fact, depending on the situation and context we are in, we use different types of English, and the discourse can change by adapting our language to a specific situation, and as consequence the way we use English to produce a discourse it’s not the same when we write in English. To make it easier to understand here there’s an example : When you write a message to a friend, the modality is different from the way you write a formal message --
the choice that you make on the level of language to use is the discourse , while the effective writing is the text. If we need to analyze a discourse we must use a discourse analysis , which name’s change depending on the type of discourse and the context in which they are created: For example, if we analyze a political text, we will make a political discourse analysis ... A discourse can be studied and meticulously prepared to transmit a message , or a concept and the discourse analysis helps find it. If we focused on a Donald Trump’s message he gave, he used “ they ”, while Obama in his conferences always used “ we ”. These two choices created different effects in their audience , in fact in the case of “they” it gives the impression of referring to an external group and a distance between Trump and his public, while the use of “we” represents a connection made in a single group. We have a sense of separation opposed to a sense of community, both created without using the actual words. That makes us understand that pronouns can tell us more than we think, in fact, it lets us understand the use of the language and the message underneath it. However, language variety in a discourse can also be used to communicate ideas, beliefs, values, and every concept that is possible to express by using words. That happened in Aladdin, in fact, the movie takes place in Arabia where the main language is Arabic , but the original language is English and that’s because of Disney nationality, and in the Italian translation many linguistic features disappear. As for the protagonists that speak like Americans' mother-tongue. While the characters from the village, they speak English with a strong Arabic accent even if they live and interact together. This is because it’s a matter of social class , and it represents the power of America, in fact the protagonists represent what is good in America , while the village people are the negatives characters that have nothing to do with America. This is typical of Americans and Disney's reality. Standard American English is the desirable accent of characters who play a positive role in Disney animated films. Despite that, also Jafar , who is the bad guy, has an American accent thanks to his high social status with whom it can represent American.
This doesn’t just happen in Disney’s but in most of the American cinematic reality such as House of Gucci, and the Simpsons. The power of constructing the discourse and the modality used are very important. For example, there are many ways to say no, in fact we can see it in a negative sentence like: “I can see how that would be helpful for you. Unfortunately , we can't do that , at this time, but I will definitely share your suggestion with my own team.” In here, 4 good words and 2 bad words are used, in this way the message it’s clearly a no but it’s said in a more polite way. PRAGMATICS The philosopher H. P. Grice is well known in pragmatics research for his maxims of conversation t hat underline the efficient cooperative use of language. Pragmatics is the study of the choices that we make when we use a language , the reason for those choices, and the effects that those choices convey. These choices are made depending on the context, and to relying on pragmatics, in order to understand what it’s said, we need to share the same context. In fact, the context contributes to creating meaning , because I can say something, but if the other person, who I’m talking to, doesn’t know what I’m referring to the meaning will be different. Just as in the following example: “Could you please put the moka pot on?”, when what I’m meaning is “Could you please make a coffee?” As a result, this phrase makes sense just in a specific context not in a casual and inappropriate one. In addition, even the use of punctuation can change a hole significate in a phrase, as for the use of a full stop, commas.... For example, if I’m writing a message I don’t need to put a full stop at the end, because in a messaging context the full stop at the end is not required, while if I’m writing a book , or a poem the punctuation is necessary because it rules the rhythm of the sequences. A pragmatic approach is the use of the language depending on what we need to express in a specific situation emphasizing practical solutions and results over theoretical and grammatical rules as in the prescriptivism tradition when people were obliged to follow them in order to avoid social criticism. Being pragmatic doesn't mean that we can say anything we want, without any type of limitation, but it means that we need to use quality (need to be truthful, and support what we say by evidences), quantity (be as informative as we can), relation (what we say needs to be relevant ), manner (to be as clear as possible to avoid ambiguity and obscurity). TEXTS AND VARIETIES The notion of using English involves much more than using our knowledge of linguistic structure (using a correct grammar, vocabulary...) to create and interpret sequences and sentences and conversational interactions. It also involves being aware of the range of situations in which English can be used in a distinctive and predictable way and of the possibilities available to us when we wish to produce or respond to creative uses of language. Language can be spoken or written and both ways have important differences:
The use and the selection of a given regional dialect depends on the context , if you are talking with an American customer, it would be more appropriate to use an American English instead of British, not because he expects it but to create a connection. You can opt for a regional variation according to the context. Dialect is a regional variation, and it’s characterized by specific features of grammar and vocabulary which convey information about a person’s geographical origin , and it can be useful to know your client. It’s not a matter of how we pronounce words: but it’s about different words and vocabulary: lift-elevator, cookie-biscuit…. When two different dialects and different regional backgrounds meet, they try to get closer and become similar linguistically. This way of interacting is a regional accommodation : they converge in the way they speak. It has been demonstrating that accent’s accommodation is very frequent in accents (just a matter of pronunciation), which is different from dialect (it’s about different words, vocabulary, lexical choices...). Accent is that part of your voice which tells which country you came from and how you were educated referring to the way the same words are pronounced. The accents are very flexible , you can switch from one to another. We can switch accent depending on what situation we are in --> like Prince Harry did. He changed his accent from British English to American English: experts say he’s becoming “less posh”. Harry Styles too, he changed his accent to a mix of Australian and American accent. Experts say that they are losing their accent , and this is because of a specific situation in which they are living in in this period. Dialect and accent are strictly related , there’s a correspondence between the word you choose and the way you pronounce them. There is a standard way of pronouncing English words, that is based on Royal British English, the one spoke from Queen Elizabeth II, which is called STANDARD ENGLISH. That is why British English is associated with standard English. Received pronunciation When we talk about received pronunciation, we refer to accent which doesn’t give any geographical information at all, ‘Receive’ is because it has been passed down by the elite groups in Britain, church, royalty , considered a more prestigious way of using English in terms of pronunciation, it reflects the “ educated accent ”. Despite the falling numbers, received pronunciation is still the chief prestige accent of the country. There is a great variation of English dialects , actually there are almost 40 different varieties, like Cockney English spoken by working class Londoners, or Scouse English, which is an accent, and a dialect associated with the city of Liverpool. Another variety of English is the Estuary English spoken in the south of England in the area around London, which shares many features with Cockney. They say it will dominate general British pronunciation within 50 years. Culture talking Culture is a big part of a dialect, in fact it influences directly the language, it means using very specific references while you’re speaking to convey meaning: it’s about using specific cultural references , of a specific region, area.
For example, in: “It was like Clapham Junction in our office today”. The expression Clapham Junction refers to a very busy railway station in south London = “It was very confusing/a mess in our office today”. As we have seen it covers some less explicit meanings in some sentences that we may say. In some cases, culture talking goes beyond the cultural references, it covers some less explicit meanings in some expressions, and it belongs to a less specific, but more general regional talking. Here there are some examples:
- I hear what you say = I totally disagree and do not want to discuss this further
- Quite good = not very good
- I might join you later = there’s no chance I might coming out
- That’s not bad = poor Social variation - Chapter 21 Regional language variation provides a geographical answer to the question: “ Where are you from in the English-speaking geographical world ?” In fact, it gives you an idea of the origin of the speaker. Social language variation provides an answer to a different question: “ Who are you in the English-speaking society to which you belong? ” In this case the answer can be more than one. The same person can play different roles, may belong to different social groups , different social language varieties. For example, a mother that can also be a doctor, a political activist... and depending on these roles she can use different social language variation. It is usually language , more than any other element which is the symbol, element of our role in society. Depending on the role we play in a determined context, we use different varieties of English and language: role of friend, student, athlete, sibling, smart. So, language is directly connected with social identity. Social variation depends on many aspects , and some of them seem to have a bigger linguistic impact then others: age , sex , socio-economic status. o Sex : research has shown that man and woman use language in different way.
- For example, in the use of adjectives to describe colors. Women use nuances of color, while men’s color vocabulary is smaller than that used by women.
- Another example is the way men and women identify someone that has power. Actually, both sexes use the term “ boss ” (negative connotation), or “ leader ” (positive connotation) to describe a man in charges. While to describe a woman in charge, both sexes use the adjective “ bossy ” inclined to domineer, dictatorial, with only a negative connotation. The percentage of women using bossy to describe other women in charge is higher than men. And this is not a linguistic reason, in fact language reflects the way we think, and we see the world. To know more about that Sheryl Sandberg, chief operating officer for Facebook, has led a campaign to ban the word bossy when describing women as leaders (bossy is worse than boss). o Socio-economic status : Standard English is the product of a process of language standardization, by which conventional forms of a language are established by members of a community. The process of establishing a set of rules to regulate a language, and it can be natural or artificial.
But honorifics like Mr., Mrs., Miss. , are not gender neutral because they care about the social condition , like the choice between using one option and not another because of the age. But it’s not simply about women and men but also about situation with an unknown gender: in this case, the right thing to do is to change singular into plural so that is gender neutral sex. For example: “Every lawyer should bring … ( his/her ) briefcase.” How to make it gender neutral. “ Lawyers should bring their briefcases.” In a more colloquial and informal context we can use the pronoun “ they ” as a third person singula r form for an unspecified gender. Other contex in with we should adopt geeder neutral language is job titles, like in job advertisement in Britain this obligatory while in Italy is not. Some examples are bartender instead of barman or barmaid, or firefighter instead of fireman. Another example is how the word teacher was commonly translated by Google Translator as “ maestra ” while professor as “ professore ”. Gender neutral language applies also to humans. The word men and man originally referred to both sex while now is replace by human or people, for example mankind and now humankind. Another field of gender neutrality is, as we said before, honorifics : on one side we have the use of an unknown gender: instead of using Mr., Mrs., Miss, it’s used Mx, = /meeks/. On the other side, to use an honorific for women regardless their marital status t’s used Ms. = /mIz/ The difference between Mrs. Or Miss is also in the age of the person so depending on the age we can choose what to use but is preferred the option Ms. Gender neutrality is also legally requested on the CV that has to be sent to a company in order to not know the gender in any sort of way, and to be impartial, same for the age but of course it depends on the field. Occupational varieties Occupational varieties of language are very different from features of language which identify our geographical or social origin, in fact they tend not to vary. On the contrary, occupational varieties of language may be as distinctive as regional or class features, but they are only for a temporary use. An exception , of course, can be made for those whose work is so much part of their personality that it permanently influences their behavior, linguistically as well as socially. For example: doctors use them related to medicines when they are practicing their profession, but they may also use them during talks with colleagues about job topics, still with it but temporary ‘til they get home: the more professional is the job, the more the language is specialized. But usually, when we stop work, we stop using the language of work. Any domain could be used to illustrate occupational linguistic distinctiveness or identity, without any exception or distinction. Occupational dialect: language associated with a particular way of earning a living (job). For example, factory workers have to master an array of technical terms and administrative vocabulary in order to carry out their tasks, and in doing so they develop slang and jargon (specialized vocabulary) which set them apart from outsiders. Jargon is needed because it allows the group to function and work in a more efficient way , quickly and precisely , they have specific words for specific meanings.
Also, the more an occupation is part of a long-established tradition the more it is likely to have accreted linguistic rituals which its members accept as a criterion of performance. TL (TRANSLATING SPECIALISED LANGUAGE (S) Sometimes translators have to translate specific jargon , and to do it correctly they should domine completely the jargon involved. With that we are talking about not simply a list of words, but also a string of words linked together: they are called collocations. *Collocation refers to how words go together or form fixed relationships. Example: “ fast ” and “ quick ” are close synonyms but in English we say: “ quick shower” and not fast: if you use that people will understand you are not a native English speaker, or “ fast food” and not quick food: they are words that work together in the same place. --> COLLOCATION = Co (together) + location (place) We have strategies to translate these:
- Looking online for the collocation between a double quotation mark (“) or using an asterisk () Examples: search for possibilities, “radici della carriera” (we use asterisk because it is a possessive object --
radici something della carriera). There are also collocation dictionaries available. to understand the style , we should adopt while translating , we could be helped by finding similar text in Italian , to have a model to rely on and to refer to. Example: you must translate this: https://contently.com/2014/12/05/how-airbnb-is-using-content- marketing-to-stay-on-top/ So, take a look at this: https://www.nextredigital.it/airbnb-un-esempio-eccezionale-di-content-marketing/ This is native text, written in a language they are interested in, not a translation. They are comparable because they are used as models to translate our text. How to get a corpus of comparable text: A corpus is a collection of written texts , especially all the works of a particular author or a body of writing on a particular subject. In order to compare them it’s important to:
- Analyze the text in English.
- The source and context of the text ( is it a social media? ) = the type of text
- Audience , that means readers: are they specialized? General public? Etc.
- Understand the purpose that is the goal.
- Main topic , what's the text about ( text about linguistic.. .)
- Find the specific subtopics (... focusing on pragmatic ) These elements can guide the search to find KEYWORDS (examples: blog, Airbnb, localizations.). How comparable texts can be used as translators? To find collocations To analyze style To understand terminology
When we talk about personal variation there’s a very important element to mention: Deviance: When something does not conform to rules or norms with a given language it is known as deviance. It produces instances of language which are totally unacceptable and considered as a mistake : cat the // the cat. Deviance doesn’t mean that everything goes, it works if we produce something deviant for a communicative purpose. A deviant or strange use of language may be highly effective and widely appreciated. Here there are some forms of deviance:
- We create an adjective by taking a noun: “ Ritz” in this case a proper noun, and we add se suffix of an adjective to make it very peculiar (ritzy) “v ery ritzy Ritz ” --> “l’unico Ritzeggiante Ritz ”; “ caulilipower ” --> “ cauliflower” ,” romaintic” --> “ romaine ”. What is realized it’s a combination of words for a noun that doesn’t exist, in order to create something more attractive.
- Verbal Humor is another type of linguistic deviance which occurs in informal conversation, which comes from a humorous use of wordplay. In translating this type of instance, you have to keep the same word play and match the image. Play with words in order to maintain the same effect. The archetypal of humor are jokes, that are said or done to provoke laughter. Especially a brief oral narrative with climatic humorous twist. While puns are the most sharply focused kind of verbal humor: they consist in bringing together in a single word two unrelated meanings are suddenly and unexpectedly, and the incongruity makes us laugh --> “W anna be with you all the lime ” (time).
- Dialect humor : regional dialects and (especially in England) class dialects are rich source of conversational language play. The regional dialect within a community readily leads themselves to comic exaggeration. “ What beautiful fla-wer ” --> They use the exaggerated and incongruous use of varieties of English.
- Nonce words : new words created “for a particular occasion” when a word is on the tip of the tongue, but you can’t remember it, so you create a new word to get our meaning cross : “ He is very Jack-in-the-boxy " (word) because you have near the store Jack in the box. You come up with a different word to fill the gap. Peculiar use of the language, culturally specific. Artificial Intelligence - Seminar 25/ In recent years, the intersection of artificial intelligence (AI) and language technologies has created a dynamic shift in the job market, leading to the emergence of new professions and the evolution of existing roles. As AI continues to play a central role in enhancing linguistic processes, professionals with expertise in both linguistics and technology are in high demand. However, the job titles in this domain are still in flux, as
companies look for individuals capable of bridging the gap between AI systems and human language needs. "Artificial Intelligence": a definition Artificial intelligence is a label of an expanding " family " of software systems that is capable of performing specific tasks by collecting, analyzing, and interpreting data. It’s a Generative AI, a subset of artificial intelligence that focuses on creating new content, such as images, text, music, and more, often mimicking human creativity. It’s capable of perceiving the environment to make decisions and take actions with a certain degree of autonomy. Its increase is crucial in everyday life and social relations. Emerging Job Titles in AI and Language Technologies The landscape of AI-driven language work is evolving rapidly. Some of the most prominent job titles currently emerging in this field include: AI Research Scientist : A professional responsible for investigating and advancing AI methodologies, especially in language processing. AI Scientist : A role focused on the development of new AI algorithms with an emphasis on linguistic applications. AI Engineer : Engineers who specialize in creating AI systems and applications, including those used in translation and natural language processing (NLP). AI Developer : Professionals who build and implement AI solutions tailored to language tasks. AI Semantics Engineer : Focused on the semantic aspects of language, including meaning, context, and interpretation in AI models. AI and ML (Machine Learning) Specialist : A professional proficient in both AI and ML, driving advancements in language-related AI tools. AI Language Specialist : An expert in applying AI to linguistic tasks, improving the interaction between humans and AI systems. AI Project Manager : Responsible for overseeing AI projects related to language technologies, ensuring timely delivery and integration. Chief AI Officer : A senior leadership role in a company overseeing all AI strategies, particularly those integrating language technologies. Conversational AI Expert : Specializes in building AI systems that facilitate human-like interactions, such as chatbots and virtual assistants. Prompt Engineer : Develops and refines prompts for AI systems to ensure high-quality and contextually appropriate responses. ML Data Linguist : Combines expertise in linguistics with machine learning to analyze and optimize language data used in AI models. These job titles reflect the growing integration of AI , linguistics , and machine learning (ML) across various industries, including technology giants like Amazon and Apple , as well as more specialized linguist platforms like LINGUIST List and Career Linguist. To adapt to these new job requirements, language students and professionals are being encouraged to upskill in areas such as data annotation , technical writing , and language training. Universities are
Written/verbal communication skills; attention to detail. analytical skills; problem solving; ability to thrive in a fast-paced environment Tasks in AI Language Roles The analysis of job advertisements also highlighted key tasks that professionals in the AI language field are expected to perform: Data and Research : Analyse data; collect data; perform data annotation; update/create annotation guidelines Conduct research; develop policies and procedures. Technical : Improve OR develop NLP tools; build language models; train models; analyse, test OR improve tools' performance. Miscellaneous : Perform quality controls/assurance; ensure consistency of localizations; provide (feedback on) translations. Perform phonetic transcription. Interestingly, while employers in the translation industry increasingly require proficiency in translation technology , they do not expect candidates to be able to create new translation tools themselves. Instead, they are looking for professionals who can collaborate with these tools , manage data , and evaluate AI- generated translations to ensure their quality and accuracy. AI in Education and Its Impact on Language Learning AI-powered applications are making significant strides in the field of education. Through platforms such as Duolingo , Babbel , and various virtual tutors, AI is providing personalized learning experiences and real- time feedback. These tools use Natural Language Processing (NLP) to engage students in conversational settings, helping them develop their language skills in a more natural, interactive way. The application of generative AI —which creates new content such as text—has transformed how learners practice and improve their language abilities. The rise of Conversational AI technologies, such as chatbots and virtual assistants, is reshaping language learning. These systems are designed to mimic human conversations and offer users an opportunity to practice language in real-world contexts, thus enhancing their skills through continuous, dynamic feedback. One area where AI has already shown considerable promise is in language learning. AI-powered chatbots and virtual tutors are being used to help individuals practice English and other languages. These tools provide learners with the opportunity to engage in conversational practice at their own pace, receiving explicit feedback on their language use. The advantage of AI in this context is that it is available anytime and anywhere, reducing anxiety and offering immediate corrections. The interactionist approach to language learning—where language is learned through practice and interaction—has been enhanced by AI, especially in the form of generative AI (like ChatGPT) which can provide personalized, real-time responses. The effectiveness of these tools in improving language skills, particularly in speaking, listening, and reading, is undeniable. While there are still challenges, such as the
need for good instructions and feedback mechanisms, AI is becoming a valuable tool in language acquisition. The Debate: Is AI Truly Intelligent? While AI has become an integral part of everyday life through applications such as digital assistants (e.g., Siri, Alexa), recommendation systems , and facial recognition , there remains an important distinction between weak AI and strong AI. Weak AI (such as ChatGPT) is designed to perform specific tasks, like generating text or translating languages, but lacks general cognitive abilities. It can simulate human-like responses but does not possess true autonomy or understanding. Strong AI , also known as Artificial General Intelligence (AGI) , would have cognitive abilities on par with human intelligence. However, AGI has not yet been achieved. The current generation of AI, including ChatGPT and other language models, operates as weak AI —it generates plausible responses based on vast datasets but lacks the ability to truly "understand" or exhibit human-like creativity. Ethical Concerns: Bias and Exclusion in AI Systems Despite the many advantages of AI in language applications, there are significant ethical concerns. A significant issue raised during the seminar was AI bias, especially in natural language processing (NLP). AI systems, particularly those based on large language models , can perpetuate algorithmic bias —systematic preferences or prejudices embedded in the training data. These biases may include: Gender Bias : AI may use gendered language that reflects societal stereotypes, such as using terms like "housewife" or "fireman” or “chairman”, or assume a gender based on stereotypical job roles. Ableism : AI may inadvertently favor certain abilities or exhibit negative biases against people with disabilities. Describing people with disabilities or health conditions as "wheelchair-bound" or "victims" of a disease; " "physically challenged," Ethnic and Religious Bias : AI models can perpetuate harmful stereotypes or biases about different ethnic, religious, or cultural groups. Referring to members of one religion as "fanatic" and to those of another as "devout". Indigenous Peoples: outdated or pejorative terms such as "Eskimos" The term "tribe" is derogatory "minorities"; "people of color" ("minority" can have negative connotations about being "less than" other populations) The normative bias in AI refers to the ways in which AI systems can inadvertently reinforce societal norms and exclude certain groups, often without explicit discriminatory intent. AI systems trained on sociocultural data can reproduce these biases, which raises ethical concerns about fairness and inclusivity. Moreover, AI systems that are not designed with accessibility in mind can exclude individuals with disabilities. For example, voice-activated virtual assistants may not recognize or respond accurately to speech impediments, and visual-based systems may present challenges for users with visual impairments. As AI continues to play a larger role in our daily lives, addressing these biases is crucial to ensure that AI systems serve all users equitably. Research is ongoing to identify exclusionary practices and develop AI systems that are more inclusive and sensitive to diverse user needs. To give a resume: Observed Implicit Discrimination
Un altro esempio è il caso Balocco di Chiara Ferragni, il testo del video realizzato per scusarsi non lo ha scritto lei, bensì ha studiato un testo scritto da altri basandosi sulla sentiment analysis fatta da quello che gli utenti hanno detto a riguardo nei 6/7 mesi precedenti, ed è stato studiato per scrivere quel testo per contrastare quel sentiment. Ad esempio, la parola “frode” non è mai stata usata in quel contesto perché avrebbe richiamato quel sentiment negativo. Quando quel testo è stato tradotto in inglese c’è stata massima attenzione nel mantenere quel sentiment. In una serie tv viene fatta la stessa cosa riguardante i personaggi, ossia vedere come i personaggi vengono percepiti dal pubblico, tant’è che i personaggi scompaiono, appaiono, o si allontanano dalla scena per un determinato tempo. Electronic variation - Chapter 23 Electronic variation is a feature born with the electronic revolution , along with chatrooms, online games... that gave definition to a new domain of study called “ Electronically mediated communication ”, or “ Electronic digital discourse ” or simply "" Netspeak ”. The electronic revolution created the emergence of a new range of language varieties , as people learned to adapt their language to cope with the linguistic constraints and opportunities provided by the new technology. Some properties of electronic variation are that: When people are in front of a keyboard their behavior is very close to the traditional task of writing , for example the term “chat”. The nature of the interaction is more reminiscent to speaking. Unlike conversations there’s no immediate feedback, or prosody but it permits simultaneous interactions (Online we can have 10 chats opened at the same time, while in real life we can’t have different conversations with 10 different people simultaneously). Yes: it has dynamic graphic effects (emoticons), it’s possible to manipulate a writer’s output and it’s dependent on non-linear reading practices. Adapting the language to space constraints: As for electronic varieties one main characteristic is that we need to adapt the language to space constraints. In fact, electronic variation: Influences by the size and shape of the screen (for newspaper in the past, now with the electronic devices). Influences the organization of the available virtual space into functional areas. Its brevity of expression is a priority – it needs to be concise. The brevity of expression is responsible for the emergence of abbreviated conventions also called “Textese” (before --> b4). This is a challenge for the electronic translation and symbolizes the fact that language can change “ from below ” (initiated by the users themselves), or “ from above ” (initiated by the technology companies). Ex. Change from --> What are you doing? --> To --> What’s happening? Another example of how language changes is lexical distinctiveness, which shows how electonic variation is also flexible and dynamic. In fact, new words are creating by different techniques:
Compounding : by putting two words togetherogether as for: mouse + pad = mousepad Affixation : by using a suffix or a prefix that comes from online communication --> Ex. Chatbot: “ – bot” is the suffix; unfriend “–un” is the prefix_._ Conversions : when a word-class changes, usually a noun becomes a verb. Ex. “ghost” --> to ghost; or unfriend --> to unfriend. Business communications targeting sentiment (sentiment analysis) Sentiment analysis is a survey of all the words in a language that expresses positive or negative attitudes. With sentiment analyisis we collect and analyse informations on how people talk considering emotions and opinions. What to do to make a social media sentiment analysis:
- Find the conversation people hare having about your topic: collect hashtags and keywords and data.
- Look for terms that indicate sentiment within your mentions (usually adjectives): positive or negative words people use to talk. Ex positive: love, like, perfect, great... Negative: bad, awful, terrible... What is the role played by translator? Should maintain the same sentiment when localizing a product. Some applications of sentimental analysis are social media monitoring , brand monitoring , voice of costumer VoC (reviews), customer service (interaction by emails, and messages), market research (analyse your competitors and know what they say about them). Corpus revolution and artificial intelligence revolution How artificial intelligence tools can help translators in the world. A Corpus is a collection of texts (è termine usato nella nomenclatura bibliografica, fin dall’antichità classica, per indicare la riunione di più opere , raccolte e pubblicate con lo scopo di fornire delle serie ordinate e complete degli scritti di uno o più autori o intorno a determinate materie), naturally occurring , and not built on purpose, so they come from the natural use of the language (texts coming from New York Times, social media, platforms…). They give the context in which the expressions are used, not like grammar books (they are not created to teach English). Corpora (plural of corpus) can represent the language in its entirety or as a variety of the language (for example only British English, or American English, or a dialect like Cockney…). We can have corpora made of written texts , spoken texts or mixed. They may represent a specific genre, texts made of articles, books, newspapers, poems… When you create a corpus, you have to be careful while selecting the text , in fact, the corpus needs to be as representative as possible, they have to represent the language you are translating. What is the main purpose of a given corpus for translators? Why are corpora important for translators? We can use a corpus to understand how a given language behaves in its natural context (if you want to know more about the language used in books you need to have a look at them). It enables us to make more objective and confident descriptions of usage, as well as comparative statements about usage in different varieties.
What are the words which are more rather to be found around the word shareholder? What is the difference between shareholders and stakeholders? How can we translate shareholder in Italian? By examinate how the word is used in specialized text written in English we can have a better idea of the meaning and the collocation of the word, and how it is used. For collocations, what comes before and after. To understand the semantic prosody, read the words that follow, if they are positive or negative. A translator must use a specialized corpus of English, we can’t translate technical text without knowing this. Bilingual corpora It occurs that exist also bilingual corpora, which are made of translations , source texts and a corresponding translated text, original texts and their corresponding translation. If we look the word “shareholders”, we will find possible translation of shareholders like “azionisti” but also wrong alignments like "imprese”, we need to be careful when comparing the original with the translation and choose the right one for our translation. Our knowledge on a specific topic is limited, we may be very good on one topic but not know anything about architecture, so if we want to translate a text about it, we need to know more , and corpora are very helpful in this case. The artificial intelligence revolution: ChatGPT as a new tool for translations. What is prompting? Also known as “ prompt design ”, prompting refers to the process of formulating a request with the aim of guiding a system to generate a response , output, or perform an action. Structuring the prompt appropriately allows for an output with the desired characteristics and ensures accurate or useful results for a given purpose. The way a prompt is formulated significantly influences the output. The way we say a prompt will influence the way ChatGPT reacts, it’s again about linguistic choices. Prompts sometimes need to be translated. They can be structured in different ways : CATEGORIES With direct questions: How can I say buongiorno in English? What does it mean viaggiare? With instructions: from a linguistic point of view, we moved from question to imperative: Please translate the following sentence from German to English; Please provide a translation of this paragraph maintaining the tone and meaning... With “ persona” prompts : you are asking them to perform a role. To do that we have to instruct the chatbot to be a professional translator, so we have to be good translator in order to make the chatbot as good as us (an engineer won’t be able to do say to the chatbot what is needed): You are a professional translator, translate the following legal document from Spanish into English ensuring accuracy and formality. --> This is a basic prompt, the more I make it specific the more the answer will be accurate. It’s like a role play but without an interaction with the chatbot: it’s an instrument not a chat. It should activate a range of skills related to that specific role. Constructing according to specific linguistic guidelines.
Of course, after receiving a chatbot translation we can’t publish it, but we have to revise it and proofread it. The more our prompt is specific and detailed the more the result will be closer to what we want, and to know want we want we need to know that topic.
